AMERICAN TELEGRAPH PRACTICE McGraw-Hill Book Company Electrical World The Engineering andMining Journal En5ineering Record Engineering News Railway Age G azette American Machinist Signal kngin4 part of the sulphuric acid which unites with the zinc. Thus when the hydrogen is liberated the zinc takes its place. Zinc sulphate and hydrogen are pro- duced by sulphuric acid and zinc. The equation representing the action which takes place is expressed in the following manner: Action continues only as long as the external circuit is closed, that is, when the positive element employed consists of chemically pure zinc. When, how- ever, commercial zinc is used, which in many instances contains impurities 14 AMERICAN TELEGRAPH PRACTICE such as carbon, tin, arsenic, lead, iron, etc., there is a local action going on which results in the consumption of zinc without the desired production of useful current. For the purpose of overcoming this local action it is customary to amalgamate the zinc elements used in chemical batteries. The process of amalgamation consists of cleaning the surface of the zinc by immersing it in an acid and then rubbing upon the zinc a coating of mercury which unites with the zinc and forms an amalgam paste. The foreign matter contained in the zinc does not dissolve in the mercury but floats to the surface and is carried off by the constantly forming hydrogen bubbles. The zinc associated with the mercury dissolves in the acid solution and the mercury coating con- tinually uniting with fresh portions of zinc results in a clean bright surface of zinc being at all times presented to the attacking acid. As before stated, the hydrogen set free is carried away in the form of minute bubbles, and while most of these bubbles reach the surface of the liquid there releasing their charge, an increasing number of them gather upon the surface of the copper electrode, the longer the cell is used. Now, as these hydrogen bubbles are electropositive in practically the same sense as is the zinc element itself, the copper element by virtue of this accumulation of hydrogen bubbles upon its surface, is converted into a positive element, at least that would be the result if the action were permitted to continue indefinitely. This action is called Polarization. There are several well-known ways of overcoming this tendency to polarize. Theoretically, the desired end might be accomplished simply by brushing the bubbles off the cathode, but in practice it is desirable to secure the same result automatically. It is customary to employ as a constituent of the cell a substance with which the hydrogen gas will readily combine. Such substances are called Depolarizers. Depolarizers may be either solid or liquid. When solid, the usual method is to shield the cathode with a substance in porous form as in the Leclanche and Fuller types. When a liquid depolarizer is used, if its specific gravity be less than that of the electrolyte, they will be kept separate by plac- ing the lighter on top of the heavier liquid as in the gravity cell. There are several depolarizers which may be used with good results, namely, oxide of copper, peroxide of manganese, nitric acid, permanganate of potash, chromic acid, bromin in caustic soda, and sulphate of copper solution. The Gravity Cell. In the gravity cell the depolarizer used is sulphate of copper (solution). Fig. 2 shows the usual type of gravity cell used in telegraph and telephone work. In the bottom of the glass containing-jar is placed a "star" of copper sheet, attached to which is a well-insulated wire extending out of the top of the jar for the purpose of making terminal connection. A portion of about 3 Ib. of blue- vitriol (sulphate of copper) is placed in the bottom of the jar, being well distributed around the copper element and almost cover- ing it. The vitriol crystals used should be broken up so that none of those PRIMARY BATTERIES 15 placed in the jar are larger than a walnut. The fine particles or dust of vitriol should not be used. The zinc which is provided with a hanger is then suspended from the upper edge of the jar and the jar filled to within a half inch or so of the top with pure soft water. Clean rain water, if available, is the best for the purpose. Impure or hard water prevents proper action of the chemicals and should not be used. When a cell is first set up it is customary to hasten its action by " short circuiting" the copper and zinc terminals by means of a short piece of wire. In a short time zinc sulphate is formed around the zinc, and a copper sulphate solution forms around the copper, the two fluids being sepa- rated because of their different specific gravities. A cell in good condition shows a fairly clear line of demarcation between the two solutions. The zinc sulphate appears to float on the top of the denser copper solution beneath. If the circuit connecting the terminals of the cell, or of the battery of which the cell forms a part, is left open and the cell not called upon to do enough work to prevent mixing of the solutions, the copper sul- phate gradually comes into contact with the zinc and is decomposed, forming cupros oxide (CuO) which completely covers and adheres to the zinc. In appearance this deposit resembles black mud. Crystallization of the zinc sulphate is evidenced by the formation of salt-like crystals which creep over the upper edges of the jar and down its sides, and unless periodically cleaned away, in a com- paratively short time make a disagreeable mess on battery shelves. To prevent the creeping of salts over the tops of jars a mineral oil of high viscosity is sometimes applied by pouring it over the top of the solution to a depth of 1/16 to 1/8 of an inch. The specific gravity (sp. gr.) of water has been assigned the value of i.co, and the density or weight of all other liquids is measured in comparison there- with. Sulphuric acid has a sp. gr. of 1.84, mercury 13.58, etc. For the practi- cal measurement of the specific gravity of a liquid an instrument called a hy- drometer is used (see Fig. 3). This instrument consists of a hollow sealed glass tube or float, weighted at its lower extremity with lead shot, the stem of the float being provided with a graduated scale. When a hydrometer is allowed to float freely in a liquid, the division of the scale on a level with the surface of the liquid indicates the specific gravity of the liquid so tested. There are two or three different makes of scale, but the one generally used is known as the Baume scale which registers divisions from o to 45. When a gravity cell is at its best the scale will sink to 20. Should a test show that the density of the solution is below 5 or above 25 the cell is not in good condition. It is when FIG. 2. Gravity cell. 16 AMERICAN TELEGRAPH PRACTICE the density of the zinc solution rises above 25 that the crystallization of the zinc sulphate is most active. Obviously, the remedy is, as indicated above, to short circuit the cell until the chemical action has had an opportunity to restore normal conditions; that is, when the impoverished condition of the cell is the result of having been left on "open circuit" for an unreasonable length of time. In the installation and maintenance of gravity batteries there are a few points of sufficient practical importance to warrant consideration here. It is of considerable benefit to have jars well cleaned and dried before being used in setting up cells. The insulated conductor leading from the copper should in each case be securely riveted, making solid contact. It is not good practice to attempt to renew impover- ished cells by occasionally dropping in a few crystals of bluestone. Cells in service should frequently be tested with the hydrometer. In case the specific gravity rises above 25, additional soft water should be added. To do this, some of the zinc solution may be removed by means of a rubber-tube syphon. The rubber tube used for this purpose should be kept clean. Battery shelves should be kept scrupulously clean. If the insulating material surrounding the wire leading from the copper becomes cracked, the wire should be replaced with one properly insulated. Cells should be renewed when the original charge of blue- vitriol is nearly exhausted or when the two solutions have become too intimately merged. Gravity cells should be disturbed as little as possible. When a cell is for any reason taken down and there is enough of the zinc remaining to warrant using it again, it should be cleaned off while wet as the deposits harden very quickly and later are difficult to remove. Coppers may be cleaned by laying the plates separately on a hard surface and hammering off the deposits. The temperature of a battery room should not be permitted to get below 60 F. The rapidity with which the materials of a cell are consumed depends upon the amount of work done by it upon the quantity of electricity per unit of time that it is required to supply. The consumption and deposition of the materials used in gravity cells per ampere-hour, in fractions of an avoirdupois pound, and from which the cost of producing a given current may be ascer- FIG. 3. Hydrometer. PRIMARY BATTERIES 17 tained, theoretically, when the price of the materials is known, 1 may be cal- culated from the following table: Material Atomic weight Pounds, per ampere- hour Zinc consumed 64. o o 0026749 Sulphate of copper consumed Copper deposited . 249-5 63 . o 0.0102810 o. 002^040 In the computation the copper deposited is a credit. The electrochemical equivalent of zinc is here taken as 0.00033696 grm. per ampere-second according to the determinations of Rayleigh and Kohlrausch. The average life of a cell used for furnishing current for "local" circuits, that is, sounders and multiplex and repeator locals, is from five to eight weeks, a main-line battery supplying one or more main-line wires two months, and a duplex or quadruplex battery about six months. These estimates presuppose intelligent and careful supervision of the batteries so employed. A space interval of at least 3/4 in. should be maintained between individual cells on a shelf. It is extremely important to have all connec- tions between cells and of battery terminals tight and secure. The Leclanche Cell. The Leclanche cell is not used in the operation of telegraph lines but it has an extensive general employment in operating signaling bells, telephones, etc., and for that reason its action and assembly should 'be understood. A familiar form of this type of battery is shown in Fig. 4. The cell con- sists of a glass containing-jar about half the size of that used for the gravity cell, a porous- cup containing a plate or rod of carbon (the in- FIG. 4. Leclanche cell. active element) and a zinc pencil (the positive element). The exciting liquid is a salammoniac solution in which the zinc dissolves, forming a double chloride of zinc and ammonia, while at the same time ammonia gas and hydrogen are liberated at the carbon plate contained in the porous-cup. The depolarizer is black binoxide of manganese, small pieces of which are mixed with powdered carbon and the mixture thus formed packed around the carbon rod within the 1 F. L. Pope, "The Electric Telegraph," p. 74. 2 18 AMERICAN TELEGRAPH PRACTICE porous-cup. As a depolarizing agent the oxide of manganese slowly gives up oxygen as required. The porous-cup does not prevent the passage of current, but protects the zinc from the action of the oxide. The constituent materials of the porous-cup are: feldspar, 8 parts; ball clay, 6 parts; kaolin, 9 parts; and 2 parts quartz the latter is added for the purpose of giving the mixture the required mechanical strength. The mass is then pul- verized, mixed with water and boiled for 24 hours, after which it is moulded into cups of the desired dimensions and baked in a dry kiln at a tempera- ture of i, 800 F., for a period of 24 hours. If left on short circuit or worked hard, it is impossible to entirely prevent polarization, but after showing signs of polarization, if the cell is left on open circuit for a short time it rapidly recuperates. It is best not to use a sal- ammoniac solution too strong, as crystals will gather on the zinc and thus reduce the surface of the zinc exposed to the solution. On the other hand, if the solution is too weak, chloride of zinc will form on the zinc element. Either of these conditions consider- ably increases the internal resistance of the cell. As in the gravity cell, soft clean water should be used in forming the solution. About 6 oz. of salammoniac is sufficient per cell, and water should be added until the jar is filled to within 3/4 in. of the top of the jar. The solution should be stirred until the salammoniac has dissolved. The upper part of the carbon cylinder or porous-cup should be kept clean and dry in order to prevent leakage of cur- rent between the poles. To renew the cell, all that is necessary is to clean or renew the zinc element and pour some fresh solution into the jar. The F IG< s . Fuller cell. carbon is the negative element of the cell and the positive pole; the zinc is the positive element and the negative pole. The Fuller Cell. There are several forms of this type of battery, but the action of and the assembly of the various forms is practically the same. This c"eU is sometimes referred to as the bichromate cell because of the employment of . bichromate of potash together with a dilute sulphuric acid solution as the elec- trolyte. The bichromate chemically unites with the hydrogen and prevents polarization. A common type of Fuller cell is shown in Fig. 5. PRIMARY BATTERIES 19 A well-amalgamated block of zinc is placed in a porous-cup which is nearly filled with a dilute solution of sulphuric acid. The cup is placed in the center of a glass jar about the size of a Standard gravity jar. A carbon plate of com- paratively large section is placed in the jar at one side of, or completely sur- rounding the porous-cup. The containing vessel is then filled to within 1/2 in. of the top with a solution of potassium bichromate". It is customary to keep a spoonful or so of mercury in the bottom of the porous-cup so that the amal- gamation of the zinc contained therein may be continuously renewed. The bichromate solution (electropoin) is made up of i part sulphuric acid, 3 parts bichromate of potash, and 9 parts water. The bichromate should be dissolved in warm water and when cool the required amount of sulphuric acid should be slowly added. The reverse process should never be attempted; that is, the bichromate solution should not be poured into the sulphuric acid, as excessive heat and distressing fumes will thereby be generated. When first set up the solution is of a light brown color and as it ages, gradually turns darker. As the internal resistance of this cell is but 0.5 ohm and its e.m.f. 2 volts, it is an unusually powerful source of current and it has many uses, especially when it can be employed where intelli- gent handling may be availed of. It is always inadvisable to allow any but careful attendants to handle destructive acids. The Edison-Lalande Cell. This cell is capa- ble of yielding a large current as much as 30 amperes on short circuit, due to its low internal resistance and relatively high e.m.f., 0.75 volt. The cell is made up of zinc and copper oxide in a solution of caustic potash. As usually con- structed the plates are hung side by side from the cover of the jar. The copper oxide is plated with a film of copper for the purpose of reducing the initial resistance of the cell, and is held in a frame suspended from the cover. To prevent the inevitable creeping of salts, a film of oil is poured on top of the solution. When the solution is being mixed the caustic potash should not be placed in the cell and left to dissolve as it is very likely to solidify at the bottom of the jar. The solution should be stirred until all of the potash is dissolved. As the solution used in this battery will burn the skin and clothing, great care should be exercised when stirring, to avoid splashing. In renewing the Lalande cell a new solution should be set up at the time the zincs and oxides are renewed. The solution should always reach to the lower colored line in the jar, after it has cooled down; usually it FIG. 6. Edison-Lalande cell. 20 AMERICAN TELEGRAPH PRACTICE is found necessary to add a little water to bring it up to this line after the cooling process. Fig. 6 shows a view of an assembled Edison-Lalande cell. The Dry Cell. Generally speaking, dry cells are modifications of the sal- ammoniac cell, in which the water is replaced by one of the various gelatinous substances available for the purpose. The Gassner cell, one of the original dry-cell products, embodied a paste made of i part oxide of zinc, i part sal- ammoniac, 3 parts plaster, i part zinc chloride, and 2 parts water, all by weight. Fig. 7 shows a typical construction of dry cell as at present manufactured in this country. In making up dry cells a zinc cylinder is lined on the inside with blotting paper or an absorbent cardboard. The exciting fluid is poured into the cylinder and left for a period of 15 minutes to soak thoroughly. The electrolyte is then poured out and the -Blotting Paper cylinder mver ted so that the surplus liquid may drain off. The carbon rod is then inserted and the space between it and the sides of the absorbent paper is filled with . 7 . Dry cell. the depolarizer which usually consists of black oxide of manganese and granulated carbon, this mixture is moistened with the electrolyte before placing it in the shell. On the surface a layer of dry sand is placed and on the top of this hot pitch is poured and allowed to harden. The function of the depolarizer is to furnish a supply of oxygen required to keep up combustion. After the cell has been short circuited for a brief period or worked hard for a considerable length of time the oxygen in the depolarizer is consumed, the latter gradually hardens and the pores close up. When the oxygen is gone the cell ceases to Pitch g Pasteboard - line Filling Carbon FIG. 8. Arrangement of dry cells on shelf. be of value as a generator of electricity. The electrical output of a dry cell, as well as its length of life for a stated output is, in large measure, dependent upon the dimensions of the exposed areas of the zinc and carbon. An important consideration in the manufacture of dry cells is the thickness of the zinc strip used in forming the'shell. The plate employed for the purpose ranges in thickness from 0.014 in - to 0.02 5 in. The former will be eaten through much more quickly than the thicker plates and the result is that the life of the PRIMARY BATTERIES 21 cell is considerably shortened on account of the moisture oozing out and soaking into the cardboard casing. If two adjacent cells are thus affected and the wet sides come into contact an effectual short circuit is established which may destroy the efficiency of an entire battery. One method of overcoming this is to use pasteboard covers which have been boiled in a mixture half beeswax and half paraffine. It is good practice in arranging dry-cell batteries on shelves to set the cells a half inch or so apart and in the manner shown in Fig. 8 in order to prevent any possibility of short circuit. Standard Cells. In determining the absolute value of a standard e.m.f., instead of depending upon the accuracy of a measuring instrument to register values, a standard cell constructed from definite specifications is used. Three types of cell which have been used for the purpose are the Clark, the Carhart-Clark, and the Weston cell. Crystals Amalgam feiH 3. Solution Crystals Paste Mercury FIG. ga. Clark standard cell. FIG. gb. Weston standard cell. The Electrical Congress held in Chicago in 1893 adopted the Clark cell as the international standard of e.m.f. In this cell the positive element is mercury, the negative amalgamated zinc, and the electrolytes saturated solutions of sulphate of zinc and mercurous sulpate. At a temperature of 15 C., the e.m.f. is 1.434 international volts. The Carhart-Clark Cell. This cell embodies the same elements as the Clark, but the solution of zinc sulphate is saturated at o C., the e.m.f. being 1.440 volts. Weston Cadmium Cell. The elements of this standard cell are cadmium and mercury and the electrolytes sulphates of cadmium and mercury. On January i, 1911, the Bureau of Standards at Washington adopted a new value for the electromotive force of the Weston cell, namely, E = 22 AMERICAN TELEGRAPH PRACTICE 1.01830 international volts at 20 C. This change was made pursuant to official definitions of values adopted by the International Electrical Congress held in London in 1908. As compared with the standard e.m.f. previously employed, the change is equivalent to an increase of about 0.08 of i per cent, in the value of the international volt. The usual form of constructidn of the glass-containing vessel of the Clark standard cell is shown in Fig. g-a, and that of the newer type of Weston cell in Fig. 9-6. This is called the "H" form of cell. This type is quite easily filled, permit- ting the contents of each leg rapidly to take on the normal temperature of the bath. A 2-mm. platinum wire is sealed in each leg, the tips being flattened flush with the glass, making good contact with the mercury or amalgams. The mercury, amalgam and paste each have a depth of -from 10 to 15 mm. Some crystals are placed above the paste, and on top of the amalgam a layer of crys- tals is laid to a depth of 10 mm. The cell is filled to the top of the cross tube with the saturated solution. The Weston cell has a temperature coefficient of about one-thirtieth of that of the Clark cell. This, however, is not of great importance as the cells are maintained in a bath the temperature of which is automatically controlled. The electromotive force of these standard cells is quite constant, but decreases slightly with age. Fifteen cells in use in the lab- oratories of the Bureau of Standards have on the average decreased one-ten thousandth of a volt in four years. CHAPTER III DYNAMOS, MOTORS, MOTOR-GENERATORS, DYNAMOTORS, VOLTAGE AND CURRENT REGULATORS Electro-mechanical Generators of Electricity. These at the present time are used quite extensively to furnish current for the operation of telegraph lines, also for the operation of terminal apparatus including main-line and local instruments. In some instances current furnished by commercial power companies is used directly, being supplied by one or more pairs of wires from the power sta- tion to the telegraph office ; there the current is distributed to the various cir- cuits by means of switching systems. In general, however, the operation of telegraph circuits requires the employ- ment of different voltages, ranging about as follows: 40 volts, 85 volts, 125 volts, 200 volts, and 375 volts. Further, it is necessary that at least the two last named voltages be available in both negative and positive polarities. Regardless of considerations of economy, the usual practice is to gen- erate on the ground (in the telegraph office) the different values of potential required in each individual installation. There are two ways in which the desired end may be accomplished. One way is to set up a number of dynamos capable of generating like values of e.m.f., and to connect a suffi- cient number of them in series to produce an aggregate voltage equal to the maximum required. If, then, the units are arranged in multiples of 40 volts, potentials of the following specified values may be tapped off : 40 volts, 80 volts, 120 volts, 1 60 volts, 200 volts, 240 volts, 280 volts, 320 volts, 360 volts, and 400 volts. That is, in order to provide voltages ranging from 40 to 400 in multiples of 40 volts, ten dynamos are required. It is obvious that a series of machines so connected would all be of one polarity, either negative or positive, and that a duplicate set of machines having identical ranges of voltage are required to supply the opposite polarity. Where this method is employed it is customary to have in readiness a third group of machines as re- serve and so connected through switches that either polarity may be availed of. Another method is to make use of generators having different individual voltage outputs. That is, one dynamo for each polarity of each voltage re- quired. In either case an external source of power is required to drive the dynamos, and the customary method of driving these machines is through the medium of electric motors mechanically connected to the rotating elements of the 23 24 AMERICAN TELEGRAPH PRACTICE dynamos. The chief advantage of this arrangement is that any available commercial voltage whether it be direct current or alternating current, or whether the potential is no volts, 220 volts or 500 volts, may be employed to operate the motors and still the e.m.f. of each dynamo driven by its re- spective motor will accord with its rated output. One dynamo delivering 40 volts, another 85 volts, another 200 volts, and so on. What follows in regard to the construction of and operation of dynamos includes only such detail as seems necessary to adequately present the subject, from a telegraph standpoint. In a dynamo an e.m.f. is induced in wires caused to move through the magnetic field near the poles of a magnet. The magnetic field is the space about the magnet within which a piece of iron would be attracted to or repelled from it. The direction and strength of the magnetic force causing this attraction or repulsion is determined, and a certain unit value selected; called a line of force; by which the intensity of the magnetic field can be measured. The total number of lines of force issuing from the magnet is called its magnetic flux. The value of the e.m.f. induced in the wires referred to depends upon the number of lines of force they cut in a certain time, that is, upon the "rate" at which the lines of force are cut. If the wires are simply held in the mag- netic field, no e.m.f. is generated. Either one must move with respect to the other if an e.m.f. is to be produced in the wires. If the wires move through the field in one direction, the e.m.f. produced will ric. 10. Dynamo field-magnet poles. , . ,. . cause the current to now in a direction the reverse of that resulting from moying the wires in the opposite direction. Also, the direction of the current in the wire moving under a north pole is op- posite to that of the current in the wire moving under the south pole of a magnet. Figure 10 shows the poles of a pair of electromagnets, one marked N (north) the other S (south). The dotted lines represent lines of force (shown thus for the purpose of clearness) streaming across the gap in a direc- tion from north pole to south pole. The armature is represented as a single loop of wire A, with its ends B and C brought out at one side. If the loop is revolved in the direction indicated by the curved arrow it passes through the lines of force and an electric current is induced in the loop flowing in the direc- tion shown by the straight arrows. As the coil is moved through a complete revolution it is evident that each side of it will come within the influence first of one pole and then of the other thus reversing the direction of the current in the loop twice during each complete revolution. The result is that alternat- ing current is produced in the armature coil, that is, current which alternates DYNAMOS, MOTORS AND MOTOR-GENERATORS 25 in polarity from positive to negative and negative to positive as indicated above. If an alternating current is desired the terminals of the armature coil (or coils) are connected with "collector" rings which are mounted on one end of the armature shaft and separately insulated from it, the current being taken off by means of carbon brushes, one in contact with each collector ring and mounted in stationary brush holders to which the wires of any external cir- cuit may be connected. When a unidirectional or direct current is required, as is usual in tele- graphy, it is necessary to use a commutator in order to take from one of the brushes a constant positive current and from the other a constant negative current. Figure u illustrates the end view of a commutator having a number of insulated segments to which the ends of individual armature coils may be attached. In the figure one coil only is shown. The brushes which serve to lead the current away from the dynamo to any desired external circuit are shown in position/ The com- mutator is built in the form of a sleeve or ring and -mounted on one end of the armature shaft. The commutator C is built up of strips of copper, forming segments S, S, all in- B sulated from each other, the entire commutator being insu- lated from the shaft upon .... . . . FIG. ii. End view of commutator, showing one which it is rigidly mounted. armature coil. Remembering that the coil W is mounted on the armature and that it is revolved between the poles of an electromagnet, if we consider that the coil terminal E is positive at the in- stant shown, a positive current may be taken from brush B. But, as the coil is moved around one-half revolution, the current in it reverses and the terminal F being then in contact with brush B delivers at that point a posi- tive current. This must be so as the brushes are stationary, and the pole pieces of the electromagnets are stationary and when the terminal F comes into contact with brush B it must necessarily be in the same position in the magnetic field as terminal E was when in contact with brush B. Dynamo Fields. In the various types of dynamos manufactured at the present time, three varieties of field-magnet iron are used, namely, cast-iron, cast-steel, and sheet-steel. It has been determined that a greater number of lines of force can be produced with a certain magnetizing force in a given sec- tion of sheet-steel than in similar sections of cast-steel or cast-iron. From the standpoint of magnetic permeability, mild steel is next in order and then 26 AMERICAN TELEGRAPH PRACTICE follows cast-iron. The coils of insulated wire wound around the pole pieces of a dynamo constitute the field winding. The purpose of these coils is to con- duct a current of electricity through them in order to inductively magnetize the poles. The amount of magnetism generated is dependent upon the num- ber of turns of wire in the field coils and upon the volume of current which is passed through these coils. As was brought out under the heading of current in Chapter I, the practical unit of current is the ampere. Although the subject of electro- magnets will be taken up in detail in a later chapter, it may here be stated that the unit of magnetizing force is the ampere-turn, which signifies that i ampere of current is flowing in i turn of wire wound around the core of a magnet. The total magnetizing force is determined by multiplying the number of turns of wire wound around the core of a magnet by the number of amperes flowing in the circuit thus formed. An equal number of lines of force are generated by i ampere flowing in 50 turns as by 50 amperes flow- ing in i turn, or as 2 amperes in 25 turns. The product of the turns and the amperes determines the total magnetism developed. The factors may have any value provided the product is equal; or, T = tXl. Where T rep- resents the ampere turns, / the number of turns of wire and 7 the current in amperes, flowing. Field Excitation of Dynamos. The field magnets of dynamos may be energized either by current from some external source or by current taken from the commutator brushes of the machine itself. When an external source of current is used, it is immaterial, theoretically, whether it is a primary battery, storage battery, or an independent dynamo, provided the current supplied is unidirectional. In the smaller makes of dynamos, such as those used in furnishing tele- graph currents, the self-excited dynamo is the type generally used. Self- excited generators may be either series wound, shunt wound, or compound wound. Series-wound generators deliver a voltage which increases with the load. Shunt-wound generators deliver approximately constant voltage. Compound-wound generators deliver constant voltage. In series-wound closed-coil-armature generators the entire field winding is in series with the armature, and consequently the field coils carry the entire current generated. Figure 12 in simple lines shows the wiring of a bi-polar series- wound dynamo. In a shunt-wound dynamo the field winding is connected to the brushes in the manner illustrated in Fig. 13. A field regulating rheostat is inserted as shown. A compound-wound generator as shown in Fig. 14 is provided with a shunt field winding connected to the brushes in series with a rheostat and DYNAMOS, MOTORS AND MOTOR-GENERATORS 27 with a second winding connected in series with the armature. At no-load the shunt winding excites the machine to normal voltage. When the external circuit or load is applied, the field excitation is strengthened because of the current then allowed to flow through the series winding. For use in telegraph service the advantages of compound- wound generators is apparent, as the auxiliary series winding automatically increases the strength of the field in response to load increases, and, conversely, reduces the field strength FIG. 12. Two-pole series dynamo. FIG. 13. Two-pole shunt dynamo. FIG. 14. Two-pole compound dynamo. as the load is decreased, thus furnishing practically constant voltage regard- less of variations of the resistance of external circuits applied to it. Most of the generators built for telegraphic purposes during the past 20 years have been shunt wound, but the superior advantages of compound- wound machines over the former are very likely to result in a more general employment of compound-wound machines in the future. It is possible that their general employment may be hastened i 1 by having some of the machines at present in .- \ service " compounded." In view of this it seems justifiable to devote some space to a con- sideration of the principles involved. Magnetic Circuit. The magnetic circuit of a dynamo is illustrated by the dotted lines in Fig. 15. The circuit includes the iron cores of the field coils, the iron yoke joining these cores, the iron core of the armature, and the air "gap" between the pole faces and the armature. The amount of m.m.f. required in FIG. 15. Magnetic circuit of two-pole dynamo. the metallic portion of this circuit is directly dependent upon the magnetic strength which it is desired to develop. This, of course, is infinitely less than that required to set up magnetism of equal density in an equal length of air space. The field winding must be designed to establish magnetism in the air-gap portion of the magnetic circuit, because that 28 AMERICAN TELEGRAPH PRACTICE is the seat of action, where the insulated loops of wire (the armature) are revolved, and where the e.m.f. is generated. As previously indicated the magnetomotive force required to establish this magnetic field is expressed in ampere-turns, and it is eVident that if a small current only is to be used for field excitation, a great many turns of fine wire will be needed in the field coils, while, if a large current is to be used, a much larger wire may be employed. The factor which in large measure determines the size of wire which should be used in the shunt winding is the heating of the conductor. In practice the allowable heat limit confines the permissible shunt current to about 5 per cent, of the line current. In calculating the correct shunt winding, a magnetic flux should be provided for which will develop the desired e.m.f. with the external circuit open. When current flows in the armature coils a counteraction takes place which has a tendency to oppose field magnetization. There is also a slight reduction in the voltage of the generator due to resistance "drop" in the armature. If the line conductor is wound around the field coils a few turns, the reverse magnetomotive force developed by the armature is neu- tralized and the strength of the magnetic field restored to a value sufficient to reestablish normal voltage at the machine terminals. That portion of the line conductor wound over the shunt winding is called the series or compound winding. By this arrangement the voltage at the brushes of the dynamo may be caused to increase as the current demand made upon the machine is increased. EVen in dynamo manufacturing plants it is found to be somewhat difficult to determine accurately the number of turns to give the series winding of compound machines to meet given conditions. Usually the machine is "over-compounded" and a portion of the current shunted by means of a variable resistance placed across the terminals of the series coil until the correct Value has been obtained. A shunt-wound dynamo may be compounded by the addition of a series field winding. One method of approximately determining the number of series turns required is to run the generator with the external circuit open, and by means of an ammeter measure the current required in the field coils to develop normal voltage, then throw on the load (close the external circuit) and alter the resistance of the field regulator until the desired voltage is obtained. If simple compounding is the object this will be the no-load voltage, but if the machine is to be over-compounded the voltage will be proportionately larger. Now if the field current is again measured and the difference noted in the two readings of field current multiplied by the turns in the shunt coil, this value divided by the number of amperes of current flowing in the armature will give the number of turns required in the series coil. The " shunt" winding as a rule consists of cotton-covered wire of a very DYNAMOS, MOTORS AND MOTOR-GENERATORS 29 small gage wound on the core, next to the yoke. On account of the small size of conductor used the resistance is quite high and the current volume in the shunt circuit is correspondingly small. The " series" coil must be low in resistance, generally less than half that of the armature, and should be wound on the end of the cores nearest the armature so that the maximum magnetic effect may be obtained in the air gap. Armatures. In Fig. 10 the armature is represented as a single loop of wire. In the construction of practical dyna- mos each loop consists of a number of turns. As compared with one turn, when two turns are used the e.m.f. is doubled, ' FIG. 16. Drum wound armature, showing one coil. because an equivalent voltage will be generated in each turn of wire, and in general in order to obtain a given constant potential the loops are so located and connected with respect to each other that each turn complements and adds to the e.m.f. generated by all other loops. A completed armature consists of a core, a commutator, and a winding. The core serves to support the winding rigidly in position and also acts as a conductor of the magnetic flux from one field magnet pole-face to the other. Armatures may be either drum wound or ring wound, and may have either smooth cores or slotted cores. Fig. 16 illustrates a method of placing the armature coil on a smooth- core drum armature. A loop of three turns is shown with its FIG. 17. Ring wound armature. terminals brought out to adjacent commutator segments. In ring-wound armatures the conductor turns are wound around a ring- shaped core (Fig. 17) and that portion of each turn which is on the inside of the ring is practically inactive. Drum wiring obviates this " dead" winding, as in the drum type of armature each section of every conductor is on the outside of the core. Armatures may have either closed-coil or open-coil windings. The former have all of the loops (inductors) interconnected so that with the exception 30 AMERICAN TELEGRAPH PRACTICE of that period when a certain portion of the winding is undergoing commuta- tion, the voltage induced in each inductor is adding its quota of energy to the maintenance of a constant potential in the external circuit. Closed- coil winding also effects a decided advantage in reducing sparking at brushes to a minimum. Open-coil windings provide commutator connection with the inductors in such manner that voltage produced in the coils is made use of only when each individual coil is undergoing commutation. Closed-coil drum windings may be either lap wound or wave wound, the former method generally requires that the number of coils on the armature be the same as the number of bars on the commutator, and the number of commutator bars employed "even," to permit of equal distribution midway between the brushes, while wave winding advances around the commutator similarly to the manner in which a "wave" travels progressively from point to point. The Commutator. Fig. 18 shows a view of a commutator disconnected from the armature wiring and from the shaft. As previously stated, the current generated in an individual moving armature coil alter- nates from one polarity to the other as the coil cuts the lines of force passing from north to south pole during one-half of the revolution, while, as stated in connection with Fig. 1 1 , during the other half revolution the coil cuts the lines of force in the reverse direction. The commutator consists of a number of copper strips or segments, each one insulated from the others by means of strips of a high grade of mica. Usually the number of segments is the same as the number of armature coils, and the number of each is calculated according to the desired output of the generator. The carbon brushes which rest upon the commutator serve to lead the current from each coil at the moment it attains its maximum value. It is evident that the current gathered by the brushes will be "pulsating" in character, but by employing a large number of coils and segments the cur- rent generated is so nearly uniform that the pulsations are not noticeable in practical operations. FIG. 18. Commutator re- moved from armature shaft. ELECTRIC MOTORS Direct-current Motors. What has been said in regard to the various elements of direct-current generators, in a general way applies to direct- DYNAMOS, MOTORS AND MOTOR-GENERATORS 31 current motors, as the essential elements of one are identical with those of the other. In the case of a motor the carbon brushes resting on the commutator serve to lead the current from some external source into the coils of the armature and through the field winding, thus causing the armature to rotate, and by means of a pulley mounted on one extremity of its shaft, by gearing, or by direct mechanical connection of the shaft, furnishes mechanical power for any desired purpose. Similarly to dynamos, motors are series wound, shunt wound or compound wound. In a series motor the field consists of a relatively small number of turns of large wire directly connected in series with the line and the armature. The current in the armature coils and in the field-magnet winding is of the same value. A shunt-wound motor has field magnets wound with a large number of turns directly connected with the brush terminals of the machine or across the terminals of the external circuit supplying the motor with current. In the case of a shunt motor the strength of the field current is independent of the current strength in the armature. Compound- wound motors may have the two field windings connected so as to form cumulative winding, or differential winding. In the former case the magnetizing effects of both windings are in conjunction, while in the latter they are in opposition. With the cumulative winding, increasing the load of the motor increases the magnetic strength of the field, while with the differential winding an increase of load decreases the field strength. In most towns and cities throughout the country commercial electric power is available for the purpose of driving motors. In the majority of places there is but one potential and one kind of current at hand. In some cases no- volt direct current, in others no-volt alternating current may be the only power available. In some cities there is no other choice but to use 200- or 2 50- volt alternating current, or 5oo-volt direct current. It is possible to obtain electric motors designed to operate with a given potential or character of current, whether direct current or alternating current, and as in the operation of telegraph circuits four or five different voltages are needed, all that is required is to have each separate dynamo mechanically connected with and driven by a motor which may be operated by the available commercial voltage. Thus the various individual motors are operated by, say no volts direct current, or 200 volts alternating current, while the individual dynamos driven by these motors may have outputs ranging from 40 volts to 400 volts, or any other potentials for which they may be designed. In practice it is best to employ the lower voltages for the operation of motors as it is found that when, for instance, 500 volts direct current is used to operate motors, heating and sparking difficulties are quite frequent and annoying. When there is a choice between 500 volts direct current and an alternating- current potential of no volts or 220 volts, it is common practice to employ an alternating-current motor to operate on the alternating current available, 32 A M ERIC A N TELEGRAPH PRA CTICE this motor in turn being directly connected to a no- volt direct-current dynamo, the latter furnishing current to operate the various motors which drive the dyna- mos generating the various telegraph currents. In this way it is possible to get away from the use of the objectionable 500 volts direct current, and at the same time effect a saving in primary-current consumption. Also, it is usual to arrange for "auxiliary" or "reserve" power to provide against interruption to service in case the external source of power fails. Where both alternating-current and direct-current sources of commercial power are available it is common practice to use them alternately, or to use one of them regularly and maintain the other as reserve. Alternating-current Motors. 1 There are three types of alternating-current motor in commercial use, namely : Single-phase series type. Synchronous type. Induction motor. The latter is the type usually employed in telegraph service for the purpose of driving small direct-current dynamos. It has been shown in the case of a direct-current motor that the armature revolves between the pole faces of stationary field magnets. To comprehend the forces at work in the operation of an induction motor, consider a direct-current motor armature with current traversing its coils. When the field magnets are charged, the armature will turn ; but suppose the brushes which rest on the commutator are removed and the terminals of the armature coils connected to a copper ring. Then, if instead of the field magnets remaining stationary they be revolved around the armature, it follows that the magnetic lines will travel around in a circle, at the same time setting up an electromotive force in the armature coils. As these coils are short circuited by the copper ring, the current induced in the armature coils sets up a reaction, which results in a drag which pulls the armature around in a direction the same as that taken by the rotating magnetic field. In the induction motor, instead of the field magnets being revolved mechanically, the magnet windings are so connected with the external circuit that the active field moves around the circle in a given direction, thus causing the armature to rotate in unison with the constantly moving field. The induction motor has two essential elements, the stator (Fig. iga) and the rotor (Fig. igb). The stator consists of a stationary framework of circular construction which serves to support the primary winding, which is connected in the manner shown theoretically in Fig. 20. The projecting cores shown are di- 1 It has been thought best not to take up in this work the subject of the generation of polyphase alternating currents. While commercial alternating current is used to operate motors directly connected to direct-current dynamos for telegraph requirements, alternat- ing currents are used only in a very li mi ted way in the operation of telegraph lines. DYNAMOS, MOTORS AND MOTOR-GENERATORS 33 vided into groups. In the section illustrated the poles of one group are marked M, (xz) and (#3), the poles of a second group (3/1), (y 2 ) and (y 3 ) and those of a third group (zi), (z 2 ) and (z 3 ). It may be observed that the consecutive poles FIG. iga. Stator of induction motor. FIG. igb. Rotor of induction motor. of any group are separated by a number of poles corresponding to the number of groups employed, and that the winding around each pole of any group alternates in direction from pole to pole, thus producing north and south poles consecu- 3 34 AMERICAN TELEGRAPH PRACTICE lively. If a three-phase alternating current be connected across the terminals, the polarity of the' magnet poles of any group will be reversed twice during each cycle, and the active magnetic field progresses around the circle due to the fact that any three poles located consecutively are magnetized to a maximum one after another in order around the frame. Thus is produced the rotating magnetic field. FIG. 20. Stator magnet poles and winding connections of induction motor. The Rotor. The core of the rotor is made of laminated steel punchings mounted on an iron spider. Each slot in the core contains a single copper bar which is made fast to a short-circuiting ring at either end of the bars. The e.m.f. produced in the copper bars of the rotor is very low, but the force ex- erted upon the rotor by the revolving field is such that the induction motor is quite efficient as a source of power. Having no brushes or commutator, this type of motor is simple to operate and to maintain. The Motor-generator. On page 31 reference is made to the use of alter- nating-current motors directly connected to direct-current dynamos for the purpose of generating in the telegraph office no volts direct current to be used to operate the various motors which in turn are mechanically connected to the dynamos having different voltage outputs. This type of machine is known a? a motor-generator, having been given this name to distinguish it from the motor-dynamo to be described presently. One of the standard makes of motor-generators is shown in Fig. 21. The induction motor is on the left, while on the right is shown the direct- current generator. In making up these units to meet operating conditions, the motor may be designed to operate on either alternating current or direct DYNAMOS, MOTORS AND MOTOR-GENERATORS 35 current, and on whatever voltage is available, while the generator end may be designed to produce any required voltage. Motor-dynamos. Motor-dynamos have the two machines on one base directly connected by a steel shaft and each machine has its own field coils, see FIG. 21. Motor generator. * V FIG. 22. Motor dynamo. Fig. 22. Some of these machines, notably the Crocker- Wheeler type, have the two armature windings on the same shaft but separated electrically. Another type of machine known as the dynamotor is extensively used in telegraph work. This machine has a single field for both motor and dynamo 36 AMERICAN TELEGRAPH PRACTICE ends, see Fig. 23 . The armatures are of the iron-clad type with slots or grooves around the outside in which the armature wires are laid and held fast by re- taining wedges, readily removable when repairs are necessary. The armature core consists of sheets of soft annealed steel punchings. The copper con- ductors in the armature are insulated with mica strips or with oiled muslin and fibrous materials. Regulation of the voltage of the dynamo is accom- plished by varying the speed of the motor by means of a controlling rheostat. Both in the motor-dynamo and the dynamotor each end of the machine that is, the motor end and the dynamo end have their own commutator and brushes. The capacity of a double field motor-dynamo is determined by the capacity of the motor end and, since the dynamo operation is independent of the motor, all the methods of control practised with dynamos may be used in FIG. 23. Dynamotor. these machines without affecting the motor. The dynamo field may have a shunt winding or a compound winding to insure constant pressure regardless of variations in load.. In single field dynamotors, the motor and dynamo ar- matures are combined in one, thus requiring a single field only; that is, the pri- mary armature winding in association with the common field which operates as a motor to drive the machine, and the secondary or dynamo winding which operates as a generator to produce the secondary current are upon the same armature core. The armature reaction of one winding neutralizes that of the other and saves energy required for magnetizing the fields. Dynamotors can stand a somewhat greater overload than motor-dynamos but their e.m.f . drop cannot be compensated by compound winding, as in the case of the motor-dynamo. Also, since both windings of dynamotors are on the same core and under the influence of a single field the ratio of trans- formation cannot be varied or adjusted. Any regulation of the field strength DYNAMOS, MOTORS AND MOTOR-GENERATORS 37 will simply make the machine run faster or slower, as the ratio of the turns of wire on the dynamo end of the armature to those on the motor end is un- changeable. The voltage of the dynamo end must, therefore, remain in the same ratio as the voltage on the motor end. The voltage of a single machine may be regulated by using a rheostat in series with the motor armature for reducing its speed; but this wastes energy as much as when resistances are used to cut down the secondary voltage and interferes with the constant speed of the machine to the disad- vantage of the regulation of the dynamotor for constant pressure. Figure 24 shows a view of a dynamotor armature, with the motor commutator on one end of the shaft, the dynamo arma- ture on the opposite end, and the windings of each in the center. Motor Current Regulation. When an electric motor is at rest and the switch controlling the supply-current circuit is closed in the act of starting the motor, the initial rush of current through the low-resistance arma- ture coils is excessive unless a protective resistance is placed in series with the supply mains. In order to limit the amount of current permitted to traverse the armature conductors, a rheostat or starting box is used. When a motor is started, a reaction takes place in the armature which produces a counter- electromotive force, due to the action of the inductors in the armature cutting through the magnetic lines of force produced at the field magnet poles. The counter-e.m.f. developed opposes in direction that of the supply e.m.f., thereby in a sense automatically controlling the current volume in the armature windings. FIG. 24. Dynamotor armature. By employing Ohm's law (/ = ^) for the purpose, the amount of current in the armature may be determined, thus : Where / is the current in the armature, E the impressed e.m.f. (supply voltage), R the resistance of the armature, Then: Therefore, to reduce the current volume in the armature coils at the instant 38 AMERICAN TELEGRAPH PRACTICE of starting the motor, additional resistance must be inserted in the circuit, and with given factors we have : 7- r R+Ri Where / is the current in the armature, E the impressed e.m.f., E x the counter e.m.f., R the armature resistance, R! the resistance of the starting box. FIG. 25. Two-pole shunt motor-starting rheostat connections. The starting resistance in the rheostat is gradually reduced until all is cut out. The counter-electromotive force of a motor builds up as the speed of the armature increases in the same way as the voltage of a dynamo increases with increase of speed; therefore, as the starting-box resistance is gradually cut out, the counter-e.m.f. increases to its maximum, reaching that value when the full voltage from the supply mains is impressed on the motor terminals. DYNAMOS, MOTORS AND MOTOR-GENERATORS 39 Figure 25 shows the wiring and connections of a 2 -pole shunt motor with starting box inserted. The connections shown are the same as those used in wiring the motor end of dynamo tor sets, motor-dynamo sets, and motor- generator sets, where direct-current motors are used. An important feature of the starting box is the electromagnet shown on the right of the drawing and at the end of the row of resistance contact disks. This magnet serves to hold the rheostat arm in the "running" position as long as the magnet coil is energized by current from the supply mains. Should the supply circuit be interrupted the magnet coil is demagnetized, and, due to the force of gravity, the arm drops, thus breaking the main-line contact at C. 1 It is evident that the automatic-release feature of the starting box protects the motor armature from the injurious effects of sudden rushes of full line volt- age when the supply circuit is restored. A momentary interruption to the supply circuit does not cause the automatic release to operate as, due to momentum, the motor armature continues to run a short time after the supply current is shut off and generates an e.m.f., which furnishes current for the fields and the cut-off magnet. Starting Boxes. The resistance of the starting rheostat to be used in con- nection with a motor depends upon the amount of current required to start the motor, at full-load. The resistance should be of sufficient capacity to safely carry the current indicated by the normal rating of the motor. The word rheostat is derived from greek pstu, to flow, and err arcs, fixed. A device for regulating the flow of current. In those cases where it is necessary continuously to use additional resistance in the armature circuit of direct-current motors for the purpose of controlling the speed, the rheostat coils must have large current-carrying capacity and the construction of the rheostat must be such that it will have ample radiating surface in order, as far as possible, to avoid excessive heating. Most modern rheostats have the resistance wire wound on hollow asbestos, clay or porcelain bobbins, each bobbin after being wound with the desired amount of resistance wire is entirely covered with an insulating enamel which protects the unit against mechanical injury and prevents short-circuiting of turns. The various resistance units are assembled to form a complete rheostat. In case a unit becomes defective, it may be replaced without disturbing the remaining units. Figure 26 shows the wiring of a commercial type of starting box much used in telegraph work. It may be seen that the connection between each resist- ance unit is brought out to a contact disk. The contacts are arranged in the arc of a circle, so that the rheostat arm pivoted at the center may make a sliding contact with the connections, cutting in or out the amount of resistance required for regulation of speed. In using a starting box in connection with a shunt- 1 In some types of starting boxes the rheostat arm is withdrawn from the deenergized magnet through the action of a spring attached to the arm. 40 AMERICAN TELEGRAPH PRACTICE wound or compound-wound motor, the first contact to which the arm is moved places all of the resistance in series with the armature and the series field wind- ing, and each successive step cuts out a portion of the resistance until the arm is moved into contact with the disk on the extreme right, when all of the resist- ance is cut out and the rheostat is said to be "set" in the running position. It should be borne in mind that starting resistances are designed and intended for momentary use only, and the rheostat arm should never be stopped on any intermediate step longer than a second or two, as the excessive heating of the coils is likely to cause burn-outs. FIG. 26. Connections and wiring of motor-starting rheostat. The total resistance of motor starting boxes usually is such that when the rheostat arm is moved to the first contact, one and one-half times the full torque current of the motor will flow through the armature winding. To start a motor from a position of rest and accelerate it to full speed within a reasonably short time requires one and one-half times the full torque current of the motor at full-load. Although the duration of this excess of current is brief, it is necessary to protect the armature by employing a device which will open the motor circuit when a current 50 per cent, above its normal rating is permitted to flow for more than a fraction of a minute. An enclosed fuse is generally employed for the purpose. In a given case, for instance, a fuse may be employed which has been designed to "open" when a current 50 per cent, above its rated capacity flows through it during 30 seconds time. If the fuse employed is of the 2o-ampere class, a current of 30 amperes flowing through it for a period of 30 seconds would cause a temperature rise in the fuse wire which would melt it and thus open the circuit of which the fuse forms a part. Fuses in Motor Circuits. The necessity for employing starting resistance DYNAMOS, MOTORS AND MOTOR-GENERATORS 41 in motor circuits is apparent from the knowledge that the resistance of the arma- ture winding is very low, about 0.4 ohm. In cases where no- volt supply current is used to operate the motor, if no starting resistance were inserted, theoretically, there would be a current of 275 amperes in the armature at the instant the main switch is closed. This would cause destructive sparking and possibly melt the soldered connections. When a starting resistance is employed, and gradually cut out by means of the rheostat arm as the motor armature accelerates and generates an opposing e.m.f., the disastrous effects of excessive initial current are^ avoided. In view of the above, the necessity for careful handling of the starting resistance is apparent. In order to provide against mishap to the starting resistance or to the motor overload it is customary to "fuse" the motor circuit so that a current large enough to cause heating of the conductors or connections will not be permitted to exist long enough to do damage. The cartridge type of enclosed fuse (the form approved by the fire under- writers) consists of a short length of wire made of lead in combination with a certain percentage of tin. Tin fuses (melts) at a temperature of 235 C., and lead at 325 C. In making fuse wire a squirting process is employed, similar to that used in making incandescent lamp filaments. The fuse wire is packed in an asbestos wrapping and enclosed in a pasteboard tube equipped with brass thimbles at either end, each end of the fuse wire being soldered to one of the thimbles. Fuse blocks are equipped with spring brass clips set a sufficient distance apart to accommodate a particular length of fuse; this construction is quite convenient for quickly replacing defective fuses. Motor circuits are fused on both sides, otherwise, one side becoming grounded at one point and the other side at another point, a circuit might be established with no fuse in action. Overload Motor- starters. The overload attachment to a motor-starter consists of a magnet the coils of which carry the total current consumed by the motor. When the current becomes excessive, the magnet attracts its armature and completes a short circuit around the terminals of the retaining magnet which holds the rheostat arm at the "full on" position. The short circuit is closed by means of two brass posts conveniently mounted on the face of the starter. The holding magnet is thus demagnetized, the rheostat arm flies back to the "off" position, inserting the total resistance of the starter, opens the circuit and stops the motor. Underload Release. Diagram 26 shows the wiring of a standard type of starting box including a no-voltage release attachment. A type of "remote-control" motor-starter used in telegraph work is illus- trated in the photographic reproduction, Fig. 27. In the larger telegraph centers it is necessary to have a number of 5o-volt 42 AMERICAN TELEGRAPH PRACTICE and loo-volt potentials available for intermediate battery purposes. For, while ordinarily the regular battery arrangements are sufficient to take care of circuit requirements, there are occasions when, temporarily at least, addi- tional battery is required to maintain currents of a requisite strength to satis- factorily operate lines. Inasmuch as these extra battery facilities are for emergency service and are used only for short periods during the day, a switching arrangement is provided whereby the switchboard attendants in the main operating-room may, at will, start and stop any one or all of the machines intended for intermediate battery purposes, even though the machines are located in a part of the building remote from the operating-room. FIG. 27. Solenoid motor-starter. Figure 270 shows the connections of the automatic starter used for this purpose. The motor supply mains are connected to the switch sw which regularly is left closed. The coils C are solenoids having double windings, which when energized pull up the plungers P. To the lower extremity of each plunger is attached metal or carbon contact plates P\ which act as a switch to open or close the armature and field circuits of the motor in response to the operation of the plungers. If the circuits are traced it may be noted that the resistance coil 5 remains in series with the armature for a short time after the motor circuit is closed, and is short-circuited only after the armature has speeded up sufficiently to avoid the effects of the initial inrush of current. Four wires are shown leading from the engine-room, or dynamo-room to the operating-room switchboard, where they terminate in a specially constructed double-contact pin-jack. DYNAMOS, MOTORS AND MOTOR-GENERATORS 43 The insertion of a double-conductor plug in the jack closes both the motor and dynamo circuits. The "wedge" end of the connecting cord is inserted in series with the line at the spring-jack as shown on the right. Alternating-current Motor-starters. Alternating-current motors take a large current when started under load, so large in fact that unless proper provision is made against it, serious fluctuation of line vol- tage occurs when any motor on the circuit is started-up. To avoid this a means is employed whereby a reduced voltage is applied to the motor at starting, this is gradually increased until the full-line voltage is applied to the motor terminals when the rotor has reached full speed. The usual type of controller employed is called an auto-starter, which consists of a specially de- signed switching system, operat- ing in conjunction with two auto- transformers. The transformers are in circuit with the supply mains. Each transformer consists of a winding from which a series of taps are made, each tap providing for a different voltage. As these taps are successively connected with the motor by means of the switch an increasing voltage value is ap- plied to the motor terminals until finally the full line-voltage is thrown on. The losses due to transform- ing continue only during the start- ing process because when the full line-voltage is applied the auto- transformers are disconnected from the circuit. In moving the auto-starter handle from one notch to the next the circuit is for an instant entirely interrupted. Under ordinary con- ditions breaking and making the circuit would cause violent sparking, the con- tacts, however, are made and broken in a chamber filled with oil, thus materially reducing the liability of sparking. FIG. 270. Connections and wiring of solenoid motor-starter. 44 AMERICAN TELEGRAPH PRACTICE Supply a b c d FIG. 28a. Four- FIG. 286. Three- FIG. 280. Four- FIG. 28^. Three- wire two-phase wir e two-phase wire, two-phase wire, two-phase motor connections. motor connections, motor connected motor connected through an auto- through an auto- transformer, transformer. To Line Two-Phase Motor Auto-Starter in Starting Position Three- Phase Motor Auto-Starter in Starting Position FIG. 29. FIG. 30. ,n Off Position Starting Positions \ f\ j Running Position FIG. 31, FIGS. 29, 30, 31, 32. FIG. 32. DYNAMOS, MOTORS AND MOTOR-GENERATORS 45 Figure 280 shows the connections of a two-phase squirrel-cage motor; the wires a and b belong to one phase and the wires c and d to the other. In some instances the two wires belonging to one phase are marked a and the two wires of the other phase marked b, but in general any two wires which are found to have voltage between them belong to one phase or the other. Fig. 2 8b shows the connections of a three- wire two-phase system, the wire b or " common return" being connected jointly to the two center terminals. Fig. 2$>c shows a two-phase four- wire motor connected through an auto-trans- former. Fig. 2%d shows a two-phase three-wire system connected through an auto-transformer. After a four-wire two-phase motor has been connected up, should it develop that rotation is in a direction the reverse of that desired all that is necessary is to interchange the two wires of one phase; that is a and b } or c and d. Figure 29 shows the connections of a two-phase motor and auto-starter. Figure 30 shows the connections of a three-phase motor and auto-starter. Figure 31 shows the connections and internal wiring of a four-wire two-phase auto-starter of the oil-immersed type. The motor terminal markings are shown on the right, namely, Ai, A2, Bi, 82. Figure 32 shows a view of the auto- starter complete with the handle in the "off" position, while the starting and running positions are shown in dotted lines. Dynamo Current Regulation. The output of dynamos may be controlled by regulating the speed of the dynamo armature, or by regulating the current strength in the field winding of the dynamo. When motor-generators or motor-dynamos are employed, the first-named method may be availed of by inserting a field regulating rheostat in the field circuit of the motor as explained in connection with Figs. 23 and 24. Regula- tion of voltage through control of the field strength of the generator is accom- plished by inserting a rheostat as shown in Fig. 13, which shows the internal wiring and terminal connections of a shunt- wound 2 -pole generator. The connections of a compound- wound 2-pole dynamo are shown in Fig. 33. FIG. 33. Complete wiring connections of a two-pole compound dynamo. CHAPTER IV STORAGE BATTERIES CURRENT RECTIFIERS; MERCURY-ARC AND ELECTROLYTIC The names storage battery, secondary cell, and accumulator have been given to that type of battery which consists of elements capable of absorbing electrical energy and storing it in the form of chemical energy. The type of storage cell most generally employed in telegraph work con- sists of a number of lead plates immersed in a dilute solution of sulphuric acid. Alternate plates of a cell are joined together making up the positive element; the balance of the plates similarly joined constitute the negative element. See Figs. 34, 340 and 346. FIG. 34. FIG. 340. FIGS. 34, 340, 346. Storage cells. FIG. 346. The negative plate consists of a lead grid in the interstices of which is fixed a litharge paste consisting of sulphuric acid and oxide of lead, this, in forming, changes to metallic gray lead. The positive plate of red lead in forming changes to peroxide of lead. The electrolytic solution is made up of i part sulphuric acid (H 2 SO 4 ) and 5 parts distilled, or rain water. Externally the direction of current is from peroxide of lead plate through the connecting circuit and back to the negative, or lead "sponge" plate. This is the condition when the cell is discharging. Internally, or when the cell is being charged from an external source of e.m.f. the current is in the reverse direction. There are several methods of charging storage batteries, and the method employed in a given installation is dependent upon local conditions, as re- gards available sources of charging current. In many instances it is econom- ical to use the existing no- volt lighting current for the purpose. Where 46 STORAGE BATTERIES 47 commercial or private lighting circuits are not at hand, a gas-engine-driven electric generator may be used to charge the cells. In considering the in- stallation of a storage battery plant, the first thing to be determined is the desired output or capacity of the plant. Obviously, this depends upon the number of lines and circuits to be fed and the amount of current in am- peres, or fractions thereof, required to operate such circuits. If, for instance, there were 10 wires to feed, each requiring 40 m.a. current, the 10 wires would require 0.4 ampere, and if the wires were to be operated 24 hours per day, the required ampere-hours per day would be 0.4X24 = 9.6 ampere-hours per day. This would be the maximum demand made upon the battery, as the above calculation is based on the possibility of the circuits being closed all of the time. In practice it is found that, considering the average of a large number of lines, circuits are closed a little less than half the time. Two types of storage cell in common use are the "couple" type, and the " multiple " type. In the couple cell there are but two plates. The multiple cell has three or more plates. The terminals of the plates in multiple cells are either burned together or bolted with lead-covered bolts. Storage battery plates are made up in several sizes which are considered standard. Type " E " plates are 7 3/4 in. X 7 3/4 in. =60 sq. in. Type "F" plates are 10 1/2 in. Xio 1/2 in. =no sq. in. Type "G" plates are 15 1/2 in. Xi5 1/2 in. =240 sq. in. The normal charging rate of the "E," "F," and "G" types of cell is ap- proximately 0.08 ampere per square inch of positive plate. Thus to find the normal charging rate of a type "G" cell having 6 positive plates and 7 negative plates, calculate as follows: 6X240X0.08 = 115.20 amperes. When no- volt lighting mains are used to charge storage cells a suitable number of incandescent lamps should be inserted in the lighting circuit, as shown in Fig. 35, in order to take from the circuit a current sufficient to charge the battery. An automatic bre^k switch is inserted in the circuit so that in case the primary current is interrupted the charging circuit will be opened and the storage battery prevented from discharging back through the charging mains. The connections of this switch are such that an electromagnet energized by current from the charging mains actuates an armature which when the current is flowing acts as a switch to keep the circuit closed. When the charging circuit is interrupted, the electromagnet is deenergized, thus releasing the armature and opening the circuit. There are several different types of automatic switch which may be used for the purpose. One device in common use operates in the reverse way to that just described, that is, the armature mounted above 48 AMERICAN TELEGRAPH PRACTICE the magnet core is adjusted to remain "open" when normal current traverses the winding of the magnet, while the increment of current due to short circuit, attracts the armature and opens the circuit. When a cell has been fully charged the active agent in the positive plate consists of lead peroxide, and the negative plate has been converted into "spongy lead." During the time a storage cell is being discharged the active material in both positive and negative plates is being converted into "lead sulphate," due to the extraction of the sulphion from the acid of the electrolyte. After discharge, when a cell is recharged, the positive Line Line Lamps ferfMUUta CWpiufa*. TottKirfk- ratuaLupi BT CT PT ET 1-9 110 1-16 c p. 8-16 c P. 3-32 c. P. 5-32 c. P. 10-18 110 1 82 c. p. 2-32 c. p. 4-32 c. P. 6-32 C. p. 19-27 110 1-82 c. P. 5-16 c. P. 5-32 c. P 7-32 C P. 28-33 110 2-32 c P. 4-32 c. p. 8-32 c. p. U-32C.P. 60-60 1-16 c. P. 2-16 c P. 4-16 c. P. 6-16 c. P. Load FIGS. 35, 36. Storage battery "charging" and " discharging "circuits. plate is converted back into lead peroxide, and the negative plate into spongy lead. It is only the material of which the .buttons are made that undergoes chemical changes, as the supporting grid, consisting of a lead- antimony combination is acted upon only in a small degree. A peculiarity of lead (due to the formation of a coating of sulphate) is that it is insoluble in sulphuric acid. The statement above made, to the effect that during discharge the active material in both positive and negative plates is converted into lead sulphate, is based on the latest and most generally accepted theory of the action that takes place. Figures 35 and 36 show the connections of storage battery charging and discharging circuits as usually arranged. INSTALLATION AND MANAGEMENT OF STORAGE CELLS Location of Battery. The proper location of the battery is important. It should preferably be in a separate enclosure or compartment, which should be well ventilated, dry and of moderate temperature. STORAGE BATTERIES 49 The ventilation should be free, not only to insure dryness, but to prevent chance of an explosion, as the gases given off during charge form an explosive mixture if confined. For this reason never bring an exposed flame near the battery when it is gassing. To obtain the best results, the temperature should be between 50 and 80 F. If the temperature is very high, that is, over 80 F., for any great length of time, the wear on the plates is excessive. If the temperature is low, no harm results, but the available capacity is reduced during the period of low temperature. Installing Battery. Place the jars, after they have been cleaned, in posi- tion on the sand trays, which should previously be filled evenly with the top with fine dry bar sand. The trays, which should be separated by an air gap, rest on glass insulators, which in turn rest on stands or shelves. The cells should be so located in the room that they will be easily accessible and if practicable should be in one tier; where two or more tiers are necessary, ample head room over each tier should be allowed for. If sand trays are not pro- vided, the jars may rest directly on a board or plank, in sections of not more than 10 cells each, the plank being insulated from the stand or shelf by glass insulators, and an air gap left between the section rests. Plates of opposite polarity, except the terminal plates, are burned to- gether by a connecting strap, forming a "couple, " and are placed in adjoining jars; the positive plates are of a brownish color, the negatives of a light gray. Before placing the couples in the jars, the straps should be bent over a piece of wood 3/4 inch thick, the top edge of which is rounded. After removing from the form, the straps should be still further bent until the lower edges of the plates touch; then by gently springing them apart when putting into the jars the plates of adjacent couples will not have a tendency to get together and short circuit. In bending, care should be taken that only the connecting strap is bent, as the burned joints must not be subjected to undue strain. The plates must be placed in the jars, so that in each there will be both a positive and a negative plate and the sections of the battery must be connected, pref- erably by lead tape, so that the positive and negative terminals, which are the single plates, will be connected a positive and negative together in each case. If couples or sections are installed in the wrong direction, the plates will be seriously injured. Rubber separators are used only in Type "BT" cells; in other types no separators at all are used. Connecting up the Charging Circuit. Direct current only must be used for charging. If alternating current alone is available, a current rectifier must be used for obtaining direct current. Before putting the electrolyte into the cells the circuits connecting the battery with the charging source must be complete, care being taken to have the positive pole of the charging source connected with the positive end of the battery, and the negative pole with the negative end of the battery. If a suitable voltmeter is not at hand, the polar- 50 AMERICAN TELEGRAPH PRACTICE ity may be determined by dipping two wires from the charging terminals into a glass of water to which a teaspoonful of table salt has been added, care being taken to keep the ends at least i in. apart to avoid danger of short circuits. Fine bubbles of gar, will be given off from the negative pole. Electrolyte. The electrolyte used for filling new cells is dilute sulphuric acid of a specific gravity of 1.180 or 22 Baume (except Type "ET," see note), as shown on the hydrometer at a temperature of 70 F. When new electro- lyte is required it can be made by mixing pure sulphuric acid (1.840 sp. gr., or 66 Baume) and distilled water in the proportion of i part acid to 5 1/4 of water, by volume, for 1.180 sp. gr. When mixing, pour the acid slowly into the water (not the water into the acid) and thoroughly stir with a wooden paddle. The final specific gravity must be read when the solution is cool. A metal vessel must not be used for mixing or handling the solution; a glazed earthenware crock or a lead lined tank is suitable, or a wooden vessel which has not been used for any other purpose, such as a new wash tub, can be used for mixing, but not for storing, the electrolyte. The electrolyte must be cool when poured into the cells. NOTE. For Type "ET" cells, when being first put into commision, electrolyte of 1.210 sp. gr., or 25 Baume must be used. If the electrolyte is to be mixed on the ground, the proportions of acid (of 1.840 sp. gr., or 66 Baume) and water are i part acid to 4 1/2 of water (by volume). With the battery properly installed and the charging connections made ready, the electrolyte can be poured into the cells, filling until the plates are covered 1/2 in. Initial Charge. The charge should be started at the normal rate (see " Table of Ratings,") as soon as the electrolyte is in the cells and continued at the same rate, provided the temperature of the electrolyte is well below 100 F., until there is no further rise or increase in either the voltage or specific gravity and gas is being freely given off from all the plates. Also, the color of the positive plates should be a dark brown or chocolate, and the negatives a light slate or gray. The temperature of the electro- lyte should be closely watched, and if it approaches 100 F. the charg- ing rate must be reduced or the charge stopped entirely until the tem- perature stops rising. From 30 to 40 hours at the normal rate will be required to complete the charge; but if the rate is less, the time must be pro- portionately increased. The specific gravity will fall somewhat after the electrolyte is added to the cells, and will then gradually rise as the charge progresses, until it is up to 1.210, or thereabouts. The voltage for each cell at the end of the charge will be between 2.5 and 2.7 volts, and for this reason a fixed or definite voltage should not be aimed for. It is of the utmost im- portance that the initial' charge should be complete in every respect. If there is -any doubt, it is better to charge too long than risk injury to plates by stop- ping the initial charge before it is complete. STORAGE BATTERIES 51 After the completion of a charge (initial or with the battery in regular ser- vice) and the current off, the voltage will quite rapidly fall to about 2.05 volts per cell and there remain while on open circuit, falling to 2 volts when the discharge is started. Operation; Battery in Service. Excessive charging must be avoided, nor must a battery be undercharged, overdischarged or allowed to stand com- pletely discharged. The battery should be preferably charged at the normal rate. It is im- portant that it should be sufficiently charged, but the charge should not be continued beyond that point. Both from the standpoint of efficiency and life of the plates, the best practice is the method which embraces what may be called a regular charge, to be given when the battery is from one-half to two-thirds discharged, and an overcharge to be given weekly if it is necessary to charge daily, or once every two weeks if the regular charge is not given so often. The regular charge (at or as near normal rate as possible) should be con- tinued until the voltage across the battery has risen to a point which is 0.05 to o. i o volts per cell below what it was on the preceding overcharge, the charg- ing rate being the same in both cases; for instance, if the maximum voltage per cell attained on the overcharge is 2.52, the voltage per cell to be reached on the regular charge is fyrom 2.42 to 2.47 volts per cell. In cases where it is possible to accurately determine the amount of discharge in ampere-hours, the following method is permissible and may be found more suitable, partic- ularly where there is difficulty in reading the voltmeter closely: charge at the normal rate until the number of ampere-hours charged exceeds the preceding discharge by from 5 to 15 per cent. The overcharge (at the same rate as regular charge) should be continued until the voltage across the battery has been at a maximum for one hour, five successive i5-minute readings showing no further rise and all cells are gassing freely. If rate is less than normal, the time at maximum must be propor- tionately increased. On discharge the voltage should not be allowed to fall below 1.75 volts per cell, with current at normal rate; the limiting voltage, however, is higher if the rate is less than normal, and lower if the rate is more than normal. Inspection. Once every two weeks, on the day before the overcharge, a specific-gravity reading 1 of all cells should be taken, and likewise all cells should be carefully examined to see that the plates are not touching each other or otherwise short circuited and have normal color. Near the end of the over- charge all cells should be looked over to see that they are gassing freely. Low Cells; Indications and Treatment. Falling off in specific gravity or voltage relative to the rest of the cells. Lack or deficiency of gassing on 1 On Type "BT" cells an individual cell- voltage reading, taken just before the end of overcharge, may be substituted for the specific-gravity reading, taking, however, a gravity reading at least once every three months. 52 AMERICAN TELEGRAPH PRACTICE overcharge as compared with surrounding cells. Color of plates markedly lighter or darker than the surrounding cells. In case of any of the above symptoms being noted, inspect the cell care- fully for the cause and remove at once. Short circuits are to be removed with a thin strip of hard rubber or wood; never use metal. If, after the cause of the trouble has been removed, the readings do not come up at the end of overcharge, the battery as a whole, or preferably the section in which the low cell is located, should receive a separate or extra charge. Impurities in the electrolyte will also cause a cell to work irregularly. Should it be known that any impurity has gotten into a cell,' it should be re- moved at once. In case removal is delayed and any considerable amount of foreign matter becomes dissolved in the electrolyte, this solution should be replaced immediately, thoroughly flushing the cell with water and putting in new electrolyte of 1.210 sp. gr. Sediment. The accumulation of sediment in the bottom of the jars must be watched and not allowed to touch the plates, as, if this occurs, rapid de- terioration will result. To remove the sediment, the simplest method is to lift the couples out of the jars after the battery has been fully charged, draw or pour off the electrolyte, clean out the jars and get the couples back and covered with electrolyte again as quickly as possible, so that there will be no chance of the plates drying out. Some new electrolyte (1.210 sp. gr.) will be required to replace that lost. When work is completed charge until volt- age has been at maximum for five hours and adjust gravity to standard. Evaporation. Do not allow the surface of the electrolyte to get down to the top of the plates; keep it at its proper level (1/2 in. above the top of the plates) by the addition of pure water only, which should be added at the be- ginning of a charge, preferably the overcharge. To transport or store the water, use clean, covered glass or earthenware vessels. Restoring Lowered Specific Gravity. It will not be necessary to add new electrolyte, except at long intervals (once every year or two), or when cleaning. When the specific gravity, with the cells in good condition and at full charge and normal temperature (70 F.), has fallen to 1.190, it should be restored to standard (1.205 to 1.215) by the addition of new elec- trolyte instead of water when replacing evaporation. To correct to normal temperature, subtract one point (o.ooi sp. gr.) for each 3 F. below 70 and add one point for each 3 F. above 70; for instance, electrolyte which is 1.213 at 61 and 1.207 at 79 will be 1.210 at 70. Battery used but occasionally; Putting the battery out of commission and in again. If the battery is to be used at infrequent periods, then a refreshing charge should be given once every two weeks. If the use of the battery or any of its cells is to be discontinued for a considerable time, then it must be treated as follows : After thoroughly charging, siphon or pour off the electrolyte (which STORAGE BATTERIES 53 may be used again) into thoroughly cleaned carboys or other glass receptacles which can be covered to keep out impurities, and as each cell becomes empty, immediately fill it with fresh, pure water. When water is in all the cells, allow the battery to stand 12 or 15 hours; and then draw off the water and the battery can then be allowed to stand without further attention. To put into service again, proceed as in the case of the initial charge; but use for all types, either new electrolyte of 1.210 sp. gr., or if the old electrolyte has been saved, add enough new of 1.210 sp. gr. to replace loss. If the gravity after the first charge is low, it should be restored to standard. Obtaining Additional Life. When the condition of the battery as a whole is such that, due to normal wear on the plates, it will not do its regular work, considerable additional life can be obtained from the plates by removing the couples from the jars and bending the connecting strap in the reverse direction, so that the sides of the plate which were against the jar will face each other in the same cell; in other words, the insides of the plates become the outsides. TABLE OF RATINGS Type LT BT CT PT ET Size of plates . . Sf'Xs" 7f"X7f" Normal rate (amperes) charge and discharge. J 1 it 3 41 THE EDISON NICKEL-IRON STORAGE CELL While on the subject of storage batteries, it may be well to give a brief description of the Edison nickel-iron storage cell recently brought out, and which is the latest development in this country in the manufacture of secondary cells. So far, the new Edison battery has been employed chiefly in operating the motors of electric vehicles, but inasmuch as its construction is a new depar- ture in storage-battery engineering and as its performance has been quite satisfactory, its possibilities as an efficient and economical source of e.m.f. may in the course of time insure it a more extended use commercially. The latest type of the Edison cell is known as "type A." Two sizes of cell, known as A~4 and A-6, have four and six positive plates respectively. Instead of employing a lead-peroxide and acid-electrolyte combination as is usual in the construction of lead storage cells, the Edison cell employs active materials con- sisting of nickel and iron oxides for the positive and negative electrodes, in combination with an alkaline electrolyte, the latter being a solution of caustic potash in water. The retaining vessels are made of sheet steel, all seams being 54 AMERICAN TELEGRAPH PRACTICE welded by the autogenous method. The retaining-cans are electroplated with nickel, which protects the steel from rust. A type A~4 cell contains four positive and five negative plates. Each positive plate consists of a grid of nickelplated steel supporting the active mate- rial which is contained in two rows of tubes, 15 in each row. The tubes are made from thin sheet steel, perforated and nickelplated, each tube being rein- forced by eight ferrules which preserve correct alignment and prevent expansion of the tubes. The active material in the tubes is intermixed with thin flakes of pure metallic nickel which are produced by an electrochemical process. The negative element or plate consists of 24 rectangular pockets supported in a nickelplated steel grid, in three horizontal rows. These pockets are the same as the tubes in the positive element except in shape and dimension. Each pocket is filled with oxide of iron or iron rust. When the pockets have been assembled in the negative grid, the whole is subjected to a heavy pressure which produces a solid and compact unit. The plates are assembled in the container in a manner similar to that employed in assembling lead cells. The electrolyte consists of a 2 1 per cent, solution of caustic potash in distilled water. It is claimed that the Edison cell does not deteriorate when left uncharged and that it is not injured by overcharging. CURRENT RECTIFIERS The Mercury-arc Rectifier. Current from an alternating-current source may be changed to direct current by means of current rectifiers. The diagram, Fig. 37, shows theoretically, the connections of the " mercury" rectifier. The alternating current to be rectified is supplied through the transformer shown at the top of the diagram. When a current of electricity is made to flow in a given direction between two points in a circuit separated by a gap containing vapor of mercury, should the direction of current be changed suddenly, the current will be interrupted due to a peculiar characteristic of mercury which in effect opposes a change in direction of current. Referring to Fig. 37, represents a glass tube or globe containing a deposit of mercury and exhausted of air. Terminals A, A', B and C are sealed in the glass. If at a given instant the terminal X of the alternating-current supply circuit is positive, the terminal A is then positive and the arc will flow between the terminal A and the mercury terminal J5, continuing on through the storage battery F, through reactance coil DI and back to negative terminal Y of the transformer. An instant later when the impressed e.m.f. has dropped to a value insufficient to maintain the arc against the counter-e.m.f. of the arc and the load, the reactance coil DI which has been charging now produces an induc- tive discharge in the same direction as formerly, which assists in maintaining the arc until the e.m.f. of the supply circuit has passed through zero; reversed, and built up in the opposite direction sufficiently to strike an arc between A' CURRENT RECTIFIERS 55 and the mercury terminal B. The arc now being maintained between A ' and B is supplied with the combined current from the transformer and from the coil DI. Obviously the current in the alternating-current supply circuit is con- stantly changing in direction, thus tending to enter at A and leave at A', and in the reverse direction to enter at A ' and leave at A a great number of times per second. The only action, however, which can take place is that the first impulse enters at A' and leaves at B and, due to the maintenance of the arc as before explained, the next impulse will enter at A and leave at B. Therefore the current continuously flows out at B in the same direction (direct current). The choke coils D and DI obstruct alternating , current, but permit direct current to pass ' VVWWWVWV ' through. Were it not for the action of these coils a current wave coming down either side would divide and be neutralized. Electrolytic Rectifiers. One type of elec- trolytic rectifier (the "Hickley") consists of a solution or electrolyte, such as phosphate of soda, in combination with carbon and aluminum electrodes, contained in a vessel F, Fig. 38, to which are attached radiator loops R permitting circulation of the solution (necessary on account of heat developed in the cell) thereby prevent- ing the weakening of the electrolyte. The direct current supplied by the rectifier is, of course, pulsating, but owing to the condenser effect of the cells whereby a portion of the current is recovered, currents are derived which are sufficiently steady for telegraph require- ments. With this type of rectifier 80 volts direct current are procurable from no volts al- ternating-current primary voltage. The durability of the electrolyte and the electrodes of this rectifier depends upon the amount of energy delivered, but if not overworked the rectifier will not require renewal oftener than once each year, assuming daily operation. A suitable transformer T is utilized to give either higher or lower voltage than that supplied by the available alternating- current .mains, the rectifier being designed to supply e.m.fs., ranging from 6 to 1,000 volts. These rectifiers may be operated on any alternating-current fre- quency from 25 to 133 cycles. The electrolytic rectifier is based purely upon the principles of electrolytic action as utilized in various branches of the electrical arts. If two rods of aluminum are placed in a vessel containing an alkaline solution such as carbonate of soda or phosphate of soda, and an attempt is made to pass a current of electricity from one rod to the other through the FIG. 37. Mercury-arc current rectifier. 56 AMERICAN TELEGRAPH PRACTICE solution, it is found that during a brief interval current will flow and then entirely cease. If, however, one of the aluminum rods is replaced by a rod of carbon, iron, or platinum, it at once develops that current will flow from the carbon to the aluminum electrode, but not in the reverse direction. The reason is given that the carbon gives off a gas which is dissipated through the electrolyte, while the aluminum electrode (if positive) retains a portion 1 of the gases generated. These gases, hydrogen and oxygen, unite with minute portions of the aluminum and form hydroxide of aluminum. The aluminum electrode being coated with hydroxide prevents the flow of current from it; therefore, when an alternat- ing current is supplied to the two electrodes it is only when the current is positive to the -carbon that current flows. It is evident that the negative impulse is obliterated, and that the secondary cur- rent delivered through the rectifier will consist of the succeeding positive impulses only. With an alter- nating current of low frequency it would seem that the utilization of alternate impulses only would produce a secondary current so slowly pulsating that it would not be sufficiently continuous for practical requirements, but the condenser effect above referred to operates to tide over the no-cur- rent intervals, and in practice it is found that the rectified currents are quite satisfactory as direct currents. MANAGEMENT OF THE ELECTROLYTIC RECTIFIER The Solution. The solution used in the Hickley - -,-,, , , f . rectifier is non-inflammable and does not contain .TIG. 3- Jiiieciroiy tic rec- tifier and switch panel. &Cld. In setting up the solution distilled water or rain water free from foreign matter and acids should invariably be used. Evaporation and decomposition of the water of the solution should be taken care of by occasionally adding fresh water. The amount of water decomposed is proportional to the amount of current in watts passing through the solution. When the solution has become "milky" in appearance and there is de- posited a sediment in the bottom of the cell, the solution requires renewing. x The sediment deposited contains small particles of aluminum which act as conductors and, consequently, reduce the efficiency of the cell. CURRENT RECTIFIERS 57 The Electrodes. The aluminum electrodes are made of a special alloy and are fitted with glazed porcelain tops. The formation of the hydroxide of aluminum takes from the electrode minute particles of aluminum, so that the greater the demand for current made upon the rectifier, the greater is the disintegration that takes place. The porcelain cap should be kept secure and tight, else the solution will creep under it and interfere with proper rectification, and allow the electrode and the solution to become quite hot. Electrodes should be suspended freely in the center of the jar and should not be per- mitted to touch the sides. The electrodes should not be handled any more than is absolutely necessary, as the hydroxide is liable to be destroyed by* undue handling. Installing and Starting. The location of the rectifier should be a place where there is good air circulation. The cells should be supported on in- sulators as it i? important that there should be no electrical contact between the rectifier and the ground, or between tlje cells. Owing to the possi- bility of the hydroxide coating being destroyed in shipping, it is well in setting up new cells to take the precaution to pass a small current through the rectifier for an hour or so before the entire load is thrown on. Should a rectifier for any reason be retired for an indefinite period, it is well to remove the electrodes and hang them in a dry place where they will be free from handling until required for service. If inspection of the rectifier should dis- close cracks in the porcelain cap of an electrode, the electrode should be replaced immediately. The humming sound sometimes in evidence may be due to the operation of the transformer or the reactance, but should the sound increase in volume, it is probable that a defect has developed and that alternating current is passing through. The gases released by decomposi- tion of the water in the solution escape through vent holes in the top of the cells. After long continued operation it may be found that the gases have carried upward particles of the chemicals from the solution, which on coming in contact with the air have formed crystals in and around the vent holes, thus interfering with the escape of the gases. Vent holes should be kept free of obstructions. Crystals which may have formed should be brushed back into the cell, where they will quickly dissolve. Fig. 39 is a reproduction of a photograph of a type B Hickley rectifier. FIG. 39. Electrolytic rec- tifier and switch panel. CHAPTER V POWER-BOARD WIRING; BATTERY SWITCHING SYSTEMS AND ACCESSORIES It is desirable that currents furnished by dynamos for the operation of telegraph circuits should be as nearly continuous as possible. By this is meant that the currents so supplied should closely resemble the non-pulsatory currents derived from primary batteries. The internal resistance of the gravity battery is about 21/2 ohms per cell, and this resistance in itself has the effect of controlling the current derived from a given number of cells. For instance, suppose a certain battery consists of 100 cells, each cell having an internal resistance of 2 1/2 ohms, and an e.m.f. of 1.07 volts; by Ohm's law it may be shown that on short circuit the current available from the battery will be 0142 ampere, or about 420 milliamperes, for 100 0.42 ampere. 2.5 Xioo Machine generators of electricity, such as are employed to furnish telegraph currents, have a very low internal resistance; considerably less than an ohm. When these machines are employed in place of chemical batteries, it is custom- ary to insert in the potential leads a total resistance which equals about 2 ohms per volt in order to protect the generators in case short circuits occur in the telegraph apparatus, or in event of grounds occurring on line wires at points close to the home office, and also for the purpose of controlling excessive " sparking" between the contact points of instruments, where circuits carrying comparatively large currents are opened and closed continuously. l The purpose of the power-board is to provide convenient mounting for the various accessories which as a whole make up the battery switching system. In most installations the power-board has mounted upon it the switches and fuses controlling the primary or motor circuits as well as those controlling the secondary or dynamo circuits. Other accessories usually mounted on the face of the board include ammeters, voltmeters, and field-regulating rheostats. When the type of battery resistance unit employed consists of a coil of 1 As explained in a later chapter, the present tendency of American telegraph engineer- ing in this regard, is to reduce the amount of resistance inserted in the potential leads. A reduction of the number of ohms per volt of potential requires that dependence must be placed in fuses to protect the dynamo in the event of short circuits, and that improved means must be availed of to reduce the spark at "make" and "break" contacts, 58 POWER-BOARD WIRING 59 "resistance" wire the various units are mounted on "coil racks," and when the type of resistance used consists of a form of incandescent lamp, the resist- ance units are mounted in "banks." Dynamo leads go directly from the power-board to the coil racks, or to the lamp banks, as the case may be. The diagram, Fig. 40, shows the motor "supply" wires leading from the busbars, through the fuses in each side of the circuit, to a double-pole single- throw knife switch, the latter serving to close or open the circuit leading to the motor end of a dynamotor. The dynamo end is shown as having the negative terminal grounded. The positive terminal is connected through the power-board and "resistance" rack to its prearranged service assignment in the operating-room. To Power Board thence < fuses /D.P.SJ. Switch Field fo Coil Rack and Operating Room. Brush Dynamo End Brush TT^ Armature Brush MotorEnd Brush FIG. 40. Dynamotor wiring connections. The Postal Telegraph-Cable Company's dynamo arrangement provides that 4o-volt potentials supply current for the operation of sounder circuits, repeater locals, duplex and quadruplex pole-changer and transmitter key circuits, lamp annunciator circuits, etc. For the operation of Morse short single circuits, loops, and for intermediate battery purposes 85-volt potentials are employed, while the longer main-line wires operated single Morse, are fed from i3o-volt or 2oo-volt potentials. Machines supplying respectively 200 volts of each polarity (positive and negative) are allotted to duplex operation. Three hundred and eighty-five volts, positive and negative, respectively, are used for the operation of quadruplex circuits, and high-potential "leak" duplex circuits. Forty- volt mains for use in local circuits are brought from the coil racks, before mentioned, to fuse-blocks situated on the tops of instrument tables 60 AMERICAN TELEGRAPH PRACTICE in the operating-room. Forty- volt, 85-volt, and i25-volt mains for application to single main lines are brought from coil racks to disks in the main-line switch- board (see Fig. 41) and properly marked, indicating potential and polarity. Two-hundred-volt and 385-volt "plus" and "minus" leads are brought directly from the power-board to cabinets located in the aisle ends of instrument tables, there connected through the proper resistance coils to six-point switches situated on the tops of the tables. Figure 42 shows the wiring and battery connections of the type of six- point switch used for the purpose. Throwing the switch lever to the right places the lower or 200- volt potentials in connection with the "line" contacts f r \ / ~\ A: /Switcht f ; Resistance* Coils-* \ \ + \ \ M r r- r "] ] 2 v^^^/ ii o >*-^ ^ wot O I K5t o ) o o '-- "'Fuses * ~~> 85+ f ( o 40+ o 1 o o 200+ J25+ ^5+ AO+ L Ond j 11 o. IL ^f / ^ 4 u FIG. 41. Potential mains connected to battery disks in main line switchboard. of the multiplex apparatus, while throwing the lever to the left connects the higher, or 3 85-volt potentials with the line instruments. In the newer offices the plan has been followed of carrying all battery wires leading from the power-board to the main switchboard and to instrument tables, in 2-in. iron piping. Where practicable the piping is imbedded in concrete flooring in the operating-room. Cast-iron hand-holes made from standard patterns are located at the aisle end of each table, the top of the hand-hole extending up into the wiring cabinet built into the end of the table. In this cabinet the various resistance coils and fuses are located. The Western Union Telegraph Company's dynamo arrangement differs POWER-BOARD WIRING 61 only in detail from that just described. Owing to varying conditions in different localities, uniform battery arrangement has not always been possible. In some of the older Western Union installations, the arrangement referred to in the beginning of Chapter III is used, whereby a number of dynamos capable of generating like e.m.fs. are connected in series, each machine having a poten- tial of, say, 60 volts. Six dynamos, each having an e.m.f. of 60 volts, if con- nected in series have an aggregate e.m.f. of 360 volts. A "tap" taken from the first machine of the series gives 60 volts, from the second 120 FIG. 42. Six-point battery-switch for mounting upon operating tables. volts, and so on in multiples of 60 volts until at the end terminal of the sixth machine an e.m.f. of 360 volts is available. As mentioned in Chapter III, with this arrangement it is necessary to provide a series of dynamos for each polarity, and to have available a third series as spare. In some installa- tions multiples of 70 volts have been used in arranging a series of dynamos, and it is, of course, feasible to connect machines having different voltage out- puts in a series, in order to meet particular requirements. The chief objection to the "series" arrangement is that in case an individual machine of a series becomes disabled, the entire series of which it forms a part has to be shut down until the disabled machine is repaired or replaced. 62 AMERICAN TELEGRAPH PRACTICE THREE -WIRE SYSTEM Where conditions are such that a three-wire system of commercial power may be availed of to advantage, it is possible by means of power-board switch- ing arrangements to obtain potentials of different values and of both polarities. 110) 220 FIG. 43. Three- wire system. no*at fir t L OCAL ANO f*UL T/PL.CX W/R/NG fffOM 3 W/fff I/O VOL. TMA/NS FIG. 44. Fig. 43 shows diagrammatically the connections whereby two no- volt gener- ators are coupled together in some commercial power systems, for the purpose of avoiding the stringing of an out-going and a return conductor for each no- volt generator. When two generators are coupled as indicated in the diagram, POWER-BOARD WIRING 63 one return wire serves for both machines. Also there is the additional advan- tage that while each generator external circuit is separate, for all practical pur- poses, it is possible to obtain from the two outside wires a potential of 220 volts no positive and no negative. Figure 44 gives the connections usually made when the three- wire system is utilized for telegraphic purposes. By following the connections it may be seen that either the no- volt positive or negative wire may be applied direct to quadruplex locals, main lines or call circuits, by way of the single-pole double-throw switch, i/2-ampere fuses and regulation resistance coils; the latter mounted on coil racks previously referred to, while the fuses and the knife switch are mounted on the face of the power board. COILS aOO~0/tM3 CACH CO/L BOAffO SOOSrsf? W/f?/NG fOrt QUADRUPLE* f/fOAf TH/f W//?/f0 VOL TM/l/t/S. FIG. 45- By connecting the two outside conductors of the three-wire service through the double-pole single-throw switch, fuses, and resistance coils, no-volt potential is obtained for the operation of duplex circuits; that is, no volts of each polarity. The double-pole single-throw switch also places the commercial no-volt positive and negative leads, each in series with the generator terminals of "booster" motor-generators, thereby adding the voltage of the latter to that of the former. Boosters having out-puts of 130 volts positive and negative respectively and connected as shown in Fig. 44, raise the potentials to 240 volts for the operation of duplexes and short quadruplexes. 64 AMERICAN TELEGRAPH PRACTICE BOOSTER CONNECTIONS Figure 45 shows the starting-box and "booster" 1 connections necessary to obtain quadruplex potentials of each polarity from three- wire mains. Generator Pole Changing Switch for Meter. A.C. Motor No. 2 FIG. 46. Typical power-board wiring. 1 The "booster" consists of a generator driven by a motor mechanically connected to its armature shaft. The terminals of the generator are connected in series with one "leg" of the feed system. The current in the feed wire excites the field in proportion to the amount of current flowing. Inasmuch as the armature is independently rotated in the field; will produce an e.m.f., in proportion to the excitation, and this e.m.f., is added to that of the feed system. POWER-BOARD WIRING 65 POWER-BOARD WIRING Figure 46 shows the power-board wiring of an installation comprising two motor-dynamos with alternating-current primaries. The primary circuits from transformers through the double-pole single-throw switches to motors, and the secondary, or dynamo circuits through field-regulating rheostats and double- pole, double-throw switches to service mains are readily traceable. Figure 47 shows the wiring of a power-board, embracing two panel-boards, switches, fuses and resistance units of a "rectifier" installation. The double- pole, single-throw switch shown in the upper right-hand corner serves to connect the alternating-current supply circuit with the electrolytic rectifier cells. The Direct Current Rectifier / \ Rectifier FIG. 47. Power-board wiring of an electrolytic rectifier installation. secondary, or direct current derived has the negative side grounded from the double-pole, double- throw switch shown in the center of the diagram, while the positive lead is connected through resistance coils and fuses to the desired ser- vice assignment. Figure 48 shows a diagram of the wiring of a power-board installation which provides all necessary switching arrangements for a motor-generator plant consisting of: 8 motor generators with no- volt primaries. 3 of them having 385-volt secondaries. 3 2oo-volt secondaries. i 555 o s q- mil = 0.00155 sq. in. i sq. cm. = 197,300 cir. mils = 0.155 sc l- in. =0.00108 sq. ft. i sq. in. = 1,273, 240 cu ~- nrils= 6.451 sq. cm. =0.0069 sq. ft. i cir. mil-foot = 0.0000094248 cu. in. i cu. cm. = o.o6i cu. in. i cu. in. = 16.39 cu - cm - The microhm = 0.0000001 ohm. Microhms per inch cube = 0.393 7 X microhms per centimeter cube. Pounds per mile-ohm = 5 7. 07 X microhms per cm. cu. X specific gravity. Ohms per mil-foot = 6.01 5 X microhms per cm. cu. The square of the diameter of a given wire expressed in mils (o.ooi in.) gives the circular mils. 78 AMERICAN TELEGRAPH PRACTICE Problems involving comparisons of conductors having unequal resistances, require that: The square of the diameter of each wire be multiplied by the length of the other. The ratio of the resistance of one wire to that of the other is obtained by dividing the products thus obtained. The resistance value sought is determined by multiplying the ratio by the known resistance. For example: A given wire of 50 mils diameter and 20 miles long has a resistance of 200 ohms. An- other wire has a diameter of 40 mils and is 40 miles long. Assuming that the wires are made of the same material what is the resistance of the second wire? Solution: * 5 o 2 X 40= 100,000, relative resistance of first wire. = 32,ooo, relative resistance of second wire. 100.000 --- = 3. 125, ratio of second wire to the first. and 200X3.125 = 625 ohms. As an example of calculating the required diameter in mils of a conducting wire_, assume that a wire i mile (5,280 ft.) long has a diameter of 100 mils and a resistance of 10 ohms; what would be the diameter of a wire made of the same material of which a length of i mile has a resistance of 40 ohms? Solution: = 4 (ratio of resistances) 5,280 - = 1,320 ft. and 2,500, which is the square of the diameter of the second wire, and \/2,5oo = 5o (mils) CIRCUITS AND CONDUCTORS 79 SPECIFIC RESISTANCE, RELATIVE RESISTANCE, AND RELATIVE CONDUCTIVITY OF CONDUCTORS Pertaining to Mathiessen's standard Resistance i n microhms at o /~* Relative Relative l\/Tof olo ' t j , , IVietaiS resistance, conductivity, Centimeter Inch cube per cent. per cent. cube Silver, annealed i-47 o-579 9 2 -5 108.2 Copper annealed i ^ o. 610 97 5 IO2.6 Copper (Matthiessen's standard). 00 1-594 0.6276 y / o IOO IOO.O Gold (99 . 9 per cent, pure) 2. 2O 0.865 138 72.5 Aluminum (99 per cent. pure).. . 2.56 i .01 161 62.1 Zinc 5-75 2.26 362 27.6 Platinum, annealed 8.98 3-53 565 17.7 Iron o . 07 3.. C7 57 17.6 Nickel v / 12.3 O O / 4-85 778 12.9 tin I 3 I 5.16 828 12. I Lead 20.4 8.04 1,280 7.82 Antimony 3^2 13 O 2,210 4-53 Mercury O 94.3 O V 37-1 5,930 1.69 Bismuth 130 o 094 40. 25. 79 10 101 .9 10,381 3 1 - 32. i. ii 90.7 8,234 25- 40. i . 26 12 80.8 6,530 20. 50. i-59 13 72.0 5,i78 15-6 64. 2.0O 14 64. i 4,107 12.4 81. 2-53 82 AMERICAN TELEGRAPH PRACTICE DIMENSIONS AND RESISTANCE OF PURE COPPER WIRE (Continued) (Specific gravity 8.9; resistance at 75 F.) American or B. & S. gage Diameter d in mils, i mil = . ooi in. Circular mils d 2 ) Pounds per 1,000 ft. Feet per pound R ohms per 1,000 ft. 15 57-1 3,257 9.8 IO2. 3-2 16 50.8 2,583 7-8 128. 4-0 17 45-3 2,048 6.2 161. 5- 18 40.3 1,624 4-9 204. 6-4 19 35-4 1,252 3.78 264. 8.0 20 32.0 I,O2I ' 3-09 324- 10. 2 21 28.5 810 2-45 408. 12.8 22 25-3 643 1.94 SIS- 16.2 23 22.6 509 1-54 650. 2O.4 24 20. 1 404 I. 22 819. 25-7 25 17.9 320 97 1,033. 32-4 26 15-9 254 77 1,302. 40.8 27 14.2 2OI .61 1,642. 51-5 28 12.6 1 6O .48 2,071. 64- 2 9 n-3 127 .38 2,611. 82. 30 IO.O IOO 30 3,294. I0 3 . 31 8.9 80 .24 4,152. I2 9 . 32 7-9 63 .19 5,237. I6 4 . 33 7-i 50 15 6,603. 207. 34 6-3 40 . 12 8,328. 26l. 35 5-6 32 . IO 10,500. 329. 36 5-o 25 .08 13,240. 415. 37 4-5 20 .06 16,691 . 524. 38 4.0 16 OS 20,850. 660. 39 3-5 12 .04 26,300. 8 3 2. 40 3-i 10 03 33,200. 1,050. From the foregoing it is evident that conductor resistance depends upon dimension, composition, and temperature of the conductor. So far as external conductors in aerial lines and cables are concerned the average difference in temperature between summer and winter months, say 100 F. and 22 F., results in an increase of resistance while the higher tempera- ture prevails, of about 26 per cent, for iron wire, and 18 per cent, for copper wire. Where the windings of resistance coils and of electromagnets are involved the temperature of the conducting wire, or rather the variation in the tempera- ture of the wire is always a factor to be reckoned with. The use to which resistance coils are put is such that frequently the temperature of individual coils may rise above 150 F. or 65 above normal (75 F. or 23.8 C.). Joint-resistance. Referring to Fig. 60, where a source of e.m.f. has both of its terminals joined by two conducting wires. If the several conducting CIRCUITS AND CONDUCTORS 83 wires are of equal length and cross-section, composed of the same material and at equal temperatures, the current will divide equally between the two, and the electrical resistance of the joint-path will be the same as if there were but one conductor of the same length and of a cross-section equal to the total of the cross-sections of the two individual conductors. The current, therefore, existing in each conducting wire is in the same proportion to the total current circulating as the sectional area of one branch is to the total sectional area of the joint-path. It is obvious that when a circuit is divided into two or more branches, variations in the characteristics of the individual conductors of the joint circuit, determine the amount of current flowing in each branch. Thus, one of the branches may be of greater length than the others, or the cross-sectional area of one may be greater than that of the others, or there may be involved a difference of temperature sufficient to increase or decrease the resistance of one of the conductors as compared with the electrical resistance of another of the conductors of equal dimension and of the same composition. A derived or branch circuit is in effect a shunt circuit, see Fig. 61. In a case such as that illustrated in Fig. 61, it is evident that the current R. FIG. 60. FIG. 61. . FIG. 62. from the battery is shunted by the wire A-B, and, if we assume that the compo- sition and dimensions of the wire in the longer loop and in the shorter circuit are identical, it is obvious that a greater portion of the total current circulating will exist in the shorter path. There are definite laws for determining the resistance values of shunt circuits,' where it is desired by this means to regulate the amount of current flowing in circuits in which a shunt path forms a part of the joint circuit. These laws shall be considered presently. There exists a popular fallacy in regard to the circulation of electric currents, which in the minds of those not familiar with electrical laws causes confusion, that is "that a current of electricity always takes the path of least resistance." The truth is that the major portion of the current flowing traverses the path of least resistance, but the current as a whole divides between the various paths in proportion to their electrical resistances individually. The relative strengths of current flowing in two branches of a circuit (Fig. 60) is inversely proportional to their resistances, and on the other hand, proportional to their conductances. If the resistance of RI in Fig. 60 is 20 ohms, and RZ, 30 ohms, then the portion of the total current flowing in RI will be "as 20 is to 30," or, three- fifths of the total current will flow through RI and two-fifths through R 2 . 84 AMERICAN TELEGRAPH PRACTICE In Fig. 62 the joint resistance of the divided circuit between A and B will be less than the individual resistance of either branch, as, through this portion of the circuit the current has a joint path equaling in sectional area the sum of the sectional areas of each conductor taken singly. The joint- conductance will be the sum of the two individual conductances. The state- ment of a law by means of which the joint-resistance may be determined is : The joint-resistance of a divided conductor is equal to the product of the two separate resistances, divided by their sum. In general, this is referred to as the Law of shunts. In Fig. 62 if the branch R\ has a resistance of 100 ohms* and the branch R^, a resistance of 200 ohms, then, where R equals the value of the joint-resistance in ohms, R = 100X200 100+200 2O,OOO 300 66 2/3 ohms. As illustrated in Fig. 63, the same result may be ascertained graphically. If two perpendiculars are erected at the extremities of a base line as shown , each perpendicular representing in height the value of the resistance of one 200 Ohms Ohms FIG. 63. of the conductors of a divided circuit and two diagonals drawn as illustrated, a perpendicular extending from the base line to the point of intersection of the two diagonals will indicate the value of the joint-resistance of the two branches. A line drawn parallel to the base line and extending between the two outside perpendiculars through the point of intersection will indicate on either of the latter the joint-resistance in ohms. CIRCUITS AND CONDUCTORS 85 When it is required to compute the joint-resistance of three or more conductors as in Fig. 64 or 640, the same formula applies as in the case of a divided circuit having two branches. First the joint-resistance of any two of the branches is ascertained, and the result thus obtained combined in the same way with another of the conductors, and so on until all branches have been included in the calculation. To illustrate, suppose that in Fig. 64, RI has a resistance of 40 ohms, R<>, 50 ohms, and R 3 , 60 ohms, then, combining RI and R% first, we have 40X50 : - = 22 2/9 ohms. 40+50 Combining the joint-resistance of have and R 2 with that of the third branch, we R = _22 2/9X60 16+ ohms. 22 2/9 + 60 If there were a fourth branch, the process would be continued in the same manner, that is, the joint resistance of the first three branches, 16+ ohms, FIG. 64. FIG. 640. would be combined with the resistance of the fourth branch as above explained. The graphic method illustrated in Fig. 63 also may be used to determine the joint-resistance of three or more branches of a divided circuit. The derived perpendicular indicating the joint-resistance of the first two branches considered, would then have a diagonal drawn from its upper extremity intersecting a fourth diagonal representing the resistance value of the third wire, and so on. Where it is desired to determine the joint-resistance of a large number of conductors connected in parallel it will facilitate the computation to employ the rule: The joint-resistance of any number of conductors in parallel is the reciprocal of the sum of the reciprocals of the separate resistances. Figure 64 represents a derived circuit having three branches. Let RI, RZ, and RS represent the respective resistances of the three branches, then j?-> XT* and -=- are the separate reciprocals of the resistances of each KI jci ^ 3 branch. Therefore the joint-conductivity would equal i i i T> ' TT> I T> -ft-l A2 J\-3 86 AMERICAN TELEGRAPH PRACTICE And, as the joint-resistance is the reciprocal of the joint-conductivity, Therefore 40X50X60 K =7 . . , \ . , vx , N , , ^ = io-h onms, (50X60) + (40X60) + (40X50) or the same result as was obtained by the first method. Figure 640 shows several conductors leading from a source of e.m.f. and placed in contact with the earth as also is the opposite terminal of the bat- tery. Electrical circuits arranged in this way are termed ground-return circuits, while the arrangement of conductors shown in Fig. 64 provides what is termed a metallic circuit. In telegraphic practice, the earth is generally availed of as a return conductor. There are two or three different theories held pertaining to the part which the earth plays in completing electrical circuits, but so far as present purposes are concerned none of these theories are of practical impor- tance. Suffice it that for ordinary requirements we are able to obtain a complete electrical circuit by using the earth as a part of circuits otherwise made up of metallic conductors. Shunt Circuits. In any combination of branch circuits, each branch acts as a by-pass for a portion of the current, and any branch carrying a portion of the current in a circuit is, in effect, a ~P | R ^ shunt to the other branches of the circuit thus f ^500 divided. There are instances where the ap- plication of a shunt circuit requires that a FIG. 65. definite value be given the shunt in order to regulate the amount of current flowing in a branch circuit connected in parallel therewith. For example, suppose it is required to provide a shunt "S," (Fig. 65,) having a resistance which will permit one-third of the total current in the circuit to flow through the 500- ohmcoil shown on the right, what must be the resistance of the shunt? Let R represent the resistance of the circuit to be shunted, n represent the multiplying power of the shunt, S represent the resistance of the shunt, Then s _ R ni The value of n is arrived at when it is known what proportion of the current is to be shunted. In the case before us, it is required that one- third of the total current shall pass through branch R, therefore the multiplying power is 3, and '' ' 500 3-! =25 ' 5 = 250 ohms. FALL OF POTENTIAL 87 To find the multiplying power of a shunt of known resistance, the following formula applies: R+S Suppose, for instance, that the multiplying power in the former example is unknown, and it is desired to learn the current proportions in each branch of the circuit. Then and ,5 = 250 500+250 250 = 3* E.M.F. 120 Volts 110' 100 1 90 i 80. 7d 60 r 50. 40' 30' 20' 10' 1 FIG. 66. Fall of potential. Fall of Potential in an Electric Circuit. Some little confusion at times results from the use interchangeably, of the terms electromotive force and potential difference. In practice, it is usual to regard a primary battery, storage battery, dynamo-electric machine, or other generator of electricity as having a definite e.m.f., while an external circuit connected to the terminals of a given e.m.f. will have a difference of potential which decreases in value as resistance is overcome, in a direction from positive to negative terminal of the source of e.m.f. If a circuit external to the source of e.m.f. consists of a single conductor of uniform composition and uniform dimension throughout, and consequently of uniform resistance; it is found that the potential falls uniformly in a direction as stated above. When a current flows in a conductor such as that illustrated by the heavy line A-E, Fig. 66, the difference of potential between the conductor and the earth at E decreases in the direction in which the current is flowing. If a 88 AMERICAN TELEGRAPH PRACTICE dotted perpendicular is erected at the battery end of the line representing the conductor; the height of the perpendicular representing the value of the e.m.f. in volts, and a horizontal line is drawn from the top of the former as shown, we have a means of determining the difference of potential between any specified point along the conductor and the earth. For example, if it is desired to ascertain the difference of potential between a point (C) halfway along the conductor, and the earth, the erection of a perpendicular at that point between the base line and the dotted horizontal will indicate the difference of potential in volts as measured by an identical height of the perpendicular at the end of the base line; in this case, 60 volts. Obviously the difference of potential between any point along the conductor and the earth may be determined in the same manner. 160 c^ 150 1 "\ 140i " x ^ 130, *\ 120 1 1101 \ 100. 90i --i^ 80 1 70 1 60. 50 40 1 30> 20' 101 FIG. 67. Fall of potential. Consider a circuit such as that depicted in Fig. 67, where an e.m.f. of 160 volts is applied to a grounded circuit of 340 ohms resistance, and it is desired to ascertain what the difference of potential is at a point 150 ohms "distant" from the source of e.m.f. If the distance in ohms from o to 340 is properly graduated along the base line representing the conductor, and the e.m.f. in volts properly indicated along the perpendicular representing e.m.f., then a perpendicular erected at that point on the base line indicating 150 ohms will be found to have a height identical with the height of the perpendicular repre- senting the e.m.f. at a point which indicates 90 volts, approximately; or, to be exact, 89.41 volts. Obviously the difference of potential at any point along the conductor, distant in ohms from the source of e.m.f., may be determined in the same manner. FALL OF POTENTIAL 89 The above graphical method of determining the difference of potential in a circuit while of considerable value in clearly setting forth the principles in- volved in the fall of potential, is seldom used in practice. A formula based on Ohm's law, and by which the same end may be attained, is given herewith: Where E represents the applied e.m.f., in volts, R represents the resistance in ohms, of the entire circuit, Ri represents the point distant in ohms from the source of e.m.f., where the value of the difference of potential is desired, then X = R Employing this formula to determine the difference of potential at a point 150 ohms distant from the source of e.m.f. in a circuit having a total resistance of 340 ohms (Fig. 67) and an applied e.m.f. of 160 volts, we have, =160 ^ = 340 and i6oX(34Q-i5) 340 _ 160X190 340 X = - = 89.41. ans., 89.41 volts, or the same as was determined by the graphic method. ELECTROSTATIC CAPACITY OF CONDUCTING WIRES When, as "charge," electricity is present upon the surface of bodies, it is referred to as static electricity. In the operation of telegraph lines, static electricity, is encountered; generally as a disturbing agency, due to the fact that charge is accumulated on the surface of the conductor- from both inter- nal and external sources. A knowledge of the principles of electrostatics is essential in the study of telegraphy, and while it is true that the subject is pretty well covered in text- books dealing with electricity and magnetism, the bearing which the subject has upon practical telegraphy has not always been clear to the student. If one end of a telegraph line is connected with one terminal of a source of e.m.f., while the other terminal of the battery and the other end of the line are grounded (Fig. 68) it may be shown that when the key K is closed, current traverses the entire circuit almost instantaneously, affecting the distant end 90 AMERICAN TELEGRAPH PRACTICE of the conductor at nearly the same instant that the key is closed. The first indication of electrification at the distant end, however, is quite feeble, but the current strength increases gradually until maximum value obtains. If a current-indicating meter be inserted at the distant end of the line, there will be no deflection of its pointer until the current has attained a strength sufficient to energize the coils of the instrument thereby causing the indicating needle to move from its position of rest. The more sensitive the instrument employed for the purpose, the earlier will be the indication of current passing through it. K FIG. 68. 1 After the first movement of the needle has been observed, the amount of deflection will increase gradually until when maximum current obtains at the distant end, the pointer will have moved to a definite position, distant from its position of rest. On short lines the interval elapsing between the time the key is closed and the time constant-current is indicated is, of course, very brief. Should several current-indicating instruments be inserted at different points along the line as shown in Fig. 69, when the key is closed placing the source of e.m.f. in contact with the line, the instrument nearest the battery end of the <2> CD FIG. 69. line will be the first to indicate the presence of current, the others following in order, each a fraction of a second later than the one behind it until the in- dicator located at the distant end of the line is affected. From the foregoing it is apparent that current does not arrive at the distant end of a line "all at once" as does a bullet at a target. The initial portion of current flowing into a conducting wire is retained, or accumulated on the sur- face of the conductor, the quantity accumulated depending upon the length and surface of the wire, upon its distance from the surface of the earth, and upon the nature of the dielectric intervening between the conductor and the earth. It is convenient to assume that the conducting wire absorbs the first portion of each charge sent into it, and that its capacity of absorbing electric charge has to be satisfied before constant current conditions obtain in the circuit of ELECTROSTATIC CAPACITY OF LINES 91 which the conducting wire forms a part. The effect is as if the conductor requires to be "saturated" before delivering current at the distant end of the line, in a somewhat similar manner to that of a sponge, which has to be satu- rated to capacity before water drips from it. It is found that when a circuit, such as that shown in Fig. 69, is closed by means of a key or otherwise, the indicating needles of the instruments located on the half of the line nearest the source of e.m.f. " over-shoot " the point at which they finally come to rest, while the indicating needles of the instruments located on the other half of the line have a continuously increasing angle of deflection until the conductor has become fully charged and normal current prevails, at which instant the needle has reached its maximum deflection and remains there. The conditions which obtain during the time the current is equalizing through- out the entire circuit is referred to as the variable state. The duration of the variable state varies in different circuits and depends upon the physical and electrical characteristics obtaining in any given instance. 1 The permanent state has been established in a circuit when the current strength has been equalized in the conducting wire, or when the current value has become constant. The permanent state is first established in the middle of the line, at an instant practically four times sooner than constant-current conditions prevail at either end of the line. The statement that the " quantity of charge accumulated on the surface of a conducting wire, depends upon the dielectric intervening between the conductor and the earth," in so far as aerial lines are concerned, involves an understanding of the conditions prevailing at all points along the length of the conductor. When properly suspended and insulated, the conductor is enveloped in aji insulating stratum of air, but this stratum varies in thickness as the conductor is carried past objects which are in contact with the earth. As the surface of a conducting wire increases with its radius, the inductive capacity of the wire increases proportionately. The greater the inductive or electrostatic capacity of a conducting wire, the longer time it takes to charge it the longer the duration of its variable state. The electrostatic capacity of a conducting wire in effect diminishes the velocity of electrical action along the conductor, that is, each time the circuit is closed through a source of e.m.f. electrostatic capacity has the effect of retarding or delaying the initial appearance of current at the distant end of the wire. In the transmission of telegraph signals over a wire, the circuit is closed and opened, say, four or five times per second, and in the case of long lines, the effect of electrostatic capacity is to considerably curtail the number of impulses or signals which may be sent over the wire in a given time. Where slow signal- 1 A further treatment of the subject of capacity is given in Chapter X in con- nection with electric condensers, and in Chapter XI dealing with "Speed of Signal- ing," also see "The Capacity Balance," Chapter XV. 92 AMERICAN TELEGRAPH PRACTICE ing is concerned the effects of capacity are not of much consequence, but where high speeds are concerned the electrostatic capacity of a circuit may be the criterion of speed attainable. Electrostatic Induction. Where a charge of electricity of either sign (positive or negative) exists on a conductor, it will induce in neighboring conductors a bound static-charge of opposite sign on that side of the adjacent wires nearest to it and a charge of identical sign on the sides farthest away from it. See Fig. 6ga. + +H-+ -h +. + + -f 4- + FIG. 6ga. Electrostatic induction. The upper conductor is represented as having a positive charge and the interaction which takes place between the upper and the lower wires results in a bound negative charge being induced on the upper surface of the lower wire, while on the under side of the latter a free positive charge will appear. If the current in the upper wire should be changed from positive to negative, the reverse process would take place in the lower wire; thus, if the direction of current in the upper conductor is alternated from positive to negative and back again either slowly or rapidly a continuous interaction takes place be- tween the upper and lower conductors. Electromagnetic Induction. Any conducting wire charged with current has surrounding it at all points along its length, lines of force in the form of closed rings or loops, and the space surrounding a charged conducting wire is B D FIG. 70. FIG. 71. an active magnetic field. As will be described later in connection with the theory of electromagnetism, current in a circuit while increasing or decreasing in strength exercises an inductive effect upon neighboring circuits. It is true also that due to the expansion and contraction of the magnetic rings surround- ing the conductor, as the circuit is closed and opened, there is an inductive action of the current in the conductor upon itself. Naturally this effect of self-induction is great if the circuit (as in the case of magnet coils) is made up of a coil of many convolutions, and much greater still when the turns of wire surround an iron core. If a current is caused to flow in the wire A-B, Fig. 70, in the direction indicated by the arrow, commencement of cur- rent or increase in its strength induces a current in the neighboring con- ELECTROMAGNETIC INDUCTION 93 ductor C-D in the direction indicated, or the reverse of the direction of current in the inducing circuit. In a circuit such as that shown in Fig. 71 where the conducting wire is coiled back upon itself, an increasing current flowing in the direction A-B in the outer convolution, induces a current to flow, or to attempt to flow in the opposite direction C-D. The induced current being greatly inferior to the original current in strength, results only in opposing the latter and delaying its rise to maximum strength. When the circuit is opened and the current strength in that portion of the circuit A-B is decreasing, it tends to induce a current between C and D in the same direction as that of the origi- nal current, and this results in prolonging the duration of the latter by virtue of the induced opposition to its decrease. In either case the effect of self- induction is to oppose change, and in a sense might be regarded as electro- magnetic inertia. The fact that individual impulses are thus in turn retarded and prolonged diminishes the rate at which signals may be sent over long lines, as fewer distinct impulses can be transmitted in a given time. In comparison with the effects of electrostatic induction, the effects of electromagnetic induction in lines of ordinary lengths is very slight. In very long lines the reverse is sometimes true. Where magnet coils are concerned, such as are essential in terminal apparatus, self-induction plays a prominent part in limiting the speed of operation over a given line, and in limiting the length of line over which satisfactory operation may be maintained. The length of time required for an impulse to reach the distant end of a line and rise to a strength sufficient to operate receiving apparatus, particu- larly electromagnetic devices, depends upon the distributed electrostatic capacity and the ohmic resistance of the circuit. In fact, it has been deemed good practice to consider that the limits of satisfactory operation are pro- portional to the product of these two factors, which would mean that the prod- uct should be kept at as low a figure as practicable. The electrostatic capacity of aerial conductors suspended at any height above the surface of the earth is intricately involved with the number of and relative proximity of other wires on the same pole line. A single No. 12 B. & S. gage copper wire suspended 30 ft. above the earth, with both ends grounded, was in one instance found to have a capacity of 0.009379 micro- farads per mile and an inductance of 0.003149 henries per mile. Two similar wires placed i ft. apart and suspended 30 ft. above the earth were found to have a capacity between either wire and the earth of 0.012 microfarad per mile. For a line 500 miles long this would mean a total capacity of 6 microfarads. It is important to note the difference in the interactions taking place be- tween neighboring conductors, attributable to electromagnetic induction and to electrostatic induction. Figure 72 gives a cross-section or end-view of two conducting wires carrying current. The closed loops shown perpendicular to and surrounding each 94 AMERICAN TELEGRAPH PRACTICE conducting wire represent the magnetic rings which exist during the periods that current is flowing in either direction through a conductor. If we sup- pose a condition such as that shown in Fig. 73 where current is temporarily suspended in one wire while current in the other wire is increasing or decreas- ing in strength, a current will be induced in the wire B due to its cutting the expanding and contracting magnetic rings, as they are called into being or destroyed, by the closing or opening of the circuit of which wire A forms apart. As we proceed with the study of the various factors which have a bearing on the current strength in electrical circuits, it becomes apparent that Ohm's law, strictly speaking, is not applicable except where the factors are constant. FIG. 72. FIG. 73. FIGS. 72 and 73. Electromagnetic induction. At the outset it is apparent that when a current of electricity is turned into a circuit for a brief instant and then interrupted, or when a circuit is first closed through a source of e.m.f., the current strength in the circuit for a short period is not truly represented by the formula From what has been stated in regard to the effects of capacity and induc- tance in electric circuits, it is evident that the factor of "time" plays an important part in determining the current strength obtaining in a given circuit at a given instant. Pulsatory currents of either polarity or currents which alternate in direc- tion do not have a value in accordance with Ohm's law. Helmholtz the great German physicist interpreted Ohm's law in a form which takes into consideration, the element of "time" and from this evolved ELECTROMAGNETIC INDUCTION 95 a formula which gives the current value in a circuit at any given instant, thus, Rt\ ~ L ) I-e I Where /, is the current in amperes, , is the applied e.m.f., R, is the resistance in ohms, t, is the time in seconds, L, is the inductance in henries, e, is the base of the system of natural logarithms, or 2.7183. In telegraph circuits operated at usual speed, it is obvious that the low value of the negative exponent in the above formula would give a determination practically agreeing with that obtained by means of Ohm's law, but if the value of / is reduced or the value of L is increased, a point is soon reached where the simpler law would not give a true indication of the condition. The electrical properties as well as the physical properties of a circuit may be determined without regard to the e.m.f. to be applied to it. A tele- graph circuit, including as it does the magnet winding of receiving in- struments, offers a greater resistance to the passage of currents which alter- nate in polarity than is represented by the resistance of the circuit in ohms, the additional resistance being the direct result of inductance. The resist- ance in ohms combined with the inductance in henries produces impedance (symbol Z). If L = the inductance in henries, 1 N = cycles per second, R = the resistance in ohms, then Current in amperes e.m.f. impedance Assume a circuit having values as follows: R= 1,200 ohms, N= 20 cycles per second, L= 6 henries. By means of the above formula the impedance (Z) may be shown to be 1,417, and the maximum current to be 141 milliamperes, where an e.m.f. of 200 volts is applied to the circuit; while Ohm's law 1 A definition of the value of the henry is given in Chapter I, and a method of measuring inductance is described on page 103. 96 AMERICAN TELEGRAPH PRACTICE would give the current value as 167 milliamperes. Thus by considering frequency and inductance as factors, in the same sense that resistance is considered, it is possible to arrive at the true value of the current flowing in the circuit. ELECTROMAGNETISM AND ELECTROMAGNETS The phenomenon of magnetism may be exhibited by bringing a piece of iron into the neighborhood of a natural magnet (lodestone), permanent magnet or any form of electromagnet, where it may be attracted. If free to move, the iron will come into contact with and cling to the magnet. A powerful magnet will have an appreciable effect upon a delicately poised needle (see Fig. 74) even if the two are situated a considerable distance apart, and undoubtedly the influence of the magnet extends far beyond the boundary established by the methods ordinarily employed to determine its range. It is customary to consider iron as being peculiarly subject to magnetic influence, and as stated in the chapter on Electricity and Magnetism, steel, FIG. 74. Compass needle deflected by ... the influence of a horseshoe magnet. mckel > cobalt > chromium, manganese and other substances are similarly in- fluenced to an extent varying in the case of each substance. Also it is known that all kinds of matter possess this magnetic quality in some degree. It has been shown experimentally that temperature plays an important part in determining the magnetic susceptibility of a substance. Iron, for instance, when heated to 750 is irresponsive to magnetic influence, the reason, roughly stated, being that the atomic structure of the substance is so disarranged by high temperature that the atoms are unable to "line-up" magnetically in response to the influence of the inducing magnet. Although for present purposes (investigating the properties of electric circuits), we may acquire the desired information relating to the magnetic properties of conductors carrying currents of electricity, by studying the magnetic action of solenoids, the further study of electromagnets requires that the connecting link between the two the core be treated of at the same time, or in connection therewith. A helix of wire carrying a current of electricity possesses magnetic proper- ties. When the helix consists of a coil of insulated wire and is wound around a bar of soft iron, the iron becomes magnetized when the electric circuit is energized, and the combination of helix and core is called an electromagnet. Substances in which a magnetizing force produces a high degree of magneti- zation, are regarded as possessing high permeability. ELECTRON AGNETISM 97 The intensity of magnetization, while in part dependent upon the strength of magnetic field produced by the helix, is also dependent upon the properties of the metal forming the core, that is, upon its permeability. If a conducting wire connected through a source of e.m.f. is bent into a loop as in Fig. 75 the lines of force will thread through the loop from one end to the other in a direction depending upon the direction in which the current is flowing through the conducting wire. Should an iron core M be brought close to the loop of wire thus formed, the core would tend to enter the loop lengthwise, that is, place itself with its longest axis projecting into the loop of wire, and always in a direction the same as that taken by the lines of force. If the conducting wire is coiled into a helix or solenoid having a num- ber of loops, as in Fig. 76, the lines of force around each turn or loop will reinforce those around neighboring loops, and the cumu- lative result will be the formation of numerous long lines of force as shown, which extend through the entire coil. The above statements mean that the solenoid possesses properties iden- tical with those possessed by bar magnets (permanent magnets). Inasmuch as the lines of force enter one extremity and leave at the other, the solenoid exhibits the phenomenon of polarity, having a north and a south pole. A FIG. 75. Direction of current in a completed circuit and resulting lines of force. FIG. 76. The solenoid. coil such as that shown in Fig. 76 if traversed by an electric current, and suspended in a horizontal position, will, if free to turn, come to rest pointing in a north and south direction. The polarity of a solenoid is dependent upon the direction in which a current of electricity flows through it, and upon the direction in which the conducting wire is wound in forming the coil. The presence of an iron core very greatly increases the number of lines of force passing through the coil from end to end, the amount of increase being dependent upon the permea- bility of the substance forming the core. Obviously, when no core is inserted 98 AMERICAN TELEGRAPH PRACTICE in the coil there is a considerable amount of leakage of the lines of force out through the sides of the coil as indicated in Fig. 76, the total number extend- ing all the way through being small compared with the number of lines carried to the polar extremities due to the concentrating properties of the iron core. In fact, the presence of a core not only reduces the leakage of lines of force but adds materially to those already existing. (n As was stated on page 9, permeability (/*) equals ^ where B represents the magnetic induction in lines of force per unit area of cross-section, ^represents magnetizing force. Permeability. Permeability might be referred to as that characteristic susceptible of expression through a numerical coefficient representing the ratio between the number of lines of force formed in a space containing air only, as in Fig. 76, and the number of lines formed in a space filled with a FIG. 77. The electromagnet. given quality of iron as in Fig. 77. This ratio varies somewhat with different grades of iron and steel. Magnetic resistance, or reluctance (^), is less, the higher the coefficient of permeability, and, naturally the higher the permeability of a substance, the better it is suited for the purposes of electromagnet cores. On page 25 the relative merits of various grades of iron and steel for field magnet core purposes were given in the order of their permeability; to these might be added silicon-steel, as the latter has been found to possess a high permeability and is being used to some extent in the manufacture of magnet cores. The number of magnetic lines of force that can be forced through a core of given cross-section, while in great measure dependent upon the permea- bility of the substance of which the core is made, also has to do with the degree of magnetic saturation attainable with a given core material with given ex- citation, or, in other words, increasing the excitation beyond a certain point does not always increase the number of line? of force. In each case a point is reached beyond which there will be no increase of lines. A specimen of iron when subjected to a magnetizing force (B) capable of producing, in air, 52 magnetic lines to the square centimeter, was found to contain about 17,000 lines per square centimeter. By means of the for- (T> mula /*= ~' the permeability in this instance would be 326 times that of air. ELECTROMAGNETISM 99 Good grades of wrought iron will carry approximately 20,000 lines per square centimeter, and cast iron about 12,000 lines. Figures considerably higher than these have been obtained where extraordinary magnetizing forces have been employed, but correctly plotted magnetization curves show that there are pronounced evidences of saturation when the values reach those above stated. It will be remembered that the value of B is given in gausses, the definition of the unit being stated on page 8. The following table of values of 5H and <5# from samples of first grade American wrought iron, were determined by Dr. Sheldon, and the magnetic permeability in each case may be ascertained by means of the above formula: 7< & (Gausses) 10 13,00 20 14,70 30 15,300 40 15,700 50 16,000 60 16,300 70 , 16,500 80 16,700 90 16,900 100 17,200 150 18,000 2OO l8,7OO 250 19,200 300 > 19,700 The magnetomotive force or magnetization of an electromagnet is pro- portional to the number of turns of wire wound around the core, and the current in amperes flowing through the coil. A unit pole will have 4X3.1416 lines of force proceeding from it, or to reduce to c.g.s., units 47T =1.25764, which is generally taken at the value of 1.257. Where TV = the number of turns in the coil, / = the current in amperes. Magnetomotive force & (Gilberts) = 1.257X^X1. A field of ^ units refers to one where there are 3( dynes on unit pole, and it is customary to follow the rule of drawing a number of lines of force to the square centimeter of cross-section of the core equal to the number of dynes of force on the unit pole. Unit of Work. The unit of work, the erg, refers to the amount of work done when a force of i dyne is overcome through a distance of i cm., or, in other words, the amount of work done in moving a body through a distance of i cm. against a force of i dyne. 100 AMERICAN TELEGRAPH PRACTICE A unit magnetic pole has as many lines of force proceeding from it as there are square centimeters on the surface of a sphere of i cm. radius. A sphere having a radius of 2 cm., obviously has a diameter of 4 cm., and an area of D 2 X 3. 141 6. A sphere having a radius of i cm. has a surface area of 2 2 X3.i4 I 6 = i2.5664 sq. cm., therefore, unit magnet pole of unit strength has 12.5664 magnetic lines of force. As i sq. in. is equal to 6.452 sq. em., it follows that when unit density of magnetism concerns a density of i magnetic line of force per square centi- meter, we have an equivalent value of 6.452 lines per square inch as unit density of magnetism. In a magnetic circuit, magnetomotive force corre- sponds to electromotive force in an electric circuit. Reluctance, or mag- netic resistance in a magnetic circuit corresponds to ohmic resistance in an electric circuit, or magnetomotive force Flux = reluctance Magnetic flux (The Maxwell, Symbol 0) is, therefore, dependent upon magnetic reluctance (The oersted, symbol &} the latter being a property which tends to oppose the passage of magnetic flux. As previously stated, however, the magnetomotive force may be measured in terms of ampere- turns (the ampere- turn is equal to 1.26 Gilberts). Hysteresis. When the iron or steel core of an electromagnet has been magnetized by a current of electricity flowing through the magnet winding, and the exciting current is discontinued, the core will be found to retain more or less magnetism. The magnetism remaining is termed residual, and its value is spoken of as remanence. A completely closed magnetic circuit, that is, one where the (( keeper" or ar- mature is in mechanical contact with the pole-faces of the magnet, will show much greater remanence than one having an air gap between the armature and the polar extremities of the magnet. When an electromagnet is permitted to draw its armature into actual contact with its pole-faces, the armature will not fall back instantly when the exciting current is discontinued, due to the fact that it is held fast by the residual magnetism of the cores. It is for the purpose of avoiding this that electromagnets are generally so arranged with regard to the moving element the armature that mechanical contact between the armature and the pole-faces does not occur in practice. It is common practice to set a small brass pin in the face of each magnet pole, the pin projecting about one-sixteenth of an inch, or a distance suf- ficient to prevent the armature from coming into actual contact with the pole-faces proper. The brass pin is really a safety stop, as receiving instruments used in telegraphy, which consist mainly of an electromagnet and an armature are so designed that the armature "plays" between a front and a back ''contact, " TIME-CONSTANT 101 both adjustable, and which limit the travel of the armature within any desired space . Magnetic materials manifest a tendency to resist any change either increase or decrease in their magnetic condition, a characteristic which might be regarded as a sort of magnetic inertia, and which is known as hysteresis. An effect of hysteresis is to cause a delay, or "lag" in the magnetization of the core, behind the energizing current traversing the winding of the magnet. When a circuit including an electromagnet is closed, the relation between the exciting current and the magnetic flux produced will be such that maximum magnetization will lag somewhat behind the maximum elec- trical excitation of the winding. The result being that there will be a time interval of a fraction of a second between the time the electrical circuit is closed, and the time maximum magnetic flux is produced. A part only of this delay is chargeable to hysteresis, for, as was pointed out in connection with the effects of self-induction, the latter phenomenon is directly responsible for the delay observed in the increase and decrease of electric current in the coil winding, which in turn delays the increase and decrease of magnetization of the core. Time -constant. When a constant e.m.f. is impressed on a circuit includ- ing a magnet coil possessing inductance, the current flowing in the circuit does not reach its full value instantly, as at first it is opposed by the counter electromotive force due to inductance. The counter e.m.f. gradually becomes less as the current value reaches full strength. In practice it is usual to regard the operating requirements as such that a value of 63.2 per cent, of the full strength of the current should obtain in an efficiently operative circuit. The period required to attain this value is called the time-constant of the circuit. ,. , , inductance (in henries) Time-constant (in seconds) = r-r 7-. r \ resistance (in ohms) or _ he nriesX final amperes applied volts As usually submitted the first formula is given as Time-constant = ^ * K Suppose for example that it is desired to determine the time-constant of a relay having a resistance of 300 ohms, and an inductance of 2.65 henries. For L we have a value of .265 and for R a value of 300, and 2.65 = 0.000, 300 or a time-constant of 0.009 second. 102 AMERICAN TELEGRAPH PRACTICE Assume a circuit including an electromagnet to have a resistance of 600 ohms and an inductance of 6 henries, then with an e.m.f. of 40 volts applied to the circuit the time-constant would be T =0.01 (second). 600 By means of the formula I = ^r>> the final current strength in the circuit would be 40 600 = 0.066 ampere, therefore in o.oi second the current strength in the circuit will have reached a value of 0.041 ampere, or 0.632 of its full strength. The same result could have been obtained by means of the second formula, or Tf Time-constant L- E 6X40 = ^40 = 0.01 second. 600 The several electrical and mechanical actions which govern the forward and backward movement of the armature of a telegraph relay, thus are involved with the element of time, even if but a fraction of a second is consumed with each transit. The time-constant, therefore, of the circuit refers to the time in seconds, or fractions thereof, which it takes the current strength in the circuit to build up to a value approximately two-thirds that of its final strength. On the other hand, after the armature has been attracted toward the pole-faces of the magnet, and the circuit again opened or discontinued, it requires an appreciable time for the magnet to "let go" the armature or to release it. Careful experiments have shown that after the magnet circuit has been opened, the average time required for a magnet to release its armature varies from 0.003 seconds with maximum retractile spring tension to 0.033 with minimum retractile tension. Average opera- ting adjustments obviously give "releasing" values about midway between these figures. In determining the time-constant of a circuit which includes electro- magnets by either of the above formulae, it is required that the inductance be known. In cases where the impressed electromotive force varies according TIME-CONSTANT 103 to the Sine law of alternating currents and the inductance (L) is constant, the effective value of the inductive counter e.m.f. is E = 27T fLI, where / represents " frequency" or cycles per second, 7 the effective value of the current, E would then be the inductive reactance, or the inductance of the circuit. Another method, and one more applicable in approximating the induc- tance of magnets used for telegraphic purposes, is that known as the " standard condenser" method. By means of a Wheatstone bridge and an adjustable condenser arranged as shown in Fig. 78, the inductance of a magnet, a pair of magnets, or an " instrument" may be determined quite accurately. The four arms of the bridge, namely, a, b, x, and R, are shown in their respective positions. G is a galvanometer, r an adjustable resistance, and C an adjustable condenser. 1 The " instrument" to be measured for inductance is inserted at X, then, after the bridge has been balanced; that is, after R has been ad- justed to equal the resistance of X (a condition which is indicated when the galvanometer pointer remains in the center of the scale and unaffected when the keys K and KI are closed) if tjie key K 2 is now closed, it will be found that when the keys K and K\ are closed and opened at short intervals, the galvan- ometer pointer will be "kicked" to one side or the other due to the counter e.m.f. of induction from the magnets located in the X arm of the bridge. The counter current thus produced is obviously of but momentary duration and results in the galvanometer pointer being "kicked" to the right or to the left (depending upon the direction of the flow of current through the galvanometer), the pointer im- mediately returning to "center." All that is required to determine the inductance of the coil X is to adjust r and C until there is no "kick" when the keys K and Ki are closed. Then, if the arms a and b each FIG. 78. Condenser have 100 ohms resistance inserted as shown in Fig. 78, method of measuring the inductance may be determined by means of the formula ioo)+RXioo], where h is the inductance in henries. 1 1 The resistance units which make up the values in arms a, b, and R are practically non-inductive, as the resistance wire wound on the "bobbins" making up the various units is "doubled back" on itself, so that the lines of force produced in one-half of each coil are "neutralized" by those produced in the other half, thus nullifying the inductive action of the coil as a whole. 104 AMERICAN TELEGRAPH PRACTICE PRACTICAL ELECTROMAGNET DATA Self-induction is proportional to the square of the number of turns of wire in an electromagnet, everything else being equal. Connecting the windings of a pair of magnets in "parallel" reduces the time-constant one quarter. From the formula L Time-constant = ^ J\. it is obvious that the time-constant of a circuit, including electro- magnets, may be reduced by reducing the self-induction or by increasing the resistance. With the armature mounted so that a distance of o.oio in. or more separates it from the pole-faces of the magnets, maximum pull is obtained when "flat" pole-faces are employed. When the distance separating armature and pole-faces is less than o.oio in., pointed or concave poles are more effective. The efficiency of a magnet is independent of the resistance of the winding. It is immaterial whether a thick or a thin 79 conducting wire is used, provided the thickness of the wire is sufficient to carry the required 5 1 O ^| ^i *~\ N a current, and that the same number of watts -* ^ ~^ are spent in heating it. Heat waste in a I OOP p\ N magnet coil is proportional to the square of <_ J J 837 ohms. The number of cells of gravity battery required to furnish 50 milliamperes of current through a resistance of 1,837 ohms may be ascertained by means of the formula given on page 75. In this instance the values of the various factors are ^ = 1,837 ohms, = 1.07 volts ^ = 2.5 ohms, 7 = 0.050 ampere n and as the number of cells (N) =^ L r r then ^1 -- i cells. 0.050 The reason that the law does not correctly apply for this purpose is that each cell of battery has a resistance of 2 1/2 ohms, and the 97 cells place an additional 243 1/2 ohms resistance in the circuit. However, after the value of E for the entire battery has been determined by the above method, the simpler formula Tf I=X may be employed to check the result. The example here considered, where the resistance per mile of the conductor has been multiplied by the number of miles, assumes perfect insulation of the line. The 97 cells of battery required to furnish 50 milliamperes current would not necessarily have to be located at one end of the line, but might be dis- tributed, part at one end and part at the other, in which latter case opposite battery " poles" should be placed to line at each terminal station positive at one end and negative at the other. If the source of e.m.f. availed of to furnish current to operate the line Tf above considered were a dynamo, then the formula I=n would serve the purpose, as the internal resistance of the dynamo is so low as to be negligible. SINGLE MORSE CIRCUITS 113 Suppose two terminal stations situated 500 miles apart are connected by a No. 9 copper wire, and that there are two intermediate stations in the circuit, each station, including the two terminals, being equipped with re- ceiving instruments having 150 ohms resistance (Fig. 84). If we assume the line to be perfectly insulated and it is desired to ascertain the voltage necessary to maintain 45 miliamperes current in the circuit by means of the formula E = IXR } the required potential may be arrived at thus: FIG. 84. = 0.045 ampere 2,795 ohms, and 2,795X0.045 =125.775, or 126 volts. In this example we have not allowed for the insertion of any additional resistance in applying the e.m.f. to the circuit. This imposes the require- ment that the source of e.m.f. must have very little or no internal resist- ance if 45 milliamperes current is to be maintained in the external circuit. As previously stated, several lines may be worked out of one battery, and in practice it is found that when the internal resistance of the source of i =T. 2 = -3 T 3 ij FIG. 85. Several lines fed from a common battery. e.m.f. is infinitely small in comparison with that of the several lines connected thereto, the strength of the current in each circuit will be practically the same as if it were the only line attached to the battery. In view of this when a number of lines are "fed" from a common battery it is immaterial whether but a single line is operated, or whether several lines are operated simultaneously. By reviewing the calculations in connection with Figs. 64 and 640 on page 85, dealing with joint-resistance, the student will be better able to under- stand the explanation of this seeming inconsistency in the behavior of electric currents. 8 114 AMERICAN TELEGRAPH PRACTICE When a single line is being fed from a battery, and a second and a third line are connected to the same source as in Fig. 85, the total amount of current flowing from the battery divides along three paths and when all three circuits are closed, that is to say; taking current, the aggregate sec- tional area of the conductor is so increased that the total resistance which limits the volume of current flowing from the battery is greatly reduced, and as the strength of current taken from the battery increases in the same proportion, the loss which would result from the division of the current into three separate branches is compensated for by the increased current strength due to the reduction in resistance of the circuit as a whole. The maximum current efficiency may be derived from a battery when the total internal resistance of the battery equals the resistance of the external circuit. Of course, actual conditions are such that it is not convenient, or for that matter necessary to maintain this balance of resistance between the external and internal portions of the circuit, but it is due to this that the gravity cell, with its comparatively high internal resistance per cell (2 1/2 ohms), has met the requirements so satisfactorily, and that this type of battery has been so extensively employed in the operation of telegraph lines. The longer the line, or rather the greater the resistance of the line, the better does this type of cell answer the purpose, but this is true only when each separate line has its own battery. When the battery is required to feed several lines, the internal resistance becomes an important factor. By considering the conditions which prevail in a case such as that depicted in Fig. 85 it is evident that the internal resistance of the battery will remain constant while the resistance of the circuit beyond the point X will vary considerably, depending upon the number of branches which are closed at one time. As the battery resistance then becomes an important part of the total resistance of the circuit, the former should be kept as low as is practicable, for, no matter whether one or more lines are being operated at the same time, each line should have equal current strength. When a battery composed of gravity cells is employed to feed several lines, the current volume in each separate circuit varies according to the number of circuits which are closed at the same time. Suppose, for instance, that five separate lines each having a total resistance of 1,200 ohms are fed from a gravity battery having a total e.m.f. of 100 volts and a total internal resistance of 200 ohms, then with one circuit "closed" and the other four 'open, the current value in the closed conductor would be E 1 = r+R 100 = 71 m.a. 200+1,200 SINGLE MORSE CIRCUITS 115 With all five circuits closed, the current value obtaining in each circuit may be determined by means of the formula T E 1 = - $-# TOO = - -5-5 = 45 m. a. , 1,200 200 H - Or, as the five circuits are operated simultaneously or intermittently the volume of current in each conductor will fluctuate between 45 milliamperes and 71 milliamperes, depending upon the number of circuits which are closed at one time. In this particular instance the discrepancy between maximum and minimum current values in any one circuit may not be great enough to be' regarded as unsatisfactory, but it should be noted that the conditions are such that the minimum current value is still high enough to operate the usual type of receiving instrument under favorable conditions, also the resistance of each of the five lines is identical, and further; but five lines are being fed from the battery. Obviously, if the number of lines were increased, or if the individual resistances of the various lines should be unequal, the unsuitability of primary batteries for the purpose of supplying current to many lines would be more apparent. If, for instance, the number of lines in the above example should be increased to six, then (other conditions remaining the same) the minimum current would be 25 milliamperes, a value too low for single-circuit operation. Where secondary-cells or dynamo machines are employed to furnish current to operate telegraph lines, the negligible internal resitsance of these sources of e.m.f. fulfills the condition previously referred to "where the internal resistance of the source of e.m.f. is infinitely small in comparison with that of the several lines connected thereto," and if the five lines con- sidered in the above example were supplied with current from a dynamo having an e.m.f. of 100 volts, the current strength in the closed conductor (the other four remaining open) would equal 100 = 83 m.a. 1,200 and with all five circuits closed, 100 1200 ^-5 = 83 m.a. 5 If instead of five circuits, ten are connected to the same source of e.m.f. the current value in one closed circuit would be I0 O 83 m.a. 1200 116 AMERICAN TELEGRAPH PRACTICE and with all ten circuits closed simultaneously, the current volume in each circuit would be = 8 m. a. 1200 10 Thus, when several lines are fed from a dynamo, the current values obtaining in each circuit are constant, and independent of the closing or opening of other circuits fed from the same source. The foregoing examples assume identical resistance values for the various circuits fed from a common source of e.m.f. In those applications where the individual resistances of the various circuits fed from a common battery are not "evened up," it must be remembered that the current in a circuit varies directly as the electromotive force, and inversely as the resistance of the circuit. Suppose, for example, that a loo-volt dynamo is employed to feed three circuits, the first having 1,000 ohms, the second 2,000 ohms, and the third 3,000 ohms resistance. By means of the formula given on page 84 for calculating joint-resistance, the joint-resistance of these three circuits would be 545 ohms. The total current strength in the joint circuit would equal 100 = 183 m.a. 545 and the portion of the total current traversing each branch may be ascertained thus: T T> T I0 In AI, I = loom. a. 1000 T T> T I0 In 7? 2 , / = = 50 m.a. 2OOO , ~ 100 183 m.a. MORSE SINGLE-LINE INSTRUMENTS Figure 86 gives a view of a Morse key equipped with extension "legs" to be used in fastening the key on top of the operating table. This type of sending key, which is known as the "Bunnell steel lever key," is at the present time quite generally employed in commercial and railroad telegraphy. Its construction combines lightness of moving parts, durability, and ease of adjustment. Figure 87 illustrates another form of the same type of key, designed to be fastened to the operating table by means of ordinary screw nails. In- SINGLE MORSE CIRCUITS 117 FIG. 86. Bunnell key, "leg" type. FIG. 87. Local Circuit Main Line 1 Main Battery FIG. 88. Morse relay, skeleton connections. 118 AMERICAN TELEGRAPH PRACTICE stead of the connecting wires being attached to "legs" as in the form of key illustrated in Fig. 86, two binding-posts mounted on the base of the key serve to hold the wires which connect the key into the circuit. Figure 88 shows the binding-post and internal connections, both main line and local, of a main-line "relay" of the usual type. For the sake of FIG. 89. Morse relay. k FIG. 90. Morse sounder. clearness, a few turns only of magnet wire are shown wound around the core of each spool. The way in which the movable armature tongue when attracted by the magnets connected in the main-line circuit closes the local circuit, thus operating the reading sounder, may easily be traced in the drawing. Figure 89 shows a relay completely assembled, and Fig. 90 a view of a type of sounder extensively employed in this country. CHAPTER VIII LIGHTNING AND LIGHTNING ARRESTERS FUSES GROUND CONNECTIONS For many years it was believed that lightning was simply an alternating current of very high frequency. During the past 20 years, however, a large amount of research work has been carried on with the object of learning something definite and conclusive in regard to the nature of lightning dis- charges, and it is now known that lightning may be manifested in several different ways. A lightning disturbance may occur as: A single discharge of very high potential. As an alternating current of comparatively low frequency, having greater inductive than static effects. As an alternating current of high frequency, having large capacity and high self-induction. When lightning strikes a telegraph line, a part of the line may be de- stroyed due to the charge reaching the ground by way of the poles, the latter being split and torn as if from internal explosion. When lines are provided with ground wires attached to poles and fixed close to the conducting wires, and with lightning arresters at terminals, the charge is divided, reaching the earth at as many points as are presented in the form of discharge-gaps. But even if the charge has. been quickly drained off, its presence upon the conductor even for a brief interval of time affects the electric circuit so that a disturbance more or less pro- nounced is the result. The single discharge of high potential, or the direct stroke as it is some- times called, although rare in comparison with the number of disturbances due to electrostatic induction, is more disastrous to property. The imme- diate, or local effects of the direct flash include the shattering of glass or other insulators, splintering of crossarms and poles; incidental to the passage of the discharge to ground, as referred to above. The damage may be con- fined to a short section of the line, sometimes two or three pole lengths, or may extend over a distance of a mile. In most cases the severity of the damage decreases with the distance from the point at which the discharge takes place. The fact that beyond a comparatively short distance from the center of shock there is no visible damage to the line does not mean that the discharge has been completely dissipated, but rather that the current induced 119 120 AMERICAN TELEGRAPH PRACTICE in the conducting wire has after traveling an indefinite distance along the conductor become attenuated to an extent that robs the charge of its power to do further damage of the nature cited above. In the conductor, however, there has been started a current wave progressing outward which causes a surging likely to produce indirect disturbances at distant points in the circuit. Analogous to the way in which a river may rise until dams and embank- ments give way, an induced charge of electricity mav accumulate in a circuit as a result of rain, snow, fog, or clouds of dust being driven across the line, until the difference of potential between line and ground assumes enormous proportions, and at the breaking-point discharges across lightning arresters attached to the line, and, seeking paths of least resistance, discharges to ground through the intervening dielectric. If a positively charged cloud passes over a line, an electrostatic charge may be induced in which the earth below the line assumes a negative electro- static charge. The line itself, due to its more elevated position, also takes on a negative charge, somewhat higher in potential than that of the earth, Of course, the sign of the charge on the conductor depends greatly upon the degree of insulation maintained between the line and the ground. As a positively charged cloud approaches a perfectly insulated line the latter may assume a positive charge at cloud potential, and as the potential rises with the approach of the cloud, the potential difference between line and earth may rise to a point where discharge takes place between the earth and the line. As the cloud recedes from the line, the latter then remains nega- tively charged and, inasmuch as this charge is no longer bound by the posi- tive charge of the cloud, a discharge takes place from line to earth. . When an electrostatic charge affects a line, there is a strain of contending forces potentials at opposite polarities; naturally disruption takes place when these forces meet. The enormous strain manifested is not confined to the conduct- ing wire or wires, but embraces all neighboring conductors or semiconductors, so much so that in certain instances persons standing 25 or more feet away from where the chief damage has been wrought have been severely shocked. Undoubtedly the most frequent manifestations of atmospheric electricity in line conductors are the result of electrostatic induction from passing clouds. Each readjustment between cloud and ground or between cloud and cloud in the neighborhood of a conducting circuit brings about an abrupt alteration in the electrostatic charge on that part of the line immediately in the vicinity of the disturbance. The induced impulses in the circuit increase in strength and frequency as the cloud approaches the line and decrease as the cloud recedes. The popular scientific conception of the conditions which exist, in the atmosphere when an oscillatory discharge takes place assumes that the air, the cloud and the earth, in effect, constitute a huge condenser with the air as the dielectric. It is true, of course, that the dielectric in this case is con- LIGHTNING AND LIGHTNING ARRESTERS 121 FIG. 91. "Saw-tooth" lightning arrester. stantly varying in density, purity, and humidity, and this inconstancy of the insulating medium; in a measure accounts for the variegated effects observed. When the air breaks down under the strain and becomes heated to incandes- cence, the phenomenon observed is called lightning. Many years ago it was discovered that a lightning discharge traveling through a conducting wire has, so to speak, an aversion to turning corners, insisting to its utmost upon traveling in a straight path. The excessive heating effects of these induced charges often deflagrate telegraph wires at points where the conductor has been injured mechanically (thus reducing its cross-section) or where the wire is "kinked" or bent. In the design of modern lightning arresters advantage has been taken of the fact that " kinks" or turns of wire serve to " choke" the induced oscilla- tory currents. Thus in several forms of arresters employed to protect aerial lines a choke-coil forms an important element of the arrester. The design of satisfactory protectors should provide against undue prolongation of abnormal currents in the conductor. This is accomplished by means of a properly designed "fuse." Also an air-gap or high-resistance path to earth should be pro- vided for high-potential discharges. This may consist of two metal plates, one connected with the earth and the other with the line wire, one plate being provided with pointed "saw-teeth" as illustrated in Fig. 91. This lightning arrester is seldom seen except in the older installations. Where it is required to guard against high potentials, it is customary to employ an arrester which offers, for large currents, a path to earth having a lower break-down point than is offered by the insulation of the circuit pro- tected. Most protectors designed with this end in view consist of two con- ducting surfaces, one of which is connected to the line conductor and the other to the earth. The two sides of the arrester may be separated by a gap, either in open air or in a vacuum, or they may be separated by a high- resistance material such, for instance, as carborundum. The sensitiveness of a lightning arrester depends considerably upon the width of space separating the metallic or conducting elements of the arrester, and, although in practice arresters are employed having spacings of 0.005 to o.oio, and as high as o.ioo in., depending upon locality and character of protection, it is important that accurate spacing be maintained. Figure 92 shows one form of arrester consisting of spring clips 5 and 122 AMERICAN TELEGRAPH PRACTICE FIG. 92. Carbon-block arrester. ,Line fr Apparatus FIG. 93. Vacuum-gap arrester. FIG. 94. LIGHTNING AND LIGHTNING ARRESTERS 123 -?- - -m Sij carbon blocks C and Ci, and separator M, the latter being made up of strips of mica to the desired thickness. The mica separator is perforated in several places, thus making as many air-gaps between the two carbon blocks, the latter being in contact with the spring clips which in turn are connected to line and ground respectively. In this form of arrester it is of the greatest importance that a high grade of carbon be used in making the blocks, as the poorer grades are liable to "chip" or to oxidize and form carbon dust, and thus interfere with the correct spacing of the blocks. Other forms of this type of arrester which have recently been introduced are the "vacuum gap" and the "Brach." The former is shown diagrammatically in P'ig. 93. In this make of air-gap arrester the dis- charge takes place in the form of a "brush" between two carbon plates separated by a par- tial vacuum. It is well known that an electrical m discharge will take place between two conduct- ing surfaces at a lower potential in a vacuum than in air at ordinary pressures. Thus, a greater separation of plates may be maintained when the discharge takes place in a vacuum. The opposing surfaces of the carbon blocks as in the original metal-plate arrester are ser- rated, and there is no carbon dust produced which would form a deposit likely to reduce the insulation existing between the terminals of the JT IG- 95 . Combination saw- arrester, tooth and carborundum-block Figure 94 gives a photographic view of the arrester, vacuum arrester. In the Brach arrester a direct contact path from line to earth is provided through a high-resistance block which separates the metallic surfaces of the arrester. See Fig. 95. The cut shows an arrester equipped with fuses and an auxiliary air-gap of the older form. The departure from the air-gap principle embodied in this arrester is illustrated at the lower extremity of the arrester elements and between the fuses, where M represents metallic plates, C, carbon plates, and R, blocks of a high-resistance compound. The cut shows a two-line unit. The separator blocks used in this arrester have a resistance of about 4 megohms when subjected to a pressure of 240 volts, the resistance decreasing rapidly as the potential is increased. Where arresters of the direct-contact type are distributed at intervals along a line, it is evident that the total insulation of the line will be somewhat 124 AMERICAN TELEGRAPH PRACTICE reduced, but where a high degree of insulation between line and earth is not essential, the "static" draining possibilities of this type of arrester may be of considerable advantage. As a protection against oscillatory currents of high frequency and large self-induction, a "choke" coil may be included as an element of the arrester. A length of 2 or 3 ft. of insulated wire wound into a coil J or J in. in diameter, when inserted in the line constitutes an effective barrier to the passage of high-frequency alternating currents. A well-known form of arrester which embodies the principles above referred to is that known as the Argus. A well-designed form of choke coil as employed in guarding against lightning discharges is shown in Fig. 96, in which the conductor is carried through a spirally turned pipe of small diameter. The section of the FIG. 96. Lightning choke-coil. conducting wire enclosed in the iron-pipe spiral is insulated, thus are combined two forms of impedance. The discharge-rod shown traversing the entire coil acts to carry off the static charge held back by he choke coil. Lightning arresters connected with line wires; practically are condensers of small capacity, and in proportion to this capacity present conducting paths to earth for alternating currents. After each lightning storm it is well to inspect all open-type carbon block arresters and to clean away any deposit of carbon dust which may have accumulated on the faces of the blocks or on the mountings. LOCATION OF LIGHTNING ARRESTERS Undoubtedly, the most desirable location for lightning protectors is outside of buildings, but owing to the close regulation practised and to the fact that the instruments and apparatus protected must be safeguarded from all high-tension currents extraneous to the buildings, it is customary to locate arresters inside the building. Outside or "pole" arresters in various forms are used as additional safeguards, and it is good evidence of the efficiency of these external pro- tectors that the number of instances are few where lightning and contact with high-tension circuits result in fire damage to buildings. Figure 97 shows one method of attaching a lightning ground-wire to a LIGHTNING AND LIGHTNING ARRESTERS 125 pole. With this arrangement a double-grooved insulator is required. The wire in contact with the ground is fastened along the length of the pole by means of staples, and at its upper extremity is twisted around the upper groove as shown on the right, the end of the wire being bent so as to form a .Ground Wire hook with which to clasp the bottom of the insulator. As shown on the left, a space, or air-gap, which may be regulated to suit the requirements, separates the tie-wire from the ground-rod. Compound - D- Screw Contact \;-Bfock - Separator Block Contact D- Screw W.FComp. FIG. 98. The "Brach" pole arrester. A cross-section view of the Brach arrester adapted to out-door service is illustrated in Fig. 98. The manner in which this arrester is attached to the line wire is illustrated in the reproduction, Fig. 99. When new lines are constructed, one telegraph company requires that : 126 AMERICAN TELEGRAPH PRACTICE "About 10 ft. of line wire be formed into a flat coil, and placed under the butt of the pole. The other end of the wire must be stretched up the pole and fastened thereto by twelve or more wire staples. It will be extended 7 in. above the top of the pole, and the end of the wire will then be turned back and fastened to the pole, making a projection above top of the pole 3 in. in length and doubled back, the said projection to be given three turns or twists." Practice in regard to the spacing of lightning ground-wires along a line varies somewhat. One company requires that a ground wire be attached to every fifth pole, while another company requires that a wire be attached to every sixth pole on leads carrying from i to 12 wires, 35 poles per mile, and on lines carrying 12 wires or upward, with more than 35 poles per mile, the ground-wire must be attached to every tenth pole. FIG. 99. Fuses. Protection of apparatus against abnormal currents, or currents of excessive strength, is usually accomplished by the employment of properly designed fuses. In determinating the capacity of a fuse to be used in a given case, the principal points to be considered are: 1. The amount of current the fuse must carry continuously under nor- mal working conditions. 2. The amount of current which the wiring or windings of the apparatus can safely carry during a certain period without undue heating. 3. The possible sources of trouble from foreign circuits carrying high potentials. When these requirements have been determined with reasonable accuracy, the carrying capacity of the fuse may be decided upon. In every conductor FUSES 127 there is a point above which the temperature must not be allowed to rise, and the customary method of protecting against excessive temperature is to employ a fuse which has been designed to "melt" when that point has been reached. Ordinarily, fuses consist of short lengths of wire composed of an alloy of lead and tin. The wire employed for the purpose may be of any desired diameter and length, its dimensions depending upon the degree of heat required to melt it when excessive currents flow through the conductor of which the fuse forms a part, during a given period of time. The capacity of fuses used in telegraph circuits ranges from 1/2 to 10 amperes, with intermediate steps of i ampere, 2 amperes, and so on. The half-ampere fuse generally employed will "blow" within two or three seconds after being subjected to a current of i ampere at 75 F. Owing to variations in temperature in different parts of the country, be- tween winter and summer seasons, it has not been found practicable to adjust the blowing point of 1/2- FlG I00 ._ Enclosed fuse . ampere fuses much closer than that indicated above. To adopt fuses of greater capacity than those named, for telegraph circuits, would place such circuits in the category of electric-light wires, which would be manifestly unreasonable, as the regular operating currents in telegraph circuits are infinitesimal when compared with the large currents carried in lighting circuits. In the construction of fuses, several different types of fuse-link are used. These might be classified as "straight-wire link," "air-drum link," "flat link," "multiple-link," "cylinder link," etc. FIG. 1 01. Air-drum fuse links. The "enclosed" fuse, such as that illustrated in Fig. 100, consists of a pasteboard tube, containing a non-combustible filling, in the center of which is stretched the fuse-link, or wire, each terminal being securely connected with brass or copper ferrules affixed to the end of the tube. The straight- wire link consists simply of a short length of fuse-wire of uniform diameter. In the construction of the air-drum link (Fig. 101) advantage has been 128 AMERICAN TELEGRAPH PRACTICE FIG. 102. Mica enclosed fuse. taken of the fact that the blowing time of a fuse may be rendered practically constant for any predetermined overload, regardless of the temperature of the filling, by enclosing a section of the fuse-wire in an air-tight casing. In the simpler form of fuse the porous filling completely envelops the wire throughout its length, and it has been found that the blowing time varies considerably due to the fact that the material of which the filling consists dissipates the heat generated in the fuse-wire, which to an appre- ciable extent, makes the blowing time of the fuse dependent upon the temperature of the filling. The air-drum link is more regular in action, owing to the fact that the air space around a portion of the fuse metal permits of 'a more definite relation between the temperature of the fuse-wire and the current value in the circuit. The other types of fuse-link mentfoned are modifications of the two described. The filling used in packing the fusable element must be non-combustible, and preferably should be non-absorptive of moisture, chemically inert, porous, and have no tendency to solidify. Figure 102 shows a form of "fuse" wherein a short length of fuse metal is enclosed between two strips of mica, the fuse element being stretched between two flat copper terminals which may be inserted between spring clips, or held fast by cross screws extending through the "slot" ends. Figure 103 shows a convenient method of mounting a number of fuse units such as that illustrated in Fig. 102. When mounted in a box as shown, one terminal of each fuse is connected to a metal strip, which in turn may be con- nected with a battery or other source of e.m.f. while the other terminal of each fuse may be connected to any circuit required to be fed from that particular battery. Ground-wires, or Earths. When a material such as dry wood, fiber, or glass is filled with earth, and the por- tion of earth thus isolated used as a section of an electric circuit, it is found that the resistance of the earth follows the same laws as that of any other substance, or the ohmic resistance of the isolated section of earth depends upon the character of the earth employed, the amount of moisture it con- tains, and upon its length and cross-section. When two metal plates are buried in the earth, the resistance of that portion of the earth extending between them does not vary in the ratio of their distance apart as it does in the case of a portion of earth enclosed in an FIG. 103. Box for mounting mica fuses. box constructed of insulating GROUND WIRES, OR "EARTHS" 129 isolated box. The resistance between two separated ground plates is depend- ent upon the character of the soil in the immediate neighborhood of each plate, upon the depth to which the plates are buried in the earth, and upon the size of plate used. On account of the large surfaces exposed to the earth, water-pipes, and gas-mains make excellent "earth" connections. Where such pipes are not available, satisfactory ground connection can be had in moist earth or in a river which does not flow a long distance in a channel of rock. A sheet of zinc, or tinned copper, about tV in. thick and about 4 ft. square should be buried in a hole or trench, made deep enough to reach below dry sand or earth, and of rock. The bottom of the trench, which must be where the earth is always moist, should have a layer of coke about 2 ft. deep on which the metal plate is to rest, and above the plate should be deposited a layer of crushed coke about 2 ft. thick, after which the trench should be filled up with moist earth. Connection with the earth plate should consist of a hard- drawn copper wire of a size not less than No. 9 B. & S. gage, the earth end being soldered entirely across the surface of the ground plate. When gas-pipes or water-pipes are used in place of buried earth plates, the connection should be made by wrapping a number of turns of the ground- wire around the pipe, thoroughly soldering the joint. Connections made to pipes should invariably be made on the "street" side of all service taps, to avoid as far as possible interruptions to the ground connection when changes or repairs are being made in the pipe systems. CHAPTER IX MAIN-LINE SWITCHBOARDS FOR TERMINAL OFFICES AND INTERMEDIATE OFFICES At an office where a " one- wire" line terminates, the only circuit access- ories required in addition to the signaling instruments are a lightning ar- rester, a line "fuse," and a "ground" connection. In order that the signal- ing relay may be "cut out," that is, disconnected from the line, during the absence of the attendant, or on any other occasion when such action might be desirable, it is usual to embody a "cut-out" feature in the lightning arrester and ground-switch unit. A simple form of this type of apparatus is illustrated in Fig. 91. This same device would answer all the require- ments of an "intermediate" office on a single-wire line. Where two or more line wires are cut into an intermediate office, or terminate at an office, then, in addition to the features above mentioned, a means must be provided whereby any two wires may be quickly cross-con- nected, or looped. Also a means should be provided whereby any one of the various line wires may be connected to any one of several sets of signal- ing instruments or to several sets of instruments at the same time. From an operating standpoint, the importance of a telegraph office is closely related to the extensiveness of the switching facilities necessary to carry on the work of the office, and of the "wire district" in which the office is situated. On account of the constantly changing conditions, it is rather difficult to classify telegraph offices in an order that would predetermine the appa- ratus required to equip any particular office. For general purposes, how- ever, it is possible to gain a helpful understanding of the requirements in a given case, where offices are classified as follows: Branch Offices, meaning, in a city, branches from the main office. Way Offices, small intermediate offices, cut in on one-, two-, or three-way wires, operated simple Morse. Intermediate Test Office. An office on a trunk line having all through wires cut in for testing purposes. Repeater Station. An office on a trunk line where signals are automatically repeated, from one section to another on some or all of the wires connected into the office. Terminal Station. Offices located in large centers, where a considerable volume of local telegraphic traffic is handled, where messages are relayed 130 MAIN-LINE SWITCHBOARDS 131 by hand, to other points, where automatic repeating facilities are available, and where "battery" is applied to main-line wires radiating therefrom. A more extended classification would mean the subdivision of each of these classes into several grades, and the governing factors would include UNS & LOO PS FUSES /OO M/LS M/DA SOUNDER SWITCHBOARD EQU/PMENT AND STRAP BOARDS FIG. 104. the amount of business handled, whether or not main-line wires take battery at the office, whether main-line testing is done from the office, etc. Where the design of, and the operation of the switching apparatus are concerned, it is, of course, quite desirable to employ standard apparatus. 132 AMERICAN TELEGRAPH PRACTICE The continual improvement being made in the design and construction of switchboard equipment means that standardization and improvement must go hand in hand, and the benefits are best secured when improvements are introduced gradually throughout the entire system. A type of main-line switchboard formerly known as the Universal, now generally referred to as the strap-and-disk board, has for many years been extensively employed at both intermediate and terminal stations. Figure 104 gives a diagrammatic view of the electrical connections between line wires and office instruments 'where a strap-and-disk switchboard is used. Two separate line wires are shown "looped" into the office, both J-, JL r"w rw JL ^ JLJL U-.JL :,w e w ^wew e, w e v \__/ *-*. *-* - 6- S/* ^ ( ^ H 6- s | \ f . 1 k ( \* \_-.y :< ^-H *i < \f ' \ f \ 1 ^ X/ 1 \f > ( ^ ' ^H \__y ^ "> ^ FIG. 105. FIG. 106. FIG. 107. FIG. 108. w 8 w w E w FIG. 109. FIG. no. FIG. in. FIGS. 105 to in. Strap-and-disk switchboard combinations. sides of each loop being connected through 2o-ampere fuses, and to a lightning arrester having a separation between line and ground plates, of one tenth of an inch. It is evident that the "board" shown in Fig. 104 has accommoda- tion for four through wires, that is, four line wires may be looped into an office having a board of this size. The vertical elements or " straps" are connected with either side of a line, and each pair of straps in use represent one external circuit connected into and out of the office, while the horizontal elements, the disks (which are connected together in horizontal rows by means of metallic strips on the back of the board) are by way of binding posts connected through half-ampere MAIN-LINE SWITCHBOARDS 133 fuses, and lightning arresters having a o.oi- in. gap between plates, to signaling relays mounted on the operating tables. Figures 105 to in inclusive show various combinations which may be made at an in- termediate office, with two through circuits, extending, say east and west of the office. Figure 105 shows the horizontal elements and the vertical elements of the switchboard, so connected by means of metallic "pegs" that each circuit is connected through the office, including in one circuit the signaling relay connected with binding posts A , and in the other circuit the relay connected with posts B. Figure 106 shows wire No. i west " cross- connected" with wire No. 2 east, and wire No. i east with wire No. 2 west; each circuit so made up includes the windings of the sig- naling instrument wired to the terminals A and B, respectively. If it is desired to eliminate the winding of the instrument connected with posts A from the circuit in which it is connected, all that is necessary is to place the two center pegs in the positions indicated in Fig. 107. Simi- larly, instrument B may be eliminated from the circuit as shown in Fig. 108, and both instruments may be cut out if the pegs are in- serted as shown in Fig. 109. A horizontal row of disks is assigned to the ground con- nection G, and any wire may be "grounded" simply by inserting a peg in the hole at the intersection of the vertical strap connected with the wire to be earthed. Figure no shows the disposition of the pegs when it is desired to "loop" wire No. i east with wire No. 2 east, allowing the instru- ment A to remain in circuit, or if it is not re- quired to have the home instrument cut in, FlG . II2 ._ Improved formof strap- the pegs should be inserted as in Fig. in. and-disk switchboard. Switchboards of this type may be built large enough to take care of any number of wires. It is evident, of course, O O 134 AMERICAN TELEGRAPH PRACTICE that as a board is enlarged to accommodate a large number of wires, its dimensions increase in both directions; that is, vertically and horizontally. One difficulty experienced with the strap-and-disk switchboard is that the pegs are liable to work loose, and result either in a poor contact between to AM P. Fuses. tOO/M/LS FIG. 113. Cross-bar main-line switchboard. strap and disk (thus introducing an abnormally high resistance into the circuit) or fall out entirely and interrupt the circuit. This difficulty is more often encountered in offices located in railroad depots where vibration caused by passing trains in time causes the pegs in the switchboard to work loose. MAIN-LINE SWITCHBOARDS 135 To avoid this annoyance an improved form of strap-and-disk board has recently been brought out (Fig. 112) the construction of which provides for a more positive union between peg, disk and strap. Instead of the tapered pegs usually employed, the improved board has a straight peg which goes through a hole drilled all the way through the slate or abestos board FIG. 114. Double-conductor plugs for use with cross-bar switchboard. base, and engages spring clips which are fastened to the backs of the disks and straps. A form of switchboard known as the " cross-bar" board, in use at many offices, is illustrated in Fig. 113. In principle this form of switchboard is identical with the more common strap-and-disk board. All of the combina- tions possible with the latter may be made with the cross-bar arrangement. The only noteworthy difference being that the home relay is connected into the desired circuit by means of a " double-plug, " which completes the circuit from horizontal strip through the double- conductor cord and back to the vertical strip. When it is not required to have the home relay in circuit, a solid plug is inserted in place of the double-plug. The diagram, Fig. 113, shows the fuse, lightning arrester, FlG ' , , , . , , . and ground connections, also the connections of the main-line and local instruments required in connection with this type of switchboard. It is evident, too, that for a given number of vertical straps, the cross-bar form of switchboard will accommodate twice as many lines connected through an intermediate office as will the strap-and-disk board, owing to the fact that the lines in one direction are connected to the vertical straps and the lines in the opposite direction to the horizontal straps, or bars. Figure 114 illustrates the form of double plug used with the cross-bar board to cut in^ set of instruments. The cord used in connection with this plug is a flexible double conductor, one conductor being connected with the "tip" of the plug, while the other is connected with the metal portion of the plug back of the hard-rubber insulating strip. ii5- Split plug for use with strap-and-disk switchboard. 136 AMERICAN TELEGRAPH PRACTICE The strap-and-disk board, also, may be connected so that the lines extend- ing in one direction will be attached to the binding-post terminals of the horizontal elements, while the lines in the opposite direction will be attached to the vertical straps. When this is done, a " split-plug" of the form shown in Fig. 115 is used to cut in a set of instruments at the home station. Where switchboards are wired in this way, it is necessary to have one or two spare JL_ FIG. 116. Double porcelain base spring-jack. horizontal and as many spare vertical straps in order that line wires may be "looped" when required. The unit type of strap-and-disk switchboard illustrated in Fig. 112, in connection with the spring-jack arrangement shown in Fig. 116, has within recent years been introduced for the purpose of meeting the need for a more flexible and rapid switching system. FIG. 117. Single and double conductor wedges for use with spring-jacks. The "wedges" (Fig. 117) used in connection with spring-jacks to make the various combinations of circuits required in practice are made up either as single conductors or as double conductors. The "single" wedge has a length of flexible single-conductor cord attached to a brass strip on one side of the wedge, the other side of which is of hard rubber, while the "double" wedge has a brass strip on each side, separated by an insulating strip of hard rubber. Each of the metal strips has connected with it one of the conduct- ors of a flexible twin-cord. The spring- jack, permitting as it does of the insertion of several wedges MAIN-LINE SWITCHBOARDS 137 in various relations to each other, provides an excellent means of meeting main-line telegraph switchboard requirements. Figure 118 shows a front and a side view of one unit of the arrangement referred to. The line wires are shown entering through the "fuse" and FIG. 1 1 8. Front and side views of switchboard unit including strap and disk, and spring-jack connections. FIG. 119. lightning protector, from there connected to the inside or stationary element of the spring-jack, the spring-actuated, or movable element (the shank) of which is in turn connected with a vertical strip of the switchboard proper. Figure 119 shows the back-of-the-board wiring of a five-line intermediate 138 AMERICAN TELEGRAPH PRACTICE switchboard of the strap-and-disk type equipped with spring-jacks. It may be noted that the connecting wires leading from the office side of the protective device to the heel of the spring-jack, are "cabled" instead of being brought down separately as shown in Fig. 118. The lower portion FIG. 120. of Fig. 119 plainly shows the theory of the connections of this convenient switching arrangement. One of the advantages of the unit type of board is that the switching facilities of an office may be increased to take care of additional lines, simply by adding additional units to the existing switchboard. Several years ago telegraph engineers recognized the possibilities of the MAIN-LINE SWITCHBOARDS 139 telephone type of jack (the pin-jack) for telegraph purposes, and the pioneer work along this line, done by Mr. J. F. Skirrow, Associate Electrical Engineer of the Postal Telegraph- Cable Company, New York, has resulted in the development of a line of switching apparatus which embodies all of the ad- vantages of this compact and useful device. Pin-jacks are made to meet various requirements, and are known as " open-circuit jacks," " closed- circuit," "patching," "grounding," "series," "multiple" jacks, etc. In construction, several of these forms of jack are identical, but the different forms are variously designated as stated. Several of these Blocks may be used together where Cut-outs on/y, withouf Switching Facilities are required. Looping Cord for Connecting "Line" to "Table*. Main LineSounder, K.O. B. \5dr. FIG. 121. Figure 120 shows several styles of pin-jack, each designed to meet a different requirement. If in each case the dark sections are regarded as consisting of insulating material, the uses to which each may be put is self- evident. No. i, for instance, is a series or closed-circuit jack intended for use in a wooden shelf. No. 2, a series or closed-circuit jack for use in a porcelain block. No. 3, an open or multiple-jack for use in a wooden shelf. No. 4, an open jack for mounting in a porcelain block. Nos. 5 and 6 , patching jacks for mounting in wood and porcelain respectively. Figure 121 shows a switch "block" having the line and instrument circuits connected through pin-jacks. The pin-jacks are mounted in a porce- lain block on a common base with the fuse holders and the lightning arresters. 140 AMERICAN TELEGRAPH PRACTICE Line ^Arrester Figure 122 shows the theoretical connections of the five pin-jacks. It will be seen that the two "line" and two "table" jacks are of the closed- circuit or series type, while the grounding-jack is of the open-circuit type. The insertion of a solid metal plug in the grounding-jack connects !j^ the line wires to "earth" on the ||j ground side of either line- jack, while aa _ ==== = L j^^ n Adi tne insertion of double-conductor [T^^ '-Jl iP^ss^n ^|| , , , , . . . . 1 Line li-L-Jj Line | P lu S s ( sh wn on the right in Fig. 121) in either of the line-jacks con- nects the line in series with which- ever table-jack the double plug on the other end of the flexible cord may be inserted into. FIG. 122. Connections of the five pin- The mam _li ne instruments in the jacks mounted in the switch block. Fig. 121. . Al . , , ,, office are permanently wired to the binding-posts on the lower edge of the switch-block, the binding-posts, in turn, being connected with the table-jacks. It is evident that provision is made for operating one or two main -line Line Arrestrl FIG. 123. Switch block equipped with cross-connecting facilities. MAIN-LINE SWITCHBOARDS 141 instruments upon a "loop," and also for "splitting" or dividing the loop into two single grounded circuits, the latter being accomplished by the insertion of a solid metal plug in the ground-jack. This switching arrangement, however, cannot be used where cross- connecting facilities are required. As a branch office "cut-out" this arrangement meets the requirements -admirably. When the office instruments are to be cut out at night, the only operation necessary on the part of the attendant is to withdraw the plugs from the pin-jacks. A switch-block similar to the above in construction and appearance, but having facilities for cross-connecting wires, is depicted in the diagram, Fig. 123. Intermediate switchboards intended for several wires may be made up by assembling a number of these units. To ground a wire in either direc- tion, a solid metal plug is inserted in the ground-jack. Cross-connections are made by means of flexible con- ducting-cords having solid metal plugs on each end. One plug is in- serted in the patching-jack of one wire and the other plug in the patch- ing-jack of the other wire, east or west, north or south, as desired. To test a patch by grounding the line, one plug of a patching-cord is held in contact with the ground-post below the lightning arrester, while the other plug is held in contact with the line wire where it enters the fuse. The office instruments may be cut in through the looping-jacks by means of the double-conductor cords and plugs shown on the right and left, Fig. 123. Figure 124, shows theoretically the connections through the various jacks. These switch-blocks are fire-proof and practically indestructible. In most of the offices of the Western Union Telegraph Company, and in the majority of railroad offices throughout the United States and Canada, the strap-and-disk switchboard is used at intermediate offices. In the offices of the Postal Telegraph- Cable Company, as well as in the railroad offices operated in connection with the Postal Company's system, although there are a large number of strap-and-disk switchboards in use, the pin-jack type of switch is extensively employed, and such equipment is regarded as standard. At intermediate offices having not over six main wires, the "Postal" FIG. 124. Connections of the five pin-jacks mounted in the switch block, Fig. 123. 142 AMERICAN TELEGRAPH PRACTICE employs a switching system made up as shown in Fig. 125. In the diagram is shown all necessary circuit equipment for six through wires. The view at the top shows the course of the circuit from where the line wire enters from the west through the pin-jack contacts and fuses to the point where the line east leaves the office. The lightning arrester and fuse equipment in each case is mounted on separate porcelain blocks. Also, the six pin-jacks are mounted in a porce- lain block. Thus each wire connected into and out of the office passes through a three-block unit consisting of two fuse and arrester blocks and a pin-jack block. Two of the jacks are looping-jacks, one to cut-in east, the other to cut-in west. Two of the jacks are patching-jacks east and west, it/era*** BfrouGt/T THirovfftt uuc Ft/sea. FIG. 125. Pin-jack switchboard equipment for offices having not over six lines. and the remaining two are grounding-jacks east and west. Where this type of switchboard is used, the following directions apply to its operation : To Cut in or Loop an Instrument upon a Wire. Place the instrument plug in one of the jacks of the wire it is desired to loop into, under the word "loop" in the brass guide plate. To Open a Wire. Place a solid plug in the jack of the wire it is desired to "open" under the words "open or patch" in the guide plate. To Ground a Wire. Use the same plug as for opening, but place it in the jack under the word "ground" in the guide plate. If a grounding- plug and an opening-plug are used upon the same side of a wire at the same time, the wire will be opened upon that side and grounded upon the other. Looping, opening or grounding may be done north, south, east or west according to which jack of the two provided for that purpose is used, in accordance with the marks on the guide plate. To Patch a Wire. Use a cord with a solid plug on each end. If it is de- sired to patch No. i west to No. 2 east, place one plug in the jack No. i west MAIN-LINE SWITCHBOARDS 143 under the words "open or patch" in the guide plate and the other plug in the jack No. 2 east under the words "open or patch." All other patches are made in a similar manner. JMCKS \ JACKS lilll! Stiff* 3 C/fOSS CONNECTING f?AC/f ANO TE&M/NAL BAR FIG. 126. Pin-jack switchboard for offices having six or more through wires. To Connect a Loop into a Main Line. Loops are brought to the switch- board in the same manner as line wires, that is, one side of the loop comes in at each side of the board. Use a double-cord with a double, or "looping " 144 AMERICAN TELEGRAPH PRACTICE plug on each end. Place one plug in a jack of the loop under the word "loop," and the other plug in a jack of the line under the word "loop." Another method: use single cords and patch one side of the loop to one side of the line, and the other side of the loop to the other side of the line, using the patching jacks as explained under "to patch a wire." When neces- sary to place more than one loop upon the same line, connect the loops together, using a double-cord from a looping-jack of one loop to a looping- jack of the other loop. Then connect the line to one of the loops as described above. When an attendant is asked to ground a wire "out side of the board" for a test, the procedure outlined in connection with Fig. 123 for "testing a patch by grounding the line" may be followed. All of the "fuse" and switchboard connections of a wire can be "bridged" out (to test fuses, jacks, etc.) by using a single cord, placing one plug against each of the line terminals at the fuse blocks. A double plug attached to a double-conductor cord connected with the office instrument is inserted in a looping jack when it is desired to cut the instrument in circuit. At intermediate offices having six or more through wires, in those in- stances where pin-jack equipment is used, the fuse and arrester blocks, pin-jack blocks, etc., are mounted on an angle-iron frame of substantial con- struction and finished appearance as indicated in Fig. 126. The circuit -connections are, of course, identical with those shown in the smaller switchboard, Fig. 125. To the larger board there is added a "ter- minal bar," and a cross-connecting rack, mounted on the back of the switch- board as shown on the right, Fig. 126. These additional features make possible a systematic distribution of cable conductors and provide means whereby cross-connections may be made between the line wires on the back of the board when so desired. TERMINAL OFFICE SWITCHBOARD EQUIPMENT The highest development in switchboard construction is found at the larger terminal stations, where on account of the large number of lines to be cared for, it is necessary to employ thoroughly systematized methods of circuit identification, and to provide facilities for making alterations and additions to the wire plant in such a manner that regular service will not be interrupted. Figure 127 shows an arrangement of fuse and arrester equipment at a terminal office. The line wires from the aerial or underground cable are shown coming from the street in a cable. Each line wire has a circuit through one of the terminal blocks, to the cable which leads to the cross-connecting frame shown on the left. Usually the terminal frame shown on the right is MAIN-LINE SWITCHBOARDS FOR TERMINAL OFFICES 145 located as near as possible to the point where the cabled line wires enter the building. The terminal bars mounted in this frame consist of porcelain blocks, each one of which has mounted in it four pin-jacks as shown in outline ooooooooooooooooooooo FIG. 127. at the top of the frame. These pin- jacks provide a means whereby " ground- ing," "patching" and "looping" may be done, in a way identical with that described in connection with the smaller intermediate switchboards. 10 146 AMERICAN TELEGRAPH PRACTICE MAIN-LINE SWITCHBOARDS FOR TERMINAL OFFICES 147 The cross-connecting frame shown on the left is located in close proximity to the main switchboard in the operating room; generally it is convenient to mount the frame immediately in the rear of the switchboard. A more complete plan of the wiring between the point where line wires are brought from the terminal room to the fuse and arrester frame in the rear of the switchboard proper, and the cross-connecting frame, is shown in Fig. 128. The arrangement of conductors from cross-connecting frames to pin- jacks and spring- jacks in the main switchboard and from cross-connecting frames to instrument tables is clearly shown in the diagram. The pin-jack locations A provide a means for transferring circuits from one section of the main board to other sections. All of the connections are made with flexible conducting cords the terminals of which are equipped with single or double plugs, single or double wedges as required. The terminal room equipment is in reality a switching system, practically a duplicate of that installed in the main operating room, but constructed with the object of providing for flexibility and utility rather than for fine appearance. The advantages of having a complete switching system close to the point where the lines enter the building from underground and aerial cables are that the jacks in the terminal room frame serve both for cable terminals and for temporary cross-connecting purposes by means of cords when cables are in trouble. From this frame may be made quick tests of cable interruptions, and the equipment may be used as a temporary switch- board in case the main switchboard in the operating room should be de- stroyed by fire or disabled from any cause. The function of the cross-connecting frame mounted in the rear of the main switchboard is to act as an intermediate connection between the pin- jack and spring-jack connections of the switchboard and the instruments located on operating tables. In the offices of the Postal Telegraph-Cable Company all frames are of angle-iron, including switchboard frames and all switchboard connections, cross-connecting frame, and terminal frame connections are mounted either upon slate, porcelain, or asbestos board. CONDUCTORS BETWEEN CROSS-CONNECTING FRAMES AND OPERATING TABLES All of the conducting wires required between cross-connecting frames and instrument tables, and between the power switchboard and instrument tables are, in all up-to-date installations, carried through floor ducts or trenches. The trenches are from 4 to 8 in. wide and of about the same depth, and, as usually arranged, have a conveniently remcvable cast-iron top or lid, laid flush with the surface of the floor. In this trench are laid all of the battery 148 AMERICAN TELEGRAPH PRACTICE if o n n n n n U U U U U U MAIN-LINE SWITCHBOARDS FOR TERMINAL OFFICES 149 wires and cables leading from the cross-connecting frames to the various operating tables. In some instances, instead of using open top trenches, iron pipes i\ in. in diameter are embedded in concrete flooring and so distributed that all parts of the operating room are served as indicated in Fig. 129. The diagram shows a skeleton main-line switchboard with cross-connecting frame in the rear. An iron-pipe conduit is shown laid beneath the office flooring and stretching from the wiring frame to an operating table, there terminating in a hand-hole with surface outlet under the table. From the hand-hole a 'Cable Transfers 'ity Loop FIG. 130. Line wire connections between underground or aerial cables and main-line switchboard at a terminal office. short lateral duct provides- access to the instruments mounted on top of the table through a wire chute situated between the type-writer lockers. In the more recent installations a wiring cabinet has been built into the aisle end of each operating table, the hand-hole outlet from the floor duct being built in the floor within the cabinet. This latter arrangement provides for accessible mounting of all fuses, resistance coils, cable terminal strips, etc., constituting that part of the equipment of the table. The plan and construction details shown in Fig. 129 are self-explanatory. 150 AMERICAN TELEGRAPH PRACTICE That portion of the wiring of a terminal office between the cross-connect- ing frames and the switchboard, so far as main-line wires are concerned, is shown in Fig. 130. The course of a line wire from the aerial or underground cable may be traced through the line fuse, lightning-arrester block, cross-connecting wire to terminal bar, thence through a closed-circuit or series pin-jack to the spring-jack mounted on the face of the main switchboard. If the main- line wire shown is regularly assigned to a particular service, the connection is made by means of a short flexible cord with plug terminals, the plugs being inserted in the pin-jacks as suggested in the diagram. The employ- ment of short cords for regular circuit assignments greatly reduces the amount of conducting cord necessary to make a given number of connections. FIG. 131. Sections of a main- line switchboard at a terminal office. If the line wire shown were required to be transferred to a distant part of the board, the plug on one end of the cord would be inserted in the closed- circuit jack, while the other plug would be inserted in one of the transfer- jacks leading to that section of the switchboard where the connection is desired. Figure 131 shows two sections of a large main-line switchboard built up of slate panels mounted on angle-iron framework. From the circuit diagrams heretofore shown, the reader will recognize the strap-and-disk arrangement, as well as the spring-jack and pin-jack equipment mounted underneath. NEW WESTERN UNION SWITCHBOARD EQUIPMENT In new construction, and in the reconstruction of switchboard equipment, the Western Union Telegraph Company plans to make extensive use of the MAIN-LINE SWITCHBOARDS FOR TERMINAL OFFICES 151 FIG. 132. Western Union distributing frame. 152 AMERICAN TELEGRAPH PRACTICE pin-jack. The aim is to abandon the use of the strap-and-disk equipment, also the spring- jack and wedge accessories now universally employed for main-line switching purposes. The new type of switchboard consists of an angle-iron frame, having mounted on its face porcelain panels, each panel containing 16 telephone- type pin-jacks, similar to those previously illustrated and described. A switchboard containing ten panels will have 160 pin-jacks, and where four jacks in series are required to take care of the various operations of " grounding," "looping," and " patching," a lo-panel board will accommodate 40 main-line wires. DISTRIBUTING FRAMES In new installations, the Western Union Company consolidates the " terminal room," and "cross-connecting" frame, features as utilized in the "Postal" Company's service, forming a common "distributing frame" of the type employed in telephone service. Figure 132 shows a perspective view of a section of the distributing frame, which is located near the main switchboard. All line wires cabled into the office from underground or aerial lines are brought directly to the distributing frame as also are all cables from instrument tables, the various house circuits being completed by means of short cross-connecting wires extending through the frame. The terminal block units for mounting on the frame are made of porcelain, each block having 10 wing-nut binding posts. The distributing frame as a whole is made up of "base units" and "top units." The illustration, Fig. 132, is that of a base unit. Both sides of a standard base unit will accommodate 24 terminal blocks, and both sides of a standard top unit, the same number. The function of the distributing frame is identical with that of the cross- connecting frame previously described, that is, to provide means for making any required connection between the different sets of instruments in use, or between line wires and signaling instruments mounted on operating tables, without interfering with the cabled conductors, or disturbing the permanent wiring of the switchboard proper. CHAPTER X ELECTRICAL MEASURING INSTRUMENTS TELEGRAPH LINE AND CIRCUIT TESTING The satisfactory operation of telegraph circuits is almost entirely depend- ent upon the efficiency of the methods of testing practised, upon thoroughness of inspection, and upon the standards of line maintenance observed by the operating department of a telegraph administration. From the beginning of the art and until a comparatively recent period, the Tangent Galvanometer was used almost exclusively for the purpose of making tests and measurements. In recent years, however, the quickening of the service has created a demand for more rapid methods of circuit testing, and at the present time the direct-reading instruments, such as the voltmeter, ammeter, and milammeter, are entensively employed in telegraph testing. Even the Wheatstone bridge, so long the standard measuring instrument, is now used only where accurate figures are necessary. The demands of fast service are such that modern practice recognizes the value of qualitative as distinguished from quantitative measurements; so much so that we find the simple telephone receiver gaining favor as an indicator of faults. True, the great growth in the practice of cabling con- ductors has created conditions favorable to the employment of the telephone receiver as a fault finder. In what follows, various practical and laboratory methods of making all tests and measurements required in practice are explained in sufficient detail to enable the practical telegrapher to familiarize himself with the procedure customary in each case. The Galvanometer. The term galvanometer might correctly be applied to any indicating instrument which measures the magnitude of, or indicates the direction of electric currents. While there are many makes of galvanometer in use, practically all such instruments are either of the moving-coil or moving-needle type. The former is known as the d'Arsonval type of instrument, in which a small coil of wire is suspended between the poles of a magnet, with its axis normally at right angles to the lines of force in the magnetic field. The moving-needle instrument has a magnetized -steel needle or pointer delicately suspended with its axis horizontal, and having a movement in a horizontal plane. Normally the indicating needle points in a north and 153 154 AMERICAN TELEGRAPH PRACTICE FIG. 133. d'Arsonval portable galvanometer. FIG. 134. Mirror galvanometer. ELECTRICAL MEASURING INSTRUMENTS 155 south direction due to the influence of the earth's magnetic field, or to the field of artificial magnets mounted near it. Close to the center portion of the needle, generally surrounding it, is mounted a coil of insulated wire with its axis at right angles to the normal north and south direction of the needle. When the coil is energized from a source of electric current, the needle tends to move into a new position to a point somewhere between the original field and that of the axis of the coil, the distance through which the needle moves being dependent upon the strength of current in the circuit of which the coil winding forms a part. The d'Arsonval ty'pe of instrument is the one generally used for commer- cial measurements. So far as the principles involved, and operation are concerned, the galva- nometer and the ammeter are identical. The former, however, may be used for detecting currents of a much lower value. Figure 133 shows a make of portable d'Arsonval galvanometer used in connection with Wheatstone-bridge measurements. Where exact measurements are necessary, a galvanometer of high sensi- bility, such as that illustrated in Fig. 134, is used. This form of d'Arsonval galvanometer has the moving element mounted between magnet poles and suspended from a point near the top of an upright tube by means of a very fine plated phosphor-bronze, silver, or steel wire. A small round mirror is fastened to the moving element, reflecting outward. The deflections of the coil, resulting from the presence of electric current in the winding, are meas- ured by means of a telescope and suitable scale. The image of the scale in the mirror may be read through the telescope, or as variously used, the move- ment of a spot of light on a stationary scale mounted a short distance away from the mirror may be directly observed while measurements are being made. Due to the fact that the moving coil and its suspension are non- magnetic, and that the magnetic field in which the moving element turns is very strong, the readings of this current indicator are not appreciably affected by the earth's field, or other neighboring magnetic disturbances. Differential Galvanometers. For comparing the relative strengths of two currents, a galvanometer is sometimes employed in which the coil con- sists of two separate identical windings, mounted side by side. If equal currents are at the same time sent through both windings, there will be no deflection of the indicating needle, but should the currents be unequal in strength the needle is deflected, due to the influence of the stronger current; to a degree corresponding to the difference in the two current strengths. When the current strengths are equal, the effect of one coil upon the needle is completely neutralized by that of the other. The Ballistic Galvanometer. Ballistic galvanometers are employed to measure currents of momentary duration, such, for instance, as flow in a circuit when a condenser is discharged through it. With this instrument 156 AMERICAN TELEGRAPH PRACTICE the oscillation period of the needle must be long as compared with the dura- tion of each discharge. As the needle which is long or heavy swings slowly around, the amount of deflection is additive, that is, the intermittent indivi- dual impulses impressed on the circuit result in a cumulative effect upon the needle. Where no damping of the needle is resorted to, the sine of half the angle of the first swing is proportional to the quantity of electricity that has flowed through the coil. Galvanometer Shunts. In cases where it is necessary or desirable to use a high sensibility galvanometer in making measurements requiring a con- siderably lower sensibility, the galvanometer coil may be shunted by a resist- ance having a definite ratio to that of the galvanometer coil. A formula for determining shunt values was given on page 86 in connection with Fig. 65. Where galvanometers are not equipped with regular shunt-coils any ordinary resistance coil or coils of the correct value may be used for the pur- pose. As usually furnished with galvanometers, shunts are adjustable to 1/9, 1/99, or V999 f the galvanometer resistance, so that i/io, i/ioo, or i/iooo part of the current only passes through the galvanometer coil. Constant of a Galvanometer. The constant or sensibility of a galva- nometer refers to the value of the resistance in ohms through which i volt will produce a deflection of one degree on a standard scale, or to apply a general rule: Multiply the deflection by the multiplying power of the shunt and by the resistance in the standard resistance box expressed in megohms or fractions thereof. In Fig. 135 the necessary connections are shown for taking the constant of a galvanometer. R, is a standard resistance of 100,000 ohms. Closing the key K will cause the galvano- meter to be deflected d degrees. If then the shunt employed has a multiplying power of 1,000, obviously had no shunt been used the amount of deflection of the needle would FIG. i 3 5.-Taking the constant of have been I ' OO times as 8 reat This at least a galvanometer. would have been the case theoretically. Had a resistance of 1,000,000 ohms been used in place of 100,000, the deflection would have been but one-tenth of d; so that the deflection K, through 1,000,000 ohms in series with the galvano- meter coil with no shunt applied would have been i,o A = IO Where the multiplying power of the shunt is represented by m, the deflection ELECTRICAL MEASUREMENTS 157 in degrees by d, and the resistance in megohms, or fractions thereof by R, then K = Rmd. The terms "constant," " figure of merit," and "sensibility," when used with reference to galvanometers, have the same meaning. MEASURING THE RESISTANCE OF GALVANOMETERS Half -deflection Method. Connect the galvanometer as shown in Fig. 136. The resistance R and the source of e.m.f. B should be so regulated that the deflection of the galvanometer needle is over one-half of the scale. Note the deflection, then increase the resistance R until the needle moves R -WAAA- FIG. 136. Half-deflection method of measuring the resistance of a gal- vanometer. FIG. 137. Kelvin's method of measur- ing the resistance of a galvanometer to a point on the scale exactly midway between zero and the point of first deflection. Disregarding the battery resistance, the resistance of the galvanometer will be the resistance of R measured at "half deflection," less twice the original resistance of R, or G = r-R2. Kelvin's Method. Connect the galvanometer in the X arm of a Wheat- stone bridge as shown in Fig. 137 and adjust R until the deflection of the needle is the same whether key K is closed or open, then :-*! Of course, where two galvanometers are available, the instrument whose resistance is desired may be inserted in the X arm, and the " bridge" balanced in the usual manner by means of the other galvanometer. THE VOLTMETER Measuring instruments which indicate the value of the e.m.f. in volts impressed upon their terminals are called voltmeters. When a voltmeter is connected across the terminals of a source of e.m.f. a 158 AMERICAN TELEGRAPH PRACTICE current will flow through it which is directly proportional to the impressed voltage. Attached to the moving element (which in principle is the same as that of the d'Arsonval galvanometer) there is a light pointer moving across a scale which has been empirically graduated into divisions to indicate the value of the impressed e.m.f. Contained within the voltmeter casing there is a non-inductive resistance ranging in ohms from 10 to 2,000 times the full scale reading in volts. This resistance is in series with the winding of the movement coil, and it is customary to insert 100 ohms or more for each volt as indicated on the scale. The higher the series resistance per volt, the greater will be the accuracy of the indications. There are various types of voltmeter available for different needs, among which might be mentioned the alternating-current voltmeter for measuring currents of a given frequency, one make of which has a mass of soft iron so placed that it will be moved into a solenoid, or from the center of a solenoid to one end, the movement of the soft iron plunger controlling the travel of a scale pointer. Also there are voltmeters based upon the principle of the electrodynamometer which may be used for either direct-currents or alternating-currents, and which are independent of variations in current frequency and of wave form. Hot Wire Meters. Hot-wire voltmeters and ammeters are used in which the passage of current through a length of thin wire causes a rise of temperature with consequent expansion of the metal conductor. As the wire expands, the slack is taken up by a spring, the resulting movement of which causes a pointer to travel across a properly graduated scale, thus indicating the strength of the current traversing the hot wire. MULTIPLIERS FOR VOLTMETERS The range of a given voltmeter may be increased by employing a suitable multiplier in the form of an additional external resistance placed in series with the voltmeter. With a low-reading voltmeter and a set of multipliers it is practicable to measure voltages covering a large range of values. As- sume, for instance, that the only meter available is a 5o-volt instrument, having 5,000 ohms resistance, and that it is desired to use the meter for measuring higher voltages. A multiplier with a value of 2 would measure 10,000 ohms, which would give a scale value for the meter of 100 volts. A multiplier with a value of 10 used with the 50- volt meter would measure 50,000 ohms, which would give the meter a scale value of 500 volts. It is, of course, understood that 50,000 ohms would represent the total resistance of meter and multiplier in series. A formula applicable to any requirement might be stated thus: ELECTRICAL MEASURING INSTRUMENTS 159 where R is the resistance of the voltmeter, R r the multiplier resistance to be connected in series with the meter, V the highest reading of the meter nor- mally, and V the highest reading desired. V The scale reading observed must be multiplied by -~ to obtain the correct value of the e.m.f. CURRENT METERS The ammeter and the milliammeter are instruments for measuring the current strength in circuits, the indicating scales being marked off into divi- sions representing amperes and milliamperes respectively. In the series ammeter the entire current to be measured traverses the coil winding of the instrument, and as in the case of the galvanometer, variations in the current strength cause the indicating needle to be deflected to a greater or less degree from its position of rest, the amount of deflection being dependent upon the value of the current in the circuit. For currents of any considerable volume, the shunt ammeter is generally employed in which a small portion only of the current is carried by the instrument coil. In portable ammeters the shunt coil is mounted in the base of the instrument. BATTERIES FOR TESTING PURPOSES Where line tests are made from terminal or intermediate test offices, generally there is available current from motor-generators or gravity bat- teries which may be applied in any desired manner, but where measurements are to be made from manholes giving access to underground cables, from aerial cable boxes, or from any point where regular battery is not available, it is necessary to employ a portable battery for the purpose. Modern Wheatstone bridge sets have a self-contained dry battery source of e.m.f. which yields a current of sufficient strength for making all ordinary "bridge" measurements. They are also equipped with a conveniently mounted battery switch which makes possible regular main-line switch- board battery connections when the bridge is used for line testing from a terminal office. Where the resistances involved are not great, ordinarily, dry cells serve the purpose satisfactorily, but in cases where high resistances are to be dealt with in making certain tests, potentials of at least 50 or 100 volts are best suited to the purpose. Formerly the chloride of silver battery was extensively used for testing purposes. This form of battery met the requirements admirably, and has only recently given way to the more efficient storage battery, which is now available in light and compact units. 160 AMERICAN TELEGRAPH PRACTICE One excellent make of portable storage battery designed for testing pur- poses is known as the "Witham," or Marcuson battery. Several small storage cells are assembled in boxes to form batteries having certain ranges of voltage. There are four standard sizes, having maximum voltages as follows : 100 volts, 140 volts, 1 68 volts, and 256 volts. The boxes containing the battery complete weigh 17 1/2, 24 1/2, 29, and 42 Ib. respectively. These batteries are divided into two or more sections, which by means of a commu- tating switch may be connected in series or in parallel, thus giving at the terminals either the full voltage of the battery or a fraction of same. THE WHEATSTONE BRIDGE This instrument consists of an arrangement of conductors as shown theoretically in Fig. 138. One terminal of the battery is led to the point d, where it divides into two paths which are united again at the point c, so that a portion of the total current flowing passes through the point e, and a portion through the point /. The four conductors a, b, R, and X are called the " arms " of the bridge. When the electrical resistances of three of the arms are known, the resistance of the fourth may be calculated according to the proportions of their relative values. What was said on page 87 in re- FIG. i 3 8.-Theoretical connections of the gard to f dl of poten tial along a con- ductor" (Fig. 66) has a direct bearing upon the underlying principle of the Wheatstone bridge. Referring to Fig. 138: It is obvious that there will be a fall of electric potential between the battery terminal and the point d; also that there is a further drop along the upper branch d,f, c, and that the potential of the lower branch falls along the path d, e, c. If the point e and the point/ are equally distant, electrically, from the point d, and in the same sense equally distant from the point c, then the potential will have fallen at e to the same value it has fallen to at the point/, or if the ratio of the resistance a to the resistance R be equal to the existing ratio between b and X, then the points e and f will have equal potentials. Connecting a galvanometer between the points e and/, as shown, furnishes a means whereby it is possible to observe whether or not the points e and/ are at equal potentials. When such is the case, there will be no deflection of the galvanometer needle, or when the resistances of the four arms are in "balance" we have by proportion a:b: :R:X, and if we know the resistance values of a, b, and R } X may be determined thus: THE WHEATSTONE BRIDGE 161 The unknown resistance or the resistance to be measured is inserted at X, that is, between the points d and e. When the resistance to be measured is not greater than the total resistance of R, the ratio arms a and b may be made equal, then if the rheostat arm (R) of the bridge be adjusted until the galvanometer indicates a "balance," it is plain that the unknown resistance (X) has a value equal to that of R. When it is desired to measure a resistance greater than the total value of R, the ratio arm b should be given a higher value than a, and to measure very low resistances the ratio arm b should be given a value less than that of a. For example, let a = 10 b i ooo 5 = i,ooo khen-X' = - = X54o = 54,ooo. : > then X = X 540 = 5.4. Let a = 1,000 = io COMMERCIAL WHEATSTONE BRIDGE INSTRUMENTS There are several large instrument houses that manufacture high-grade measuring instruments and bridge sets, and when such apparatus is given proper care after being delivered by the manufacturer, generally it will be found to be quite constant and reliable in performance, and will have long life. One of the newer makes of bridge is illustrated in Fig. 139. In this particular set the battery is inclosed within the casing and consists of four dry cells of a stock size. The rheostat, or R arm of the bridge system is composed of four dials of ten coils each, which have values of units, tens, hundreds and thousands ohm coils. The ratio arms a and b have values of i, 10, 100 and 1,000 ohms in each arm. The galvanometer, which is of the d'Arsonval type, is mounted flush with the floor of the box containing the resistance coils. The scale has 30 millimeter divisions with center zero. A zero adjustment is provided which enables the tester to bring the needle exactly on zero, or on any point of the scale desired. Instead of metallic plugs being used to cut-in or cut-out the various resistance units, radial brush contacts are provided which swing in a complete circle in either direction, going from the highest value to the lowest value coil in any decade, without having to be turned back over the intervening contacts. By means of extra binding-posts and accessible switches it is possible to substitute an external source of e.m.f. in place of the dry-cell battery con- tained within the box, and to employ a separate galvanometer in place of the one mounted as a part of the set. Included as a part of the equipment of the set is an Ayrton shunt which allows full current, i/io part, or i/ioo part of the current to 11 162 AMERICAN TELEGRAPH PRACTICE flow through the galvanometer of the set or external galvanometer, which- ever is used with the bridge at the time. The set described, in common with other modern Wheatstone bridge sets, may be used in making the following tests: THOMPSON teVEff/NGCO PHILA.PA ._ FIG. 139. Commercial form of wheatstone bridge, including galvanometer and self- contained dry-cell battery. Measuring resistance by the bridge method. Measuring insulation resistance by the direct deflection method. Comparing e.m.f.'s by the fall of potential method. Checking up voltmeters. Measuring battery resistance. Making the Murray loop test. Checking up ammeters by using a shunt of known value. Making the Varley loop test. Testing out "grounds." The galvanometer and battery can be used in series. THE ELECTRIC CONDENSER The fact that both long and short metallic conductors used in form- ing electrical circuits possess capacity, means that when the electrostatic capacity of a conductor is to be measured, or when the static discharge from line conductors is to be compensated for, it is sometimes necessary THE ELECTRIC CONDENSER 163 to have available for these purposes electric condensers having variable capacities. In considering the theory of the electric condenser, it might be stated that the factors involved are a source of electric charge, a conductor of electricity, and a dielectric (insulator). One of the most comprehensive generalizations relating to electro- statics, established by Faraday, was that all electric charge and discharge is essentially the charge and discharge of a Leyden jar (a form of electric condenser). The original form of condenser, the Leyden jar, owes its name to the city in which it was invented. The appearance of this condenser is illustrated in Fig. 140. Later forms of the Leyden jar were made with an inside and an outside coating of tin-foil reaching from the bottom to within 2 or 3 in. of the top of the glass jar. A shellaced wooden top fitted into the neck of the jar, served as a support for a brass knob mounted on the upper end of a brass wire; the latter extending through a hole in the top had affixed to its lower or inside end a length of brass chain reaching to the bottom of the jar and in contact with the in- side foil coating: the efficiency of the jar as a con- denser being considerably increased by the applica- tion of a coating of shellac, due to the fact that shellac very materially retards the dissipation of the charge *** T 4 -Original form . of Leyden-jar condenser. over the uncovered surfaces of the jar. Condensation of moisture on the surface of the glass interposes a conducting path, even if of very high resistance, which permits gradual equalization of the opposite charges gathered on the tin-foil surfaces attached to the inside and to the outside of the jar. From this brief description it is evident that a Leyden jar condenser, like any other form of electric condenser, consists simply of two conductors separated by an insulator. The capacity of any form of condenser, that is, its ability to retain a greater or less quantity of charge, is dependent upon the area of the conducting surfaces and upon their distance apart. COMMERCIAL, OR STANDARD CONDENSERS The capacity of commercial condensers is either fixed or variable, or as more commonly stated, non-adjustable or adjustable, respectively. A non-adjustable condenser has its conducting surfaces, or leaves, arranged as shown in Fig. 141, in which the leaves are represented by vertical lines and the connecting metal strips by horizontal lines. 164 AMERICAN TELEGRAPH PRACTICE Alternate leaves are connected with the metal strip A, while every other alternate leaf is connected with the metal strip B, as shown. Conden- sers constructed so that their capacity may be adjusted or varied, have alternate leaves connected with a common terminal as shown in Fig. 142, while the remaining alternate leaves are connected in groups which may be placed in contact with the other condenser terminal B, at i, 2, 3, 4, etc., as desired, by inserting metallic plugs in contact plates distributed at these points. It is important to note that an adjustable condenser when completely assembled and connected for full capacity, is simply a number of non- adjustable condenser units arranged in parallel. I LU iiu i in iiu i ui 12345 B FIG. 141. FIG. 142. Capacity of Condensers. The joint capacity of condensers connected in parallel is equal to the sum of their respective capacities, or in the case of two condensers Total capacity = Ci+C 2 The joint capacity of two condensers connected in series is equal to the prod- uct divided by the sum of their respective capacities, or c xc Total capacity = >r Where three condensers are connected in series, the joint capacity -4- + r ~" ' r ' r 1 **'i < - / 2 ^3 Or for any number of condensers in series, the joint capacity is equal to the reciprocal of the sum of the reciprocals of their respective capacities, thus following the law of the joint resistance of parallel circuits as explained in a previous chapter. When combined in multiple series, the same law applies, the total capacity of each group of condensers connected in parallel being regarded as the capacity of a single condenser in order to obtain values for the purpose of computing the total capacity available from a given arrangement. MEASURING CAPACITY Occasionally it is necessary to determine the capacity of a condenser, a line wire, or cable conductor. One method of obtaining the desired in- MEASURING "CAPACITY" 165 formation is that known as the direct deflection method, and the procedure is as follows: charge a standard condenser Ci Fig. 143, from a source of e.m.f. for a period of, say, 30 seconds, then discharge the condenser through a galvanometer (preferably a ballistic, or an astatic galvanometer); note the deflection and call it d. Next charge the condenser to be measured from the same battery and for the same period of time, and discharge it through the galvanometer in the same manner. Note the deflection of the needle, and call this di Then and CuCiidid di l d' C, K II- FIG. 143. Measuring the capacity of a condenser by the direct-deflection method. Bridge Method. Connect the two condensers to be compared as shown in Fig. 144. Ri and R% are non-inductive resistances of about 2,000 ohms each, G a galvanometer, E a source of e.m.f., and K a key. Adjust the resistances Ri and R 2 so that there is no deflection of the galvanometer needle when the key K is manipulated. Then ~\ *t (\\(i. K E mm mm c, FIG. 144. Comparing the capacity of condensers by the bridge method. Modern commercial adjustable condensers are made up of a number of small units assembled within a sheet-iron or tin case, equipped with sliding contacts controlled by a revolving knob by means of which the capacity of the condenser is varied. The sliding contacts take the place of the peg-and-hole connections formerly used. Insulation Resistance of Condensers. It might be supposed that the insulation resistance between the two terminals of a condenser is infinitely high, but it is found that condensers, as usually manufactured, sometimes have an insulation resistance so low that when the condenser is inserted 166 AMERICAN TELEGRAPH PRACTICE between line and ground, the circuit through the condenser in reality forms a high-resistance leak. In telegraph practice, a condenser which is found to have an insulation resistance between terminals of not less than 500,000 ohms is regarded as satisfactory for all practical requirements. Using a standard milammeter, with an applied e.m.f. of 375 volts, one division 1 deflection on the lower scale of the milammeter represents an in- sulation of 1.8 megohms. Therefore a condenser tested with 375 volts pressure which shows more than 3 1/2 divisions deflection on the lower scale of the milammeter has an insulation resistance too low for satisfactory service. MEASURING THE INTERNAL RESISTANCE OF BATTERIES The well-known half-deflection and direct-deflection methods of measur- ing the internal resistance of batteries, formerly extensively employed, V.M. 4- B A.M. FIG. 145. Measuring the internal re- sistance of a battery by the voltmeter- ammeter method. FIG. 146. Measuring the internal re- sistance of a battery by the bridge method. have been superseded in modern practice by the employment of the volt- meter-ammeter, and the bridge methods of making these measurements. Voltmeter-ammeter Method. In making this test the ammeter is con- nected through a resistance in series with the battery, the internal resistance of which is sought, while a high-resistance voltmeter is connected in shunt with the ammeter circuit as shown in Fig. 145. With the key K open, the volmeter reading E is noted. Then with the key closed simultaneous readings may be taken of the voltmeter E\ and the ammeter 7, then the resistance of the battery Bridge Method. The battery to be measured is connected in the X arm of the bridge as shown in Fig. 146. With a and b equal, adjust R until the galvanometer deflection with the key K open or closed is the same. 1 Five divisions on the lower scale represent i milampere. EARTH CURRENTS 167 On account of the necessity for altering the regular bridge set connections when making this test, it is generally preferable to employ a portable gal- vanometer and separate resistance boxes. EARTH CURRENTS In those measurements where the ground is used as a portion of the completed circuit, occasionally earth currents introduce errors which make the readings unreliable. Also it is of considerable importance in certain operations to determine the value of the difference of potential between terminal offices due to earth currents. In the case of grounded- circuit measurements, where the earth current is fairly constant in potential and polarity, its effect may be compensated for by making first a measure- ment with the negative pole of the home battery to line, and then with the positive pole to line. If the readings have an appreciable difference in value, their average should be taken as the correct result. Measuring Earth Potentials. There are several methods of determining the potential of the earth between two points, but for practical requirements the simplest way, and which requires no calculation, is to use a voltmeter for the purpose. If the meter has a double scale, use the one giving the greatest deflection with minimum potential difference. To make the test, remove all regular battery from the line. Ground the line at both ends, that is, at each terminal of the wire under test. Connect the voltmeter into the line at the switchboard by means of a "wedge" or otherwise. This will connect the voltmeter direct from line to ground. If when the meter circuit is closed while the binding-post marked (+ ) is to line, the indicating needle moves to the right, then the ground at the distant sta- tion is positive to the home ground. If the needle swings to the left, reverse the wedge so that the post marked (+) connects with the home-station ground in which case the ground at the distant station is negative to the home ground. In noting the readings, the polarity as well as maximum and mini- mum potential readings should be recorded. MEASURING THE RESISTANCE OF EARTH CONNECTIONS One method of ascertaining the value of the earth resistance between two stations which may be applied where two line wires are available between the stations, is shown in Figs. 147 and 148. ' The two line wires are "looped" together at the distant station, and at the home station are connected in the X arm of the bridge, as illustrated in Fig. 147. The looped resistance is noted. Assume, for example, that it is found to be 6,000 ohms. Call this RI. Then, as in Fig. 148, measure the resistance of one of the wires between the home station and the ground at the distant 168 AMERICAN TELEGRAPH PRACTICE station. In like manner measure the resistance of the other wire. Say that one was found to measure 3,204 ohms, and the other 2,812 ohms, or a total of 6,01 6 ohms. Call this value R 2 . Then, resistance of the distant ground (assuming that the resistance of the home ground is nil) or 6,016 6,000-7-2 Ans. 8 ohms. FIG. 147. FIG. 148. FIGS. 147 AND 148. Measuring the resistance of earth contacts. In view of the variable conductivity of contacts made through "peg" switchboards, where this method of grounding is employed, it is essential that positive contacts be made, otherwise errors will be introduced which produce misleading results. It is, however, an excellent and quick method of determining whether or not a suspected ground connection has a resistance abnormally high. Another method, and one which takes into consideration the value of the voltage impressed on line wires due to earth potentials, is shown theo- retically in Fig. 149. In this test, the regular galvanometer of the bridge set is replaced with a voltmeter, all other connections remaining the same. The resistance of the line wire extending between the home station and the station where the ground resistance is to be determined is measured by means of the loop method, after which this wire is "grounded" at the distant station. The R arm of the bridge has all of its resistance plugged out, then with the arm MISCELLANEOUS TESTS 169 a or b opened temporarily the voltmeter indicates the value of the earth poten- tial from the distant ground. In most cases this value will be found to fluc- tuate somewhat, and it will be necessary in such cases to note the mean deflec- tion. The key should be left open while this reading is taken. Now close the key and with positive pole of battery to line raise the resistance of the R arm of the bridge until the deflection is of the same value as that first observed. Call this reading A. Reverse the battery terminals so that the negative pole will be to line. Close the key and adjust the resistance R until the volt- meter again shows the same value. Call this reading of R, B. FIG. 149. Measuring the resistance of earth contacts, where earth currents exist. Then the resistance of the line wire plus that of the ground connections at each end will be AXB A + B 2 ' If the line resistance as at first calculated is now deducted from the result obtained, the remaining figure will represent the resistance of the' distant ground connection. Assuming, of course, that the resistance of the home ground connection is known to be nil. MISCELLANEOUS TESTS In what follows, various methods of testing " opens," "grounds," "crosses," "escapes," "insulation," "conductivity," "resistance," "capac- ity," etc., will be explained. It might be deemed sufficient to give one ap- proved method of making each measurement or test, but as it is not likely that the testing equipment of a given telegraph administration, or of a given railroad telegraph system is the same at all of its testing offices, it has been thought best to submit alternative methods covering each test, so that no matter what the conditions are the attendant may have at hand a method of making the desired test which will meet the requirements. It is well, too, for the younger wire chiefs to seek an understand ing of the various standard methods of making all necessary measurements, for, by virtue of possessing such knowledge they are better able to grasp the principles involved and to understand the subject generally. 170 AMERICAN TELEGRAPH PRACTICE WHEATSTONE BRIDGE MEASUREMENTS Where strap-and-disk main-line switchboards are employed, the practice of the Postal Telegraph- Cable Company provides permanent connections between the testing set and disks in the main-line board, as shown in Fig. 150. Main 5wifth board FIG. 150. Wheatstone bridge permanently connected to main line switchboard. It may be seen that the line binding-posts (the X arm) of the bridge are connected to separate disks in the switchboard, and that a third wire is brought' to another disk in the switchboard for the purpose of grounding one side of the bridge when so required in making certain tests. The electrical connec- tions of the bridge shown are identical with those shown in the theoretical bridge diagram, Fig. 138, with the exception that short-circuiting switches are provided for the purpose of permanently closing the gal- FIG. 151. Theoretical circuits of bridge connected to vanometer and battery keys, main line switchboard. Also, a battery . reversing switch and a grounding switch are added for convenience in making required tests. Theoretically the con- nections would be as shown in Fig. 151. Practically all of the troubles to which telegraph lines are subject may be investigated with the bridge circuits arranged as shown in Fig. 152. THE MURRAY LOOP TEST 171 In those cases where the resistance of a " cross" varies within wide limits, and a third wire is taken for the purpose of making the desired measurement, the battery is "grounded" as shown in Fig. 153, the other bridge connections remaining the same. FIG. 152. Arrangement of bridge connections for locating faults in telegraph circuits. Wheatstone bridge measurements are divided into two general classes, usually spoken of as the "Murray" and the "Varley" methods. In making certain tests, the Murray method offers an excellent means of obtaining quick results. In other cases this method of testing is not as applicable, owing to FIG. 153, the requirement that the conductors under test must be of the same size and length, and of the same material. The Varley method is, generally speaking, more applicable to everyday telegraph requirements than is the Murray method. THE MURRAY LOOP TEST Referring to Fig. 154. Where it is desired to locate grounds or crosses in open lines or in cables, when the Murray method is employed it is not neces- sary that the ohmic resistance of the conductors under test be known, so long as both wires comprising the loop formed are of the same length, size, and kind. For the purposes of telegraph line testing, generally 1,000 ohms is the 172 AMERICAN TELEGRAPH PRACTICE best resistance to have cut-in in the a arm of the bridge. As shown in Fig. 154, the arm b is short circuited; the adjustable resistance (arm R) now in effect taking the place of the arm b. The testing battery is connected to the dividing point D, through the key BK; the galvanometer and its key GK, to the bridge terminals GW and FW. Good Wire I FW Fault T FIG. 154. Bridge circuits arranged for locating a "ground" by the Murray loop test. TO LOCATE A GROUND Loop the faulty wire with a good conductor of the same gage and length at the distant station. Connect the good wire to the terminal GW, and the faulty wire to the terminal FW. Obtain a "balance" by closing both bat- tery and galvanometer keys for a moment, repeatedly. At the same time adjust the resistance of the rheostat until there is no response of the galva- nometer needle to the manipulation of the keys. Then, the distance to the fault may be determined by applying the following formula: RXL x= B+R FW Fault FIG. 155. Bridge circuits arranged for locating a "cross" by the Murray loop test. Where X represents " distance" to fault. B represents resistance in a arm of bridge. R represents resistance in rheostat after balancing. L represents length of loop in feet or miles. To illustrate: Suppose the conductor under test is 3,700 ft. in length, then the VARLEY LOOP TESTS 173 length of the loop formed by two similar lengths would be 7,400 ft. If when a balance has been obtained it is found that the unplugged resistance in the rheostat is 42 ohms, and 1,000 ohms resistance in arm a, then the distance from the testing station to the fault is : 7400x42 = 310800 ft 1000+42 1042 CROSSES Should the fault be a cross instead of a ground, ground one of the crossed wires as shown in Fig. 155. CORRECTION FOR LEAD -WIRE RESISTANCE It is well to avoid the employment of connecting wires between the bridge and the conductors under test. When necessary to do so, use wires of the same gage, or of the same aggregate dimension as the conductors, and add the combined length of the connecting wires to the length of the loop formed by the conductors proper. After obtaining the result by the formula given, deduct the length of the short wire connected to the faulty conductor, and the remainder will be the distance to the fault. For in- stance, if in the previous example, it were necessary to use connecting wires 10 ft. in length, the formula would resolve into: (7400 + io + io)X42 7420X42 3 1 1640 _ 200. 1000+42 1042 1042 299-10 (length of one connecting wire) = 289 ft., distance to fault. VARLEY LOOP METHOD In those instances where faults are to be located on loops formed of conductors having sections of different dimensions, the Varley method may be used to good advantage. FIG. 156. Bridge circuits arranged for the Varley loop test. In Fig. 156 the faulty wire is looped with a good conductor at the distant station, after which the resistance of the loop thus formed is measured by the regular Wheatstone bridge method. When the resistance of the loop has been determined, the connections for the Varley test are made as shown in Fig. 156. 174 AMERICAN TELEGRAPH PRACTICE In practice the best results are obtained when the arm a has 10 ohms resistance, and the arm b 100 ohms. Obtain a balance by closing both keys repeatedly while the resistance of the rheostat is adjusted until the galva- nometer needle is not deflected. The resistance to the fault is determined by means of the formula: (Ri+R)b Where x represents resistance to fault. RI represents resistance of loop. R represents resistance in rheostat. a represents resistance in arm a. b represents resistance in arm b. To illustrate: If the loop has a resistance of 420 ohms and a balance is indicated when the rheostat has a resistance of 2,560 ohms, with a 10 ohms, and b 100 ohms, then the resistance to the fault will be: (420 4- 2560) X 100 298000 --- 2,560= - - 2,560 = 2,700 2,560 = 149. 10 + 100 no If the faulty wire is a No. 10 B. & S., gage copper conductor, it may be ascertained by referring to the table of wire gages (see tables in appendix) that its resistance is 5.28 ohms per mile at 60 F., which for all practical purposes is sufficiently accurate at all ordinary temperatures. If then the resistance to the fault (149 ohms) be divided by the resistance per mile, of the wire (5.28 ohms), the quotient 28.2 miles will be the distance to the fault. If short wires are used between the bridge and the conductors tested, ascertain the resistance of the short wire connected to the faulty wire and deduct its resistance from the total resistance to the fault. For example: had connecting wires been used in the above instance, and that connected to the faulty wire found to measure 2 ohms the end of the formula would be 149 2 = 147 ohms actual resistance to the fault. And T ^g = 27. 8 miles to fault. Obviously, any measurement made in the above described manner, may be checked for accuracy by reversing the individual "legs" of the loop in the bridge, and computing the resistance of, or the distance to the fault along the good wire and back along the faulty wire. In those instances where there are indications of current in the crossed wire due to foreign battery, which cannot conveniently be eliminated, remove the regular battery from the testing set and substitute a ground connection in its place. In making Wheatstone bridge measurements, the battery key should be closed a moment or so in advance of the closing of the galvanometer key ; RESISTANCE MEASUREMENTS 175 this to avoid momentary false indications of the needle. Also, care should be taken to have current in the bridge coils during the shortest possible time, in order to avoid charring of the silk insulation of the wire. TO MEASURE THE CONDUCTOR RESISTANCE OF GROUND RETURN CIRCUITS Arrange the bridge connections as shown in Fig. 148. Adjust the bridge arms, and balance as previously explained. To accurately measure the resistance of a wire, where two other wires between the same points are available, proceed as follows: Suppose it is desired to measure the resistance of the wire X, Fig. 157. l Measure separately the resistance of loops made up as follows : Y Wire X with wire F, Wire X with wire Z, _ Wire Y with wire Z, If the first loop measures 90 ohms, the second 93 ohms, and the third loop 99 ohms, then the total resistance of the three loops is 282 ohms. As each of the wires was used twice in making up the loops measured, the total resistance of the three wires measured once, obviously would be one-half of 282, or 141 ohms; and as the resistance of the. loop formed of the two wires Y and Z is known to be 99 ohms, the resistance of the wire X is obtained by deducting 99 from 141, leaving 42 ohms as the resistance of X. Once the resistance of one wire is known, the resistance of each of the other wires is determined by looping the wire of known resistance with the wire to be measured and then subtracting the resistance of the first wire measured from the resistance of the loop. METHOD OF LOCATING "OPENS" IN CABLES An excellent test in cases where the insulation is normal, may be carried out by connecting the bridge as shown in Fig. 1 5 8 . G is a source of alternating current. Where current from an alternating-current dynamo is not available, a small induction coil may be used to supply the desired current. Another convenient method of providing an alternating-current source, is to connect two double-contact relays, or transmitters, as indicated in Fig. 159, where a source of direct current is shown connected in series with a vibrating bell and the windings of two transmitters. A second source of direct current is con- nected with the local contact points of the transmitters in the manner illus- trated. As the circuit through the coils of the transmitters is continuously opened and closed, due to the operation of the vibrating bell, it is evident that the current sent out on the lines connected to the respective armatures of the transmitters, will be alternating in character. 1 "Examples from Postal Telegraph- Cable Co.'s book of instructions." 176 AMERICAN TELEGRAPH PRACTICE For cables approximately 1,000 feet in length, the alternating-current generator should have an e.m.f. of from 40 to 130 volts, and the arm a of the bridge may have a resistance of 100, or 1,000 ohms. The capacity of a con- ductor increases as its length is increased: the greater the capacity of the conductor under test, the lower should be the voltage of the testing battery, and the lower should be the resistance of arm a of the bridge. M Good Conductors M Broken or Open Conductor K (rood Conductor adjacent to Broken L Conductor. FIG. 158. Method of locating "opens" in cabled conductors. To locate a break in a cable conductor, pick out three good conductors having the same gage as the open wire and connect them as shown in Fig. 158. The conductors selected should have relations as shown in Fig. 158, that is, L must be adjacent to K, and N adjacent to M. To obtain a balance, open the four wires under test at the distant end, close keys GK and BK and adjust the resistance in the R arm of the bridge until no sound, or at least until minimum sound is heard in the telephone receiver when placed to the ear. D.C. FIG. 159. Convenient arrangement for supplying an alternating current for testing purposes where an alternating current dynamo is not available. Then the distance to the fault LXa R Where X represents the distance in feet, to fault, L represents the length of the cable in feet, a represents resistance in arm a, R represents resistance in arm R. THE BLAVIER TEST 177 For example: Suppose we have a cable 5,280 ft. in length, and that a balance of the bridge is obtained when the unplugged resistance in the R arm is 1,872 ohms, while a has a resistance of 100 ohms. Then the distance from the testing station to the fault is: 100 - = 282 ft. 1872 THE BLAVIER TEST A method known as the Blavier, sometimes used for locating a partial ground or an escape, where there is no good wire available for looping pur- poses, may be carried out as follows: Let r 2 represent the total resistance of the conductor under test (this must be known from previous measurement, obtained from a wire table, or calculated from the length, size and conductivity of the wire). Let R repre- sent the resistance of the wire with the distant end open, and ^ the resist- ance of the wire with distant end grounded. Then, the resistance from the testing point to the fault x = ri - By dividing x by the resistance per unit length of the conductor, obtained as above suggested, the distance to the fault is arrived at. If L represents the length of the conductor in feet, and r 2 the normal resistance of the faulty wire to the distant end of the line, the distance in feet to the fault _xL rz The accuracy of measurements made by this method depends upon the resistance of the fault remaining constant during each measurement. There are instances where the resistance of the fault is so high or so variable that the Blavier method is not reliable, and in general it is found that the Murray arrangement (Fig. 158) is more satisfactory, where additional good conductors are available for the test. THE FISHER LOOP TEST This test may be used in cases where there are two good conductors avail- able which terminate at the same point as the faulty wire. It is not necessary that the resistances of the conductors be equal, so the good wires used for the test may be in another cable, or may be open aerial wires. In Fig. 1590 the faulty wire C and the good wires D and E are shown con- nected at the distant station Y. It is then necessary to make two separate tests. First, one side of the battery is connected to the sheath as shown at 12 178 AMERICAN TELEGRAPH PRACTICE X. The resistance 6* is adjusted until the galvanometer needle shows no de- flection. The resistance values of R and 5 are noted. Then as in 159^ the battery is connected to the good conductor E, and a balance taken, the re- sistance values being noted as RI and Si. Then if L equals the length of the faulty wire, the distance to the fault is determined by the formula In cases where connecting wires are used between the testing set and the actual conductor connections, the connecting wire entering into the measure- c * V' f FIG. 159. Fisher loop test. ment is the one extending from the bridge to the faulty wire. In the Fisher tests, the same rules apply in locating crosses and shorts as in the Murray tests. ROUGH TESTS When a wire chief has become thoroughly familiar with the electrical and physical characteristics of the various line conductors in his division he is in a position to apply certain rough tests, which although they do not produce accurate figures, often serve to restore circuits quickly, especially where trunk-line facilities are limited. In switch-board parlance, each main-line circuit has its "feel," and a wire chief familiar with the peculiarities of a particular circuit, can ROUGH TESTS 179 tell by "feeling" it whether conditions are normal or abnormal. Consider, for instance, a line wire extending between two terminal stations 200 miles apart, and that there are 10 intermediate offices connected into the circuit, each intermediate office having inserted in the line a relay of 150 ohms resistance. If the circuit is opened at the distant terminal (by opening the line key or disconnecting the ground wire) then, provided battery is applied to the line at the home station, when the home key is closed at intervals, as in making "dots" slowly, there will be a pronounced "static" kick as current momentarily traverses the coils of the home relay, causing the relay armature to be attracted to an extent directly dependent upon the capacity of the line wire, and upon the number of relays included in the circuit. As the circuit is open it is evident that the effect on the home relay is due to the capacity of the line wire, which means that the longer the line and the greater the number of relays in circuit, the greater will be the force producing the kick of the home relay armature. Therefore when a ground return circuit opens at an un- known point, if the "kick" of the relay armature has about the same strength, or "feel" as when the line key at the distant terminal is opened, it is likely that the line is open at a point near the distant terminal. If, however, the kick is feeble, the line is open near the home station. And generally the strength of kick indicates to the tester the approximate distance to the fault, enabling him to call in an intermediate office near the fault on another wire. Thus, time is saved which otherwise would be consumed in tracing the fault from station to station along the line from the terminal office. Had the trouble which developed on the line been the result of an acci- dental ground contact instead of an "open," the wire chief's knowledge of the normal operating characteristics of the circuit enables him to apply a rough test to determine the approximate location of the ground. Under normal conditions a circuit has a regular e.m.f. applied to it. This, with the regular resistance of line plus relay resistance, permits of a definite current value in the circuit. Normal operating current produces what might be called "normal pull" on the armature of the relay. In the case under con- sideration, where the line wire is supposed to be "grounded" at an unknown point between the two terminals of the line, if it is found that the magnetic "pull" of the testing relay is abnormally strong, it is evident that the wire is grounded at a point not far distant from the home station. If on the other hand the pull is about normal, the ground in all probability will be found not far from the distant terminal of the line. And, in general, the strength of the current flowing through the home relay indicates to the tester approx- imately the distance to the accidental ground contact. When two wires entering the same switchboard become "crossed" some- where out on the line, it is not always immediately apparent which two are in contact. When a wire shows a cross with another circuit carrying current, the identity of the latter may be ascertained by removing the regular battery 180 AMERICAN TELEGRAPH PRACTICE from the first wire, and grounding that circuit at the home station through a test relay at the switchboard. The relay thus inserted in the line has its winding energized by a current which enters the wire at the " cross," and the procedure of the rough test is to open consecutively each wire which follows the same pole route, until one is found which when opened (thus removing its battery) opens also the wire in which the test relay is inserted. The point at which the cross exists may be located by having intermediate stations with this wire cut into their switchboards, open the circuit for a few seconds. If the test progresses from station to station away from the home station, the first station called in, whose open key fails to open the test relay, is, in fact, the first station beyond the cross, and the point at which the fault exists has been located between two certain stations. The above method of identifying crossed wires is applicable only where the wires involved take main battery at the testing station only, or in cases where main battery supplied at the- ^ "\^ distant terminal is temporarily removed from the wires affected. FIG. 160. When two wires are crossed, and until the cross is cleared by a lineman detailed for that purpose, one good circuit may be made up by having a station on each side of the cross (in each case as close to the cross as possible) open the least important circuit. This creates a condition such as that shown in Fig. 160, permitting one good circuit to be restored to service between the terminals of the line. ALL CIRCUITS INTERRUPTED It occasionally happens, due to storms, sleet, fire or other cause, that all wires on a route are at a certain point grounded, crossed, or open. Such a condition is usually referred to as a wreck. When this happens, the wire chief having jurisdiction over this particular section, is called upon to make good as many circuits as possible and as quickly as he can. In the language of the "board" he is required to "dig a hole through." The arrangement shown in Fig. 161 furnishes a means for obtaining quick results in case of a general wreck of wires. It may be seen that one side of the testing relay is grounded. The test requires that the other side of the relay circuit be connected in turn with each of the line conductors involved. While the relay is connected in series with a particular wire, all other wires are opened at the home switchboard, and bat- tery applied to the wire under test. The closing of the relay armature tongue indicates a cross between the wire connected through the relay and the wire to which the battery is applied. With this arrangement it is possible quickly to ascertain which wires are crossed, which open, which grounded, etc. ROUGH TESTS 181 After repairmen arrive at the wreck, and the restoration of circuits begins, it is of considerable advantage to employ an automatic arrangement at the testing office, which will announce to the wire chief when a fault has been cleared. The arrangement illustrated in Fig. 162 is extensively employed for this purpose. Suppose, for instance, that the wire shown connected to the relay through the main switchboard and spring-jack is grounded through a battery. Switches S and Si are thrown to the left, thus placing the vibrating bell in circuit through the Frc. 161. back contacts of relay R. The office first beyond G (the ground contact) is instructed to leave the grounded wire open. As long as the circuit remains grounded at the point G the relay R is energized and its armature remains in the closed position, which leaves the bell circuit open. As soon, however, as the ground at G is lifted, relay R opens, due to the fact that the circuit is open at the station first beyond G, and immediately the signal bell announces the removal of the ground contact. T FIG. 162. Wire chief's test relay and signaling bell connected to announce the removal of a ground contact or the closing of a break. Switches 5 and S\ are then thrown to the right, which places the regular sounder in circuit in place of the signal bell. Had the wire under observation been open tinsead of grounded, the distant terminal station would have been instructed to keep his end of the wire to ground until advised further, and the switches at the home station would have been disposed, 5 to the right, and S\ to the 182 AMERICAN TELEGRAPH PRACTICE left. This provides that when the repairman closes the break, relay R will be energized as a result of completion of the circuit from the home battery to the ground at the distant terminal of the wire. In this case, with the switches disposed as above stated, the vibrating bell sounds the closing of the break. VOLTMETER TESTS Voltmeters haying a self-contained series resistance of about 2,000 ohms per volt, are used to a considerable extent for line-testing purposes. The various circuit arrangements employed in practice are shown herewith. MEASURING A GROUND CONTACT Connect the voltmeter as shown in Fig. 163, with one side of the testing battery to ground. A permanent deflection of the voltmeter pointer indicates a "ground." To ascertain the value of the resistance to ground, note the FIG. 163. Voltmeter method of measuring a ground contact. reading in volts with key K open. Call this figure V. Close key K, note the altered reading in volts, and call it Vi. Then, where R is the resistance of the voltmeter, the resistance to the ground contact on the line MEASUREMENT OF HIGH RESISTANCE As in Fig. 164, connect the resistance to be measured at X and close the switch K (thus short circuiting X) and note the deflection of the voltmeter I 1 1 iwwv- l~ X FIG. 164. Voltmeter method of measuring high resistances. pointer. Call this Vi. Open the switch K and note the altered deflection. Call it F 2 . Then the resistance value desired \/" * ~D A = f^ - R representing the resistance of the voltmeter. TESTING WITH VOLTMETER 183 CAPACITY TEST Connect the voltmeter with a standard condenser C as shown in Fig. 165. First, move the switch to position i, and note the throw in degrees of the instrument pointer. Call this figure Fi. Then place the switch lever on 2 FIG. 165. Voltmeter method of measuring the capacity of a conductor. and again note the deflection of the pointer. Call this figure F 2 . Then, the capacity of the line =C ~> where C is the capacity of the standard condenser in microfarads. MEASURING ORDINARY RESISTANCES For the purpose of measuring ordinary resistance values, such as instru- ment windings, lines, etc., connect the unknown resistance at X as shown in Fig. 1 66. Shunt the voltmeter with a resistance S having a value such that -/vwwv FIG. 1 66. Voltmeter method of measuring ordinary resistances. the combined resistance of the voltmeter and the shunt will have some con- venient value, say 200 ohms. (See page 86 for calculating shunt values.) The measurement is then made in the same way as in the case of a high resist- ance, and the value is Fi-F 2 X= ^ 200. ROUGH VOLTMETER METHOD OF LOCATING A CROSS Several ingenius arrangements have been suggested from time to time, with the object of developing a satisfactory voltmeter loop test. It is found in practice, however, that the voltmeter is not as adaptable for making accurate measurements with looped conductors as is the bridge method previously described. Referring to Fig. 167. If the potential at the point L where the line con- 184 AMERICAN TELEGRAPH PRACTICE ductor enters the switchboard is 150 volts, obviously there is a gradual drop of potential along wire A until the point G\ is reached where the potential has fallen to zero. If a cross between wires A and B exists at the point F the drop of potential at that point may be ascertained by connecting the volt- meter in series with the home end of the wire B and ground. Then the distance from the testing station to the fault: Y _C-DE C Where C represents the voltage at L D the voltage at F (or P) E the distance from L to G. The wire B must be left open at the distant terminal station, or at an office beyond the cross. It is evident that errors will be introduced, due to any FIG. 167. Voltmeter loop test. difference of potential from earth currents between the ground connections G and Gi, to possible high resistance at the fault F, to the resistance of the wire B, and to leakage. The extent to which these factors introduce error is dependent upon the resistance of the voltmeter. If a meter having a re- sistance of 2,500 ohms per volt is used for the test, results are fairly accurate. INSULATION RESISTANCE OF LINES With any form of construction commercially practicable, perfect insula- tion is not possible. On open aerial lines although electrostatic and electro- magnetic induction takes place, there is no current leakage from wire to wire through the air, but at every point at which wires are supported, even with the best construction there will be some leakage from wire to wire and from wire to ground. At every pole there exists a leak to earth. The electrical resistance of this leak is high if the wire is well insulated, and low if the insulation is poor. At the point of support the wire is separated from the cross-arm or pole by an insulator, and the effective insulation of the line is dependent upon the construction, shape, material and condition of these insulators; also upon the INSULATION RESISTANCE OF LINES 185 space along the cross-arm separating the insulator from the pole. Glass of certain grades offers the highest insulation to electrical conduction through its mass of any commercially available material. For the purposes of telegraph insulation, glass does not ideally meet the rquirements, due to the fact that surface conduction plays an important part in leakage from line to wooden support. Glass is highly hygroscopic, and in almost every state of the weather and of the atmosphere it becomes coated with a film of moisture 1 or of gross matter. Certain grades of porcelain, in this regard, meet the requirements more satisfactorily, as porcelain is not as hygroscopic as glass, and rain runs readily from its highly glazed surface. Many attempts have been made to explain the peculiar behavior of leak- age of electric currents over the surface of insulators on which moisture has condensed due to exposure to ordinary atmospheric conditions. Whether the potential applied to the conductor is of one polarity, or is alternating from positive to negative continuously or occasionally, seems to play an important part in varying the electrical resistance of the film* of moisture deposited. In some cases it is found that the resistance is enormously greater when the current passes in one direction than when it passes in the opposite direction. Apparently the nature of the oxide formed on the con- ductor as a result of electrochemical action between the metal of the con- ductor and the moisture film, undergoes a change as the current in the conductor is reversed. Duration of contact of either polarity, as well as rapidity of reversal, probably are the factors which determine the resistance between the wire and its support, across the surface of the insulator, assum- ing, of course, that a film of moisture is present. The hygrometric state of the surrounding atmosphere, varying, as it does, naturally accounts for the variations in the thickness of the film of moisture, and this in turn has a direct bearing upon the initial resistance when battery is applied to the line, irrespective of polarity. The American Telegraph and Telephone Company has considered a clear weather insulation of 10 megohms per mile as satisfactory. The Western Union Telegraph Company has a standard of 50 megohms per mile, while the Postal Telegraph- Cable Company aims to maintain an insulation of 100 megohms per mile in clear weather. Wet weather conditions, however, greatly reduce these figures, and when a drizzly rain and fog prevails for any considerable length of time, it is found that the insulation resistance of lines may drop lower then one megohm per mile. A low-lying dense fog has a most pronounced effect in reducing the insulation resistance of a line, and the hygroscopic characteristics of glass insulators are clearly evidenced when 1 The large amount of common salt (chloride of sodium) floating about in the form of fine particles in the air results in condensation upon the surface of all exposed bodies. Where these deposits are made upon the surface of insulators, the saline film thus formed is a much better conductor of electricity than is the insulator. 186 AMERICAN TELEGRAPH PRACTICE wires are thus weather-bound, by the fact that during a dense fog, should there be a fairly heavy rain-fall lasting a few minutes, the insulation resistance of the line rapidly increases. So much so, that while the rain continues the insulation has been known to closely approach clear weather values. It is hardly likely that the dripping of the rain from the surface of the insulator produces a hydro-kinetic effect which clears the moisture condensed on the inner surface of the petticoat insulator, so that, so far as investigation accounts for the phenomena observed, nothing is explained except that the exterior surface of the insulator has been washed clean. In addition to insulator leakage, other causes bring about a lowering of insulation resistance, such as contact between wires and limbs of trees, kite strings (the latter when wet sometimes causing leakage from wire to wire), broken insulators, permitting direct contact between wire and cross-arm, surface leakage along the surface of bridle wires resting against cross-arms, etc. All of these avenues of escape are more effective as leaks during wet weather. The foregoing has been introduced at this time so that an understanding of the various causes which permit leakage of current may be gathered. Mil-AM-Meter B C FIG. 168. The insulation resistance of a line wire may be measured by ascertaining the resistance at the terminal A while the distant end X is open, or insulated, as in Fig. 168. Under such conditions, the insulation resistance observed does not equal the sum of the several insulation resistances from B to G\, C to Gz, D to 6*3, E to 4, and F to G 5 . Correctly considered the insulation resistance ob- served is that of the circuits ABGi, BCG 2 , CDG 3 , DEC*, and EFG 5 , and at once it is apparent that as each pole or support unavoidably constitutes a leak to earth, the total insulation resistance of the line has the same relation to the joint-conductivity of the various leak paths to ground, as obtain in all other problems concerning joint-conductivity, and which have been considered in an earlier chapter. ME A S URING INS ULA TION RESIST A NCE 1 87 MEASUREMENT OF INSULATION RESISTANCE The voltmeter is used quite extensively in making insulation measure- ments, but it should be remembered that, inasmuch as the resistance to ground Vr-V* X= y R } as explained in connection with Fig. 164, the resistance per volt of the meter is of the first importance. A voltmeter having a range of 100 volts, and a resistance of 100 ohms per volt, could not be used satisfactorily in measuring resistances as high as one megohm. The highest resistance that can be measured with such a meter is TOO- i - i x = ~~ 10,000 = 990,000 ohms. It follows that with a loo-volt meter having a resistance of 206 ohms per volt, the highest resistance which can be measured is 1,980,000 ohms. INSULATION RESISTANCE MEASUREMENTS WITH MILAMMETER One method of measuring the insulation resistance of lines, which has been used with success, makes use of a milammeter in connection with the quad- ruplex "long-end" potential of 375 volts negative, as shown in Fig. 169. M. A. Meter 375-T H?r rl Line FIG. 169. Measuring the insulation resistance of a line. Milammeter method. To measure insulation resistance, the lower scale of a standard milammeter (five divisions equal to one milampere) is used. The meter is inserted in series with a 25,ooo-ohm resistance unit and connected directly to the line to be measured, as shown. The 25,ooo-ohm coil serves to protect the meter from damage in case the line under test is grounded close by. Its presence also minimizes the effects of induced currents, and this results in steadier action of the indicating needle. With a potential of 375 volts and a resistance of 25,000 ohms the milammeter needle travels exactly the full length of the scale, in accordance with Ohm's law. The insertion of any additional resistance, such as that of a line, reduces the amount of deflection. To minimize the work in computing the insulation resistance in ohms', the reference table given herewith, is used. The column of figures at the extreme left, reading from 30 to 500, refers to length of line in miles, while the row of figures at the top, reading from i to 75, refers to divisions deflection on the 188 AMERICAN TELEGRAPH PRACTICE 10 O o O o o O NO H * *, * - H M H M M d d M CO CO co O 10 , * !* H - d d d CO CO CO -* * to NO NO to "* M M d d CS CO co CO * * 10 t- 00 ON O CO CO * * 10 NO - 00 ON M M H to CO H M H d CO * * to NO - OO o CO M M M H d CO * to NO - 00 00 ON H CO 10 ^ CM d CO * to 00 ON H CO 10 00 o d ro 10 CM CM CM CM d *> v 00 ON H d ^ 10 00 M ,j. ^ ?> " ;f O d g 10 tx M Jt CO H CO CO M CO *> * * H CO 10 tj. ON M VO CO WJ 00 CO to M CO * V) 00 CO 10 00 CO 10 10 o O M * *> 00 w CO tj. M .0 ON *. 10 d 00 O to ** 00 s M d oo d CO CO M ? 10 10 co M 00 0\ to t*. -o M oo M oo d d CO co M * to to N? 5 - Line 200 Miles 1000 Ohms AL 1000 Ohms several thousand ohms and where differential relays are used in duplex operation, in order to insure that equal current values obtain in each coil of the relay when the home bat- tery is applied to the line and the FIG. 218. Resistance of the artificial line distant end of the line is grounded, balancing the resistance of the line wire to it is nec essary to have at the home distant station. ,. , . A , . station an adjustable resistance through which the other coil of the relay may be connected to ground. Obviously if this resistance is adjusted to have a value equal to that of the line wire to the distant station, like current values will exist in both coils of the relay and there will not be any magnetism produced in the cores of the relay. The adjustable resistance used to equate the resistance of the main-line wire is generally called the artificial line, Fig. 218. THE ARTIFICIAL LINE As has been pointed out elsewhere in this work, all line wires possess electrostatic capacity. The quantity of electric charge accumulated upon the surface of the conductor depends upon the superficial area of the con- ductor, upon the distance intervening between the conductor and the earth (or between the conductor concerned and other conductors in electrical contact with the earth), and upon the nature of the insulating medium inter- vening between the line wire and the earth. In any line of considerable length, a portion of the current is bound up in the form of static charge. The first rush of current into the line at the instant the battery is applied thereto, (sometimes referred to as the current of charge) for an instant pro- duces a much greater magnetic effect upon the armature of the home relay, than obtains when the entire line has been fully charged and permanent con- ditions established in the circuit. The result of the initial inrush of current, greatly exceeding in volume, as it does, the final current, is that a false signal or "kick" of the relay ar- DUPLEX TELEGRAPHY 253 mature is produced. The energy of the kick depends upon the electrostatic capacity of the line, being greater where the capacity is high, and less pro- nounced as the static charge taken on by the line wire is less. Also, there is to be considered the effect of static discharge which occurs at the instant the line wire is shifted from the battery connection to the ground connection upon opening the key controlling the operation of the trans- mitter. At this instant the electrostatic charge which has been accumulated upon the surface of the conductor flows back to ground by way of the ground contact of the transmitter, passing through the main-line coil of the differen- tial relay, again producing a kick of the relay armature. In view of these considerations, therefore, it is necessary if the false signals which are produced at the beginning and the end of each intended signal are to be neutralized or nullified, that the artificial line be made to possess properties identical with those of the main-line wire, i.e., resistance and capacity. The application of the electric condenser as an adjunct of the artificial line gives to the latter the desired property of electrostatic capacity. A condenser path to ground via the artificial-line coil of the differential relay results in an initial rush of current through that coil at the instant battery is applied to the line, which, by means of adjustable ''timing" resistances in series therewith may be made to exactly equal in strength and duration, the corresponding rush of current which takes place at the same instant through the main-line coil of the relay, thus at the critical moment insuring identical current values in both coils of the relay. And, further, when the line wire is shifted from battery contact to the ground connection at the moment the key is opened, the discharge from the condenser associated with the artificial line takes place through the relay coil forming a portion of the artificial-line circuit at the same instant that the main line discharges through the relay coil forming a portion of the main- line circuit, thus again at the critical moment insuring equal current values in the two coils. .To understand the import of the above remarks, one must have in mind the positions of the main-line circuit and of the artificial-line circuit through the windings of the respective relay coils, also that the magnet made up by the artificial-line relay coil, and the magnet made up by -the main-line relay coil both control the same armature. When the relay is operated by current from the distant station its opera- tion is due to a surplus of current in the main-line coil over what may be in the artificial-line coil of the relay. When the signaling keys at each end of the line are closed and like poles of battery are applied at both ends of the line, the desired signal is made by the home battery on the home relay, and is the result of a surplus of current in the artificial-line coil of the relay over what may be in the main-line coil. 254 AMERICAN TELEGRAPH PRACTICE When, due to electrostatic charge or discharge of the main line the current in the main-line coil of the relay is augmented above that traversing the artificial-line coil of the relay, a false signal will be produced unless at that instant the current flowing in the artificial-line side of the relay is increased to an equal value. This is what is accomplished by using condensers and retardation resistance coils in connection with the artificial line. Figure 219 shows the theoretical arrangement of the artificial-line circuits. On the right is shown three adjustable resistance groups used in balancing the "ohmic" resistance of the main-line wire. With the values shown it is possible to avail of resistances ranging from 10 ohms to 11,100 ohms, variable within steps of 10 ohms. Main Line Ten 1000 Ohm Steps FIG. 219. Circuits of the artificial line, showing the adjustable resistance on the right and the condensers and condenser timing resistances on the left. On the left, two adjustable electric condensers C 1 and C 2 are shown, each having a maximum capacity of 3 microfarads. Each condenser has in series with it an adjustable resistance (see Timing the Condenser Discharge). DOUBLE -CURRENT DUPLEX SYSTEMS As a result of the development of more efficient and satisfactory duplex systems, the single-current duplex is rarely used in this country, except where it is combined with the polar duplex in forming the differential quadruplex system of telegraphy by means of which two messages are sent in each direction over a single wire, simultaneously. Further treatment of the single-current duplex will be deferred until it is considered as a component part of the quadruplex. DUPLEX TELEGRAPHY 255 THE POLAR DUPLEX The essential elements of the polar duplex are a battery pole-changer, a differentially wound polarized relay, an artificial line rheostat, and an artificial capacity. THE POLE-CHANGER The transmitter shown in connection with the single-current duplex, Fig. 215, has connected to one of its contacts the positive pole of a main- line battery and to the other contact a circuit to ground. If to the latter tne negative pole of a main -line battery were connected instead of the ground wire, closing the signaling key would send to line a positive impulse, and opening the key would send to line a negative impulse, in which case the trans- mitter might correctly be regarded as serving as a pole-changer, inasmuch as the polarity of the battery placed in contact with the line wire changes from positive to negative and vice versa each time the transmitter tongue is caused to break contact with the positive battery terminal and make contact with the negative battery terminal. " When dynamo currents were introduced in the operation of telegraph lines it was found that the form of transmitter here considered, and which had previously answered the requirements where gravity batteries were universally employed in telegraph work, failed to give satisfactory results owing to the momentary short circuit which exists when the line contact is shifted from the lever to the tongue of the transmitter, and again when the opposite movement takes place, in the act of signaling. Irr* the Stearns duplex the resistance presented to the individual battery used, was, at the instant the transfer of contact took place, made approximately equal to that of the total internal resistance of the battery. When the dynamo with its negligible internal resistance is applied to the operation of the polar duplex, two machines of equal potential and of opposite polarity are separately con- nected to the contacts between which plays the armature tongue carrying the main-line contact. At the instant, therefore, that the transfer of con- tact takes place unless there is an appreciable length of air-gap main- tained between the opposing battery terminals, there will be established a momentary short circuit between the two dynamos (where two 2oo-volt machines are employed this amounts to an aggregate potential of 400 volts) which might result in serious damage to the machines. Even where arti- ficial resistances are inserted in series with each machine, excessive spark- ing occurs when the line contact is shifted from one pole to the other. The introduction of the double-current duplex called for the substitution of a transmitter in place of the type of instrument used with the single-current duplex, which would meet the changed conditions. 256 AMERICAN TELEGRAPH PRACTICE The new form of transmitter, or pole-changer as it has since been called, provides for the maintenance of ah air-gap as the main-line contact is shifted from one pole of the battery to the other. One form of pole-changer employed by one of the commercial telegraph companies, is that known as the walking-beam pattern, see Fig. 220. Another type of pole-changer extensively employed is that illustrated in To Line FIG. 220. Walking-beam pattern of pole- changer. FIG. 221 . Double-contact relay type of pole- changer. Fig. 221, which will be recognized as a simple double-contact relay form of instrument, a photographic view of which is shown in Fig. 222. With either of these instruments, it is evident that connection cannot be made between the line and one dynamo until contact has first been broken between the line (the lever) and the other dynamo connection. FIG. 222. Pole-changer. So far as the polar duplex is concerned, the same necessity does not exist for the employment of a continuity-preserving transmitter as was the case with the single-current duplex, the reason for which will be explained presently. The combination of the polar duplex with the Stearns duplex, in forming the differential quadruplex, transferred to the latter the alternative of using DUPLEX TELEGRAPHY 257 on the single-current half of the system a continuity-preserving transmitter, or a pole-changer type of transmitter which at all times maintains a small air-gap between the opposing dynamo terminals. From what has been stated in regard to the possibility of short circuits where the continuity- preserving instrument is used in connection with dynamo machines, it is apparent that the employment of the latter mentioned instrument is not practicable. In practice, therefore, it is customary to use an open-gap trans- mitter on the single-current side of the quadruplex, similar to that used on the double-current side. In the operation of the single-current side of the quadruplex, the objec- tions previously mentioned in connection with the employment of the air-gap transmitter in Stearns duplex operation, still exist, as it is evident that there are constantly recurring periods of "insufficient current" while the lever of the transmitter (to which the line conductor is connected) is traveling from one battery contact to the other in the act of signaling. Inasmuch, however, as the employment of the open-gap transmitter is imperative, it has been necessary to avail of other means of bridging over these periods, and to employ circuit accessories which act to prevent, or at least to minimize the tendency to produce false signals on the reading sounder operated by the receiving relay on the single-current side of the quadruplex. These accessories are variously referred to as "bug-traps," " uprighters, " etc., and their application and action will be described in connection with quadruplex systems. THE " POLAR" RELAY All of the inherent difficulties experienced in the operation of single Morse lines, are encountered in the operation of the single-current differential duplex system. During favorable weather and where a high degree of line insulation is maintained, both of these methods of telegraphy are satisfactory. But, when, due to excessive leakage conductance the current values at the re- ceiving end are low, considerable difficulty is experienced in maintaining satisfactory operation. The polar duplex overcomes this difficulty to a great extent, and by means of this system lines may be worked satisfactorily long after adverse "weather conditions have rendered single Morse, and single-current duplex systems inoperative. Figure 223 shows a theoretical view of the magnetic circuit of the polar relay. It will be seen that the windings are identical with those of the ordinary single-current differential relay. Current from the battery flows through 17 258 AMERICAN TELEGRAPH PRACTICE Line the windings in opposite directions, the action of one coil neutralizing that of the other, the result of which is that the core is not magnetized so far as any action due to the current from the battery is concerned. The fundamental difference between the two instruments is that in the polar relay the tongue is held on either side due to the magnetic pull of the permanent magnet which constitutes the cores of the electromagnets. In the case of the common differ- ential relay the armature tongue is held in the closed position by the action of either or both magnet coils, and in the open position by the action of a retractile spring which with- draws the armature from the closed position when the coils are not ener- gized. The armature of the polar ?%'Li'ne relay is held in the closed position and in the open position by the at- i- -i- traction of one pole of the permanent FIG. 22 3 .-Theory of the differential polar- magnet, and it is necessary of course ized relay. that the armature be drawn into contact with the open or the closed pole due to magnetism in the cores, resulting from the action of current in either coil of the instrument. The important feature is that after the arma- ture has once been attracted toward either contact, it will remain there whether current remains in the coil winding or not (provided there is no current in the opposite coil). Referring to Fig. 223 : When the key K is operated, the armature lever of the pole-changer is caused to make contact, first with the negative battery terminal and then with the positive battery terminal. If the ohmic resist- ances of the real line and the artificial line are equal, current from whichever dynamo is con- nected with the armature lever will go through the companion windings of the relay differentially, with the result that there is no electromagnetism produced in the cores facing the relay armature. It matters not whether the out-going current is from a positive source or from a negative source: owing to the fact that it passes through the windings of the relay differentially there will be no mag- FIG. 224. Permanent magnet and armature suspension of the Siemens polar relay. DUPLEX TELEGRAPHY 259 netism produced, and this, irrespective of the polarity of the current flowing in the circuit. If the key K is manipulated, there will be sent out a series of impulses alternating in sign, from positive to negative each time the key is closed and opened, and if the resistance of the artificial line side of the relay balances that of the line side, the armature of the relay will not be affected. More- over, it will be found that if the relay tongue is moved by hand into contact with its closed contact or with its open contact, it will still remain passive to the out-going reversals from the pole-changer. In one of the older standard patterns of polar relay, which is known as the Siemens, or " camel-back" relay, Fig. 224, a comparatively large per- manent magnet has mounted on one end two cross-pieces made of soft iron which form the cores C and C' of the main-line and artificial-line coils. From the other extremity of the permanent magnet the armature A is sus- pended, being pivoted in a brass casing, that is, it is pivoted in bearings, which, being non-magnetic, introduce a gap between the pole-face of the large permanent magnet and the armature. The latter is magnetized inductively across the existing air-gap, to a degree sufficient to create the desired attrac- tion between the free end of the armature and the cores of the magnets, both of which are of identical polarity, and opposite to that of the extremity of the permanent magnet at which the armature is pivoted. In the completed instrument, the cores C and C', carry the coil windings of the main-line mag- net and the artificial-line magnet respectively. If it is assumed that the end B of the permanent magnet at which the armature is pivoted is the north pole, then the free end of the armature is also of north polarity, and, owing to the fact that both cores, of the electro- magnets are attached to the south pole of the permanent magnet, and taking that polarity, it is evident that the armature will cling to whichever core it may be placed in contact with. That is, it will cling to either the open contact, or to the closed contact when no current traverses the windings of the electromagnets, or when current flows through the windings differentially. The magnets of polar relays are usually so wound that when current from the distant station flows through the main-line coil, it is given a path through an auxiliary winding in the opposite coil, in the reverse direction (Fig. 225), which results in the permanent induced magnetism in one of the cores being neutralized, while the magnetism existing in the other core is intensified, causing the armature to be attracted toward the opposite con- tact. The reverse action takes place when the battery poles at the home station and at the distant station are in opposition (like poles to line) in FIG. 225. Coil windings of differential polarized relay. 260 AMERICAN TELEGRAPH PRACTICE which case the artificial-line coil of the home relay has its magnetism in- creased, and the line coil has its magnetism neutralized. Thus, due to the action of the current in the coils, the armature is caused to move into con- tact with the open or the closed contact as desired. The office of the auxiliary winding in each case is to act as a " clearing out" agency. There are several distinct types of polar relay used by the various tele- graph administrations, each relay having its peculiarities of design, but the principle upon which all polar relays operate is the same. Figure 226 illustrates the form of polar relay used by the Postal Tele- graph-Cable Company. It differs from the type of instrument previously described in that the permanent magnet used to magnetize the armature, or rather the "armatures" in this case, is situated under the base of the instrument. FIG. 226. Type of polar relay used by the Postal Telegraph- Cable Company. The permanent magnet is mounted under the base. Among the other types of polar refay in use might be mentioned the "Krum" relay, which instead of employing permanent magnets to hold the armature in connection with the open or the closed contact of the local sounder circuit, has an extra pair of magnets, one beside the main-line and one beside the artificial-line magnets, which are constantly charged from a separate source of current, serving the same purpose as the permanent magnet used in other types of relay. Another efficient type of instrument is that known as the Wheatstone polar relay, in which the pivot of the vertical armature rests on one end, thus effecting a considerable reduction in the mechanical inertia of the moving element. Also, the magnet coils are somewhat longer than in the ordinary types of relay. The windings have a comparatively high resistance, but as they are connected in multiple for high-speed work, the total resistance is reduced to one-fourth, and the time-constant of the relay, as an indirect result is also reduced. In this type of relay there are two armatures, both mounted on a common shaft, and so situated that their lower ends are under the influence of a permanent magnet. Each relay is equipped with DUPLEX TELEGRAPHY 261 FIG. 227. Binding-post and internal connections of artificial line rheostat Postal type. 262 AMERICAN TELEGRAPH PRACTICE four magnets, and when current traverses the windings; two poles repel and two poles attract the armature. The artificial line rheostat, and the artificial capacity used in connection with polar duplex apparatus to "balance" the resistance and capacity of the actual line are illustrated theoretically in Fig. 219. Figure 227 shows the actual internal and external connections of the rheo- stat and the condensers which make up the artificial line. The particular type of rheostat illustrated is that used by the Postal Telegraph-Cable Company. It will be seen that the artificial line side of the polar relay is connected to the binding-post L of the rheostat, from which point there is a circuit to ground, via the resistance coils, marked "tens," "hundreds," and "thou- sands, " by means of which the total resistance of the artificial line may be varied from zero to 11,100 ohms in order to balance the resistance of any line wire which may be connected into the set. It will be noticed also, that from the binding-post L there are two condenser circuits to ground, the first and second condensers being in series with variable timing resistances. The first and second group of resistance coils are connected with binding- posts C 1 and C 2 respectively, via rheostat arms which may be moved from one contact button to another; so to insert any desired value of timing resistance in series with the condensers. Each of the condensers has a capacity of 3 microfarads, and as they are connected in parallel, there is available a total capacity of 6 microfarads with which to balance the static charge and discharge effects of the actual line. The capacity of the condensers being variable, any desired capacity may be obtained simply by turning a knob mounted on the top of each condenser. The "ground" switch shown to the right of the transmitter, or pole-changer, when thrown to the right, places the home battery to line. For the sake of clearness the main-line battery connections are omitted from Fig. 227, but it is understood that when the armature tongue of the transmitter is in the closed position as shown, main-line battery of one polarity is connected to line, and when the tongue is in contact with the back-stop, battery of the opposite polarity is connected with the line by way of the ground switch and the polar relay. When the ground switch is thrown to the left, it is evident that the home battery is disconnected from the line, and that the incoming signals after passing through the relay have a path to ground via a 600 ohm resistance coil' contained within the rheostat box. The location of this ground-coil should be kept in mind, as, presently we shall again refer to it in connection with "balancing." OPERATION OF THE POLAR DUPLEX Figure 228 shows the connections of the main-line and local circuits of the polar duplex. DUPLEX TELEGRAPHY 263 Complete equipment at both ends of a duplexed circuit are shown so that the various operations may be treated with regard to their effects upon the apparatus at both ends of the line. PC and PC' are the pole-changers at stations A and B respectively while PR and PR' are the polar relays, K and K r the signaling keys, locally controlling the movements of the pole-changer armature in each case. The dynamos which furnish current for the operation of the main-line relays, are shown, two at each end. In each case one of the dynamos has its positive terminal connected with the back-stop of the pole-changer, while FIG. 228. the other dynamo has its negative terminal connected with the closed-con- tact of the pole-changer. The resistance coils and condensers which comprise the artificial line, are, at each end of the main-line circuit; marked AL. In Fig. 228 the pole-changers at each end of the line are closed, that is, the armature levers of the pole-changers in each case are in contact with their front-stops, due to the fact that the signaling keys K and K' are closed, This places to line at each end a 200 volt negative battery. As the batteries are of equal potential, no current will flow over the main line. At the instant both pole-changer armatures make contact with their back-stops thus placing opposing battery to line the levers of the polar relays at each end are moved into contact with their back-stops, thereby opening the local reading sounder circuits at each end. At this point it is important to gain a correct understanding of why the armatures of the polar relays at each terminal station are attracted toward either contact when the main-line batteries at each end of the line are in opposition. The explanation is that when the terminals of a wire are at 264 AMERICAN TELEGRAPH PRACTICE equal potentials, no current will flow in the wire. Therefore, when like poles of identical potential are to line, as in the case before us, it is apparent that the terminals of the main-line wire are at equal potential. An entirely different condition, however, exists with regard to the artificial line at each end. As, in each case, one end of the artificial line is connected with the earth (which is at zero potential), there is presented to the outgoing currents from each station a path to ground via the artificial line magnet of the polar relay. On each occasion, therefore, when like poles are to line at each end, current from the home battery flows through the artificial line and the armature of the polar relay is attracted toward its back-stop if the opposing batteries are positive and toward its front stop if the opposing batteries are negative. In order to carry on transmission in both directions at the same time it is necessary that the operator at A shall be able to control the movements of the armature of the relay at B regardless of which pole of his battery B has to line. Also that the operator at B shall be able to control the movements of the relay armature at A regardless of which pole of his battery A has to line. Suppose the operator at B should depress his key (while the key at A is open), thereby placing the tongue of his pole-changer in contact with the negative pole of the main-ine battery at B, the result will be that the main-line coil of the relay at A will be energized and its tongue attracted toward its closed-contact, thereby operating sounder S. It is evident, of course, that current continues to flow through the artifi- cial line coil of the relay at A, but owing to the fact that the current strength in the main-line coil of the relay is twice that in the former, and in the opposite direction, it is plain that the magnetism in the core of the relay at A is reversed, and the armature, as a result thereof, moves into contact with its front-stop. If what has previously been stated is true, the armature of the relay at B should have remained passive to the reversal of current sent out from B when the key at B was closed. That this is so is apparent, for, although the magnetism in the artificial line magnet of the relay at B has now been neutralized due to the presence of current in the main-line coil of the relay, the armature is held in the open position by the action of the permanent magnet associated therewith. In other words, nothing has happened so far to cause the armature of the main-line relay at B to change its position, therefore, it remains in the position taken when last it was caused to move by a surplus of magnetism in one coil over that obtaining in the other magnet coil. Similarly, when A alone closes his signaling key, the relay at B responds, while the relay at A does not. When the signaling keys at both ends are depressed, the line currents once more are in opposition, and, as in this case the currents flowing through the artificial lines at each end are in the reverse direction of that taken when both keys were open, the relay armatures at each end are caused to move into contact with their front-stops. DUPLEX TELEGRAPHY 265 In effect, therefore, when the operator at A attempts to register a "dot" on the relay at B, at the same instant that the operator at B intends to register a "dot" on the relay at A, each station causes to be produced in his own relay the signal intended to be transmitted from the distant end of the line. Or, the foregoing might be paraphrased thus: the relay at A will be closed whenever the key at B is depressed, regardless of whether A is sending or idle; and the relay at B will close whenever the key at A is closed whether B is sending or idle, but in neither case will the signals transmitted from either end conflict with those originating at the distant station. SEVERAL DUPLEX SETS WORKED FROM ONE PAIR OF DYNAMOS In the duplex circuit diagrams heretofore given, two dynamos have been shown at each station as an integral part of each set. It should be under- stood, however, that in practice, two machines, one delivering a positive potential and the other a negative potential, are used to supply current for a number of lines. In Fig. 229, two 200- volt dynamos of oppo- site polarities are shown connected to separate busbars. Instead of four wires leading there- from to pole-changers of duplex and quad- ruplex sets, any number of sets may be con- nected thereto, depending upon the capacity of the dynamos employed. The four wires shown in Fig. 229 leading from the positive busbar are connected to the back-stops of four different pole-changers, and the four wires leading from the negative FIG. 229. Several duplex sets busbar are connected to the front-stops of the worked from one pair of dynamos, same four pole-changers. One pair of ma- chines, therefore, serves to operate four or more duplexes. Each separate branch is fused, and has in series with the fuse F, a protec- tive resistance coil C, or a lamp which in any case may have a resistance of 200, 300 or 600 ohms, depending upon the value desired in any circuit that may be connected thereto. -200 "LOCAL" CIRCUIT CONNECTIONS In the preceding text-matter describing duplex-circuit operation, in several instances reference is made to the "open" and "closed" contacts of relays, transmitters and pole-changers. It might be here stated that instead of employing one dynamo to operate 266 AMERICAN TELEGRAPH PRACTICE each pole-changer, each sounder, etc., one machine having an output of 6 volts, 24 volts, 40 volts or any desired e.m.f. (depending upon the resistance of windings of local instruments, and upon the current values desired) may be availed of to feed a large number of such circuits. Figure 230 shows four separate leads from a 40-volt positive busbar, two of which are shown connected to local circuits. The upper wire is connected through the magnet winding of a pole-changer, via a circuit controlling key. The opposite terminal of the winding is connected to ground through a current regulating resistance coil C which may have any desired value. Inasmuch as the negative terminal of the dynamo is permanently grounded, closing the signaling key establishes a completed circuit through the winding of PC Relay C; FIG. 230. Several duplex and quadruplex local circuits operated from a common source of e.m.f. and ground coil C, thus energizing the magnet of PC, and causing the arma- ture tongue to move into contact with the front-stop, or closed pole. When the key is opened the circuit is broken, permitting the retractile spring to pull the armature tongue into contact with the back-stop or open pole. The second wire (Fig. 230) leading from the busbar of the local circuit dynamo is shown connected with the armature tongue of a relay. When the electromagnet of the relay is energized due to the presence of current in the main line connected through it, the armature is attracted toward the closed contact, meaning that the circuit starting at the local dynamo is extended through the relay armature, closed-contact, and on through the magnet windings of the sounder to ground through the resistance coil C. THE BATTERY DUPLEX Figure 231 shows the theoretic connections of the main-line instruments used to operate a polar duplex by means of gravity battery. In this duplex arrangement the pole-changer consists of two double-con- tact relays, or transmitters. The transmitters are connected in series, that is, one signaling key controls the operation of both instruments, so that both DUPLEX TELEGRAPHY 267 armatures are in the closed position at the same time, and in the open position at the same time; depending upon whether the key is open or closed. It will be seen at a glance, that when both armature levers are in contact with their back-stops the positive pole of the row of gravity cells is connected Line Battery I-H-H-H-H-! FIG. 231. Theory of the gravity battery duplex. to line via the tongue of transmitter No. i, and at the same time the negative pole of the battery is "grounded" via the tongue of transmitter No. 2. Conversely, when the signaling key is closed and both tongues are against their front-stops, the negative pole of the battery is connected to line, and the positive terminal of the battery to ground. The operation of the key; con- FIG. 232. Actual connections gravity battery duplex. trolling as it does simultaneously the operation of both transmitters, results in alternate positive and negative impulses being sent to line, the same as when two dynamos of opposite polarities are used. In other respects the connections are the same as in the dynamo polar duplex. Figure 232 shows the actual circuit connections of the battery duplex. THE " BRIDGE" DUPLEX The single-current duplex, and the polar duplex, being based on the differ- ential principle are dependent upon producing an equality of current strengths, 268 AMERICAN TELEGRAPH PRACTICE while the bridge duplex which is based upon the well-known Wheatstone bridge principle is dependent upon producing an equality of potentials. Figure 233 shows two stations A and B at either end of a line wire equip- ped with bridge duplex apparatus. B and B f are the main-line batteries at A and B respectively. AL in each case represents the artificial line at either end. R and R f are two artificial resistances of equal value, likewise r and r' at station B. At each end of the line the relays are connected between the points c and d of the "bridge" formed by the line wire and the artificial line resistance. Closing the key at A sends out a current which divides at a, half passing over the line wire to station B and reaching earth via the apparatus at that end of the line, while the other half passes through the artificial line at A , reaching the earth Line Ti FIG. 233. Theory of the bridge duplex. at that end of the circuit. Inasmuch as the points c and d are equidistant, ohmically, from the point #, their potential values are identical, and no current will flow through the windings of the relay at A. This is true, of course, only when the resistance of the artificial line at A is made equal to the re- sistance of the actual line to ground at the distant end. The relay at A, therefore, is not affected when A sends to B. The same condition prevails when B alone sends to A . Signals from A operate the relay at B because the incoming signals have a joint path made up of the branches c-d and c-a, thus setting up a difference of potential between the points c and d sufficient to operate the relay. The operations which take place with different key combinations at either end of the bridge duplex may be traced without difficulty. Since the line relay employed in the bridge duplex does not need to be differentially wound, it is evident that any ordinary relay may be used with this method of duplexing. It is apparent, also, that the outgoing currents do not pass through the windings of the home relay, and, as the currents pass directly to line, there is a minimum amount of retardation in the send- ing circuit. And, further, it is claimed for the bridge duplex that its line relays, on account of their position in the bridge, are not as responsive to induced line disturbances or to earth currents as are the line relays in the DUPLEX TELEGRAPHY 269 differential duplex. This is due to the fact that in the bridge system only a portion of the line currents pass through the relay, no matter whether the currents are the result of an impressed e.m.f., of induction, or of con- duction from neighboring circuits, while in the differential duplex all currents existing in the main line pass through the windings of the line coil of the relay. The bridge duplex has been more highly developed in Europe than in America, and several of the refinements applied to its operation there are particularly noteworthy as having a bearing on the general subject of high- speed signaling. These refinements include the application of the signaling condenser and the reading condenser, Fig. 234. sc Line 5C FIG. 234. Signaling condensers and reading condenser applied to. the bridge duplex. SIGNALING CONDENSERS As has previously been stated, the electrostatic capacity of the line wire must be satisfied in any given case before final-current values obtain in the circuit. Although the time required for the current to reach its maximum value is independent of the value of the e.m.f. employed, the time required for the current to reach a certain percentage of its final value is directly dependent upon the value of the potential applied to the circuit. It is well known that when a terminal potential of, say, 250 volts is ap- plied to a line the required operating current strength will actuate the relay at the remote end of the line in approximately half the time required to produce the same effect with an applied e.m.f. of 125 volts. Calculations of this kind, of course, require that the current resulting from the lower value of e.m.f. will have sufficient strength to operate the relays satisfac- torily. If, now, we consider the circuit conditions prevailing when the signal- ing condensers SC (Fig. 234) are connected in shunt with the bridge resist- ance, it may be seen that the presence of these condensers, in effect, create 270 AMERICAN TELEGRAPH PRACTICE a momentary short circuit around the 3,ooo-ohm bridge resistances. This interval, although brief, is sufficient to permit of the application of maximum battery potential to the line, which results in an initial current value at the distant end of the line, equal to that which would obtain if the 3,ooo-ohm resistance were not a part of the circuit. After the condenser and the line have become charged by the initial impulse, the final-current strength builds up through the circuit which includes the 3,ooo-ohm resistance as a portion thereof. The ultimate value of the current in the circuit will, therefore, be less than that at first prevailing at the receiving end. Obviously the final current strength will have an ^ value. Inasmuch as the 3,ooo-ohm bridge coil on the artificial line side also is shunted with a condenser having a capacity adjusted to a value equal to that shunting the 3,ooo-ohm bridge coil in the line side, it is plain that exactly like conditions exist in each branch of the circuit at the same time. When the line current is reversed it is obvious that the condensers will discharge in a direction coinciding with that due to the alternate battery pole. Thus the total value of the e.m.f. actuating the circuit will be that of the terminal battery plus that existing as charge in the condensers. On each occasion, therefore, that the condensers are taking on or giving up their charge, the initial portion of the signaling impulses in either direc- tion has a path other than that presented through the 3,ooo-ohm bridge coils. It may be observed that the effect of the condenser discharge is to greatly expedite the discharge of the line wire, and in this regard it is found that the best results are attained when the capacity of the condenser is made equal to that of the line. THE READING CONDENSER The reading condenser, or " shunted" condenser as it is sometimes called (RC Fig. 234), in British Post Office practice consists of a group of three resistance units having individual values of, 2,000, 4,000, and 8,000, ohms or a total of 14,000 ohms, shunted by an adjustable condenser having a total capacity of 7 1/2 microfarads. The function of the shunted condenser is to balance the effects of self- induction of the signaling relay. In a preceding chapter it was pointed out that when the direction of current flowing in a coil of wire or a magnet is reversed the effect of self- induction between the turns of wire in the magnet is, in the first place, to retard the rise of current strength in the circuit of which the winding forms a part and when on each occasion the circuit is opened the effect of self- induction is to delay the fall to zero current. The presence of the shunted reading condenser provides that the com- DUPLEX TELEGRAPHY 271 mencement of the reversal of magnetism in the cores of the relay will take place at the instant the transmitter tongue at the distant end of the line leaves either the positive or the negative battery contact, so that the process of reversing has progressed to a certain extent by the time the tongue of the distant transmitter reaches the opposite battery contact, or the (''ground") contact, as the case may be, and an effect is produced which balances the effects of self-induction by hastening the rise and fall of the operating cur- rent in the circuit at the instant desired. The amount of capacity and resistance which yields the best result in a given case, naturally is dependent upon the particular properties of the line conductor under consideration, and can be determined only under working tests. THE WESTERN UNION BRIDGE DUPLEX Figure 2340 shows the theory of the bridge duplex recently adopted by the Western Union Company. In this duplex the bridge arms consist of Line FIG. 2340. Western Union bridge duplex. the companion windings of an impedance coil (5 U) each arm having an ohmic resistance of 500 ohms (see the impedance coil, Fig. 279). In Fig. 2340, the main-line circuits at each end of a duplexed line are shown, in which D represents the main-line dynamos, L, resistance lamps, MA, milammeters, PR, polar relays, PC, pole-changers, r, retardation resistances, 5^7, impedance coils, SC, spark condensers, MR, main-line adjustable resistances, CR, compensating-circuit adjustable resistance, AL, regular artificial-line adjustable resistances, C, static compensating con- densers. The operation of this duplex will be better undertsood after the reader has gone through the matter describing, the Western Union quadruplex (Fig. 276). 272 AMERICAN TELEGRAPH PRACTICE THE HIGH-POTENTIAL "LEAK" DUPLEX In those telegraph installations where the only dynamos in service are those required to operate the long quadruplexes, the "leak" method of re- ducing high potentials to values sufficient to operate circuits duplex, is sometimes employed. This method is due to Mr. Minor M. Davis and was introduced on the lines of the Postal Telegraph-Cable Company several years ago. Figure 235 shows the theoretic arrangement of circuits of the leak duplex. An artificial circuit to ground is built up of coils having resistances of 800 plus 2,200 ohms, or a total of 3,000 ohms. Where the machines available for quadruplex working have potentials of 380 volts, positive and negative, respectively, it is apparent that with _800 1600 Line Condenser FIG. 235. Theory of the high-potential "leak" duplex. an internal resistance of 600 ohms in series with the leak resistance, the nearest ground is 3,600 ohms distant from the battery, at least that would be the case when the tongue of the pole-changer is midway between its back- stop and front-stop. As the pole-changer is operated its tongue is caused to make contact with the leak circuit at a point either - of the total ohmic distance to 3600 ground, or in case the 8oo-ohm resistance is short circuited at a point , of the total ohmic distance to ground. Thus the available voltage at the pole-changer contacts is reduced to a value considerably below that available at the brushes of the machines. The exact value in either case may be calculated by means of either of the methods described in a preceding chap- ter for determining the difference of potential at any point along a conductor possessing resistance between that point and ground. A leak path to ground is provided for each of the high-voltage gen- erators, so that the reduction of voltage may be made equal in the case of both positive and negative machines. The possible connections are such that three different potential values may be availed of as desired. When the circuit-controlling plugs are removed from the 2,2oo-ohm DUPLEX TELEGRAPHY 273 coils, and the 8oo-ohm coil in each circuit is short circuited, the full quadru- plex battery is available. With the 2,2oo-ohm coils in circuit while the 8oo-ohm coils are short circuited, the next lower potential is available, and when the 2,2oo-ohm coils are in circuit and the shunt circuit removed from around the 8oo-ohm coils, a third potential value is available. Figure 236 shows the actual binding-post connections of the main-line wiring of a high-potential leak duplex. FIG. 236. Actual connections of the high-potential leak duplex. HIGH EFFICIENCY DUPLEXES Within recent years a demand has been created for the development of a high efficiency duplex. Among the causes which have brought about this demand, the more important are: the increasing amount of line dis- turbance experienced due to induction from other wires of the same system IMF. \ Line FIG. 237. Theory of the high-efficiency duplex employed by the Postal Telegraph ^Company. and from neighboring conductors carrying high potentials, decrease in the efficiency of transmission attributable to the employment of semi-automatic transmitters which are not as regular in action as simple hand transmission by means of the Morse key. Also, the call for fast and dependable leased 18 274 AMERICAN TELEGRAPH PRACTICE wire service and for high-speed automatic and printer circuits has resulted in a systematic critical investigation of the duplex at the hands of several well-known experts. Fig. 237 shows the circuits of an improved differential duplex recently brought out by Messrs. Davis and Eaves. The principles of operation are the same as in the ordinary differential polar duplex previ- ously described, but several capacity and resistance units have been com- bined and applied in -such relation to the regular duplex circuits that not only do they serve to correct the inherent weaknesses of the duplex, but to bring about action in certain places and at certain intervals which materially increases the operating efncieny of a duplex to which these adjuncts are applied. , By referring to Fig. 237 it will be seen that two 5oo-ohm non-inductive coils have been introduced at the " split" behind the relay, so that the out- going current has a joint path, on the one hand through a 5oo-ohm coil and the main-line winding of the relay, and on the other through the companion 5oo-ohm coil and the artificial-line winding of the relay. The presence of these coils introduces only the property of resistance into the circuit, as owing to the fact that they are non-inductively wound there is no retardation introduced. The insertion of the resistance back of the relay steadies the balance somewhat, due to the fact that a considerable proportion of the total resistance of the circuit is inserted between the coils of the relay and the ground connection via either dynamo. The presence of the two 5oo-ohm coils causes the condenser connected in shunt therewith to take on a charge due to the difference of potential which exists between the points a and b when the pole-changer at the distant end of the line is operated. The function of this condenser is to hasten the "turn-over" of magne- tism in the cores of the home relay when the distant station sends out current reversals. The condenser anticipates, as it were, the action which will result in the home relay when the tongue of the pole-changer at the distant station reaches the negative or positive battery contact, as the case may be. With the ordinary arrangement of duplex circuits, the armature lever of the home polar relay remains in contact with the closed contact of its sounder circuit as long as the tongue of the pole-changer at the distant station is in contact with its front-stop and until the pole-changer tongue again touches its back-stop. The action of the condenser here considered is to cause reversal of magnet- ism in the relay at the instant the tongue of the pole-changer at the distant station departs from either its front- or b'ack-stop. It is apparent that the charge which the condenser has accumulated while the tongue of the distant pole-changer has been in contact with either battery pole will discharge through the windings of the home relay in a direction coinciding with that taken by the current resulting from the next succeeding battery contact at DUPLEX TELEGRAPHY 275 the distant end. Thus, the current arriving from the distant station com- pletes the work already begun by the condenser. The two 6oo-ohm non-inductive coils connected around the relays, in series with one-half microfarad condensers, present to in-coming inductive disturbances a path to ground which does not lead through the windings of the relay, thus in large measure making the relay immune to induced currents, especially from alternating-current sources, and also to electrostatic and electromagnetic induction from neighboring wires of the same system. Another benefit derived from the 6oo-ohm, i-m.f. shunt circuit is that the discharge due to self-inductance of the relay magnets takes place through the loop circuit thus formed around the relay coils, preventing its interference with line currents. ZYNAMff ZUFLEX HA/N CIRCUITS FIG. 238. Main line connections of the Postal Company's duplex. Another decided advantage resulting from the employment of the latter circuit is that the first part of each out-going current wave gets to line and to the distant station earlier than it would if required to travel through the inductive winding of the home polar relay. The desired movement of the armature of the distant polar relay, therefore, is well under way by the time the Ohm's law current arrives. Other important features have been introduced in connection with this high efficiency duplex, among which might be mentioned the use of an improved "spark-killer" arrangement to control the sparking which occurs at pole-changer points as contact is alternately made betweeen the tongue and the positive or negative potentials. Also, a reduced internal-resistance value is inserted between the dynamo 276 AMERICAN TELEGRAPH PRACTICE and the pole-changer line contacts, and an improved form of polar relay is used. These features are referred to in detail further along. Figure 238 shows the instrument main-line connections of the high efficiency duplex just described. CITY LINE DUPLEX Figure 239 shows the theoretical connections of a short-line duplex, which may be operated over a single wire with main battery of one polarity, at one end of the line The line relay at the main office is an ordinary differential non-polarized instrument, the same as that used in connection with the ordinary single- current duplex, or on the second side of a quadruplex. At the branch office Main Office Branch Office Leak fUJWl r>^K~n ^^ B ~^ TR. . , A - High Res. Magnet FIG. 239. City line duplex. a special non-polarized differential relay is employed, the artificial-line coil of which has a winding of higher resistance than that of the main-line coil. A glance at the circuit arrangements will show that battery is to line at all times, full potential when the main-office transmitter is closed, and a reduced potential when the main-office transmitter armature is in contact with its back-stop: the value of the reduced potential depending upon the resistance value of the leak circuit which forms one path of a joint circuit to ground, the other path consisting of the main line and the apparatus at the branch office, and the path to ground via the artificial line at the main office. When the armature lever of the relay at the branch office is resting against its back-stop, the only path presented to the incoming signals is through the coils of the relay to ground via the artificial line at the branch office. When the main office only is sending, it is evident that inasmuch as the out-going currents pass through the relay at the main office differentially, the armature of that relay is not affected, while the relay at the branch office responds each time the armature tongue of the main-office transmitter closes, and opens each time the tongue of the main-office transmitter is withdrawn into contact with its back-stop, because then the current which is sent to line is not of sufficient strength to operate the branch-office relay. Ob- viously, the tension given the retractile spring attached to the armature of 'DUPLEX TELEGRAPHY 277 the branch-office relay must be such that the magnetism produced by the reduced current volume will not be strong enough to attract the armature. With the armature tongue of the branch-office transmitter in the closed position, and at rest, the incoming signals have a joint-path to ground at the orr/cc TO mOH SW/TCHES TO LCfT - Z SlftlfST5. FIG. 240. Main office connections city line duplex. branch office, but still the armature of the relay at the branch office will re- spond each time the main-office transmitter is closed, and release each time the latter is opened, for although a greater current volume exists in the main line because of the shorter path to ground presented, the total amount of ro RIGHT DUPLEX SWITCHES TO LEFT Z S/M6LE STS. FIG. 241. Branch office connections city line duplex. magnetic pull on the armature of the branch-office relay is no greater than before owing to the fact that the high resistance (and the most effective) coil of the relay is practically short-circuited by the newly presented path to ground via the tongue of the transmitter. 278 AMERICAN TELEGRAPH PRACTICE It is apparent, however, that when the armature lever of the branch- office relay is in contact with its front-stop the increased current strength in the main-line coil of the main-office relay results in the armature of that relay being attracted the very result desired, for it is when the tongue of the branch office transmitter is moved into contact w r ith its front-stop that the armature of the main-office relay should be moved into contact with its front- stop. The reason why this type of duplex is not suitable for long lines is that the ratio of operating to releasing current is such that the margin of current strength between these two values is not very great, in fact, not great enough to permit of fluctuation of current strengths such as experienced in the operation of long lines. Figure 240 shows the actual main-line and local connections of the city- line duplex at the main office, while Fig. 241 shows the connections at the branch office. At the main office a regular duplex rheostat is used; the artificial-line circuit being made up of the regulation artificial-line coils, and the " leak " cir- cuit to ground being made up of the coils ordinarily used as the first and second condenser circuits. As the arrangement is intended only for short lines, it is not necessary to employ static compensating condensers. SHORT -LINE DUPLEX, WITH BATTERY AT ONE END ONLY A short-line duplex requiring battery at one end only, which has been employed with considerable success on the lines of the Western Union Tele- graph Company, is that known as the Morris duplex. Battery Station FIG. 242. Double current duplex with battery at one end only. The system was originally devised by Mr. Gerritt Smith, later improved by Mr. Morris, and still more recently has been equipped with static com- pensating accessory apparatus which permits of its successful employment in the duplex operation of lines 150 miles or more in length. DUPLEX TELEGRAPHY 279 Figure 242 shows the theoretical main line, and the local connections of the latest arrangement of apparatus. The rheostat at the main, or battery station, makes possible the insertion of a resistance value equal to that of the line "wire plus the resistance of the rheostat at the distant office. The proper resistance value required to be inserted at the distant office may be determined by measuring the line current with the distant key open, and again when it is held closed. The resistance of the rheostat should be such that with the key closed the line current will be three times that obtain- ing in the circuit when the key is open. The proper resistance value to give the rheostat at the battery end of the line may be determined by measuring the resistance of the line to the distant ground including the resistance of the distant relay and rheostat. The operation of this duplex may easily be traced by observing what takes place when the keys are operated, and when the transmitter tongues at either end are in the various possible positions. SPARKING AT CONTACT POINTS In the operation of relays a troublesome spark is produced as contact is made or broken between the movable armature lever and the stationary front-stop each time the local circuit- is closed or opened. It is the effect of the extra current of self-induction, and is strongest at the instant the circuit is broken. Naturally, it is more pronounced in wet than in dry weather, owing to the fact that the forward and the backward movement of the arma- ture of the relay is then more sluggish. The same is true in any state of the weather of relays operating in lines which are not maintained at a high degree of insulation. ; It has been found that the more rapid the movement of the armature, the less pronounced will be the resulting spark at make and break of contact. The effects observed in the operation of telegraph apparatus are in con- formity with the general theory of the subject as enunciated by Faraday, and point to the conclusion that if connections and disconnections could be made rapidly enough "sparkless" make and break might be accomplished. Ray- leigh has shown that when a circuit is broken at velocities of the order of one meter (39.37 in.) per second, there is no evidence of sparking between contact points. It is found that with a quick-moving armature, a much closer adjustment is possible than with a slow-moving armature. If the current traversing the magnet windings of the instrument is weak, thus necessitating a weak retractile spring, it is found that a wide adjustment between tongue and contact point is necessary in order to avoid sparking, but with a strong magnetic pull on the armature, and a strong retractile spring, insuring quick 280 AMERICAN TELEGRAPH PRACTICE movement of the armature in each direction, points may be set much closer together without danger of excessive sparking. By far the greatest amount of trouble experienced due to sparking at contact points is that encountered in the operation of transmitters and pole-changers used in duplex and quadruplex telegraphy. In the case of a pole-changer, the negative terminal of a 375-volt dynamo may be connected to the armature front-stop, while the positive terminal of another 375-volt dynamo may be connected to the armature back-stop of the instrument, and as the armature lever (which is connected to the main line via the windings of the line relay) plays between these contact points, there is an ever present danger of arcing, due to the difference of potential (amounting to 750 volts), existing between the front- and back-stops sepa- rated by the air-gap traversed by the lever in its movements to and fro. When an arc forms between the opposite contacts, the great heat devel- oped quickly destroys the metal points and renders them unfit for use. Pole-changers and hand-operated keys which are directly connected into main-line circuits usually have contact points constructed wholly of, or tipped with platinum. Platinum is the heaviest and least expansible of the metals, is harder than iron, very ductile, undergoes no alteration in air, and resists the action of acids. Silver also 1 has been used to a considerable extent in making contact points, and while it is true that silver undergoes changes in air, it is claimed that the oxide of silver formed on the exposed surfaces is a better conductor of electricity than the silver itself and that the same necessity does not exist for maintaining clean bright surfaces of contact as is the case with other metals. It is probable, however, that in cases where the oxide of silver film is allowed to exceed minute thickness, there is danger of the accumulation of foreign matter of low conductivity in association with the somewhat irregu- lar deposit of oxide. This means that where silver contact points are employed, it is the part of wisdom to clean the points with a fine steel file, usually provided for the purpose, as is customary with platinum contacts. When it is remembered that in the operation of transmitters and pole- changers, the duration of contact between the armature lever (the line) and the stationary contact points (the main-line battery) is very brief, if the full voltage of the dynamo is to be impressed upon the line at each contact, it would seem to be important that the abutting contacts should be free of foreign matter, have smooth regular surfaces, and that the area of surface contact should be such that no appreciable resistance will be introduced at the instant connection is made. 1 Quite recently wrought tungsten has been introduced as a substitute for platinum in the manufacture of electrical make and break contacts. DUPLEX TELEGRAPHY 281 It is found in practice that contact points having even regular surfaces and which are kept well polished, do the work required of them more satis- factorily and cause less trouble from sparking than points which are neglected in these respects. Quite a number of meritorious arrangements have been proposed, having in view the prevention of, or the control of sparking at contact points, several of which methods are described in what follows. The Postal Telegraph-Cable Company has recently adopted a type of pole-changer which is equipped with a permanent magnet taking the place of the retractile spring formerly used to draw the armature tongue into contact with the back-stop when the magnet coils of the instrument are de-energized. With a spring retractile, when the local key circuit is closed and the pole- changer coils energized, the pull against the forward movement of the arma- ture increases as the armature moves toward the front-stop, and in the reverse movement of the armature the pull is greatest at the instant the 5mf. Line 5mf. FIG. 243. Form of pole changer in which the forward and backward movements of the armature are controlled by electro-magnets. FIG. 244. Spark-controlling arrangement formerly used by the Postal Telegraph Company. local key circuit is opened, the strength of pull decreasing as the armature travels toward the back-stop. When a permanent magnet is employed for the purpose, the retractile pull against the armature rapidly decreases as the armature moves toward the front-stop, and rapidly increases as the armature moves toward the back-stop. It is believed that where the permanent magnet is employed, it is possible to maintain a retractile pull more nearly equivalent to that of the forward pull produced by the electromagnets, thus insuring an equal speed of armature travel in either direction. As in the case of the spring retractile, it is necessary that the permanent magnet be so mounted that it may be adjusted with respect to its proximity to the armature, so that ageing of the permanent magnet may be compensated for, and that the retractile force exerted may be made to equal that of the electromagnets under any given conditions of current strength. 282 AMERICAN TELEGRAPH PRACTICE One decided advantage of the permanent-magnet retractile is that the "pull" is constant, thus preventing any tendency the armature lever may have to rebound from either back or front contact point. Where the spring is used it is claimed that the reflex action progressing while the coils of the spring are in motion, produces a rebounding movement of the lever which results in sending to line a current impulse somewhat wavy in form. The type of pole-changer illustrated in Fig. 243 is so designed with regard to the disposition of electromagnets on either side of the armature that when it is desired to have the lever move into contact with, say the positive pole of the battery, the electromagnetic force holding the lever against the opposite battery contact is instantly neutralized, thus the armature is permitted to move in the desired direction without being restrained by an opposing force. Fie. 245. FIG. 2450. FIG. 245^ FIGS. 245, 2450 and 245^. Johnson coil spark curbing arrangement. And conversely when it is desired that the lever shall move into contact with the negative battery pole, the armature again moves in the desired direction without being restrained by an opposing force. l It will be observed that there are no springs or permanent magnets employed in the operation of this instrument. The electromagnet on the right has two equal windings connected differentially, while the magnet on the left has an ordinary single winding. With the highest speed possible with any type of pole-changer where the armature must start from a position of rest, contact is broken at a relatively low velocity, and as a consequence a considerable amount of sparking takes place. Various combinations of resistance coils and condensers have been employed successfully in limiting the amount of spark formed at contact points. The arrangement illustrated in Fig. 244 was for a time used by the Postal Telegraph- Cable Company, in connection with pole-changers operat- ing in multiplex circuits. It will be seen that while the armature lever makes and breaks contact with the individual dynamo terminals in the 1 This is aside from the natural opposition to movement, due to gravity, to inertia, and to bearing friction. DUPLEX TELEGRAPHY 283 usual manner, a condenser discharge path is at all times maintained around the contact points, the armature being connected to the middle of the discharge circuit by way of a 4oo-ohm resistance coil, wound non-inductively. The above arrangement was displaced on the lines of the Postal Company by a form of induction coil known as the "Johnson" coil, see Figs. 245, 2450 and 2456. This arrangement consists of three separate windings of german silver wire of small gage, wound on a wood bobbin with an air core, the spool thus formed being about 7 in. long and i in. in diameter. The coils, although wound one on top of the other, are thoroughly insulated from each other by a double cotton covering saturated with paraffine. As indicated in Fig. 245, one end of each winding is left open, while the opposite ends of the wind- ings are connected to the battery contact points and the armature of the pole-changer as depicted in Fig. 2456, the center winding (provided with a red covering to distin- guish it from the top and bottom windings) being connected with the armature, while the top and bottom windings are connected with the positive and negative battery terminals of the pole-changer. The inductive action that takes place between the contiguous windings has the effect of absorbing and dissipating the energy of the spark. In cases where the tendency toward sparking is exces- sive it is helpful to connect two of these "coils" in parallel, similarly to the way in which two condensers are connected in parallel. Recently the Postal Telegraph Company has adopted the spark-killing arrangement of shown in Fig. 246, in which each battery FIG. 246. Present method of spark control used by the Postal Company. .25 mf. FIG. 247. Present method spark control used by the Western terminal of the pde-'changer is provided with Union Company. . a discharge path to ground through a one- half microfarad condenser. The arrangement used by the Western Union Telegraph Company to limit sparking at pole-changer contact points is illustrated in Fig. 247, in which a i/4-m.f. condenser connected in series with a 2o-ohm lamp of the 284 AMERICAN TELEGRAPH PRACTICE incandescent pattern is placed across the battery terminals of the pole- changer. THE "MAKE" SPARK It has been stated that the spark which occurs at the instant contact is broken, is due to the extra current of self-induction of the circuit. It might here be stated that the spark which occurs between the armature contact and the battery terminal of a pole-changer at the instant contact is "made" is due to the static discharge from the main and artificial lines, which takes place during the brief instant that actual contact is being made. By means of a circuit arranged as in Fig. 248, the production of the "make" and of the "break" spark may be observed, and the cause IR w^ ',1 Wlth 2 m COntaCt Wlth i als m FIG 248,-Circuit arrange- of each determined. ment for demonstrating pro- duction of the "make" and the "break" sparks. contact with a produces a strong spark at the instant contact is made, while no spark ap- pears as contact between a and i is broken. No perceptible sparking takes place as a is moved into contact with i, but at the instant contact is broken between a and i a pronounced spark appears. FIG. 249. Multiple gap and multiple contact pole-changer. In the first case the "make" spark which develops is due to the dis- charge of the circuit possessing capacity, and in the second case the spark observed is due to the extra current of self-induction. DUPLEX TELEGRAPHY 285 It should be remembered, of course, that when arcing takes place be- tween the contact points of a pole-changer, the arc is the result of difference of potential between the battery terminals of the instrument, and that the FIG. 250. Multiple contact pole-changer employing three transmitters. only part played by the make or the break spark when an arc is " struck" is that of reducing the resistance of the air-gap to a degree which permits the formation of the arc. The heat of the arc which ranges from 2,000 to 5,000 C. is very destructive to the metallic terminals. FIG. 251. The Field multiple-gap pole-changer. The ordinary make and break sparks, if excessive are liable to heat the air of the gap (thus reducing its electrical resistance) to a point where arcing is likely to occur. In place of the shunt discharge path, a plan has been tried which con- sists of providing a multiple gap as illustrated in Fig. 249, wherein it may be 286 AMERICAN TELEGRAPH PRACTICE noted that each dynamo terminal is brought to two separate contact points. The armatures of two separate pole-changers are controlled by an individual battery and key circuit, which when closed places both armature levers in contact with the negative pole of the battery and when opened places both levers in contact with the positive pole of the battery. An instrument designed on this principle is known as the Berry pole-changer, being the invention of Mr. T. H. Berry. It is evident that as each battery contact is made at two separate points, the sparking tendency at each contact is halved. A similar arrangement employing three ordinary pole-changers for the purpose is illustrated in Fig. 250. Figure 251 shows a pole-changer having a multiple gap, which has been designed by Mr. Stephen D. Field. In cases where high potentials are employed, and where high signaling speeds are not essential, the a oil" break has been used with success. With this arrangement the pole-changer is inverted and the contact between armature and battery terminals is made to take place in a chamber filled with thin oil, in which case the oil serves to extinguish the spark as soon as formed. CHAPTER XIV THE QUADRUPLEX THE JONES SYSTEM. THE FIELD KEY SYSTEM. THE POSTAL QUADRUPLEX. THE SINGLE DYNAMO QUADRUPLEX. THE METALLIC -CIRCUIT QUAD- RUPLEX. THE GERRITT SMITH QUADRUPLEX. THE WESTERN UNION QUADRUPLEX. THE B.P.O. QUADRUPLEX. Quadruplex telegraphy consists of a method of sending two messages simultaneously over an individual wire in one direction, while at the same time two additional messages are being transmitted over the same wire in the opposite direction. A wire equipped at each end with quadruplex apparatus may be used to transmit one, two, three, or four telegrams at the same time. That is, when the wire is equipped for quadruplex working, one message at a time may be sent over it, or, if required, four telegrams (two in each direction) may be transmitted simultaneously. The system of quadruplex telegraphy generally employed is based on a combination of the Stearns, or single-current duplex, and the differential polar duplex, both of which have been described in the preceding chapter. One message in each direction may be transmitted by means of the single-current half of the system due to changes effected in the strength of the line currents without regard to the polarity of said currents, while one message in each direction may at the same time be transmitted by means of the polar half of the system due to alterations in the polarity of currents impressed upon the line, which alterations are effected through the agency of ordinary transmitting keys and pole-changers as described in connection with the differential polar duplex system. Figure 252 shows the theoretical wiring of the main-line circuits, and the pole-changer and transmitter local circuits of a quadruplex arranged to operate with gravity battery. In the diagram the circuit arrangements at two terminal stations are shown, the two stations X and Y being connected by a line wire. For the sake of clearness the reading sounder circuits which are operated locally through the action of the armatures of the polar relays and the neu- tral relays have been omitted. The letters o and c, however, are used as indices to denote the open and the closed positions of the respective relay armatures. In each case the closed position of the relay armature implies that the signaling armature lever of the reading sounder connected thereto would also be in the closed or marking position. 287 288 AMERICAN TELEGRAPH PRACTICE It is to be remembered that the armature of the polar relay will be drawn into the closed position when current traverses the coil windings of the relay in a given direction, and into the open position when current travels through the coils in the reverse direction. It is immaterial whether the respective currents are weak or strong. A weak negative current, for instance, will cause the armature to move in one direction, while a weak positive current will cause the armature to move in the opposite direction. Polar relays work satisfactorily with currents varying from 3 milliamperes to 200 milliamperes, which means that although the movement of the armature in one direction may be the result of a strong positive impulse, the armature THE QUADRUPLEX 289 may be moved in the opposite direction by a weak negative impulse, pro- vided, of course, that the positive current has been disconnected or sup- pressed. Also, it will at once be apparent that should a wire carrying a current of, say, 50 milliamperes from a positive source be connected to one terminal of the coil winding of the relay, while a wire carrying a current of 55 milliamperes from a negative source is connected to the other terminal, the surplus of 5 milliamperes negative current would be sufficient t> move the armature in the direction which a negative current of any strength would move it. Further, as stated in connection with the operation of the relay used in the polar duplex, the armature tongue of the relay, due to the attrac- tion of the permanent magnets associated therewith, remains in connection with the open or the closed contact once it has been moved there, until the direction of the current through the coils of the instrument has been reversed, whereupon the tongue instantly moves over to the opposite contact. That half of the quadruplex which is operated by means of current re- versals is called the polar, A, or first side of the system, while the half which is operated by raising and lowering the strength of the current obtaining in the main-line circuit is called the neutral, common, B, or second side of the system. THE DIFFERENTIAL NEUTRAL RELAY The description of the differential relay given on page 251 applies equally to the type of relay employed on the second side of the differential quadruplex to record the signals transmitted from the distant station as a result of the operation of the transmitter connected into the line at that point, and by means of which the strength of current permitted to traverse the line is regulated. The forward and backward movements of the armature of the neutral relay are accomplished in a manner somewhat different from that which actu- ates the armature of the polar relay. The armature tongue of the neutral relay is drawn into contact with its back-stop by the action of a retractile spring which may be given a tension such that a comparatively large volume of current must traverse one or both coils of the relay before the armature will be attracted forward. Also, as is the case with the ordinary or common single Morse relay, it is immaterial whether the current traversing the coil windings of the relay is from a positive or a negative source, provided the current actuating the magnets has the required strength to overcome the spring tension which tends to hold the armature tongue against its back-stop. Thus it is seen that if the current operating the polar side of the system is kept down to a strength of, say, 2 5 milliamperes, the retractile spring of the companion neutral relay may be given a tension which will prevent it from responding to currents of such comparatively low volume. 19 290 AMERICAN TELEGRAPH PRACTICE It is customary to so adjust the neutral relay that a current strength three times, or four times, greater than that which operates the polar relay must be impressed upon the line before the neutral relay will respond. As long, therefore, as the neutral side transmitter at the distant station is not operated, and while minimum current value obtains in the main-line circuit, although the polar side may be operated, the neutral relay remains unrespon- sive. The instant, however, that the neutral side transmitter at the distant station is closed, maximum current value obtains in the main-line cir- cuit and the neutral relay at the home station instantly responds. The various electrical actions which take place when full quadruplex operation is maintained over a wire are directly dependent upon the differ- ence of potential existing between certain points in the main-line circuit A C R. 200 Ohms KR. 200 Ohms FIG. 253 The Diplex. within the apparatus at each end of the line, upon the resistance of instrument windings and accessory resistance units, and upon the direction and strength of currents flowing through relay windings at certain instants and under certain conditions. Most students find it difficult to carry in their minds a picture of the many operations taking place which, taken all together, constitute quadruplex work- ing. But, if the subject is approached with a view to mastering each detail of operation separately, it is found generally that when the various details are understood, the theory of the system as a whole will be more firmly impressed upon the mind than if this method of study were not followed. We have seen in the case of the Stearns duplex, the polar duplex and the bridge duplex, that two messages at a time may be sent over a single wire, one in each direction. In order to maintain quadruplex operation, means must be provided for transmitting four messages at a time over a single wire, two in each direction. It will be helpful ; first to consider an arrangement such as that illustrated in Fig. 253, by means of which it is possible to transmit two messages simul- taneously over a single wire, both in one direction. This provides what was at one time known as diplex operation. As transmission is carried on in one direction only, one station is equipped THE QUADRUPLEX 291 with sending apparatus, while the other station is equipped with receiving apparatus only. The particular arrangement of circuits depicted in Fig. 253 is submitted here; not that it closely resembles the circuit arrangements comprising the diplex system of telegraphy originally introduced, but because it embodies features common to the present-day system of quadruplex telegraphy which make possible the simultaneous transmission of two messages in each direc- tion over a single wire. A line wire having an assumed resistance of 1,800 ohms is shown extend- ing between stations A and B, the direction of transmission being from A to B. The main battery consisting of gravity cells, having a total e.m.f. of- 200 volts and an internal resistance of 400 ohms is located at A, as also is the pole-changer PC, operated locally by means of a key K, and the trans- mitter r, the latter in this case consisting simply of a key K z which, when open places the 3,ooo-ohm shunt coil r in series with the line wire, and when closed short circuits this coil. At the receiving end of the line two relays are connected directly into the main-line circuit as shown. One of these the polar relay PR is actuated by current reversals, that is, its armature is moved into the closed posi- tion when the negative terminal of the distant battery is placed to line, and into the open position when the positive terminal, or pole, of the distant battery is placed to line. The operation of the common relay CR is dependent upon the strength of the current traversing its coils, and not upon the direction of current. By referring to the diagram it may be seen that at the sending station the key K is depressed. This action has moved the spring contact A away from the line contact-block c, with the result that the positive terminal of the battery is connected to ground via the key K, while the negative terminal of the battery is placed to line by way of spring contact B and line contact- block c, from which point the main-line circuit to ground at the distant sta- tion is made up via the 3,ooo-ohm coil r (Key K^ now being open) the 1,800- ohm line wire through the windings of the polar relay and the common relay, thence to ground. Calculation will show that the current strength obtaining in the circuit is about 36 milliamperes. And if it is assumed that the spring S attached to the armature of the common relay has been given a tension such that a current strength considerably in excess of 36 milliamperes must obtain in the circuit before the armature of the relay is attracted, it is plain that the opera- tion of the pole-changer at the sending station will have no effect upon the common relay, while on the other hand, the armature of the polar relay is moved into the closed position each time a negative current is sent to line, and into the open position each time a positive current is sent to line from the 292 AMERICAN TELEGRAPH PRACTICE distant station. It must be kept in mind, as pointed out elsewhere, that the polar relay responds to very low current strengths. Now, if key K 2 is depressed, thus short circuiting the 3,ooo-ohm coil r, a strength of current will obtain in the circuit which is about three times greater than that which existed while the key K 2 remained open. It is self-evident, therefore, that the operation of the key K 2 controls the movements of the armature of the common relay, while the operation of the key K controls the operation of the polar relay. DOUBLE TRANSMISSION IN BOTH DIRECTIONS Having an apparatus such as the diplex by means of which two sets of signals may be sent in the same direction over a single conductor without interference with each other, it is evident that by employing differentially wound relays at each end of the line, placing one winding of each relay in the main-line circuit while the other winding of each relay is included in the artificial-line circuit, as is done in the case of the Stearns and polar duplexes, it is possible to transmit two messages in each direction simultaneously. THE GRAVITY BATTERY QUADRUPLEX Figure 254 shows theoretically the main-line circuits of a quadruplex operated with current derived from a gravity battery. The type of pole- changer shown here is different from that illustrated in connection with Fig. FiG^ 254. Postal Telegraph Company's gravity battery quadruplex. Theory. 252 (see also Fig. 255) and consists of two double-contact relays of the usual construction. .In practice an individual sending key connected through a local battery controls the operation of both of the instruments comprising the pole-changer. By this means, when the key is depressed both armatures of the pole-changer THE QUADRUPLEX 293 relays are attracted into contact with their front-stops, and when the key is opened both armatures are withdrawn into contact with their back-stops, due to the tension of the retractile springs attached to them for that purpose. It may be noted that the function of the armature of the instrument on the right is to "ground" either pole of the main battery, while the function of the armature of the instrument on the left is to place to line that pole of the battery which is not grounded. The pole of the battery which is grounded and the pole which is to line at any given instant depends upon the positions of the respective armatures. When the signaling key is closed, both armature tongues will be in contact with their front-stops. In the case before us this action has placed the negative pole of the battery to the line, while the positive pole of the battery is grounded. On the other hand, FIG. 255. Pole-changer or transmitter. Postal pattern. when the signaling key is opened, the positive terminal of the battery will be placed to line while the negative pole will be grounded. Thus, it is seen that the operation of the signaling key controlling individually the movements of the two armatures results in currents being sent to line which alternate in polarity. THE TRANSMITTER The position of the armature of the transmitter J 1 , at any given instant determines whether the whole or a portion only of the main battery is utilized. By observing the two possible positions of the transmitter armature it will be evident that when the signaling key which controls the operation of the transmitter is depressed, the armature tongue will be moved into contact with its front-stop thereby opening the "tap" connection and placing the entire battery in service. When, on the other hand, the key is opened and the armature tongue is withdrawn into contact with its back-stop, one-third or one-fourth, as the case may be, of the main battery is availed of. 294 AMERICAN TELEGRAPH PRACTICE >1 G rt O, < 44 a 2 I H sed quad THE QUADRUPLEX 295 LONG END AND SHORT END When the armature of the transmitter is in the position which places the entire battery in service, it is said that the long end is to line, and when the armature of the transmitter is in the opposite position, that is, when a part of the battery is utilized, it is said that the short end of the battery is to line. Figure 256 shows the actual binding-post main-line connections of a battery quadruplex, the theoretical wiring of which is shown in Fig. 254. In the actual instrument connections the artificial line is made up of an adjustable rheostat and two 3-m.f. condensers, the latter also being adjustable. THE DYNAMO QUADRUPLEX THE JONES SYSTEM The distinguishing feature of the Jones dynamo quadruplex is the method employed to furnish the long-end and the short-end main-line potentials. By referring to diagram 257, it may be seen that four separate dynamos are required to furnish the desired potentials two 130- volt machines and two 385-volt machines. The 130 volts plus and minus serving as the reduced potential, while the higher voltage serves as the full potential to operate the neutral relay at the distant end of the line. In the diagram the polar-side transmitting key KI is shown in the closed position, the result of which is that the armature levers of the two instru- ments comprising the pole-changer are in contact with their front-stops. This in turn connects the i3o-volt negative potential and the 385-volt negative potential with the back-stop and front-stop respectively of the transmitter T. It is plain then, that as long as the key KI is kept closed, closing key K 2 sends to line full current strength, while opening key K z thereby placing the armature of the transmitter in contact with its back-stop sends to line a current of a strength approximately one-third of that sent out when the transmitter armature is in the closed position. Where the booster arrangement (see Fig. 45, page 63) for supplying quadruplex potentials is installed, the Jones quadruplex system may be employed to advantage. THE FIELD KEY SYSTEM 1 With the Jones quadruplex it is necessary to employ two different e.m.fs., one to operate the polar and one to operate the neutral side of the system. 1 The first practical application of the dynamo as a substitute for the chemical battery in the operation of telegraph lines was made in the year 1879 by Mr. Stephen D. Field, then of San Francisco. (See U. S. patents, Nos. 223,845, Jan. 27, 1880, and 243,698, July 5, 1881.) 296 AMERICAN TELEGRAPH PRACTICE THE QUADRUPLEX 297 As the current strength required to operate the latter compared with that required to operate the polar side is in ratio 3 to i , or 4 to i , as the case may be, it is evident that with the Jones system four sources of e.m.f. are required, two of the higher value, one being negative and the other positive, and two of the lower value, negative and positive. It is, of course, understood that while four dynamos are required to operate one quadruplex, the same four machines may at the same time be employed to supply current for a number of similar quadruplex circuits. Where the Field quadruplex system is employed the number of dynamos required is reduced one-half, as two machines only are needed, one delivering positive and the other negative current. Each dynamo delivers an e.m.f. of sufficient strength to operate the neutral side, and by employing properly proportioned resistances the insertion of which is automatically controlled by operating the transmitting key associated with the second side, the poten- Line Condenser FIG. 258. Theory of the Field quadruplex. tial may be reduced to a value suitable to operate the polar side of the system. In this manner is produced what has heretofore been referred to as the "long end" and the "short end." The ratio of "long-end" to "short-end" e.m.f. determines the ratio of maximum to minimum line current, and the novelty of the Field key quadru- plex, is in the arrangement of voltage-reducing resistances which not only provide for the sending out of the two properly proportioned current strengths, but at the same time insure that the whole or joint-resistance of the terminal circuits remains the same regardless of the position of the armature of the transmitter controlling the strength of currents sent to line. Figure 258 shows the theoretical main-line wiring of the Field quadru- plex. With the armatures of the pole-changer and the transmitter in the positions shown in the diagram, the completed circuit extends from the 385- volt -dynamo to the closed-contact of the pole-changer, thence via the arma- ture of the pole-changer along a short connecting wire to the closed-contact of the transmitter, from which point the circuit extends by way of the 298 AMERICAN TELEGRAPH PRACTICE armature of the transmitter to the "split" or joint-circuit made up through the differential windings of the two relays, half of the current reaching the earth via one winding of each relay and the artificial-line rheostat, while the other half reaches the earth at the distant station via the companion windings of the relays, the line wire, and the distant end apparatus, that is, the current divides equally when the resistance of the rheostat is made to equal the resistance of the line wire plus the resistance of the terminal apparatus at the distant station. It is apparent, owing to the shunt circuit established around the 1,200- ohm added resistance by the transmitter armature, that the only appreciable resistance inserted before the differential circuit is reached is that of the internal resistance unit (in this case having a value of 600 ohms) connected into the battery lead between the dynamo and the pole-changer, and which serves to protect the machine in case of accidental short-circuit. It is evident, too, with the armatures of the pole-changer and the trans- mitter in the positions shown, that the nearest ground contact is at the dis- tant station and at the end of the home artificial-line rheostat circuit, each ground being equally distant ohmically when the resistance of the artificial- line circuit is made to equal the resistance of the circuit to ground at the distant station. As long, therefore, as the transmitter armature remains in the closed position the full potential of the dynamo is impressed upon the line, and the operation of the pole-changer results in positive and negative currents of maximum strength being sent to line as the armature tongue of the pole- changer is moved into contact with the open or closed contact respectively. Assume now that the key circuit controlling the operation of the trans- mitter is opened, permitting the retractile spring to withdraw the armature into contact with its back-stop, it will be seen that the nearest ground con- tact is distant in ohms from the source of e.m.f. 600 ohms (internal), plus 1,200 ohms (added), plus 900 ohms (leak), or 2,700 ohms; and further, that the point X is two-thirds of the distance (ohmically) to the nearest ground, for, 600 plus 1,200 is two-thirds of 2,700. This means that at the point X, while the transmitter armature tongue is in contact with its back-stop, the voltage has dropped from 385 to one-third of 385, or omitting fractions, 1 28 volts. 1 With these particular values of added and leak resistances we have a ratio of 3 to i, with a long-end potential of 385 volts and a short-end potential of 128 volts, the former being impressed on the line when the armature tongue of the transmitter is in contact with its front-stop, and the latter when the tongue is in contact with its back-stop. This in turn insures that maximum current will obtain in the main-line circuit while the signal- 1 See Fall of Potential in an Electric Circuit, page 88. THE QUADRUPLEX 299 ing key which controls the operation of the transmitter is closed, and that minimum current will obtain in the main-line circuit while the same key is open. The result, therefore, is that closing the transmitter key 1 causes a current volume in the main-line circuit of sufficient strength to attract the armature of the neutral, or common-side relay at the distant station, while opening the transmitter key circuit results in the. main-line current strength being so reduced that the armature of the distant neutral relay is withdrawn from the closed position due to the tension of the retractile spring attached to it. So far as the operation of the neutral relay at the distant station is con- cerned it is immaterial whether the armature of the home pole-changer is in contact with the +385-volt dynamo or the 385-volt dynamo, as the neutral relay responds to either polarity, provided maximum current strength obtains in the main-line circuit. The operation of the polar relay at the distant station being dependent upon current reversals, regardless of the strength of the current, is as a conse- quence under the control of the home pole-changer. As long as the armature tongue of the home pole-changer remains in contact with the negative battery connection, the armature of the polar relay at the distant office will be in one position, and as long as the armature of the pole-changer is in contact with the positive battery connection the armature of the distant polar relay will remain in the opposite position. METHOD OF DETERMINING THE REQUIRED OHMIC VALUE OF RESIS- TANCE COILS TO USE IN THE FIELD KEY SYSTEM TO OBTAIN ANY DESIRED PROPORTION Instead of a current ratio of 3 to i, it is sometimes advisable to maintain a ratio of 3 1/2 to i, or 4 to i, etc. Convenient and simple formulae, for determining the values of the added and leak resistances for any given ratio are given herewith, where R, represents ratio, B, represents internal resistance, RI, represents added resistance, R 2 , represents leak resistance, BR then R-> = -^ or 2 = B (R-i) - For the sake of clearness the transmitter local key circuit has been omitted. 300 AMERICAN TELEGRAPH PRACTICE The value of the internal resistance is usually selected with regard to the measure of protection to be given the generators, and in practice may be 600, 300, 200, 100 ohms, or less. Suppose, for instance, that with an internal resistance of 600 ohms a certain circuit is to be operated on a 3 to i basis: R, has a value of 3, B, a value of 600, then added resistance = 600 X (3 i) = 1,200 ohms leak resistance = - = 900 ohms, or the same as the 3-i values indicated in Fig. 258. When the values of the internal, leak, and added resistances are known, the ratio, according to the formula, would be 600+900+1200 _ 900 ~^ To ascertain the value of the potential at the pointZ (Fig. 258) with any given combination of added resistance, the following formula applies: y E (R-RJ R where E represents the voltage of the generator, R represents the resistance of the whole circuit to the nearest ground, RI represents the point distant in ohms from the source of e.m.f. With the resistance values shown in Fig. 258, X = ^ -=128 1/3 volts. 2700 By the aid of the foregoing formulae, the subjoined table has been com- piled showing the added and leak resistance values required with ratios of 3-1, 3.5-1, and 4-1, with internal resistance values ranging from 50 to 1,000 ohms. THE QUADRUPLEX ADDED AND LEAK RESISTANCE 301 3 t D I 3-5 to i 4 t 3 I Internal resistance Added Leak Added Leak Added Leak 5o IOO 75 125 70 150 67 TOO 200 J 5o 250 140 300 133 200 4OO 300 500 280 600 267 300 6OO 45o 750 420 900 400 400 800 600 1,000 560 1,200 533 500 ,OOO 750 1,250 700 1,500 667 600 ,200 900 1,500 840 i, 800 800 700 ,4OO 1,050 i,75 980 2,100 933 800 ,60O 1,200 2,000 I,I2O 2,400 1,067 QOO ,800 i,3So 2,250 I,26o 2,700 1,200 I,OOO 2,OOO 1,500 2,500 I,40O 3,000 1,333 TERMINAL RESISTANCE It is necessary that the whole or joint-resistance of the terminal apparatus remain the same regardless of the position of the armature of the transmitter at any instant. That this is important is evident from the fact that in the act of balancing, the distant station adjusts the resistance of the artificial-line rheostat to equal the resistance of the line plus the resistance of the apparatus at the other end of the line. Therefore, after the balance has been taken the same value should at all times be maintained, else the outgoing currents will not have identical values in the separate windings of the relays. Figure 259 shows the resistance values of the various elements of a quad- ruplex set at each end of a line extending between stations Y and Z, assum- ing a line conductor resistance of 2,000 ohms, an internal resistance of 600 ohms in each dynamo lead, a long-end potential of 385 volts, and a ratio of long to short e.m.f. of 3 to i. It may be observed that the resistance of the terminal apparatus at each station remains the same at all times. With the armatures of the pole- changer and the transmitter in the positions shown in the diagram the only appreciable resistance presented to incoming currents at station Z (in addi- tion to the relay resistance) is the 6oo-ohm internal resistance B, as it is plain that the 1,200 ohms added resistance is short circuited by virtue of the closed position of the transmitter armature. Should the pole-changer key be opened so that the armature tongue of the latter is withdrawn into contact with its back-stop there is still presented a 6oo-ohm path to ground. 302 AMERICAN TELEGRAPH PRACTICE Referring now to the conditions prevailing at station F: At first sight it would seem that owing to the armature tongue of the transmitter being in contact with its back-stop, a path is presented to incoming signals which has a resistance considerably higher than that which obtains when the arma- ture is in the closed position, but a little consideration will show that with the transmitter armature in the position shown at F the incoming signals have a joint path from the point X consisting of a goo-ohm branch and an 1,800- ohm branch, and calculation will show that the joint-resistance of these two paths is 600 ohms. RESISTANCE OF THE "GROUND" COIL When the attendant at one end of the quadruplexed circuit " takes a line balance, " that is, when he adjusts the resistance of the artificial line rheostat to equal the resistance of the line and of the apparatus at the distant end of the line, it is necessary that the main-line battery at the distant end of the line be removed until the balance is taken. Assume that the attendant at F (Fig. 259) is about to take a balance. In the process of balancing, the attendant at Z is required to "ground" the line (thereby removing the main-line battery at Z) by moving the switch lever S into contact with the ground connection. Inasmuch, therefore, as the balance is taken when a 6oo-ohm terminal resistance is presented at the distant station it is essential that when the switch S is turned back to the regular position (thereby placing battery to line) the terminal resistance presented must remain the same, namely, 600 ohms, if the balance is to hold good while the circuit is in operation. The value of the resistance of the ground coil must be identical with that of the internal resistance B y which also is identical with the joint resistance of the terminal resistance presented to incoming signals when the armature of the transmitter is in contact with the leak circuit. Undoubtedly it will have occurred to the reader that the resistance of the terminal apparatus (600 ohms in this case), in reality forms one branch of a joint circuit, the other branch of which consists of the 3,ioo-ohm circuit to ground made up via the relay windings and the artificial-line rheostat, and, of course, this is true, for in the case under consideration (Fig. 259) the joint resistance of 600 ohms and 3,100 ohms is approximately 503 ohms, but it is evident that the 3,ioo-ohm branch of the joint circuit formed a joint circuit with the 6oo-ohm ground coil at the time the balance was taken, so that the actual total resistance of the terminal apparatus presented to incoming sig- nals is the same regardless of the position of the ground switch. OPERATION OF THE QUADRUPLEX If the reader has mastered the principles of operation of the Stearns duplex and of the polar duplex described in Chapter XIII he should have THE QUADRUPLEX 303 N CV) cr TJ 'w E 0) rfl 304 AMERICAN TELEGRAPH PRACTICE little difficulty tracing out the various operations which take place in the quadruplex while two messages are being transmitted in each direction at the same time. The method of study which gives the best results is for the student to draw diagrams showing the various positions of the transmitter and pole-changer armatures at either end of a quadruplex circuit and from these gain a first hand knowledge of the operation of the relays at each end in response to the operation of the signaling keys at the opposite end of the circuit. In making up diagrams the following schedule of combinations may be used as a guide in tracing the various possible connections : / Home PC open, Distant PC open, \ Home T open, Distant T open. Home PC closed, Distant PC open, Home T open, Distant T open. Home PC open, Distant PC open, Home T closed, Distant T open. Home PC closed, Distant PC open, Home T closed. Distant T open. Home PC open, Distant PC closed, Home T open, Distant T open. Home PC closed, Distant PC closed, Home T open, Distant T open. Home PC open, Distant PC closed, Home T closed, Distant T open. Home PC closed, Distant PC closed, Home T closed, Distant T open. Home PC open, Distant PC open, Home T open, Distant T closed. Home PC closed, Distant PC open, Home T open, Distant T closed. Home PC open, Distant PC open, Home T closed, Distant T closed. Home PC closed, Distant PC open, Home T open, Distant T closed. Home PC closed, Distant PC closed, Home T open, Distant T closed. Home PC open, Distant PC closed, Home T closed, Distant T closed. Home PC closed, Distant PC closed, Home T closed, Distant T closed. THE DAVIS-EAVES, OR POSTAL QUAD 14. < 16. The Postal Telegraph-Cable Company has recently put into service a large number of quadruplex sets arranged as shown theoretically in Fig. 260. It will be recognized that the arrangement constitutes an improved Field quadruplex; the new features consisting of the 5oo-ohm " bridge" coils, the bridge-condenser circuit, the " timed" condenser circuit around THE QUADRUPLEX 305 4 20 306 AMERICAN TELEGRAPH PRACTICE the relays, the 25,ooo-ohm leak circuit from line to ground, also a reduced value of internal, leak, and added resistance. It will be noted, too, that the spark curbing device consists of two i-m.f. condensers shunting the pole-changer battery contacts and provided with a discharge path. The functions of the bridge coils and the condenser circuits have been described in connection with the high-efficiency duplex, Fig. 257, page 296. Owing to the fact that the terminal resistance of the set has been reduced to 300 ohms in place of the 600 ohms formerly used; that the resistance of the polar relay has been reduced from 300 to 200 ohms, and that of the LEAK RHEOSTAT TRJUfSfffTTER Eta * POLAR XffZJfY MAIN CIRCUITS FIG. 261. Actual connections of the "Postal" quadruplex. neutral relay from 150 to 60 ohms, the total resistance of the apparatus remains practically the same as that of the Standard Field quadruplex, and inasmuch as specially arranged paths have been provided for induced line disturbances it is not necessary to employ very high potentials to override line currents from extraneous sources, so that in many instances it is possible to reduce the potential from 385 volts to 250 volts, or less, and still maintain satisfactory quadruplex operation. Figure 261 shows the actual binding-post connections of the Postal quad including all of the new elements referred to above. SINGLE DYNAMO QUADRUPLEX Figure 262 is a sketch of the theory of the connections of a single-dynamo quadruplex arranged with a "double-relay" pole-changer so that one dynamo THE QUADRUPLEX 307 Line Cond'r 300 Ohms FIG. 262. Single dynamo quadruplex. Theory. XOE03TAT, FIG. 263. Binding-post main line connections of the single dynamo quadruplex. Line FIG. 264. Theoretical connections of a set arranged to be used either as a single dynamo quadruplex or as a double dynamo Field quadruplex. 308 AMERICAN TELEGRAPH PRACTICE will serve to furnish currents of both polarities, positive and negative, for main-line purposes. It will be noted that when the armatures of the two instruments which comprise the pole-changer, are in contact : with their front-stops, the negative terminal of the dynamo is placed to line, while the positive terminal of the dynamo is grounded. And, when the respective armatures are in contact with their back-stops, the positive terminal of the dynamo is to line and the negative terminal grounded. Otherwise the connections of the set are the same as in the standard Field quadruplex. FIG. 265. Instrument binding-post connections of combination single dynamo and double dynamo quadruplex. This arrangement is efficient and economical, but its employment is advisable only where constant quadruplex service is not required. Figure 263 shows the actual main-line binding-post connections of the single-dynamo quadruplex. Figure 264 shows a diagram of the required switching connections of a quadruplex set wired to operate either as a single-dynamo, or as a regula- tion two-dynamo quadruplex, while Fig. 265 shows the actual instrument binding-post main-line connections of a set arranged to operate as a single- or a double-dynamo quadruplex. METALLIC CIRCUIT QUADRUPLEX Figure 266 shows the instrument and battery switchboard connections of a quadruplex set arranged for metallic circuit operation. The system illustrated is arranged so that it may be used for grounded- circuit, or metallic-circuit operation. Throwing the switches to the left THE QUADRUPLEX 309 provides for metallic-circuit operation, and throwing the switches to the right provides for single-line grounded circuit operation. It happens sometimes that all of the wires along a particular route are so affected by induction from neighboring high-tension power circuits that quadruplex operation over single grounded circuits is impossible, or at best very unsatisfactory. In such cases the metallic circuit quadruplex (using two main-line wires looped) has been found to give satisfactory results. FIG. 266. Field quadruplex arranged to be operated over a ground return circuit or a metallic circuit. THE NEUTRAL-RELAY "KICK," AND THE "BUG-TRAP" METHOD OF COUNTER- ACTING ITS EFFECTS ON SOUNDER SIGNALS As the student constructs diagrams showing the various positions of the armatures of the transmitters and pole-changers at each end of a quadruplex circuit as suggested in the schedule showing the sixteen possible combinations, it will very likely occur to him that the various battery and condenser actions incident to the operation of the four signaling keys, will result in constantly recurring intervals of no magnetism in the cores of the relays. So far as the polar relay is concerned the period of no magnetism is of no consequence as its armature is held by a permanent magnet in the position into which it was last moved due to current in the relay coils, and will remain there until the current flowing through the relay coils has been reversed. Not so with the neutral relay, however, as the armature of the latter being acted upon by a retractile spring is drawn away from the electromagnet and 310 AMERICAN TELEGRAPH PRACTICE the closed contact point immediately upon the cessation of current in the coil windings, or upon reduction of the strength of current actuating the magnet. It is understood that the direction of the current is reversed each time the armature of the pole-changer is caused to move from the positive to the negative battery contact, and vice -versa. As the armature of the transmitter at Z (Fig. 267) is in the closed position, the armature of the neutral relay at Y will be in the closed position. If now the pole-changer at Z is operated, the first movement of its armature will be from the negative to the positive THE QUADRUPLEX 311 battery terminal, which results in a reversal of the direction of current in the windings of the relays. This change, of course, requires time, as the current due to the negative battery must disappear and the current due to the positive battery must build up to the strength required to hold the armature of the neutral relay against its front-stop, and this entails an interval during which the magnets of the neutral relay are not magnetized. It is evident then, that at this instant the armature of the neutral relay due to the action of the retractile spring departs from its front-stop. Fortunately this interval of no-magnetism is brief, and the armature has time to recede but a short distance before the magnetism has again built FIG. 268. Repeating Sounder "bug-trap." up to the strength necessary to attract it, due to current from the opposite battery pole, which in the interim has been applied to the line. But, although brief, the interval of no-magnetism frequently is of sufficient duration to cause a false signal to be produced on the reading sounder operated locally by the tongue of the neutral relay. This disturbance is sometimes referred to as the "B side kick" and its objectionable effects are counteracted by devices variously arranged and known as " bug-traps." The Repeating Sounder. The earliest adopted method of bridging over the period of no-magnetism was that employing a "relaying" or "repeat- ing" sounder, so called. Figure 268 shows schematically the wiring of the receiving side "local" circuits of a quadruplex set equipped with a repeating sounder. It will be seen that the armature lever of the reading sounder S, is in the marking position while the armature tongue of the neutral relay is in contact with its front-stop, and it is evident that the armature of the reading sounder will not be released until the armature of the neutral relay has been drawn into contact with its back-stop. Obviously, a time element is introduced which consists of the time taken by the relay tongue to traverse the gap maintained between its front-stop and back-stop plus the time taken for the magnetism in the sounder magnets to build up to a strength sufficient to attract their armatures. The repeating sounder was purposely equipped with a heavy armature lever in order that it would possess considerable inertia and as a result thereof, be slow acting as compared with an instrument equipped with a light lever. 312 AMERICAN TELEGRAPH PRACTICE It was found in practice that during the period of reversal when all other conditions were favorable the tongue of the neutral relay had time to travel but a minute distance away from its front-stop before being called back by the resumption of magnetism in the cores of the magnet. It might here be observed that when the short-end is to line at the distant station, there are no deleterious effects resulting from the reversals of polarity consequent to the operation of the distant pole-changer, as now the armature of the home neutral relay remains in contact with its back-stop and the armature of the reading sounder is in the non-marking position. The Gerritt Smith Arrangement. The Gerritt Smith neutral relay arrangement has in the past been used upon quadruplexes in both Western Union, and Postal Telegraph service, and at the present time is in use on a number of quadruplex sets on various railroad telegraph systems. A theo- retical sketch of the arrangement is shown in Fig. 269. In the "Postal's" service the line coils were wound to a resistance of 150 ohms, and the "extra" coil to a resistance of 400 ohms, and as an additional protection the Diehl bug-trap (to be described presently) was also employed. Line FIG. 269. The Gerritt Smith neutral relay arrangement. Figure 269 shows that in the "Smith" arrangement two 4oo-ohm coils form the "divide" for the main and artificial lines. The insertion of these coils is for the purpose of establishing a momentary difference of potential through the "extra" coil and condenser circuit which is "bridged" across the main-line and artificial-line circuits when battery is applied at the distant end of the line. When the resulting actions are traced it will be seen that while either pole of the distant battery is to line a difference of potential is established across the terminals of the condenser which results in the latter becoming "charged." At the instant of reversal of polarity at the distant station the condenser discharges through the extra coil, in a direction the reverse of that in which the operating current in the main- or artificial-line coils had been flowing. The result, therefore, is that the "turn-over" of magnetism in the cores of the relay magnets is hastened considerably, and the period of no-magnetism correspondingly shortened. THE QUADRUPLEX 313 The insertion of the 4oo-ohm coils at both ends of the line naturally increases the total resistance of the circuit 800 ohms, which is an undesirable thing to do, as the increased resistance reduces the efficiency possible where these 400-ohm resistances are not inserted in the line. One method of elim- inating the objectionable additional resistance, which permits of retaining the efficacious features of the Smith arrangement is that of transposing the positions of the polar and the neutral relays, for the purpose of utilizing the resistance of the coil windings of the former in place of the 4oo-ohm resistance coils usually connected at the divide. Figure 270 shows a view of the Smith Neutral Relay, in which it may be seen that this instrument is identical in construction and design, with the FIG. 270. Neutral relay with extra binding posts to which are attached the terminals of the extra windings. ordinary form of neutral relay, with the exception that two additional binding-posts are provided for the terminals of the extra coil, which consists simply of a third winding over the cores of the magnets carrying the main- line and artificial-line windings. The Diehl "Bug-trap." Another very effectual method of tiding over the period of reversal is that whereby a "bug-trap" relay (BT, Fig. 271) in connection with the back-stop of the neutral relay, is used to introduce a time element during which the armatures of the neutral relay and the bug-trap relay must traverse the gap separating their back- and front-stops before a signal will be registered on the reading sounder. A glance at the diagram will show that while the armature of the neutral relay may flutter uncertainly against its front-stop during reversal of the long-end battery at the distant end of the line, the armature of the reading sounder will remain undisturbed and in the marking position until the armature tongue of the relay has fallen into contact with its back-stop. The Diehl bug-trap is extensively employed in the quadruplex service of the Postal Telegraph- Cable Company. 314 AMERICAN TELEGRAPH PRACTICE The Differential "Bug-trap." Figure 272 shows a "bug-trap" arrange- ment employing a differentially wound bug-trap relay which in a somewhat different manner accomplishes the same purpose as the Diehl bug-trap relay, and the repeating-sounder arrangement previously described. Bug-trap Suitable for Use on Neutral Side of "Decrement" Quad. While general practice pro- vides for the operation of the neu- tral side relays by an " increment" of current, there may be special cases where it is advisable to keep the long- end battery to line normally, and to operate the distant neutral relay by a " decrement " of current. To make this possible the only circuit change necessary at the send- ing end is to transpose the neutral side transmitter connections so that the closed key will cause the lower potential to be placed to line and the open key the higher potential. When a quadruplex is so arranged that the operation of the neutral relays is the result of a decrement of current strength, it follows that the reading sounder must " close" when the armature tongue of the line relay makes contact with its back- stop instead of with its front- stop as with the more com- mon arrangement. Figure 273 depicts a differential bug-trap ar- rangement which provides that the armature of the reading sounder will be in the marking position while FIG. 271. The Diehl bug- trap. FIG. 272. Differential bug- trap relay. the armature tongue of the neutral relay is in contact with its back-stop, and in the non-marking, or "spacing" position while the relay armature is not in contact with its front-stop. The "Condenser" Bug-trap. A neutral-side reading sounder arrange- ment used in British Post Office telegraph practice which has been found to give excellent results, is illustrated schematically in Fig. 274. The reading sounder S, has a resistance of 1,000 ohms, but as there is a 9,ooo-ohm shunt around the coils of the sounder the actual or joint-resistance of that portion of the circuit is 900 ohms. The 5o-ohm coil is placed in the circuit for the purpose of curbing the THE QUADRUPLEX 315 sparking that would appear at the contact points of the relay due to the discharge from the condenser when the circuit is completed. The capacity of the condenser may be varied from 2 to 8 m.f. according to the require- ments of the line wire operated, while the resistance in series with the sounder may be varied from 100 to 700 ohms in steps of 100 ohms. The purpose of the variable resistance is to "time" the discharge from the con- denser and the discharge due to inductance in the coils of the sounder, thus making it possible to prolong the magnetization of the cores of the sounder magnets over the period required to maintain the armature of the reading sounder in the marking position while the armature of the relay momentarily breaks circuit during the periods of reversal at the distant station. 2to8mf. BT ; FIG. 2 73 . Differential bug-trap arrange- ment for use on neutral side of "decrement" quadruplex. FIG. 274. Condenser bug- trap. If the circuits shown in the diagram are carefully traced it will be seen that while the armature of the relay is in the closed position the condenser takes a charge due to the difference of potential which exists across the termi- nals of the sounder. Now, if while the relay tongue is in contact with its front-stop the polarity to line at the distant station is reversed, there will be a momentary break in the sounder circuit controlled by the armature of the neutral relay. At the instant, however, that this occurs the condenser discharges through the only circuit presented to it through the sounder. The discharge from the condenser and the discharge due to the inductance of the sounder magnets results in prolonging the magnetiza- tion of the cores of the sounder until current from the opposite battery pole at the distant end of the line has had time once more to resume control of the armature of the neutral relay. THE FREER SELF-POLARIZING NEUTRAL RELAY Of the various attempts that have been made from time to time, to develop a type of relay for the "second" side of the quadruplex, which would meet the requirements more satisfactorily than the ordinary types of neutral 316 AMERICAN TELEGRAPH PRACTICE relay; the product which has survived longest is that known as the Frier relay, the theoretical arrangement of which is illustrated in Fig. 275. The moving element the armature of this relay is pivoted in a socket formed in the pole-piece of an extra electromagnet which is wound differen- tially and connected into the main line, and artificial line circuits in the same manner as the regular line coils C and C\. A current in the extra coil Cz will result in the core of that magnet having north polarity at one end and south polarity at the other end. Now if the end of the core upon which the armature (Fig. 275) stands, is at a par- ticular instant a north pole, the armature extending upward between the pole-faces of the two regular magnets will be magnetized inductively and have a north polarity, and, as obviously the windings of the two regular coils are so connected that when current flows in the circuit made up through them, the pole-pieces facing the arma- ture will at all times have opposite polarities; it is evident that the polar- ized armature will move toward the pole-piece which at that instant is a south pole. The office of the extra coil is to maintain the polarity of the armature at all times in opposition to the coil Ci Local FIG. 275. Theory of the Frier neutral relay. regardless of the direction of the current flowing in the circuit. In Fig. 275 the magnet on the left is shown as presenting a north pole, while the magnet on the right presents a south pole to the armature, and as the latter possesses north polarity the result is that it has been repelled from the left-hand magnet and attracted by the right-hand magnet. Should the line current at the distant station be reversed while the armature of the relay is in the position shown, the right-hand magnet will present a north pole to the armature; but as at the instant the magnetism in the right-hand magnet is reversed, the magnetism of the core of the extra coil also is reversed (and consequently, also the polarity of the armature) the armature remains in connection with the closed-contact as long as the long-end battery at the distant station is to line. The method by which the armature of the relay is made immune to short-end reversals from the distant station, is the same as that employed in the operation of the ordinary neutral relay; namely by attaching a retract- ile spring to the armature, and giving it a tension sufficient to hold it away from the closed-contact when the short-end only is to line. Adjustment of the Freir Relay. Under ordinary line conditions, the best adjustment to give the armature of the Freir relay is such that the gap between the armature and the left-hand magnet will be about twice as THE QUADRUPLEX ' 317 wide as that separating the armature from the pole-face of the right-hand magnet. OTHER METHODS OF TIDING OVER THE PERIODS OF REVERSAL From what has hereinbefore been stated in regard to the necessity of establishing a time interval during which the armature of the second-side relay may be made to ignore, as it were, the period of " no-current" which exists while the long-end current at the distant end of the line is reversed, it would seem of the utmost importance that the "reversal" should be made with the greatest possible speed, and that anything which can be done to hasten the movement of the armature of the pole-changer between back- and front-stops during operation, will have a directly beneficial effect upon the operation of the distant neutral relay, since reducing the period of no- current correspondingly minimizes the task set for the bug-trap arrangement- employed in any given system. It might here be restated, as covered more fully elsewhere herein, that close adjustment of the points of the pole-changer between which the arma- ture tongue plays, will accomplish more in the way of reducing the interval of no-current, than anything else that may be contributed. The best prac- tice in this regard is that wherein the adjustment of pole-changer and trans- mitter points brings the opposite contact-points as close together as sparking will permit. THE SHORT-CORE OF THE NEUTRAL RELAY In the design of nearly all forms of neutral relays, advantage has been taken of the fact that by reducing the length of the iron core the magnetism builds up to its full strength more rapidly than where comparatively long cores are employed in winding electromagnets. NEUTRAL RELAYS WITH HOLDING COILS When the Jones quad was the standard of the Postal Telegraph-Cable Company, prior to the adoption of the Field key system, the neutral relay was equipped with a third coil which at the instant of " reversal" was charged by means of a form of induction coil known as the inductorium. THE INDUCTORIUM The inductorium consisted of an iron core upon which three coils were wound, one coil connected in series with the main line coils of the two line relays, another in series with the artificial line coils of the relays, while the third 318 AMERICAN TELEGRAPH PRACTICE coil had its terminals connected directly to the winding of the third coil of the neutral relay; the latter when energized serving to hold the armature of the relay in the closed position for a brief instant, or during the reversal of the distant line-battery. Reversal of the battery at the distant station resulted in an induced current being set up in the third coil (the secondary winding) of the inductor- ium, which in turn energized the " holding-coil" of the neutral relay at the critical moment, thereby tending to hold the armature of the relay in contact with its front-stop during the moment of no-magnetism in the line coils of the relay. HOLDING COIL OF THE NEUTRAL RELAY EMPLOYED IN THE PRESENT WESTERN UNION QUADRUPLEX The quadruplex which at the present time is the standard in the service of the Western Union Telegraph Company, includes a form of neutral relay which is equipped with a holding coil, somewhat similar in action to the in- ductorium. The coil is placed in series with a condenser and is connected across the main and artificial lines, being energized at the instant the distant pole-changer "breaks" contact with either battery pole. The effect of this holding coil upon the efficiency of the second side of the Western Union quadruplex, will be described more in detail, when, presently, the principles of that quadruplex are explained. THE WESTERN UNION QUADRUPLEX Figure 276 shows a theoretical diagram of the quadruplex recently a- dopted as standard by the Western Union Telegraph Company. The improvements incorporated in this system include a form of pole- changer invented by Mr. S. D. Field and illustrated in skeleton in Fig. 277. Like the pole-changer used in connection with the Postal Telegraph- Cable Co.'s quadruplex, this instrument may be used also as a " transmitter " on the second side of the quad. The magnet on the left of the armature has a solid iron core, while the magnet on the right has a laminated iron core and is somewhat shorter than the former. Each magnet is wound to a resistance of 4 ohms, making a total of 8 ohms for both magnets in series, and the re- spective coils are so wound that the magnetism developed in each pair of coils when the transmitting key is depressed, is of such polarity that the ac- tion of one magnet opposes that of the other. The armature is situated between the opposing pole-faces of the electromagnets and has attached to it a retractile spring which holds the armature normally toward the left-band magnet. When the transmitting key which controls the operation of the pole- changer is depressed, both magnets are energized, but the magnetism builds up THE QUADRUPLEX 319 cr T3 3 T3 d 320 AMERICAN TELEGRAPH PRACTICE much more rapidly in the right-hand magnet than in the other, due to the fact that the former is considerably shorter and that it has a laminated iron core, and to the further fact that in practice the coils of the longer magnet are shunted with a resistance coil, the total result of which is that the left-hand magnet does not acquire its maximum magnetic strength until the armature has been drawn into contact with the main-line battery contact on the right; remaining there until the signaling key is released or opened. -o P 8 Ohms- FIG. 277. Form -|of pole-changer and transmitter used in connection with the W. U. quadruplex. At the instant the signaling key is released thus opening the battery circuit through the coils 6f the pole-changer the laminated core of the right- hand magnet instantaneously loses its magnetism, permitting the retractile spring aided by the slowly disappearing magnetism in the long core, to rapidly draw the armature into contact with the opposite battery contact. The object aimed at is to hasten the transit of the armature, thereby reducing to a corresponding degree the interval during which the main-line battery is not applied to the line. THE W. U., NEUTRAL RELAY Figure 278 shows in skeleton the usual differential windings of the main- line and artificial-line circuits, each wound to a resistance of 350 ohms, while situated immediately above the line magnet is shown a "holding" magnet. It is evident from the circuit arrangements illustrated in Fig. 276 that this quadruplex embodies principles common to the "bridge" and to the " differential" multiplex systems, inasmuch as the polar relay occupies a position the same as that occupied by the relay used in the bridge duplex, while the neutral relay has one of its windings connected in series with the main-line wire, and the other in series with the artificial-line circuit. THE QUADRUPLEX 321 The Holding Coil. As shown at H, Fig. 276, the holding coil is con- nected across the main and artificial lines in series with a condenser HC which accumulates a charge while battery is applied to the line at the distant station. As the distant pole-changer armature leaves either battery contact, the home condenser almost immediately thereafter discharges through the path provided for it through the holding coil thus tending to hold the armature of the neutral relay in the closed position during a period which approximates in duration that required by the armature of the distant pole-changer to travel from one battery contact to the other. FIG. 278. Neutral relay used in connection with the W. U. quadruplex. The Impedance Coil. In the bridge duplex, Fig. 233, page 268, the four arms of the bridge consist of the line wire, the artificial line, the arm r f , and the arm r. The latter two resistances have identical values. In British Post-office telegraph practice each arm is given a value of 3,000 ohms. In the quadruplex -under consideration the corresponding arms are represented by the two windings of an impedance or retardation coil ($U, Fig. 276) which has a circular core built up of iron wires forming a closed magnetic circuit. The windings are differential, each coil having a resistance of 500 ohms. In the bridge duplex, employing non-inductive resistance units to form the two "bridge" arms, the operation of the polar relay is dependent upon the existing difference of potential across the terminals of the relay, which necessitates that in order to have an operating current of sufficient value to insure quick responce of the relay armature, the resistance of the "arms" of the bridge must be high enough to maintain the required difference of potential. When the non-inductive bridge arms are replaced by arms possessing a considerable amount of retardation or impedance, the operation of the polar 21 322 AMERICAN TELEGRAPH PRACTICE relay is not dependent solely upon the difference of potential existing across its terminals, as, in this case (see Fig. 276) the received current finds a less obstructed path through the relay than through the upper arm of the bridge coil, because the latter presents " impedance" which "retards" the flow of current into it: at least, the first part of each received current wave is diverted through the relay, causing it to actuate its armature without having to wait for the current due to difference of potential across the bridge arms. The Ohm's law current, of course, builds up gradually as the magnetic inertia of the retardation coil is overcome, and, if the cir- cuit is properly timed, in this regard; comes along in time to hold the armature of the relay in the desired position. In view of the fact that the first part of the received impulse is diverted through the relay circuit it is not necessary (on account of the inductance possessed by the alterna- tive path) to have high ohmic resistance in the bridge arms. In the Western Union arrangement, there- FIG. 279. 5-U impedance coil. fore, each coil has a resistance of 500 ohms instead of the 3,000 ohms per arm used in the old form of bridge duplex. Indeed, it has been found prac- ticable in ocean cable duplex operation to employ, bridge arms having a resistance of 1 5 ohms each, but in this case the coils are of large dimensions and have an inductance of 15 henries each. 1 EFFECT OF THE $-U COIL UPON OUT -GOING CURRENTS The fact that the windings of the coil are differential, and that the action of one coil neutralizes the action of the other; so far as the magnetic effects produced in the iron are concerned, provides that to out-going currents there will be no appreciable retardation. It is obvious that with equal current strengths in each coil of the bridge, no magnetism is produced in the core. This being the case the situation is the same as if no iron core were inserted within the coils; or, as if the coils were simple solenoids. We are dealing with the characteristics of the magnetic field set up in the space immediately surrounding the coil windings, and inasmuch as there is no magnetism in the core, the " extra current" due to self-induction will be of low value in comparison with that which would be produced due to the stronger magnetic field were the core magnetized. Consequently, as the 1 The "bridge-resistance" used in ocean cable work is known as the Brown magnetic bridge, being the invention of Mr. S. G. Brown. THE QUADRUPLEX 323 impedance presented to the out-going impulses, would consist mainly of the reverse, or opposing currents due to self-induction of the coil, it is at once apparent that the impedance will be less when the core is not magnetized. Naturally there will be a certain amount of "choke" due to the con- tiguous turns of the winding of the coil in either arm, but this is small in comparison with what it would be in one coil, were the circuit through the companion coil opened; or, if the windings were not differential. The amount of retardation natural to the coil winding graduates the rise and fall of the out-going currents, and this in connection with the spark con- denser (SC, Fig. 276) reduces considerably the strength of the electrostatic induction which otherwise would take place between the line wire and neighboring conductors on the same pole line, or in the same cable. THE EFFECT OF THE 5 -U COIL UPON THE HOME RELAYS It is needful now to look to the effects produced in the home receiving relays due to the action of the bridge coil. Obviously, the only time when there is no magnetism in the core of the bridge coil is when identical current values obtain in both main and arti- ficial lines. The operation of the distant pole-changer, naturally, causes magnetic variations in the core of the retardation coil at the home station, and the extra currents produced as a result thereof traverse the coils of the polar relay in a direction which aids the incoming currents from the distant station in effecting the movement of the armature of the relay in the desired direction. Also, under certain conditions during the period of reversal at the distant station, the instant the armature of the pole-changer leaves either battery contact, magnetic variations in the home bridge coil produce extra currents, which, having a path across the main line and artificial line by way of the holding coil of the neutral relay, aid the discharge current from the condenser HC in holding the tongue of the neutral relay in contact with its front-stop during the critical period. OPERATION OF THE W. U. QUAD By referring to Fig. 276, it will be seen that the line currents are furnished by two dynamos, the negative pole of one machine being connected to the " closed" contact of the pole-changer, while the positive pole of the other dynamo is connected to the "open" contact of the pole-changer; the opposite pole of each dynamo being grounded. The ratio of long-end to short-end current is 3 to i, and the method employed to obtain the desired proportions is the same as in the Field quadru- plex. In Fig. 276, the added resistance is shown at AR and the leak resistance at LR. 324 AMERICAN TELEGRAPH PRACTICE At the station shown on the left the pole-changer PC is represented as sending a positive current to line, the strength of which has been reduced to the short-end value on account of having a joint-path; on the one hand via the added resistance AR, to ground at the distant station, and at the end of the artificial line at the home station : on the other hand from the pole- changer armature to ground at the home station via the leak resistance LR. None of the out-going current will pass through the coils of the polar relay provided the resistance of the artificial line has been adjusted to equal that of the main-line and distant apparatus to ground, for the reason that under such conditions there is no difference of potential across the terminals of the relay. The same is true of the holding coil H, of the neutral relay. In the case of the neutral relay and the retardation coil, it is seen that each of these instruments has wound upon its iron core one coil in series with the main-line wire, and one coil in series with the artificial line circuit. In other words each instrument is wound differentially, with the result that when equal current strengths obtain in each winding, the tendency of one coil to magnetize the core is nullified by the action of the companion coil. It is plain, therefore, that under well-balanced conditions the home relays are unresponsive to out-going signals. The action of the received current may be traced by assuming that a short-end impulse has been transmitted from the station on the left to the station shown on the right (Fig. 276). Obviously, the current passes through the main-line coil of the neutral relay NR, and through the holding-coil H (in the winding of the latter the current exists only while the condenser HC is taking on its charge), but as the retractile spring attached to the armature of the neutral relay has previously been given a tension which holds it in contact with its back-stop until the long-end potential has been applied at the distant end of the line, the relay is unaffected. After passing through the line coil of the neutral relay the received impulse finds two paths open to it; one through the polar relay PR, and one through the coil 5 U. In attempt- ing to enter the line coil of the latter, the current upsets the magnetic balance of the coil by magnetizing the core (due to the increased current volume now flowing in one winding over that flowing in the other) with the result that the extra current of self-induction thereby created, opposes the flow of the line current; momentarily at least, or until the head end of the received current wave has been diverted through the path containing the polar relay. As the armature of the polar relay is moved into the closed or marking position, the current builds up through the line winding of the impedance coil, and joins that portion of the current which has passed through the polar relay and the other winding of the impedance coil, passing thence to ground via the transmitting instruments and dynamo. Figure 280 shows the actual instrument binding-post connections of THE QUADRUPLEX 325 Line Lightning Arrester FIG. 280. Instrument main line binding-post connections W. U. quadruplex. 326 AMERICAN TELEGRAPH PRACTICE the Western Union quadruplex. The "Line resistance box" shown in the upper right hand corner of the diagram (an enlarged view of which is shown in Fig. 281) is made up of two independent variable resistances, of 1,250 ohms total each, equipped with a common double-lever switch for the purpose of throwing equal amounts of resistance into the main line and into the arti- ficial line simultaneously. One use to which this resistance is put, is to increase the resistance of comparatively short lines which it is desired to operate quadruplex with the regular long-line potentials. Inserting additional resistance in both main and artificial lines adds to the electrical length of line wires which in FIG. 281. "Line" resistance box W. U. quadruplex. themselves would be so low in resistance that the currents from the regular dynamos would be excessive. The insertion of added resistance in the line immediately in front of the home apparatus is of considerable benefit in caring for quick changes in the insulation of the line wire during wet weather, for the reason that with 800 or 1,000 ohms of the external circuit perfectly insulated from the ground (as would be the case when that much added resistance is inserted in series with the line) the insulation of the entire line, per electrical mile, is considerably higher, and as a consequence permits of greater variation in the resistance of the exposed section of the line, without seriously affect- ing the "balance." It is necessary, of course, when such additional re- sistance has been inserted in the line at one end of a quadruplexed circuit, to notify the distant office of the fact, so that the line balance at the distant end may be changed to compensate for the added resistance. THE QUADRUPLEX THE MILAMMETER 327 The milammeter shown in the diagram, Fig. 280 (an enlarged view of which is shown in Fig. 282) is connected in series with the polar relay in the bridge circuit for the purpose of facilitating the operation of balancing. FIG. 282. Milammeter used in "balancing" duplexes and quadruplexes. tO FIG. 283. W. U. artificial-line rheostat. 328 AMERICAN TELEGRAPH PRACTICE THE QUADRUPLEX 329 Figure 283 shows the binding-post connections of the artificial-line rheostat. Figure 284 is a diagram of the connections of both main-line and local circuits of the W.U. quadruplex/ The spring-jacks SJ represent the loop- board terminals of the receiving and sending " sides" of the polar and com- mon sides of the set. THE BRITISH POST-OFFICE QUADRUPLEX The quadruplex system used in British Post-Office telegraph service, is, in theory, practically the same as the original American quadruplex systems. A theoretical diagram of the main-line circuits is shown in Fig. 285, a consideration of which will show that the " increment key" IK serves the same purpose as the transmitter employed on the second side of the American systems; namely to place either the short- or the long-end battery to line, while the "reversing key" RK serves as a pole-changer. Line FIG. 285. In the diagram submitted herewith the transmitting keys are shown in American conventional outline, for the purpose of clearly portraying the action of each instrument. The actual appearance of one key is about the same as that of the other. Both are larger and more massive than the American key, and the contact points which are mounted at the end of the key remote from the hard-rubber knob are enclosed in a dust-proof metal cylinder with a glass top, see Fig. 2850. The reversing key, or A key as it is termed in Great Britain has its battery contact point so arranged that the continuity of the circuit from the battery is preserved while the lever passes from positive to negative battery terminal or -vice versa. Obviously a momentary short circuit exists during the reversal of current, the same as in the case of the continuity pre- serving transmitter employed in connection with the Stearns duplex. The increment, or B key is so constructed that the battery is never cut off from the A key. In order that this may be satisfactorily accomplished 330 AMERICAN TELEGRAPH PRACTICE it is necessary to so adjust the contact points that the lever will make con- tact with one battery terminal before breaking with the other. For the purpose of curbing the sparking at contact points when full potential is to line a loo-ohm resistance coil is connected as shown at R } Fig. 285. It is necessary that the resistance of the battery should be the same regardless of the position of the keys at any instant, otherwise the balance FIG. 2850". Transmitting key, B. P. O. quadruplex. at the distant station would be disturbed when the positions of the keys are altered. It is necessary, therefore, to insert a resistance r in the tap wire equal to the resistance of the long-end battery, plus the loo-ohm spark coil. With the key IK in the position shown in the diagram the internal resistance; including that of the short-end battery, is the same as that obtaining when the key is depressed. The type of polar relay used is that known as the Wheatstone relay. The differential neutral relay, or B relay, has its local connections arranged as shown in Fig. 274. CHAPTER XV "BALANCING" DUPLEXES AND QUADRUPLEXES If the reader will review that part of Chapter XIII, dealing with "The Differential Relay" (Fig. 217), and "The Artificial Line " (Fig. 218) he will recognize the necessity for a correct "ohmic" and "static" balance of lines operated duplex or quadruplex. In describing the various duplex and quadruplex systems in the preceding text matter, "methods" of balancing each system have been purposely omitted, for the reason that with all systems the requirements are the same, and that the author during a fairly extensive teaching experience has found that no little confusion exists in the minds of students, when the idea pre- vails that each system so-called of duplex or quadruplex telegraphy can be balanced only by some particular process or method. It is true that the various telegraph administrations furnish "rules" for the guidance of employees in balancing the particular duplex and quad- ruplex employed in each case, some of which will be incorporated herein ; but it should be borne in mind that the procedure in every instance has the same purpose in view, viz., that of establishing an "ohmic" and "static" balance between the main-line wire and the artificial line. THE RESISTANCE, OR "OHMIC" BALANCE In Fig. 286, the artificial line rheostat is shown as having a resistance of 2,400 ohms. Although of secondary importance so far as the balance is Line FIG. 286. Balancing the main and artificial lines. concerned it is well to learn what the unplugged resistance of the rheostat represents. In the typical case presented in Fig. 286, as in all similar installations, the 2,400 ohms in the rheostat at station A represents the resistance of the 331 332 AMERICAN TELEGRAPH PRACTICE line wire beyond the home relay, the resistance of the main-line coil of the relay at B, plus the joint-resistance of the artificial-line circuit to ground (including the resistance of the artificial-line coil of the relay) and the circuit to ground via the battery or dynamo ; or, 600 X (200 + 2400) v =487 ohms 600 + (200 + 2400) and 487+200+1,713 = 2,400 ohms. The indicated resistance of the artificial-line rheostat at A, therefore, represents the 1,713 ohms line resistance, plus the resistance of the mairL-line coil of the relay at B (in this case 200 ohms) plus the joint-resistance of the battery circuit and the artificial-line circuit from the point X at B. The reason why the resistance of the relay at A does not enter into the calcu- lation is that the resistance of one side already balances the resistance of the other. If, for instance, the resistance of the arti- ficial-line coil were regarded as a part of 1 MMMf ^ e tota ^ res i stance f th e artificial line, [ then the resistance of the main-line coil of the relay would have to be regarded as a FIG. 287. Principle of the bridge P art of the total line resistance, and the in- balance. .dicated resistance of the rheostat would re- main the same. That this is true may readily be seen by considering the conditions of current obtaining in a wheatstone bridge circuit, see Fig. 287. The "bridge" coils a and b occupy the same positions in the circuit that the AL and ML coils of the differential relay occupy in a duplex or quadruplex circuit, while the bridge arm R represents the line wire, and the arm X the artificial line of a duplex or quadruplex circuit. The galvanometer G is connected across the terminals of the coils a and b for the purpose of indicating the presence of current in the galvanometer circuit. As was explained in describing the principle of the Wheatstone Bridge (Fig. 138) no current will flow through the galvanometer circuit when the resistance of a equals the resistance of b and the resistance of R equals the resistance of X. Therefore, if in the duplex or quadruplex circuit the resist- ance of one winding of the relay equals the resistance of the other winding, the circuit will be balanced when the resistance of the rheostat is adjusted to equal the resistance of the line wire to ground via the distant apparatus. It is apparent also, that the resistance of the rheostat will remain the same regardless of the resistance of the coils of the home relay, provided the CAPACITY, OR STATIC BALANCE 333 resistance of each coil is the same. In Fig. 286, for instance, the rheostat resistance (2,400 ohms) would remain the same and the home balance would be unaffected if the resistance of the home relay were changed to, say, 20 ohms instead of 200 ohms per winding. The resistance balance may be established by adjusting the resistance of the artificial-line rheostat until the armature of the differentially connected polar relay, neutral relay, or galvanometer remains passive to the operation of the home pole-changer while the line is grounded at the distant station, and in the case of a "bridged" polar relay or galvanometer; until the armature and pointer, respectively, of those instruments indicate "no current." THE CAPACITY, OR "STATIC" BALANCE As pointed out elsewhere herein, when battery is applied to a line the conductor has to be "charged" before the recording instrument at the distant end of the line will indicate the presence of current in its coil windings. When the key K is closed (Fig. 286) a rush of current takes place into the circuit which possesses capacity the line wire passing through the main-line coil of the relay, magnetizing its core momentarily, thereby pro- ducing a kick of the armature which seriously interferes with intended signals. Also, upon opening the key the line "discharges," again, momenta- rily energizing the main-line winding of the relay, thereby producing a false signal. Line tf) > n ii J 8o v9 < Et -<, J2 4>^ G 54 o "S safes. '01 a K g. sar Q ^ u_J^ o.r ^33 *i SET o > 1 /~^Sl 11 i?)^ o S i| gg z LJ o 3TS "S ^ w z t 4NW ^ < 1 H|N| l-ilf^v *>* i n *^ Lnn ' L o 0^ o tltm J l hih. yf isi ' 11 a* 3 1 1 i 7 t S < -JN ^ ^ _c fl if pK 1 5. V5 ^ f i? * i,"i 4,- j ran o ?fff A I'J &* X uJ 5! Jas u. ^ Q TK [R~ r v . 1 U c^ ho Q < * ' ^ VC^T o 1/5 cy a, i 2 1 : g Q LJ X LJ c 0) z 13 o> CO BRA NCH-OFFICE A NN UN CIA TORS 365 5 4 a (v .. (0 c^ J *- a. _i Z o 5 2-3 Z^l C LU * 2 tl H "2 HIGH-SPEED AUTOMATIC TELEGRAPHY 409 controlled by the perforated holes in the paper slip, as the latter is moved along from right to left by the star- wheel W, above the arms 5 and M. The transmitter is constructed so that it may be connected directly to line, the contact points Cu-Cd and Zu-Zd acting as the duplex battery terminals and the divided lever as the pole-changer tongue connected to the main-line wire, but it has been found that the ranges of adjustment are not so limited where the automatic transmitter is employed to operate locally a polar relay, the tongue of which is connected to line and the local contact points of which carry the main-line potentials, plus and minus. The upper and lower halves of the divided lever D-U are mechanically connected, but separated electrically, that is, one is insulated from the other, so that either the lever D, or the lever U, in connection with the lower or upper contact points respectively Cd-Zd, or Cu-Zu, may be used to operate the line instrument. In case the operation is transferred from the upper to the lower contacts, or vice versa, the only alteration in connections required is that the ground contact be transferred from transmitter binding-post U to D or D to U, as the case may be. The significance of the letters D and U may be borne in mind by noting that U refers to the upper pair of contacts, and D "down" or lower pair. The rocking beam is equipped with two pins P, P f , which project out- wardly. The revolution of a driving wheel (within the case of the instru- ment and not shown) which is fitted with a projecting pin near its periphery, causes the rocking beam to move up and down alternately upon a central pivot. The pivoted cranks A and A ' are held against the under side of pins P and P' by springs attached at right angles to the lower extremities of the cranks. Rising from the ends of the two cranks are the rods 5 and M. Actually, the rods- are side by side, one on each side of the star- wheel W. In the sketch the position of one of them has been changed somewhat in order to show both rods. Two adjustable screws B and B f regulate the dis- tance backward at which the rods may be set, the springs S f and S 2 holding the rods against the screws. In their upward movement the rods pass through slots cut in a brass platform. As the perforated tape is moved along the platform by the star-wheel, the rods continuously moving up and down enter the holes in either side of the tape directly as these holes appear over the rods. Above the star- wheel is mounted another wheel a trifle wider than the tape which acts to hold the tape down and permits the projections of the star-wheel to enter the center row of holes in the tape and thus propel it forward. With the transmitter running free, that -is without tape, rods S and M, in response to the movements of the rocking aim, rise and fall alternately. The lower extremity of the upright section of crank A moves to the right when the rod S moves upward; this action pushes the lever acting between the contact points to the right by means of the rod and boss K. The up- 410 AMERICAN TELEGRAPH PRACTICE ward movement of the rod M in the same manner causes K to push the lever U to the right. Were the transmitter connected directly to line, this action would mean that a "make" or marking current would go to line at the instant the rod M rises, and a spacing current would be sent to line when the rod S rises. Thus, with the transmitter running without tape, a series of reversals are sent out producing "dots" in the distant polar relay. Inserting a strip of perforated tape in the transmitter results as follows : Assuming that the marking rod M has risen and entered a hole in the tape and that the tape moves forward three or four spaces before a perforated hole appears above the rod 5, then the marking current will be continued until the spacing rod S has an opportunity to rise. It is obvious that the rod 6" has in the meantime continuously bombarded the tape, awaiting the first opportunity to travel over its full course in response to the tension of the spring 53 and which it has been prevented from doing by having presented before it a portion of the tape in which no perforations have been made. As in the regulation polar relay, the lever of the transmitter must remain on either closed or open-contact point. In the polar relay this is brought about by employing permanent magnets to hold the armature in either position. In the transmitter the same thing is accomplished by the jockey wheel /. It is evident that as the lever moves to the right or left it is held in either position o o o o o ^ by the action of the spring bear- ing down the jockey wheel. Figure 370 shows a sketch FIG. 37o.-Specimen of perforated tlpTbea'ing the of the Plated slip required word "and." to transmit the word "and." The upper holes are those en- gaged by the rod M and the lower ones by the rod S. When the tape is in proper position in the transmitter the lower holes are on the outward side, or toward the attendant, the tape moving from right to left. When unpunched paper is inserted, both rods S and M are pressed down- ward and the pins P, P f , in their motion do not actuate the crank levers A, A'; the lever DU, consequently, does not move and a permanent current is therefore sent to line. 00 If now, slip, perforated, say, with the letter (a) be inserted, then, when rod M rises, it will be free to pass through the first upper hole, and the lever DU will be moved and will send out a "marking" current. When the reverse movement of the rocking beam Y takes place, rod S will be free to pass through the first lower hole, and the current sent by DU will be reversed; a dot will therefore have been sent. On the next movement of the rocking beam, M will be free to pass through the second upper hole, and the length HIGH-SPEED AUTOMATIC TELEGRAPHY 411 of the " spacing" current is consequently precisely equal to that of the pre- vious " marking" current (dot). The " marking" current being now to line, when the rocking beam leaves S free to rise, it is prevented from so doing by the paper, which is not perforated below the second upper hole. In this case, therefore, the " marking" current is kept on until the rod S is again free to rise, which it can do through the second lower hole, and the current is then reversed. It will be seen that the "marking" current is kept to line during movements equal to two dots and the space between, this being the established length of a dash. It is clear, therefore, that when correctly per- forated slip is run through the transmitter any required Morse signals dots, dashes and spaces can be automatically sent to line. Adjustment: One end of the flat spring which carries the jockey wheel /, is attached to a brass piece F, which is in turn screwed rigidly to the frame of the gearing. The upper side of F is V-shaped, and the tension of the spring is adjustable by means of the two screws which fasten it to its support. It should have sufficient tension to enable it to push the lever DU suddenly to the right or left when either of the collets K or K r push it beyond the center of the jockey wheel. The collets K and K' can be adjusted by being screwed forward or back- ward; their correct position may be found by running the transmitter with a blank slip, when the bar should remain unaffected, whether resting in its right or left position. The collets must, however, be sufficiently close to push the bar over the center when the slip is removed, so as to allow the jockey roller to complete the movement. In order to insure reliable action at high speed, it is essential that the spiral springs 5-3 and 5-4 be strong enough to easily overcome the tension of the flat spring acting through the jockey wheel upon the lever. The amount of play allowed between the contact screw C-d and the lever D when it is resting on Z-d, or vice versa, is about 5 mils. The contacts C-u and Zu should be adjusted to suit, so as to preserve similar distances with respect to U. The exact positions of the vertical rods 5 and M are regulated by the screws B, B' ; each of the rods should be so adjusted that it commences to enter a perforation in the slip when the left-hand edge of the perforation is sufficiently clear of the left-hand edge of the rod to allow it to pass through freely. If the screws P or P' are screwed too much either way out of their correct position, the rods will catch against the edges of the perforation, and the mechanism will not act properly. The springs Si and S-2 pull the rods S, M, back against the screws P, P', when they have become sufficiently withdrawn to be just clear of the slip. Although these springs are very light, they must be strong enough to cause the rods to return to their normal positions promptly. 412 AMERICAN TELEGRAPH PRACTICE THE MOTIVE POWER OF THE WHEATSTONE TRANSMITTER Until recently, high-speed transmitters have been operated by weight- driven gears, and while this method permitted the employment of the high- speed system at small offices not equipped with sources of electric power when upon occasion a small office was called upon to handle for a few days a large volume of business, in large offices where automatic equipment is permanently located it is desirable to have transmitters which are driven by electric motors, first, to obviate winding up the weight, and second to obtain constant speeds. Transmitters are equipped with small direct-current motors which are run at constant speed, approximately the maximum speed of the motor. No motor-control rheostat is used. An extension of the motor shaft is fitted with a metal disk which acts as a friction plate. On the face of this friction plate rests the edge of another small disk made up of compressed rawhide held rigidly between two brass plates by means of which the disk is securely attached to its axle. The end of the armature shaft remote from the friction plate is fitted with two tension-springs which act to hold the plate in contact with the rawhide disk. The axle of the disk has on one end a pinion gear which operates the driving axle of the transmitter by means of a clutch. The opposite end of the axle bearing the rawhide disk is hollowed out cone- shaped in order to engage the point of the adjusting screw which determines the position of the rawhide disk on the face of the friction plate. The method of regulating the speed of the transmitter is founded on the principle that the speed through space of various points from center to periphery of a re- volving wheel, is greatest at the periphery and least at the center. The speed-regulating screw as it moves the axle of the friction disk along, results in the friction disk being pushed nearer to the periphery of the friction plate, thus increasing the speed of rotation of the transmitter driving axle. As there is no spring used to withdraw the axle of the rawhide disk when it is desired to reduce the speed by causing the disk to take up a position nearer the center of the friction plate, it is evident that another property of the re- volving wheel is availed of to accomplish the desired end. It is well known that the upper half of a revolving disk or wheel has a motion in the reverse direction to that of the lower half and any device in frictional contact with the side of the wheel, unless restrained, takes on a motion of translation of that portion of the wheel with which it is irt contact, thus when the adjusting screw which holds the friction disk up to its work, is withdrawn the natural tendency is for the friction disk to move inward toward the center of the friction plate and the speed is gradually reduced. The clock-work gearing which drives the moving contacts of the trans- mitter proper is connected with the driving axle by means of a universal clutch. The transmitter proper is detachable from the base, the armature and battery- contact wiring being made to buffer contacts. When the transmitter gets out HIGH-SPEED AUTOMATIC TELEGRAPHY 413 of adjustment and there is a spare unit available it requires but 10 or 12 seconds to remove the defective instrument and substitute one known to be in working order. As the transmitter is set in place the buffer contacts en- gage their corresponding projecting terminal points and the driving clutch engages the driving axle without any action on the part of the attendant except that he place the transmitter in proper position on its brass bed-plate and tighten the thumbscrews. Where i zo-volt current is available, it is customary to use it for the opera- tion of the transmitter motor. The regulation of the speed of the transmitter (and consequently of the speed of transmission) is accomplished by means of the friction drive. A hard rubber knob mounted on one side of the transmitter OD Motor Terminals r D D MKC O FIG. 371. Main line and battery connections of the automatic transmitter. case, accessible to the attendant, permits of regulating the speed at which signals are sent over the line, ranging from 10 words per minute to 300 words per minute. As there is no rheostat control of the motor circuit, it is well to have a resistance of about 100 ohms in each side of the no-volt circuit to prevent heating of the motor. & Revolving the shaft of the motor causes the rocking beam of the transmit- ter to move up and down at a speed corresponding to the speed at which the star-wheel forwards the paper strip. These two related movements are accomplished by means of suitable clock-work gearing. Figure 371 shows an enlarged view of the transmitter main connections, where the automatic transmitter is employed to operate a pole-changer in the form of a standard polar relay. The terminals K, MKC, and MKZ are not used except when the duplex line potentials are connected directly to the transmitter. The terminals marked and + show where the no-volt motor leads are to be connected. When a polar relay is used to control the 414 AMERICAN TELEGRAPH PRACTICE line battery, the main- line and artificial-line binding posts of the relay are connected -to the terminals Z and C, and the terminal U or D is grounded. Either the upper or lower contacts may be used by changing the ground con- nection from U to D, or vice versa. The Wheatstone system has for many years been used on certain lines of the Western Union Telegraph Company, and within the past year or two has been introduced on the lines of the Canadian Pacific Railway Telegraph system, in the operation of a Pacific cable circuit between Montreal, Que., and Bamfield, B. C., with repeaters at Fort William, Ontario, 995 miles from Montreal, and at Calgary, Alberta, 1,256 miles distant from Fort William, also at Vancouver, B.C., 646 miles from Calgary. The distance from Van- couver to Bamfield is 115 miles, including 80 miles of submarine cable. At Montreal dynamo current is used; at Fort William, Calgary and Vancouver, storage battery is used, and at Bamfield, gravity battery. On the Pacific cable circuit, overland through Canada, the question of speed is of secondary importance, and high speeds of transmission are not aimed at. The principal object in employing the Wheatstone system is to insure accuracy. Also, a material advantage accrues from the fact that at a given speed in worols per minute, Wheatstone signals on account of their evenness and regularity, " carry " much better over long circuits than do hand signals at the same speed, resulting in fewer calls for repetition of doubtful words or letters. At a speed of, say, 40 words per minute, using Wheatstone transmission, the total amount of business handled over a circuit in a day exceeds consider- ably the amount of business that would be handled during the same period by means of the Morse key; where the sending operator does not exceed a speed of, say, 40 words per minute. This is due to the fact that in Wheat- stone working there is generally 2 or 3 ft. of slack tape which has been per- forated, between the perforating machine and the transmitter, so that the frequent stops made, from one cause or another, by the perforator operator the sender do not interrupt the continuity of line transmission, which goes on continuously as long as tape is fed to the transmitter. Wheatstone working may be applied to any polar duplex, or polar side of a quadruplex, by providing a three-point switch at each sending end for the purpose of switching the automatic transmitter, or the Morse key into circuit as desired, and by providing a similar switch at each receiving end for the purpose of switching the line wire into the automatic recorder, or the regular polar relay as desired. Where speeds above 150 words per minute are to be maintained, it is necessary to use at the terminal offices and at repeater stations the most efficient and " fastest" polar relays obtainable, otherwise the equipment and connections of the Wheatstone automatic duplex are the same as those of the high efficiency duplex (see Fig. 237). HIGH-SPEED AUTOMATIC TELEGRAPHY 415 THE POSTAL AUTOMATIC The Postal Automatic Telegraph System is identical with the Wheatstone in so far as concerns the preparation of the transmitting tape, and the trans- mission of the signals; but the reception of the signals is accomplished in an entirely different manner, being received by an electromagnetic punch, or " reperforator " which, instead of marking the dots and dashes of the letters on the receiving tape with ink, as in the Wheatstone system, perforates the characters in a continuously moving strip of paper tape, the received tape resembling the transmitting tape, inasmuch as the Morse characters appear thereon in a series of perforations. The improvement in this method as compared with Wheatstone recorder reception, is that the received tape may be passed through a local " reproducer," and the messages copied by ear from an ordinary sounder. The reproducers are motor driven and are under the control of the repro- ducing operator so that the speed of reproduction may be regulated to accord with the ability of the operator. At his convenience the tape may be stopped, pulled back and run through again for the purpose of confirming doubtful words. In practice, therefore, the reproducing operator copies from a "sender" over whom he has absolute control in the matter of speed and of repetition. Moreover, with this system, messages received at relay offices for points beyond, which are equipped with automatic apparatus, may be relayed automatically, simply by passing the received tape through an automatic transmitter of the reproducer type. In this case the reproducer operates the duplex pole-changer in the same way as it operates the sounder for local reproduction. The Reperforator. The operation of the receiving punch, or reperfora- tor, will be understood by tracing the receiving circuits shown theoretically in Fig. 372. It will be observed that here the main-line polar relay of a duplex instead of operating locally a reading sounder, as is customary in ordinary duplex working, operates an extra polar relay, the armature lever of which is grounded through a 6-m.f. adjustable condenser. Two double-spool electromag- nets, M, M', of the reperforator have circuits leading through their windings from 2oo-volt dynamos of each polarity, thence, extending to the open and closed contact points respectively of an auxiliary polar relay. The " punch " magnets control the movements of two armatures which on their free ends are equipped with steel punches, P, P } about 1/16 in. in diameter, and i in. long, which when the magnets are energized are driven through holes (h, h, Fig- 373) in a die plate, and perforate holes in a strip of paper which is being drawn through a slot past the holes in the die plate, the slot being just large enough to permit free passage of the tape. The tape is moved forward continuously by means of a tape-transmission 416 AMERICAN TELEGRAPH PEACTICE and take-up gear, operated by an electric motor the speed of which is regu- lated by a small hand rheostat. It is customary to adjust the receiver from hand sending at the distant station, before the automatic transmitter is connected to line. A closed key, sending a marking current from the distant station, results in the tongue of the home main-line relay moving over to its front contact, thereby pre- senting a ground contact to the 85-volt dynamo circuit by way of the front contact of the main-line relay, and the magnet EM of the auxiliary relay, which causes the latter to attract its armature to the left, permitting the 6-m.f. condenser to empty itself of the negative charge which it had accu- mulated while the tongue of the auxiliary relay was in contact with the negative battery terminal. The process of reversing the charge held by the condenser FA o o o o o o ^^M \ E T 5 G FIG. 372. Theory of the reperforator. from negative to positive, after the relay tongue makes contact with the posi- tive battery terminal, causes the magnet M to momentarily attract its armature A, and as the armature lever is pivoted at B, the steel punch P is driven through the moving strip of paper, perforating a hole near the lower edge of the tape. As the distant key is opened and a spacing current sent to line, the home line-relay " opens," thereby transferring the ground contact presented to the 85-volt dynamo circuit, through the magnet EM' of the auxiliary relay, causing the lever of that relay to move into contact with the opposite local contact, whereupon the charge held by the condenser is changed from positive to negative, causing momentary magnetization of the punch magnet M', the result of which is that the armature lever actuating the upper steel punch, drives the latter through the tape, perforating a hole near its upper edge. The horizontal distance between the two holes depends upon HIGH-SPEED AUTOMATIC TELEGRAPHY 417 the time elapsing between the instant the marking current is sent out and the time the spacing current is sent from the distant station. If the positive and negative battery contacts made by the distant pole-changer are made close together, as in forming the letter "e," the holes in the received tape appear as at "e" in the specimen slip, Fig. 372. If a greater period of time separates the positive and negative battery applications, as in forming the letter "t," the holes in the receiving tape appear as at "t," in the specimen slip. The steel punches are adjusted to travel forward just far enough to go through the paper and make a clean round hole, and backward just far enough to clear the face of the die-plate. In view of the fact that the tape is passing continuously through the slot in front of the steel punches, the act of punching the holes must be accom- plished by extremely rapid movement of the punches so that there will be no tendency to tear the tape. The speed at which the punches move forward and backward in response to the operation of the auxiliary relay is regulated by having the capacity of the condenser accurately adjusted, and by adjusting the tension of the strong retractile springs S, attached to the armature levers of the reperforator, so that when the steel punches are traveling the required distance to and fro, the action will be rapid and snappy. It is evident that the tape being perforated is stopped each time either the upper or lower punch is in the act of perforating a hole, and as each punch is operated many times per second, it is necessary so to adjust the tape-moving mechanism that these momentary stoppages are compensated for by "slip" in that part of the gear which pulls the tape through the slot. The instruments have been designed to do this satisfactorily, and it has been found that attendants can, with little practice, learn the correct adjustment. The present method of taking care of the received tape coming from the reperforator is the same as that used in caring for the original transmission tape as turned out by the Wheatstone perforator, that is, by rolling it up by hand as it comes from the receiver. The receiver when in operation requires the constant attention of an at- tendant, and it is quite convenient for him to take care of the received tape in the manner above referred to. The received tape may be parceled out in units of one message, two messages, or in any number required by traffic conditions, as the receiver attendant very quickly learns to read the tape and is able to follow the wording as perforated thereon. The end of each message is signified by a paragraph sign ( ) or by a succession of letters "a," without space between them. The code used is the Morse alphabet, except that the letter "L" is changed from "long dash" ( ), to "dot, three dashes," ( ), and the figure "nought" from "long dash" to five short dashes (-- ). The received tape is passed to the reproducing operators in whatever size bundles the traffic demands, and by them is run through local repro- 27 418 AMERICAN TELEGRAPH PRACTICE (D & m I < fO t~^ fO d fe HIGH-SPEED AUTOMATIC TELEGRAPHY 419 ducing machines at a speed to suit the convenience of the operator as before stated. The operation of the reproducers is quite simple, and may be learned by any Morse operator in a short time and without difficulty. Figure 373 shows the actual construction of the reperforator used in con- nection with the Postal automatic telegraph system, the various parts bearing the same index letters as do the Lock Nut-., same parts illustrated in the theoretical Capstan diagram Fig. 37*. The _ spring adjust- ^^ ments bA, are for regulating the retractile . tension exerted by the springs S, upon the armature levers A. The front adjust- ment screws FA act as back-stops for the FIG. 374. Reperforator bearing armature levers, and must be so set that adjustment, the steel punches fastened to opposite ends of the levers, when pulled back by the springs, will come to rest in the punch guide holes h, just clear of the face of the die-plate. The lever adjustment screws LA ex- tend between the two spools of each magnet, projecting far enough to prevent the armature striking the cores of the magnet, and also serve as adjustments for regulating the distance beyond the face of the die-plate the punches P are allowed to travel. The forward and the backward travel of the steel punches, therefore, is regulated by means of the adjusting screws FA and LA. In practice, a forward travel, from rest, of 0.006 in. is all that can be allowed where high speeds are to be maintained. Figure 374 shows an enlarged view of the armature-shaft bearing of the reperforator. The successful operation of the reperforator is largely depend- ent upon the elimination of lost motion in the shaft bearings, and the bearing employed while somewhat elaborate is the only one among those tried out which satisfactorily answers the purpose. The parts of the bearing are made of the hardest grade of Tobin bronze, and the adjustment is made as follows: To adjust bearing: Disconnect retractile spring from armature lever. Tighten screw A, leaving just space enough between its inner surface and the surface of the shaft to hold a film of oil. Tighten screw B of each bearing so that when the steel punches are properly lined up in the punch guide holes the play of the shaft will be equal in the bearing A on each side of the shaft. Lock-nut C should then be tightened, securing the adjust- ment of A, care being taken not to disturb A after being properly set. The reperforator as here described is the invention of Mr. F. E. d'Humy. Figure 375 shows the transmitter circuits, arranged so that either the high-speed automatic transmitter, or a Morse key operating an ordinary pole-changer may be switched into circuit, depending upon the position of the 420 AMERICAN TELEGRAPH PRACTICE lever of the switch on the right. The duplex " balancing" switch is shown on the left. A reading sounder circuit for the out-going signals is provided by means FIG. 375. Transmitting circuits, Postal automatic. FIG. 376. Tape take-up gear, Postal automatic. of a 2o,ooo-ohm leak to earth through a polar relay as shown in the lower left-hand portion of the diagram. After the speed of transmission is run up higher than 65 or 75 words per minute, the sounder, of course, fails to record the signals intelligibly. PRINTING TELEGRAPHY 421 Figure 376 shows the construction of the tape take-up gear. The receiving tape is fed to the reperforator by a tape transmission device, the speed of which may be regulated to suit the speed of signaling. As the perforated tape leaves the reperforator it passes between the rollers T, T', of the take-up gear, which are in light contact with each other, the degree of tension being adjustable by means of the compression-spring screws E, E f . A spring belt extends from the pulley P' to a pulley mounted on a shaft which is geared to the driving mechanism of the tape-transmission FIG. 377. Complete wiring connections of sending and receiving circuits. Postal automatic. gear (not shown), and the speed ratios are such that the rollers T, T f , of the take-up, revolve three times as fast as the feed rollers of the transmission device, which means that the "pull" of the rollers T, T f , is not positive or constant. It is necessary that there shall be considerable "slip" of the tape as it passes through the rollers of the take-up, for, if the pull were positive the tape would be torn during the brief instant that either of the steel punches of the reperforator are punching a hole in the tape. The combined "slip" of the spring belt and of the rollers T, T' compensates for the many stoppages of the tape which take place during the operations of punching. Figure 377 shows the wiring and binding-post connections of both trans- mitting and receiving circuits of the Postal automatic arranged for duplex operation. PRINTING TELEGRAPHS Although the subject of printing telegraphs is an old one with the inventor and with the promoter, the development of satisfactory printing telegraph 422 AMERICAN TELEGRAPH PRACTICE systems has not reached that stage where the subject is in shape for practical consideration in a work dealing with telegraph practice. The reason for this (so far as the employment of printing telegraph systems in America is concerned) is that the systems which have been tried out, and which at the present time are in service, have been operated by the inventors themselves, or under their direction, and in some cases by specially trained staffs, recruited, largely, from mechanics who know little or nothing about Morse telegraphy. On account of the many mechanical movements involved in the operation of telegraph printers, these machines are necessarily somewhat complicated in construction, and although in their design great ingenuity has been exercised in applying known laws and principles of mechanics, the apparatus produced, to do its best work, must be handled by competent mechanics. In most of the systems so far introduced, the purely electrical features, such as line-potential and line-current values, and main-line relay and transmitter functions, are comparatively simple, and it is with these features only that the Morse telegrapher has been concerned. When a new system is tried out in service, apparently it has been a much easier matter to teach mechanics what they need know about the electrical features involved, than to teach the expert telegrapher what he must know about mechanics, in order to operate the printer efficiently. These considera- tions, in a sense, isolate the subject of printing telegraphs from the subject of Morse telegraphy. It is not to be inferred, however, that printing telegraph systems cannot be employed to the advantage of the service, as it is cjuite possible that the time may arrive when a large portion of the telegraph traffic of this country will be handled by means of printing telegraph systems, and it is possible that within a few years, one, two, or more systems will have reached a stage of development and of standardization, that will make possible a technical treatment of the subject from a telegraphic standpoint that will be intelligible to Morse operatives. NAMES OF PRINTING TELEGRAPH SYSTEMS INVENTED, TRIED OUT, AND IN SERVICE Two different systems, known as the Rowland and the Wright, have within the past few years been tried out experimentally on the lines of the Postal Telegraph- Cable Company. Each of these systems was the product of printing telegraph inventors of great skill, and who were quite familiar with the requirements of such inventions. The performance of the Rowland system and of the Wright system was excellent under certain conditions of traffic, but both have been taken out of actual service and returned to the laboratory for further development. PRINTING TELEGRAPHY 423 The Western Union Telegraph Company has for a number of years past been using a printing telegraph system known as the Barclay printer. Formerly the system was known as the Buckingham, in which certain changes and improvements have been made by Mr. Barclay. The Buckingham-Barclay printer is at the present time employed com- mercially by the Western Union Company, but is still being studied with the object of introducing further improvements, or of making alterations, in order that the machine may more satisfactorily meet the requirements of modern telegraph traffic conditions. In British Post-office telegraph service, the following named systems are being used to a greater or less extent: The Creed, Murray, Baudot, and the Hughes. In the United States at the present time a printing telegraph system known as the Morkrum, is being tried out on certain lines 'of the Postal Telegraph- Cable Company, and of the Western Union Telegraph Company. In Canada the Morkrum system is being tried out on a Canadian Pacific Railway-telegraph circuit between Montreal and Toronto. For the information of those who may wish to study the historical develop- ment of printing telegraph systems, or who may desire to investigate the principles of operation, and the construction of printing telegraph machines, a condensed bibliography of printing telegraph literature is incorporated in the appendix, see section A. CHAPTER XXI TELEGRAPH AND TELEPHONE CIRCUITS AS AFFECTED BY NEIGHBORING ALTERNATING-CURRENT LINES TRANSPOSITION OF LINES USED FOR TELEPHONE PURPOSES AND FOR SIMULTANEOUS TELEGRAPH AND TELEPHONE PURPOSES It is well known that when current flows in any wire, there exists in the space surrounding the conductor an electromagnetic field, extending outward from the wire to an indefinite distance and gradually diminishing in strength. If the current traversing the conductor is alternating in polarity, the strength of the electromagnetic field is continually changing, and it is due to this change in strength that electromotive forces are induced in neighboring conductors. Due to the sign of the e.m.f. and its value in the first-mentioned conductor, there is an electrostatic field extending from it to an indefinite distance which decreases in strength with increased distance. The electrostatic field induces charges in adjacent conductors; the induced charge continually changing in sign from positive to negative and vice versa so that in effect there is in- duced in the neighboring wire an alternating current as a result of both elec- tromagnetic and electrostatic induction. It is true, of course, that electrostatic and electromagnetic fields exist in the space surrounding conductors carrying direct, or uni-directional currents; but in this case it is only the rise and fall of the current strength, with the accompanying rise and fall of the field strength (generally as a result of open- ing and closing the circuit) which induces disturbing currents in neighboring conductors. The disturbance created, so far as the effects of electrostatic and electro- magnetic induction are concerned, in the case of the former depends upon the distribution of the currents of charge, which is proportional to the rate at which the electrostatic field changes. In the case of the latter the neigh- boring wire has induced in it an electromotive force proportional to the rate at which the strength of the electromagnetic field changes. In the operation of telegraph circuits the induced disturbances resulting from the proximity of the conductor to other conductors carrying direct currents, generally are due to cross-fire between neighboring conductors of the telegraph system on the same pole lines (see Cross-Fire, page 209). Those disturbances due to atmospheric electrical phenomena also have been referred to in a previous chapter, see page 120. 424 INDUCTIVE DISTURBANCES 425 The most harmful induction experienced in the operation of telegraph lines is that due to neighboring high-tension power lines carrying alternating currents. Numerous plans for obviating the effects of induction from high-voltage alternating-current power lines have been suggested, some of which have been tried out in practice, the results in most cases being far from satisfactory. There are, however, a few methods of getting around the difficulty, which have considerable merit, and which when properly applied make possible the operation of circuits subject to inductive disturbances, which, otherwise would be inoperative as long as the physical relations of the two lines remained unaltered. One of these methods requires that two wires be used for each telegraph circuit, thus forming a metallic circuit in place of the usual single grounded conductor (see Fig. 266, "The Metallic Circuit Quadruplex")- Among the other methods proposed, might be mentioned those covered by U. S. patents, Nos. 955,141 and 955,142, issued to Mr. Minor M. Davis. The first covers the placing of an idle conductor in close proximity to a tele- graph conductor so that both of these conductors are subject to the induction from the disturbing source, and affected to the same extent. The currents induced in the idle conductor will react upon the telegraph conductor and this reaction may be adjusted so as to compensate or neutralize^ the action of the disturbing circuit upon the telegraph conductor. Or, stated specif- ically, a positive impulse in the disturbing wire induces a negative impulse in the telegraph wire and also in the idle conductor, the negative impulse in the idle conductor reacts tending to produce a positive impulse in the telegraph circuit. By regulating the resistance and capacity of the idle conductor, the impulses resulting from the reaction of the idle conductor may be made to counteract the disturbing impulses in the telegraph circuit. This plan of induction neutralization is especially applicable to telegraph conductors in aerial cables suspended on pole lines parallel to high-tension lines. To attain the desired ends the telegraph wires extending through the zone of disturbance may be cabled. The conductors in the cable consist- ing of two groups arranged parallel and insulated from each other. The wires of one of these groups being used for signaling purposes while the other group of wires the idle conductors are tied together at each end of the cable. At one end of the cable the joined idle conductors are connected to ground through an adjustable resistance, and at the other end of the cable to ground through an adjustable capacity. With this arrangement a positive impulse in the disturbing wire induces an impulse of the opposite sign in all of the conductors in the cable. If the resistance and the capacity of the idle conductors is properly adjusted this negative impulse will be so proportioned in its effect upon the telegraph conductors that the induced impulse from the disturbing wire will be neu- 426 AMERICAN TELEGRAPH PRACTICE tralized. Obviously, the idle conductor does not need to be as long as the telegraph wire, but by establishing the constants of the idle conductors capacity and resistance at the most effective values, an effect is produced which tends to compensate for the effect of the disturbing wire upon the telegraph conductor. The other plan aims to neutralize induction in parallel circuits by employ- ing as a conductor the metal sheath of the cable inclosing the telegraph con- ductors and a parallel compensating conductor, all of which are subject to the same inductive influence, and causing the induced impulses in the com- pensating conductor to react upon the telegraph conductor to an equal and opposite extent as compared 'with the disturbing cause. Where there are one or more insulated conductors in parallel relation, con- nected in circuit with telegraph apparatus in the ordinary manner, parallel or with these conductors, as in the same cable, are one or more compensating conductors; these are connected in a circuit susceptible of conveying induced impulses, and for this purpose there may be used a ground return circuit, or the circuit may include a condenser. One coil of a transformer is connected in this circuit and the other coil of the transf orner is connected in circuit with a conductor arranged parallel with the disturbing source, for this purpose there may be used a separate conductor or the metal sheath or armor of the cable included in a compensating circuit may be used, and the transformer so adjusted that the compensating conductors will develop a source of alternating current having an electromotive force efficient to compensate for the induc- tive effect of the disturbing source. If the disturbing source extends parallel to the telegraph conductors for a comparatively long distance there may be one or more compensating conductors arranged parallel with the telegraph conductors for a much shorter distance, but the electromotive force, intensity or current strength is increased, graduated or adjusted so as to compensate and neutralize the effect of the disturbing cause in the telegraph conductors. An effective and inexpensive method of screening the Morse relays in a grounded telegraph circuit from the effects of induction from a neighboring high-tension line is illustrated schematically in Fig. 378. At each end of the section of telegraph line exposed to the high-tension line a low-resistance choke coil is included in the telegraph circuit, and on the line side of each choke coil a condenser path to ground is presented to the induced currents. This arrangement is quite effective where the disturbance is not great, and where ground potentials have a low value and have a constant polarity. Figure 379 shows a method of screening the Morse relay at a way office on a single circuit, wherein an alternative path is presented to the induced alternating currents, enabling them to pass through the station with mini- mum effect upon the relay. Figure 380 shows an excellent terminal or way-office arrangement for use INDUCTIVE DISTURBANCES 427 Power Line Relay IOO M 100" Relay T w = IR *= = Imf = - = IR =lmf. F W T = FIG. 378. Telegraph circuit in close proximity to high-tension power line. Line 100 w IR . 100" . ,. , T R ItolSmfi FIG. 379. Method of screening way office relay on a single line. 5 K. 30 Ohm Inductance Coil 10 00 Ohms >6tvl8M.F. 5K. 50 Ohm Adjustable Nots: 5K , Coils inductively Inductance Coil (,vnaenser connected in series. Single Set FIG. 380. Terminal office or way office relay on a single line protected against induction from neighboring high-tension lines. P.O. 2 Col Is 16000 Turns No.30/ 7?5 Ohms Each, Differential 1M.F. TT Air Core' ^Impedance Co! 1 e*L ^5000 Turns No.24 JU Adjustable Non - 1L Inductive Resistance. T ^=^ Polar r^Re/ay \ Resistance of Shunt should be kept low. k 1 ^Adjustable Duplex Set 'AL. FIG. 381. Duplexed line protected against disturbances from neighboring 25-cycle high-tension line. 428 AMERICAN TELEGRAPH PRACTICE on single lines, employing standard 5^ coils for the purpose. In this arrange- ment the relay winding is shunted with a i,ooo-ohm non-inductive resistance forming one branch of a resonant circuit which provides a path past the relay for the induced alternating currents. Figure 381 shows the theoretical main-line connections of a differential polar duplex equipped with protective coils and condensers for off-setting the effects of induction from 25-cycle single-phase power circuits. In balancing duplex or quadruplex sets connected into lines that are affected by heavy induction, the constant chattering of the relays when the main-line battery at both ends of the line is removed by throwing the ground switches to the left, makes it quite difficult to determine when the polar relay has been balanced magnetically. In cases where it is important that the magnetic balance should be accu- rate it is necessary to place the lever of the ground switch in the center, and to open the artificial line and condenser circuits by throwing the rheostat switch and retardation resistance coil switch into the "open" position. TRANSPOSITION OF WIRES ON POLE LINES When two single wires of a telegraph system are joined together at terminal stations for the purpose of providing a metallic telephone circuit, it is neces- sary that the wires should be transposed at predetermined intervals along the line in order that there will be no "cross-talk" effects between the telephone circuit thus formed, and other telephone circuits on the same or adjacent pole lines. Transposing the two sides of the telephone circuit as above indicated also protects the telephone circuit from the effects of induction from neighbor- ing power lines, trolley line, and telegraph circuits. Through those sections of the line where the conductors are carried in aerial or underground cables, transposition is effected by twisting the two conductors of each metallic circuit around each other continuously, forming what are called "twisted pairs." There are several theoretical methods of determining the number of trans- positions necessary in a given distance, to protect the circuit from mutual and from foreign induction, where reasonably definite values can be determined for the various factors involved, but in practice it is found advisable not to adhere too closely to the dictates of theory in this regard, but rather to provide sufficient margin to offset any possible additions to the sources of disturbance. Standard practice in locating the position of transposition poles is to measure a distance of 1,300 ft. from the first pole of the line and mark the pole nearest to the point so measured A. Then measure successive distances of 1,300 ft. each, and letter the nearest poles B, C, B, A, B, C,B, A, B, C, etc., successively, the circuits being transposed upon the poles so lettered as TRANSPOSITION OF CONDUCTORS 429 ~V' 2 30L 7' W jpt 'rC 'ro 6s 7/r r? S~ ~^x /- -\ A B A c A B A c A B A C A B A c * B A c A B A C A B A C A B A C Lower Grosser rm FlG 382. Transposition of wires. Four-pin cross-arms. 'V 'I 30C )' "f opt y* Ore ss an 77 t> . , / \ / \ f \ t \ / \ / N / s. , / Z> ^> \ / \ s N. / / I> / s A B A C A B A D A B A C / A B- A \ E / A B A \ C A B A D A B A C. A B A 5 <. ; r-; , / Lower Crossarm JT IG 383. Transposition of wires. Six-pin cross-arms. V-' J(A ty opt ?r ( 're 66 an 77 S A B A C A B A D A B A C A B A A B A C A B A D A B A C A B A > Q -> v L o wer Crossarm T FIG. 384. Transposition of wires. Eight-pin cross-arms. - -,-~-?i300 ' Upper Crossarm Lower Cro6sarm FIG. 385. Transposition of wires. Ten-pin cross-arms. 430 AMERICAN TELEGRAPH PRACTICE shown in Fig. 382. The diagram shows the necessary transpositions, where two circuits are carried on the upper cross-arm and two on the lower arm. Figures 383, 384 and 385 show respectively the scheme of transposition for six-pin, eight-pin and ten-pin arms. All transpositions in copper circuits are made by cutting the wires on the pole side about 20 in. from the cross-arm, and slipping on each half a B pgygrea fr g^ST^gegP A FIG. 386. Method of transposing wires at supports. Mclntyre sleeve, with which the wires on the cross-arm side are dead-ended, one in the lower groove and one in the upper groove of the transposition glass insulator, allowing the ends to project. About 6 ft. of slack wire is then joined to the wires on the cross-arm side by using a whole Mclntyre sleeve. Half sleeves are then slipped on and the wires dead-ended in the upper grooves. ransposing 'insulator/.' Wire betrree insulator "c'* Wires f a" and "d" fastened *e"and h". Wires '&" and'c wire "between wires *# "A "C "ct'and "d" FIG. 387. Transposition of wires, carried on end-pins. vacant grooves of the insulators, after which the free ends are crossed and connected by half sleeves as shown in Fig. 386. Where the wires to be transposed are located on either end of the cross- arm, transposition is effected by supporting the two cross wires upon bracket pins as shown in Fig. 387. TRANSPOSITION OF CONDUCTORS 431 The transposition insulator consists of two single insulators on the same pin as shown in Fig. 388. The lower section has no top, and the pin projects through it, the upper insulator being secured on the top of the pin in the usual manner. Figure 388 on the left shows a standard insulator, and on the right a transposition insulator. Standard insulator Transposition insulator. FIG. 388. Instead of transposing wires on the glass insulators screwed on wooden pins inserted into the tops of cross-arms in the usual manner, it is now com- mon practice to employ iron / hooks driven into the cross-arm from the un- der side. A regulation glass insulator is screwed to the lower extremity of the / hook, so that the wire is carried under the cross-arm at the point of transposition. CHAPTER XXII TELEPHONY. SIMULTANEOUS TELEGRAPHY AND TELEPHONY OVER THE SAME WIRES Although in long-distance telephony metallic circuits are used exclusively, it is possible under favorable conditions to use single-line grounded circuits for telephoning over comparatively short distances. This fact is noted here in view of the possibilities in the way of joining up branch-line grounded circuits to long-distance metallic circuits, which will be explained presently. Figure 389 shows theoretically the main-line connections of a grounded- line telephone circuit, employing separate talking battery at each station. On grounded telephone lines intermediate stations may be introduced as at A, which shows the series connection, or as at B, which shows the bridging or multiple connection. At each station a 50- to 75-volt, alternating-current, Line Line Line FIG. 389. Grounded line telephone circuit. hand-operated generator is used to operate the polarized bells at the various stations, for the purpose of signaling. As the operation of any generator connected into the line operates all of the signaling bells in circuit, each station is given a call. Where but three or four stations are located upon a line, the call for one office may be two short rings, for another three short rings, for the third four short rings, and for the fourth station, five short rings. The single short ring, generally being reserved for " ring-off" pur- poses. Where a larger number of offices are located upon one line, it is necessary to form combination signals in order to reduce to a minimum the number of signals required, thus, the call for one office may be two short rings, a pause, and two short rings, or, the signal "22." Likewise other offices may be given signals such as "13," "41," "21," etc. The resistance of the ringer magnets used in grounded circuits when con- nected in series is about 80 ohms, and when connected in multiple, ranges from i ; ooo to 2,500 ohms. 432 SIMULTANEOUS TELEGRAPHY AND TELEPHONY 433 t* r~>* f >l ff Hj*>t 1 1 X* J f *=H 1 FIG. 390. Metallic-circuit telephone line. r*" ^ r' * FIG. 391. Series telephone set. FIG. 392. Bridging telephone set. 434 AMERICAN TELEGRAPH PRACTICE Figure 390 shows theoretically the main-line connections of a two-wire or metallic telephone circuit. In the metallic arrangement it is customary to connect intermediate offices in multiple, as shown at B, Fig. 390, using 2,5oo-ohm bridging bells. Figure 391 shows the wiring of a "series" telephone set, and Fig. 392, the wiring of a " bridging" telephone set. The two wires used for metallic circuit telephone operation, for satis- factory working must have identical characteristics, such as resistance, capacity, leakage, and inductance. In practice slight variations are permiss- able in any of these factors, but the talking efficiency of the circuit is reduced if considerable variations exist. Where a sufficient number of conductors of the same gage are available it is an easy matter to effect proper balances by transposing the conductors as explained in Chapter XXI. The current supplied to telephone transmitters may be derived from primary batteries of the Edison, gravity, Fuller, LeClanche, or dry-cell types, or from storage batteries. CONNECTING GROUNDED LINES TO METALLIC CIRCUITS A metallic circuit and a single grounded line may be joined together by connecting the two lines through a repeating coil R as indicated in Fig. 393. Line Y _ Z ? ? FIG. 393. Grounded line joined FIG. 394. Section of a through telegraph to a metallic circuit through a repeat- wire serving as one side of a metallic tele- ing coil. phone circuit. In those cases where a single grounded telephone wire is strung upon the same pole line with a through telegraph wire it is possible for short distances to use the telegraph wire as one side of a metallic telephone circuit without seriously interfering with the operation of either circuit. Figure 394 shows one method of accomplishing this. The telegraph wire A-B extends through the stations at which telephones Y and Z are located. Instead of using an earth return for the telephone circuit, a section of the telegraph wire may be used for the purpose by connecting the telephone directly to the telegraph line as indicated in the diagram. The annexed telephone circuit forms a shunt to a portion of the telegraph line, but owing to the high resistance of the telephone instruments as compared with the resistance of the section of telegraph line shunted, comparatively little of the telegraph current will be diverted. In fact, the joint resistance of the tele- SIMULTANEOUS TELEGRAPHY AND TELEPHONY 435 phone circuit and that portion of the telegraph conductor used will be less than that of the telegraph conductor alone, reducing to that extent the total resistance of the telegraph circuit. A similar method of tying a telephone line to a telegraph line is shown in Fig- 395, the connection being made at each station through condensers C and C', which permit the passage of the alternating current telephone impulses, but prevent the direct current telegraph impulses getting into the telephone apparatus. Figure 396 illustrates still another method of utili2ing a section of a through telegraph wire to form one side of a telephone circuit. At station Y, the junction of the two windings of retardation coil K, is connected with the telegraph line east, while the outside terminals of the windings of the retardation coil join the telegraph wire to the telephone wire as shown, Y ? *JF * * t I . B FIG. 395. Telephone wire con- FIG. 396. Retardation coil method of nected to section of through tele- tying telephone lines to telegraph lines, graph wire through a condenser. thereby forming a metallic circuit between stations Y and Z through the windings of the retardation coil at each station. At station Z the tele- graph line west is connected to the junction of the two windings of retarda- tion coil K'. For telephonic purposes, therefore, a metallic circuit is formed between stations Y and Z by way of C-D-E-F. A telegraph impulse passing from east to west or vice versa, finds an unobstructed path through the retardation coils, due to the fact that the two windings of the coils are connected so that the inductive action of one coil neutralizes that of the other, that is, when current passes through the windings in opposite directions, as is the case when the current enters at the junction of the two windings. While the inductive reactance of the two windings of the retardation coil to the comparatively slow telegraph impulses is neutralized, this is not the case with the high frequency telephone currents. The impedance of the coil to the telegraph current does not much exceed its ohmic resistance, as the impedance v R 2 -\-(2n nL) 2 = R, since the current traverses the two windings in op- posite directions around a common core. Obviously where hand telegraph signals are concerned the value of n (frequency) is quite low, say, 12 or 15 per second. 436 AMERICAN TELEGRAPH PRACTICE The core of the retardation coil forms a continuous magnetic circuit of comparatively large physical dimensions, which tends to impede reversal of its magnetism when the high frequency alternating telephone currents tend to reverse it, and as to the telephone currents the two windings of the coil are connected in series, with each winding acting as an independent im- pedance coil, while the two windings are connected in series the total im- pedance is ascertained by the formula: V(2R+2R) 2 +( 2 nnL+2nnl) z = V(R) 2 +(^nL) 2 . In which the value of n may be taken as 1,000, and of L as 6 (henries). It is evident, therefore, that the telephone currents are confined to the circuit C-D-E-F, while the telegraph currents will pass through the joint circuit without entering the telephones. Line Line FIG. 397. Circuits of the repeating-coil type of simplex. HHHH Metallic circuits built up as indicated in Figs. 394, 395, and 396, should be transposed in the usual manner and for the reasons explained in the pre- ceding chapter. It is apparent, also, that in any of these arrangements, while intermediate telephone circuits may be inserted by bridging the telephone sets, inter- mediate telegraph stations cannot be introduced without seriously interfering with the operation of the telephone circuit, as in that case the telegraph impulses would be distinctly heard in the telephone receivers. In setting up such combination circuits the material and resistance of the wire used in each side should be the same. THE SIMPLEX CIRCUIT The simplex circuit arranged as shown in Fig. 397, provides for tele- graphing over a circuit which is at the same time being used for telephony. THE'SIMPLEX 437 There are two types of simplex apparatus, one employing a repeating coil and the other a bridged impedance, the former is illustrated diagram- matically in Fig. 397, and the latter in Fig. 398. The repeating coil arrangement is not as efficient from a telephone transmission standpoint for long lines as the bridged impedance arrangement, as the introduction of each repeating coil of the usual type into the line has Line Line FIG. 398. Bridged impedance coil type of simplex. an effect equivalent to the introduction of i 1/2 miles of No. 19 B. & S. cable conductor, or 28 1/2 miles of No. 8 B. W. G. open line. The effect of the bridged terminal equipment of the impedance coil system may be con- sidered as equivalent to 1/4 mile of cabled conductor. The following table of equivalents indicates the relative transmission efficiency of the various gages and kinds of wire used: Miles of line equiva- Conductor Pounds per mile lent to i mile of No. 8 B. W. G. No. 8 B. W. G., copper... A-2C I . O No. 12 N. B. S. G., copper M-OO 173 0.46 No. 14 N. B. S. G., copper IO2 0.28 No. 6 B. W. G., B. B., iron 573 o. 19 No. 8 B. W. G., B. B., iron 378 o. 16 No. 10 B. W. G., iron 2^O o. 14 No. 12 B. W. G., B. B., iron ^ 165 O.II No. 14 B. W. G., B. B., iron 96 0.08 No. 19 B. & S. cable (0.072 m.f.) 21 0.03 No. 9 B. & S. copper 2IO o t?4 (est.) w . 3^1- ^v-oi^.y 438 AMERICAN TELEGRAPH PRACTICE In Fig. 397, A and B are two telephone stations connected by a metallic circuit through repeating coils C and C. Taps are taken from the middle of one side of each coil to telegraph sets T and T', thence to the telegraph main-line battery and ground at each station. The two line wires carry the telephone current in opposite directions, but, acting as a joint circuit to the telegraph current, the two line wires form one side of a ground return telegraph circuit. By noting the direction of the windings around the continuous iron core of the repeating coil, it will be apparent that the telegraph currents flow in FIG. 399. Intermediate telephone stations on simplex circuit. opposite directions around the core, the action of one winding neutralizing the inductive effect of the other, and if the electrical characteristics of the two line wires are the same, the telegraph impulses will not affect the tele- phone receivers. The retardation coil type .of simplex circuit shown theoretically in Fig. 398 has two terminal telephone stations A and B connected to the two line wires forming the metallic circuit through 4-m.f. condensers. The high- frequency alternating talking current readily passes through the condensers, while the slowly pulsating direct-current Morse impulses cannot enter the Telephone Exchange Telegraph FIG. 400. Simplex circuit connected through an intermediate telephone exchange. condenser circuit containing the telephones. The retardation coil bridged across the line wires at each terminal station has two 5oo-ohm windings around an iron core, giving the coil 1 a very high impedance to alternating currents. At the junction of the windings of the two coils the Morse circuit is tapped off, leading to relay, key, battery and ground. Inasmuch as the resistance and reactance of the coils and of the two line wires forming each side of the circuit are identical, the current divides equally in the two branches of the circuit, and as the difference of potential at any point along the line is the same in each wire, it is evident that inter- PHANTOM TELEPHONE CIRCUITS 439 mediate telephone stations may be bridged across the two line wires. See Fig- 399- In those cases where the line is run through an intermediate telephone exchange, repeating coils are connected into the metallic circuit on each FIG. 401. Intermediate telegraph station connected into a simplex circuit. side of the exchange switchboard as indicated in Fig. 400, the telegraph circuit being completed through the exchange by means of a wire connecting the middle point of each retardation coil as shown. Figure 401 shows a simplex circuit with an intermediate telegraph station. FIG. 402. Intermediate telephone and intermediate telegraph station connected ' into a simplex circuit. Figure 402 shows a simplex circuit with an intermediate telephone station and an intermediate telegraph station inserted. PHANTOM TELEPHONE CIRCUITS The phantom is an arrangement by which three telephone circuits may be obtained from two pairs of line wires. The combination is referred to as consisting of two physical circuits and one phantom circuit. FIG. 403. Phantom telephone circuit. A and B, Fig. 403, are two stations connected by a metallic circuit, as also are stations C and D. A repeating coil is inserted at each end of each metallic circuit as shown in the diagram. At each station the two line wires 440 AMERICAN TELEGRAPH PRACTICE from a third telephone set are connected to the middle of each repeating coil, thereby employing the two wires of each metallic circuit as one side of an additional, or phantom circuit. Special forms of transposition of the four wires are required in order to prevent cross-talk. Fig. 404 shows various transposition arrangements FIG. 404. Special forms of line transposition necessary where phantom circuits are made up. of the wires of the physical circuits and of the two sides of the phantom circuit. Intermediate stations may be inserted in each physical circuit, and there will be no interference; provided a correct balance is maintained between the two wires forming the circuit. K 1 FIG. 405. Intermediate telephone station on phantom circuit. Figure 405 shows the necessary connections for inserting an inter- mediate station into the phantom circuit. Retardation coils K and K' are bridged across the physical circuits. From the center of each of these coils a tap is taken to one side of a condenser, the other terminal of which is PHANTOM-SIMPLEX CIRCUITS 441 connected with the telephone set to be bridged into the phantom circuit. The presence of the retardation coil in the physical circuit does not appreci- ably interfere with the talking efficiency of that circuit as the inductive resistance of the coil to high-frequency currents tending to enter it at one end and leave at the other acts to prevent the currents circulating in the physical circuit from being shunted. The phantom currents, on the other hand, traveling in the same direction in wires i and 2, and 3 and 4, enter the retardation coils at both ends at the same instant the only opposition presented to the incoming currents being that of non-inductive resistance. The out-going currents from the intermediate phantom station pass through the windings of the retardation coils in opposite directions, again encounter- ing only non-inductive resistance. THE PHANTOM SIMPLEX CIRCUIT ^ In combining circuits to provide phantom simplex operation, there is a certain degree of loss in the efficiency of telephone transmission, and the satis- factory operation of the system requires that in every case the wires em- o Telegraph o FIG. 406. Phantom simplex circuit. ployed in forming the different sides of the circuit must be of the same com- position and of the same resistance. Figure 406 illustrates theoretically the manner in which a telegraph circuit is superimposed upon a phantom circuit already serving as two tele- phone circuits. Intermediate telephone stations may be connected into either physical circuit or into the phantom circuit without causing inter- ference. In the case of the phantom intermediate connection it is necessary to use condensers as indicated in Fig. 405. It will be apparent that when one side of a telegraph circuit consists of four line wires, as in the phantom simplex arrangement, the resistance of the circuit thus made up will be comparatively low and (as is also the case with the ordinary simplex arrangement) the telegraph circuit being grounded at each terminal station, some difficulty is likely to be experienced due to earth currents. The remedy is to insert artificial non-inductive resistance at one or both of the telegraph stations. In arranging phantom circuits, cabled conductors should be avoided as 442 AMERICAN TELEGRAPH PRACTICE far as possible, although if " double- twisted " pairs are used, reasonably efficient operation is possible. THE COMPOSITE CIRCUIT In arranging for composite operation it is well to remember that iron wire is much inferior to copper wire of the same size when used for telephone purposes, and also that cabled conductors are much less efficient than open conductors suspended on poles in the usual manner. Where the employ- ment of cabled conductors is unavoidable, from a transmission standpoint paper-insulated conductors are considerably more efficient than rubber- insulated wires; that is, for wires of the same size. This is owing to the rela- tively high electrostatic capacity of rubber-insulated conductors. In view of the widely different conditions encountered it is difficult to lay down d^nite rules applicable to all proposed installations of composite apparatus, so far as the length of line and possible number of stations in circuit are concerned. Except in those cases where composite service has been contemplated in the original construction of the line and stringing of the wires, .each particular telegraph line or telephone line must be considered separately before a correct determination can be made in regard to its adaptability for composite service. THE GROUNDED LINE COMPOSITE Figure 407 shows the circuits of a grounded composite installation at a terminal office and at an intermediate telegraph office. The direct-current Line FIG. 407. The grounded composite. impulses from the telegraph battery B have an uninterrupted path via the key and relay at the terminal station, through the retardation coil $-/,, over COMPOSITE CIRCUITS 443 the line wire, through the Morse relay at the intermediate station, and on to the distant terminal station, in the same manner passing through any other intermediate telegraph offices inserted between the two terminal stations. The presence of the condenser C, prevents the telegraph impulses from being shunted to earth through the telephone apparatus, while the presence of the condenser C connected around the Morse relay at the intermediate telegraph station provides a path for the high-frequency alternating telephone currents past that station, whether the Morse key K is open or closed. THE HOWLER SIGNAL In operating call-bell signals over the simplex circuit (Fig. 398) the alter- nating currents produced by the generator pass over the line in the same man- ner as the talking currents, and, ordinarily no difficulty is experienced pro- vided high power generators and condensers having a large enough capacity are used in connection therewith. In many modern telephone exchanges one side of the ring- ing generator is grounded. Obviously such an arrangement cannot be used for signaling over simplex circuits without causing chattering of the armatures of the Morse relays while the signaling impulses are being sent over the line. It is neces- sary, therefore, where the retardation coil type of simplex is used to provide a metallic generator F IG< 40 8. The howler, circuit. In the repeating coil type of simplex the grounded generator may be employed, as the two windings of the coil have no direct electrical connection with each other. In signaling over composited lines, a " howler," Fig. 408, is generally used. This instrument consists of a special form of telephone receiver equip- ped with a resonating megaphone. The diaphragm is operated by the high- frequency signaling currents produced by an induction coil fitted with an interrupter, as at 7, Fig. 407 depressing the button b closes the primary circuit of the induction coil connected with a source of direct current, causing the vibrator to act, resulting in sending out powerful high-frequency signaling currents over the line to actuate the howler connected into the distant telephone set. The sound emitted by the howler may be varied by adjusting the position of the diaphragm relatively to the electromagnets. THE METALLIC CIRCUIT COMPOSITE In all cases where it is desired to maintain telephone service over long distances, say from 100 to 1,000 miles, the metallic circuit is indispensable. 444 AMERICAN TELEGRAPH PRACTICE Conn pos No "in-termedicrt<3 phones bridged. FlG. 409. i T ~-p- - TbrougK "belephone, inter, telegraph. FIG. 410. """ . T.I. r 5 tf Separat-e felepViones an^i "*" Vhrough or separate -telegraph FlG. 411. ITT^x-tf INTERMEDIATE COMPOSITE. FIG. 412. COMPOSITE CIRCUITS 445 1 1 111 _ tt>~ ToC llh I A B -u^- = 6MF 'WOOWCTOCTHI < / SK -8MF 5L --8MF MF MF r o Phone To Phor J~^ 00 ^ "\_ \ ^fiAMMMitHI ' ^6MF 5K I _ 1 76 ' V-^mmS -6MF ToD JMF IMF -8MF --8MF 5L FIG. 413. Complete connections of the composite circuit, showing condenser capacities- ^m^ Inductance 4 SL-Co'il to bcvlcince =i= J inductance of JAdjustoble ^instr-umen'ts o.t -W Condenser' "Distant Station ' - 5-^: GOILS - so OHMS C0MF0SITE CIRCUIT. FIG. 414. Terminal office instrument and main switchboard connections where each side of the telephone circuit is used for a telegraph duplex. 446 AMERICAN TELEGRAPH PRACTICE The organized apparatus of the metallic circuit composite makes possible the employment of each side of the telephone circuit as a telegraph circuit. Figure 409 shows the schematic circuit arrangements of a composited line employing the two line wires as a metallic telephone circuit, and each line wire as a separate grounded telegraph circuit. Figure 410 shows the theoretical arrangement of circuits, where the two line wires are used for through telephone service and where each line wire is used for telegraph service between two terminal stations including an in- termediate telegraph station on each line. Figure 411 shows a similar arrangement, providing for separate telephone circuits in both directions from an intermediate office, and for either through These Cords to be kept inSpringJacff, with Dotted side out \ 5LCoil,300hms ~ 5LCoil,300hms 5L Coil, 50 Ohms ItoSMF FIG. 415. Instrument binding-post connections at a terminal office. Composite circuit. or separate telegraph service. Fig. 412 shows the switching arrangements at the intermediate office which provide for cutting any of the circuits, and for connecting them for through service between terminal stations. Figure 413 shows with fewer lines the actual arrangement of circuits at the intermediate office. Telephone service is maintained between stations A and B, while telegraph service is maintained over one wire between station A and a distant station C, and over the other wire between station A and a distant station D. In this diagram the condenser values found necessary in a particular case are noted. Figure 414 shows the required instrument and main switchboard con- REPEATING COILS 447 nections where duplex telegraph service is maintained over each of the two wires used in forming the metallic telephone circuit. Figure 415 shows the actual binding-post connections of the retardation coils, condensers, and telephone pin-jacks. The apparatus is connected to the two line wires by means of the cords bearing on one end double plugs and on the other double wedges, the plugs being inserted in the pin- jacks, and the wedges in the spring- jacks of the main-line switchboard as indicated in the diagram. REPEATING COILS AND RETARDATION COILS USED IN SIMPLEX AND COMPOSITE CIRCUITS Figure 416 shows a view of the form of repeating coil known as 37~^4. These coils have two primary windings of 35 ohms each and two secondary windings of 35 ohms each. The coils are used in phantom toll circuits and FIG. 416. 37:A repeating coil. in simplex circuits. The size of the baseboard upon which the coil is mounted is ii in. X 8 5/8 in. Fig. 417 shows the terminal markings and circuit connections of the 3*j-A type coil when used for simplex working. Figure 418 shows the terminal connections of the $-N type retardation coil. This coil has four windings of 250 ohms each. Total resistance 1,000 ohms when measured with direct current. The inductance of the coil with windings in series is 507 henries, and with the windings in multiple 3.1 henries. The $-K type retardation coil has two windings of 15 ohms each. Total resistance measured with direct current 30 ohms. The effective resistance of the coil to alternating currents having a frequency of 1,000 p.p.s. is 3,000 ohms with the windings in series inductively. The inductance with both windings in series inductively is 3.1 henries. The 5-1, type retardation coil has two windings of 25 ohms each. Total resistance measured with direct current is 50 ohms. The effective resistance measured with alternating current of 1,000 p.p.s. is 2,500 ohms with both windings in series inductively. With both windings in series opposing each 448 r AMERICAN TELEGRAPH PRACTICE other, the effective resistance is 700 ohms. The inductance of the coil with both windings in series inductively and measured with alternating current having a frequency of 1,000 p.p.s. is 4.8 henries, and under the same condi- tions of current w r ith both windings in series opposing inductively 3 henries. / = IP. 2= Of? 3= 75. 4= 05. FIG. 417. Terminal markings and connections of the 3 7- A coil. FIG. 418. Terminal markings and connections of 5-N type retardation coil. The 5- 7 type coil (which is of the same construction as the 37-^ type of coil) used in connection with the Western Union standard quadruplex has two windings of 500 ohms each. Total resistance 1,000 ohms measured with direct current. The inductance with both windings in series is 584 henries and with the windings in multiple 3.84 henries. CHAPTER XXIII SPECIFICATIONS FOR COPPER AND IRON WIRE, AERIAL, UNDERGROUND, SUBMARINE, AND OFFICE CABLES SPECIFICATION FOR HARD -DRAWN COPPER WIRE ROLLING AND DRAWING The copper bars, before rolling, shall be free from defects, and each coil shall be in one continuous length without joints. All wire furnished under this specification shall be perfectly cylindrical, uniform in size and quality, free from flaws, splits, kinks and other defects. The manufacturer shall cut off before inspection sufficient length from each end of every coil to insure freedom from defects. MECHANICAL AND ELECTRICAL REQUIREMENTS Diame- ter mils Weight per mile pounds Breaking weight pounds Minimum Maxi- B. & S. gage average must Maxi- Mini- Average must Mini- Lot must percentage elongation allowed in mum mileohm allowed not exceed mum allowed mum allowed not exceed mum allowed average at least 5 feet at 60 F. 7 144 337 328 332-5 1,020 1,050 .09 7* 137 305 296 300.5 930 950 .07 8 128.5 268 260 264 820 840 .06 9 9i 114 104 212 I 7 6 204 169 208 172-5 , 650 540 670 555 .02 .00 894.7 10 IO2 169 163 166 520 535 .00 B. W. G. 8 165 439 43i 435 1,328 i,378 1.14 TESTING AND APPARATUS The mechanical and electrical tests shall be made in a manner and with apparatus approved by the electrical engineer of the telegraph company. Tests are to be applied to sample pieces of wire, cut from not less than one-tenth of the number of bundles, as selected by the inspector of the telegraph company from the whole lot of wire under inspection. Samples selected at random by the inspector shall be used for electrical measurement. 29 449 450 AMERICAN TELEGRAPH PRACTICE REJECTIONS The inspector may reject any wire which does not meet the foregoing mechanical and electrical requirements; and, if such rejections include the entire lot, the expenses of inspection shall be borne by the manufacturer. Any imperfect material or work discovered before acceptance of the wire shall be replaced or corrected upon demand of the telegraph company, even if such imperfections were not apparent during inspection of the samples selected. COILS Each coil to be a continuous length without joint or splice, and to have an inside diameter of from 20 to 22 in. Under direction of the inspector, a lead seal of the telegraph company shall be attached to the inside of each accepted coil with a soft copper wire fastener. Such lead seals and wire fasteners will be provided by the tele- graph company. All other attachments to be made with strong twine. Weight of Coils. The length and weight of wire in each coil of the same gage to be as nearly equal as practicable, and the weight to be determined by the following: Ninety-five per cent, of the bundles of gage No. 9 B. & S. (.114 in.) or smaller, to weigh not more than 220 lb., nor less than 190 Ib. each. Five per cent, of low-weight coils will be accepted, the minimum weight to be 125 lb. Ninety-five per cent, of the bundles of gages larger than No. 9 B. & S. to weigh not more than 220 lb., nor less than 150 lb. each. Five per cent, of low- weight coils will be accepted, the minimum weight to be 125 lb. Each coil shall be securely bound with at least four separate wrappings of strong twine, and shall afterward be so protected by wrappings of burlap that there will be no danger of mechanical injury in transportation. Each coil shall have the weight, gage and length of wire plainly and indelibly marked on two strong tags, one of which shall be attached to the coil inside of the burlap, and the other outside of the burlap. Upon the inside tag shall be marked the order number of the telegraph company and the date of inspection. SPECIFICATION FOR GALVANIZED IRON WIRE All wire furnished under this specification shall be perfectly cylindrical, uniform in size and quality, free from flaws, splits, kinks and other defects. The manufacturer shall cut off before inspection sufficient length from each end of every coil to ensure freedom from all such defects. SPECIFICATIONS FOR COPPER AND IRON WIRE 451 Testing Apparatus. The mechanical and electrical tests shall be made in a manner and by apparatus approved by the electrical engineer of the tele- graph company. The wire must meet the requirements of the following table. TABLE OF MECHANICAL AND ELECTRICAL REQUIREMENTS Breaking weight pounds Twists in 6 in. Per- Maxi- mum Birming- ham gage Diame- ter mils Weight per mile pounds Lot must Mini- mum Lot must Mini- mum centage elonga- tion mile ohm allowed average at least allowed average at least allowed at 60 F. No. 4 238 787 2,200 2,120 15 12 15 5 220 673 1,880 1,820 16 13 i5 6 203 573 i, 600 1,550 18 15 15 7 180 45o 1,260 I,2IO 20 i7 14 > 4,700 8 165 378 i, 060 I,O2O 22 18 13 9 148 305 860 820 25 21 12 10 134 250 700 670 25 21 12 Tests are to be applied to sample pieces of wire cut from not less than one- tenth of the number of bundles as selected by the inspector of the telegraph company from the whole lot of wire under inspection. The twist tests to be made by properly gripping the sample wire at the ends by two vises whose jaws are 6 in. apart at the gripping points, and causing one vise to revolve at right angles to the wire at a uniform speed of about one revolution per second. The twists to be reckoned by the number of complete revolutions made by the revolving vise before the wire breaks. Test pieces taken at random, shall be used for electrical measurement, and the resistance calculated to 60 F. in international ohms, using a temperature coefficient, for' iron of 0.0029 ohm per degree. The electrical resistance of the wire in ohms per mile at a temperature of 60 F. must not exceed the quo- tient obtained by dividing the constant number 4,700 by the weight of the wire in pounds per mile. The inspector may measure the electrical resist- ance of as many samples as he desires to test. If upon test it be found by the inspector that the requirements for the electrical or mechanical properties of the wire, or for the finish, are not fulfilled when the wire is offered for acceptance, the expense of all tests made by said inspector on such defective wire shall be borne by the manufacturer. 452 AMERICAN TELEGRAPH PRACTICE GALVANIZING All wire must be thoroughly galvanized. Samples of the wire under inspection shall be dipped into a solution of sulphate of copper, saturated at 60 F. and allowed to remain for one minute, when they are to be withdrawn, washed and wiped clean. The galvanizing shall admit of the process four times without any signs of a reddish deposit upon the wire. Samples shall bear coiling around a cylindrical bar, twelve times the di- ameter of the sample without any signs of the zinc flaking or peeling off. COILS Each coil shall be of a continuous length and must not contain more than one splice which must be well soldered. The inside diameter of the coil to be from 20 to 22 in. Under direction of the inspector, a lead seal of the telegraph company shall be attached to the inside of each accepted coil with a wire fastener. Such lead seals and wire fasteners to be provided by the telegraph company. WEIGHT OF COILS The length of wire in each coil of the same gage to be as nearly equal as practicable, and to be determined by the following: No. 4 B. W. G. shall contain 4 coils per mile. No. 5 B. W. G. shall contain 3 coils per mile. No. 6 B. W. G. shall contain 3 coils per mile. No. 7 B. W. G. shall contain 2 coils per mile. No. 8 B. W. G. shall contain 2 coils per mile. No. 9 B. W. G. shall contain 2 coils per mile. No. 10 B. W. G. shall contain i coil per mile. Each coil shall be securely bound with four strong binding wires, the joints of these binding wires to be inside the coil. Each coil shall have its weight plainly and indelibly marked on a strong tag which shall be firmly attached to the inside of the coil. SPECIFICATIONS FOR STRANDED GALVANIZED STEEL WIRE Gage B.W.G. Diameter wire Diameter strand Lay in inches Breaking weight Maximum elongation 24 in. Minimum elongation 24 in. Pounds per 100 ft. 8 0.165 i | 4! 11,000 20 ii 52 12 o. 109 & 3^ 4,860 18 ii 22 14 0.083 \ 3 3,050 17 9 13 16 0.065 A 2 2,000 15 9 8 SPECIFICATIONS FOR AERIAL CABLE 453 INITIAL STRAIN No piece of wire under test shall be subjected to more than 10 per cent, of its required breaking weight before its elongation is considered. GALVANIZING Each wire must be thoroughly galvanized. Samples of the wire under inspection shall be dipped into a solution of sulphate of copper, saturated at 60 F. and allowed to remain for one minute, when they are to be withdrawn, washed and wiped clean. The galvanizing shall admit of this process four times without any signs of a reddish deposit upon the wire. Samples must bear coiling around a cylindrical bar twelve times the diameter of the sample, without any signs of the zinc flaking or peeling off. SPECIFICATION FOR AERIAL TWISTED PAIR (RUBBER COMPOUND DIELECTRIC) CABLE All conductors to be of No. 14 B. & S. (64 mils diameter) thoroughly annealed copper, 98 per cent, pure, according to Matthiessen's standard, equal in strength, finish, and pliability to the best market grade, well tinned and uniformly coated to a diameter of 158 mils with a high-grade rubber per- manent insulating compound, which shall adhere closely to the wire, and which shall not deteriorate under ordinary conditions. Each insulated conductor, before being laid up into cable form, must have its dielectric subjected in water, after 24 hours immersion, to a strain of not less than 1,000 volts alternating current between the conductor and the water, applied for one minute from a suitable generator or transformer; and must show while in the tank, after such immersion, an insulation of not less than 300 megohms per mile at 60 F., with not less than 100 volts applied for one minute. Test to be made by standard testing instruments in the presence of an inspector of the telegraph company. All conductors for test, either in coils or reels, must have tags securely attached, giving in plain figures the coil or reel numbers, the number of feet in each coil or reel, the gage of wire, and diameter of insulation; and such coils must so far as practicable be in uniform lengths corresponding to the length of the cable. Each insulated conductor in the cable must be protected by a closely woven cotton braid of not less than 15 mils thickness, thoroughly saturated with a compound which is not soluble in water, which does not act injuriously upon the permanent insulating compound, braid or tape, and which is not objectionable to handle. The two wires of a pair shall be twisted together, the length of the lay not to exceed 6 in. 454 AMERICAN TELEGRAPH PRACTICE One condcutor of each pair and one pair in each layer to be corded for tracing. The wires in each length shall be each of one piece and free from joints. The twisted pairs shall be laid up into a cylindrical core with the layers in reverse directions. Each pair of wires to be laid up with cushioning strands of saturated jute yarn of proper size. The cable must be wrapped over all with flexible cotton tape of first-class quality, saturated with first-class weatherproof compound. The tape must not be less than 20 mils thick, must have a lap of one-half its width, and firm adherence where lapped, so that it will not readily come apart. Over this the cable must have a durable protection of circular loom, braid, or tape, covering acceptable to the telegraph company, thoroughly saturated with the aforesaid compound. The finished cable must not be sticky or objectionable to handle. All cable made up as above, prior to shipment from factory, and after being placed upon reels, must have each length again tested for insulation by the inspector of the telegraph company; and each conductor under test must show an insulation (when all of the other conductors of the length are grounded) of not less than 500 megohms per mile at 60 F., with not less than 100 volts applied for one minute in the usual manner. 1 The conductors of the completed cable must also be tested for continuity, and the inspector shall make such tests for capacity and conductivity as he thinks advisable. The contractor will be required to 'furnish a table of coefficients of the resistance of the dielectric, showing its decrease above and its increase below 60 F., within the limits of variation of temperature to which the cable may be subjected during test. The reels upon which the cable is shipped must be strong and well pro- tected, and the cable neatly wound thereon with both ends so arranged that tests of the conductors on the reels may readily be made. A tag must be securely fastened to each reel, upon which the contractor must record the exact number of feet from end to end of the cable upon the reel, the number of conductors in the cable, and the date of shipment to the telegraph company from the contractor's factory. The contractor must give the usual guarantee that the cable will remain in good condition for one year after delivery, provided it is not used for currents of over one ampere, or having an electromotive force of over 500 volts; and must agree to repair or to reimburse the company for any expendi- tures incurred in repairing defects that may appear during that period, not caused by mechanical or other extraneous injury. The cable must conform in quality and manufacture to a sample pre- viously approved by the telegraph company. 1 Omitting immersion. SPECIFICATIONS FOR PAPER CABLE 455 SPECIFICATION FOR LEAD COVERED AERIAL OR UNDERGROUND SATURATED PAPER CABLE CONDUCTORS Each conductor to be No. 14 B. & S. gage (0.064 in.) soft-drawn copper wire, in one piece and free from joints. INSULATION Each conductor to be insulated with three wrappings of the best grade manila paper to a diameter of 158 mils. The whole core to be served with not less than three thicknesses of the best grade manila paper to a total thickness equal to that on the conductors and comprising not less than two wrappings. All wrappings to be thoroughly saturated with a high-grade insulating compound. LAYERS AND MARKING The conductors to be properly laid up with a marking wire in each layer. SHEATH Sheath to be not less than one-eighth inch thick and to contain 3 per cent, of tin by weight, to be uniform in composition and thickness and free from holes, splices, joints, porous places, or other defects, and to fit so closely as to leave no space between the core and the lead. ELECTRICAL TESTS The finished cable shall be immersed in a tank of water for 24 hours, at the end of which time the dielectric of each conductor shall be subjected to a strain of not less than 2,000 volts alternating current applied for one minute from a suitable generator or transformer. The insulation of each wire shall then be tested against all other wires and the sheath of the cable and must show a minimum insulation of 200 megohms per mile at 60 F, with 100 volts applied for one minute. Each conductor of the finished cable shall have a resistance of not more than 14.5 ohms per mile at a temperature of 68 F. The above tests to be made by an authorized inspector of the telegraph company in the presence of the manufacturer's representative. The Company requires the manufacturer to furnish a reliable table of coefficients of the dielectric's resistance, showing its decrease above and increase below 60 F., within the limits of variations of temperature to which 456 AMERICAN TELEGRAPH PRACTICE the cable may be subjected during tests, and the minimum of 200 megohms per mile will be modified accordingly. One conductor in each layer, as a tracer, to be well tinned, wrapped with dark blue paper and wound spirally with medium weight black cotton thread. REELS The finished cable to be free from mechanical defects and to be furnished in lengths as specified and wound on reels of suitable diameter. These reels are to have iron bushings of sufficient strength to safely carry the cable, and the cable to be wound thereon with both the inner and outer ends so arranged as to enable electrical tests to be made of the conductors while on the reel. A tag shall be securely attached to the reel, upon which shall be recorded the manufacturer's name, the exact number of feet of cable upon the reel, the number of conductors in said cable, and the date of shipment to the tele- graph company from the factory. Immediately after the cable has been tested and inspected by the tele- graph company's inspector, the ends of the cable shall be sealed with solder, and the inner end properly protected to prevent mechanical injury while in transit. When the manufacturer is required to draw the cable into a subway, it is to be installed with the proper splices free from all mechanical injury. Within thirty days after being laid, the cable shall be tested as herin described by an authorized inspector of the telegraph company and must show the insulation herein required. If the diameter of the cable called for in this specification is too great to admit of its being pulled into the duct of the subway provided, the Company is to be notified prior to the manufacture of the cable. MANUFACTURER'S GUARANTEE The manufacturer to guarantee the perfection of the cable and that the cable will remain in good working condition during a telegraph or telephone service of one year after it is delivered. During the first year after the cable is purchased the manufacturer to repair any defects due to faulty materials or manufacture, or to reimburse the company for expenditures incurred in repairing such defects. The manufacturer not to be responsible for defects caused by mechanical injury. SPECIFICATION FOR LEAD -COVERED TWISTED-PAIR PAPER SUBMARINE CABLE CONDUCTORS Conductors to be of gage as ordered, of soft-drawn copper wire, prefer- ably in one piece, free from joints; when joints are made they must be" so brazed that there will be no reduction in the tensile strength of the wire. SPECIFICATIONS FOR SUBMARINE CABLE 457 INSULATION Each conductor shall be insulated with not less than three spiral wrappings of the best grade manila paper, so that the total thickness of the insulating wall will be between 46 and 48 mils. TWISTS OF PAIRS AND MARKING The two wires of a pair shall be twisted together, the length of the lay not to exceed 6 in. The conductors of each pair to be one red and one white, and one white conductor in each layer to be spirally wound with coarse black thread for a layer tracer. CORE The twisted pairs shall be laid up into a cylindrical core with the layers in reverse directions. The whole core to be served with not less than three thicknesses of the best grade manila paper to a total wall thickness equal to that on the conductors and comprising not less than two wrappings. The cable must be laid up very compactly. For this reason the paper insulation should be applied loosely so that when the pairs are tightly cabled together, the interstices between the conductors shall be filled with dense paper. In the event of the number of pairs called for not permitting the construction of a compact core, the number of pairs may be exceeded by the amount necessary to insure such construction. SHEATH The lead covering must be of uniform composition and thickness, con- taining 3 per cent, of tin by weight, and free from holes, splices, joints, porous places or other defects; and shall fit so tightly as to make the core compact. The thickness shall be in accordance with the following sizes of armor wire : For No. 4 B. W. G., sheath to be 3/16 in. thick. For No. 6 B. W. G., sheath to be 1/8 in. thick. For No. 8 B. W. G., sheath to be 1/8 in. thick. JUTE COVERING OVER SHEATH The lead sheath shall be covered with three layers of jute, tightly and spirally wound in reverse directions, to a total wall thickness of 3/16 of an inch, thoroughly saturated with a compound which shall be impervious to water and resist .electrolytic action between the lead sheath and the zinc coating of the armor. 458 AMERICAN TELEGRAPH PRACTICE ARMOR Outside of the jute specified above, an armor of Ex. B. B. galvanized iron wire, gage as ordered, shall be laid on without twisting, with a lay of ten times the diameter of the cable over the armor. GALVANIZING OF ARMOR WIRES The galvanized wire used for the armor must comply with the specifica- tions of the telegraph company for galvanized iron wire. JUTE SERVING OVER ARMOR Outside of the armor shall be placed two wraps of jute laid on in reverse directions and thoroughly saturated with a preservative compound consisting of 65 parts of mineral pitch, 30 parts of fine sand and 5 parts of tar. The completed cable shall have a coating of soapstone to prevent the turns from sticking to each other on the reel. ELECTRICAL TESTS After the sheath has been put on, but before it is served with jute and armored, the cable shall be immersed in a tank of water for 24 hours, at the end of which time the dielectric of each conductor shall be subjected to a strain of not less than 1,000 volts alternating-current applied for one minute from a suitable generator or transformer. Each wire shall then be tested for insulation against all other wires and the sheath of the cable, and must show a minimum of 1,000 megohms per mile at 60 F. with 100 volts applied for one minute. Each conductor shall have a resistance and an electrostatic capacity which shall not exceed those specified in the following table: Gage B. & S. Resistance per mile in ohms at 68 F. of any wire Mutual capacity per mile in m'f'ds of any pair Average mutual capacity per mile in m'f'ds of all pairs 12 13 9.1 "5 0.150 0.125 O. 121 0.108 14 14-5 o. 100 O.OQ5 The finished cable on the shipping reel shall again be immersed and tested as above, and meet the same requirements. The above tests to be made by a regularly authorized inspector of the telegraph company in the presence of the manufacturer's representative. SPECIFICATIONS FOR TWISTED-PAIR CABLE 459 FILLING OF ENDS The ends of each length of cable must be filled with an insulating material which will seal the cable for a distance of 2 ft. or more from each end. REELS The finished cable to be free from all kinds of mechanical defects and to be furnished in lengths as specified, and wound on reels of suitable diameter. These reels are to have iron bushings of sufficient strength to safely carry the cable which is to be wound thereon with both the inner and the outer ends so arranged that electrical tests of the conductors can be made while on the reel. A tag shall be securely attached to the reel, upon which shall be recorded the manufacturer's name, the exact number of feet of cable upon the reel, the number of conductors in the cable, the reel number, and the date of ship- ment to the telegraph company from the factory. Immediately after the cable has been tested and inspected by the tele- graph company's inspector, the ends of the cable shall be sealed with solder, and the inner end properly protected to prevent mechanical injury while in transit. MANUFACTURER'S GUARANTEE The manufacturer to guarantee the perfection of the cable and that the cable will remain in good working condition during a telegraph or telephone service of one year after it is delivered. During the first year after the cable is purchased, the manufacturer to repair any defects due to faulty material or manufacture, or to reimburse the telegraph company for expenditures incurred in repairing such defects. The manufacturer not to be responsible for defects caused by mechanical injury. SPECIFICATION FOR LEAD-COVERED TWISTED-PAIR PAPER CABLE CONDUCTORS Conductors to be of gage as ordered, of soft-drawn copper wire, prefer- ably in one piece, free from joints; when joints are made they must be so brazed that there will be no reduction in the tensile strength of the wire. INSULATION Each conductor shall be insulated with not less than three spiral wrap- pings, with a half -width lap, of the best grade manila paper, so that the total thickness of the insulating wall will be between 46 and 48 mils. 460 AMERICAN TELEGRAPH PRACTICE TWISTS OF PAIRS AND MARKING The two wires of a pair shall be twisted together, the length of the lay not to exceed 6 in. The conductors of each pair to be one red and one white, except those of the tracer pair in each layer, which are to be one white and one dark blue, the latter spirally wound with coarse white thread. CORE The twisted pairs shall be laid up into a cylindrical core with the layers in reverse directions. The whole core to be served with not less than three thicknesses of the best grade manila paper to a total wall thickness equal to that on the conductors and comprising not less than two wrappings with a half-width lap. The cable must be laid up very compactly. For this reason the paper insulation should not be applied too tightly, so that when the pairs are tightly cabled together, the interstices between the conductors shall be filled with dense paper. In the event of the number of pairs called for not permitting the construction of a compact core, the number of pairs may be exceeded by the amount necessary to insure such construction. SHEATH The lead covering must be of uniform composition and thickness, contain 3 per cent, of tin by weight, and be free from holes, splices, joints, porous places or other defects. It shall be not less than i/8-in. thick and shall fit so tightly as to make the core compact. OVER-ALL DIAMETER The manufacturers must ascertain the permissible over-all diameter of the finished cable when order is placed. FILLING OF ENDS The ends of each length of cable must be filled with an insulating material which will seal the cable for a distance of 2 ft. or more from each end. ELECTRICAL TESTS The finished cable shall be immersed in a tank of water for 24 hours, at the end of which time the dielectric of each conductor shall be subjected to a strain of not less than i ,000 volts alternating current applied for one minute SPECIFICATIONS FOR TWISTED-PAIR CABLE 461 from a suitable generator or transformer. Each wire shall then be tested for insulation against all other wires and the sheath of the cable, and must show a minimum of 1,000 megohms per mile ar 60 F. with 100 volts applied for one minute. Each conductor shall have a resistance and an electrostatic capacity which shall not exceed those specified in the following table : Resistance per mile in ohms at 68 F. of Mutual capacity per Average mutual mile in m'f'ds of any wire any pair m'f'ds of all pairs 12 9.1 o. 150 O . I 2 I 13 n-5 o. 125 o. 108 14 14-5 O. TOO 0.095 16 23-5 0.092 0.087 REELS The finished cable to be free from all kinds of mechanical defects and to be furnished in lengths as specified, and wound on reels of suitable diameter. These reels are to have iron bushings of sufficient strength to safely carry the cable which is to be wound thereon with both the inner and the outer ends so arranged that electrical tests of the conductors can be made while on the reel. A tag shall be securely attached to the reel, upon which shall be recorded the manufacturer's name, the exact number of feet *of cable upon the reel, the number of conductors in the cable, the reel number, and the date of shipment to the company from the factory. Immediately after the cable has been tested and inspected by the tele- graph company's inspector, the ends of the cable shall be sealed with solder, and the inner end properly protected to prevent mechanical injury while in transit. MANUFACTURER'S GUARANTEE The manufacturer to guarantee the perfection of the cable and that the cable will remain in good working condition during a telegraph or telephone service of one year after it is delivered. During the first year after the cable is purchased, the manufacturer to repair any defects due to faulty material or manufacture, or to reimburse the company for expenditures incurred in repairing such defects. The manufacturer not to be responsible for defects caused by mechanical injury. 462 AMERICAN TELEGRAPH PRACTICE SPECIFICATION FOR AERIAL (RUBBER COMPOUND DIELECTRIC) CABLE All conductors to be of No. 14 B. & S. (64 mils diameter) thoroughly annealed copper, 98 per cent, pure, according to Matthiessen's standard, equal in strength, finish, and pliability to the best market grade, well tinned and uniformly coated to a diameter of 158 mils with a high-grade rubber permanent insulating compound, which shall adhere closely to the wire, and which shall not deteriorate under ordinary conditions. Each insulated conductor, before being laid up into cable form, must have its dielectric subjected in water, after 24 hours immersion, to a strain of not less than i ,000 volts alternating current between the conductor and the water, applied for one minute from a suitable generator or transformer; and must show while in the tank, after such immersion, an insulation of not less than 300 megohms per mile at 60 F., with not less than 100 volts applied for one minute. Test to be made by standard testing instruments in the presence of an inspector of the telegraph company. All conductors for test, either in coils or reels, must have tags securely attached, giving in plain figures the coil or reel numbers, the number of feet in each coil or reel, the gage of wire, and diameter of insulation; and such coils must so far as practicable be in uniform lengths corresponding to the length of the cable. One of the conductors in each layer of the cable must be suitably corded for tracing. Each insulated conductor in the cable must be protected by a closely woven cotton braid of not less than 15 mils thickness, thoroughly saturated with a compound which is not soluble in water, which does not act injuriously upon the permanent insulating compound, braid or tape, and which is not objectionable to handle. In case the number of conductors do not permit of layers in the mathe- matical ratio of i, 7, 19, 37, 61, 91, etc., small strands of semi-saturated jute are to be used to render the lay-up of the cable symmetrical and also to protect the insulation of the conductors when the cable is subjected to bends and twists. The cable must be wrapped over all with flexible cotton tape of first-class quality, saturated with first-class weatherproof compound. The tape must not be less than 20 mils thick, must have a lap of one-half its width, and firm adherence where lapped, so that it will not readily come apart. Over this the cable must have a durable protection of circular loom, braid, or tape covering, acceptable to the telegraph company, thoroughly saturated with the aforesaid compound. The finished cable must not be sticky or objectionable to handle. All cable made up as above, prior to shipment from factory, and after being placed upon reels, must have each length again tested for insulation by the inspector of the telegraph company. SPECIFICATIONS FOR OFFICE CABLE 463 Each conductor under test must show an insulation (when all of the other conductors of the length are grounded) of not less than 500 megohms per mile at 60 F., with not less than 100 volts applied for one minute. This test to be made without immersion. The conductors of the completed cable must also be tested for continuity, and the inspector shall make such tests for capacity and conductivity as he thinks advisable. The contractor will be required to furnish a table of coefficients of the resistance of the dielectric, showing its decrease above and its increase below 60 F., within the limits of variation of temperature to which the cable may be subjected 'during test. The reels upon which the cable is shipped must be strong and well pro- tected, and the cable neatly wound thereon with both ends so arranged that tests of the conductors on the reels may readily be made. A tag must be securely fastened to each reel, upon which the contractor must record the exact number of feet from end to end of the cable upon the reel, the number of conductors in the cable, and the date of shipment to the company from the contractor's factory. 'The contractor must give the usual guarantee that the cable will remain in good condition for one year after delivery, provided it is not used for currents of over i ampere, or having an electromotive force of over 500 volts; and must agree to repair or to reimburse the telegraph company for any expenditures incurred in repairing defects that may appear during that period, not caused by mechanical or other extraneous injury. The cable must conform in quality and manufacture to a sample pre- viously approved by the telegraph company. SPECIFICATION FOR OFFICE CABLE All conductors to be of No. 19 B. & S. (36 mils diameter) thoroughly annealed copper, 98 per cent, pure, according to Matthiessen's standard, equal in strength, finish, and pliability to the best market grade, well tinned and uniformly coated to a diameter of 101 mils with a high-grade permanent insulating compound, which shall adhere closely to the wire, and which shall not deteriorate under ordinary conditions. Each insulated conductor, before being laid up into cable form, must have its dielectric subjected in water, after 24 hours immersion, to a strain of not less than 1,000 volts alternating current between the conductor and the water, applied for one minute from a suitable generator or transformer; and must show while in the tank, after such immersion, an insulation of not less than 300 megohms per mile at 60 F., with not less than 100 volts applied for one minute. Test to be made by standard testing instruments in the presence of an inspector of the telegraph company. 464 AMERICAN TELEGRAPH PRACTICE All conductors for test, either in coils or reels, must have tags securely attached, giving in plain figures the coil or reel numbers, the number of feet in each coil or reel, the gage of wire, and diameter of insulation; and such coils must so far as practicable be in uniform lengths corresponding to the length of the cable. One of the conductors in each layer of the cable must be suitably corded for tracing. Each insulated conductor in the cable must be protected by a closely woven cotton braid of not less than 1 5 mils thickness, thoroughly saturated with a compound which is not soluble in water, which does not act injuriously upon the permanent insulating compound, braid or tape, and which is not objectionable to handle. The conductors must be so laid up as to make the completed cable sufficiently flexible to permit it to be bent without buckling to the diameter given in the following table. Diameters of drums on which office cables must bend without buckling: 5 conductor 2 in. 10 conductor 5 in. 25 conductor 9 in. 50 conductor 14 in. The cable must be wrapped over all with flexible cotton tape of first-class quality, saturated with first-class weatherproof compound. The tape must not be less than 20 mils thick, must have a lap of one-half its width, and firm adherence where lapped, so that it will not readily come apart. The finished cable must not be sticky or objectionable to handle. All cable made up as above, prior to shipment from factory, and after being placed upon reels, must have each length again tested for insulation by the inspector of the telegraph company; and each conductor under test must show an insulation (when all of the other conductors of the length are grounded) of not less than 500 megohms per mile at 60 F., with not less than 100 volts applied for i minute in the usual manner. 1 The conductors of the completed cable must also be tested for continuity, and the inspector shall make such tests for capacity and conductivity as he thinks advisable. The contractor will be required to furnish a table of coefficients of the resistance of the dielectric, showing its decrease above and its increase below 60 F., within the limits of variation of temperature to which the cable may be subjected during test. The reels upon which the cable is shipped must be strong and well pro- tected, and the cable neatly wound thereon with both ends so arranged that tests of the conductors on the reels may readily be made. A tag must be securely fastened to each reel upon which the contractor must record the exact number of feet from end to end of the cable upon the 1 Omitting immersion. SPECIFICATIONS FOR OFFICE WIRES 465 reel, the number of conductors in the cable, and the date of shipment to the telegraph company from the contractor's factory. The contractor must give the usual guarantee that the cable will remain in good condition for one year after delivery, provided it is not used for cur- rents of over i ampere, or having an electromotive force of over 500 volts; and must agree to repair or to reimburse the telegraph company for any ex- penditures incurred in repairing defects that may appear during that period, not caused by mechanical or other extraneous injury. The cable must conform in quality and manufacture to a sample previ- ously approved by the telegraph company. SPECIFICATION FOR OFFICE WIRES CONDUCTORS Conductors to be of gage as ordered, of soft-drawn copper wire not less than 98 per cent, pure, preferably in one piece free from joints. When joints are made, they must be so brazed that there will be no reduction in the tensile strength or conductivity of the wire. Wire must be tinned and uniformly coated with high-grade insulating compound to the thickness specified. BRAID Colored braid, when required, must be of the specified thickness, closely woven and with smooth surface. It is to be made of good quality strong cotton thread fast colored with non-injurious dyes. Braid on office wire must be made of closely woven, fire-proofed, strong cotton thread. Saturated braid must be filled with a first-class, water-proof, insulating compound which will give a smooth surface, but which will not be injurious to braid or rubber, become tacky at the highest summer or crack at the lowest winter temperatures. Each wire of twisted pairs must be braided separately and one of the wires suitably marked for tracing. ELECTRICAL TESTS The wire after being braided (except in the case of office wire) shall be immersed in a tank of water for 24 hours at the end of which time its dielectric shall be subjected to a strain of not less than 500 volts alternating current for two seconds for 5o-megohm wires, and 1,000 volts alternating current for five seconds for the 500- and the 75o-megohm wires. 30 466 AMERICAN TELEGRAPH PRACTICE The insulation of each length shall be tested in the usual manner with 100 volts applied for one minute at 60 F. Office wire must be immersed and tested as above before the braid is put on. After braiding, samples must be submitted for approval. COILS Coils under test must be serially numbered and so far as practicable in uniform lengths of 500, 1,000 or 2,000 ft. Subsequently the wire must be cut into lengths to conform with the order. Each accepted coil must be neatly laid up and wrapped in paper or burlap. After wrapping, a tag giving length, weight, gage, kind, color, and manu- facturer's name must be securely fastened to the inside of each coil. The coils wrapped and tagged as above must be packed in barrels which are to be plainly marked, showing the size or gage, kind, color and total number of feet in each barrel. TABLE OF STANDARD RUBBER COMPOUND INSULATED WIRES Conductor Minimum diameter Braid Minimum megohms Name B. &S. gage wire with- out braid mils 1 Thickness mils Color and finish per mile 60 F. 100 volts Office 16 113 i 3 Gray flame proof COO Bridle 14 158 20 Black saturated 750 Pothead and battery stems. 14 140 No braid 750 Call circuit 16 II3 20 Black saturated CO Outside twisted pair. 14 158 20 Black saturated 750 Annunciator 18 IO2 IIT Glazed cotton CO color as ordered. Annunciator twisted pair. 18 IO2 15 Glazed cotton, color as ordered. 50 1 Maximium allowable diameter of insulated wire without braid 5 mils above minimum. CHAPTER XXIV ELECTROLYSIS OF UNDERGROUND CABLE SHEATHS When two pieces of metal are immersed in an electrolyte consisting of slightly acidulated water, and a current of electricity is passed between tjiem, minute particles of one of the metals are decomposed and deposited upon the surface of the other metal. The action is the same as that which takes place in the process of electroplating. The detached particles are carried from one metal to the other in the same direction the current travels from positive to negative plate. A similar action takes place between the sheaths of buried cables and the tracks of electric railroad systems. The positive pole of the railroad power dynamo is connected to the feeder system and the negative pole of the dynamo to the steel rails, making a circuit via the car trolley or shoe equip- ment through the car motors and back to the generating station over the track. Where the rails are in contact with the earth either directly or indirectly (by way of metal supporting structures) the current in the rails has a tend- ency to leak away from the track and travel back to, or in the direction of, the power station. Obviously, wherever metal pipe systems are buried in the earth adjacent to electric railroad tracks, the former in many cases will form a branch of a joint-circuit constituting the return path of the railroad current. If the rails were perfectly insulated from the earth or were laid in perfectly dry ground this action could not take place, but in 'most localities there is a sufficient amount of moisture beneath the surface of the earth to act as an electrolyte. The earth serves the purpose of an electroplating tank, as it has all the elements required, namely, the car tracks and cable sheaths sepa- rated by a more or less moisture-saturated compound. The fact that the sheaths of the buried cables act as conductors for a portion of the return current is not of much concern, but if while serving as a conductor the sheath is immersed in an electrolyte, the danger is that a por- tion of the current will leave the conductor and pass through the electrolyte to another conductor, for in that case it will carry particles of the sheath to the surface of the second conductor, and while this action does not particu- larly menace the usefulness of the latter it soon results in disintegration of the cable sheath with the result that moisture is permitted to enter the cable causing leaks, crosses, short circuits and grounding of the conductors con- tained in the cable. 467 468 AMERICAN TELEGRAPH PRACTICE The amount of current which escapes from a track to adjacent pipe lines depends upon the relative electrical distance of the track circuit, the earth and the pipe line, with regard to the location of the power house. A very small current volume, however, will produce electrolytic action. For example, one ampere flowing one hour dissolves 0.035 oz - f cast iron, 0.105 oz - f wrought iron, and 0.125 oz. of lead. In practice, of course, the current does not often flow from any one point, but is spread out over a considerable length of cable or track; but it will be realized that even where the surfaces affected have considerable area, where the action continues for any great length of time the result will be disastrous. The ideal remedy is to provide a return circuit for the railway current of sufficient capacity to reduce leakage to the lowest possible degree, by properly bonding the rails or by installing an auxiliary return wire of large section bonded to the rails at intervals, but even where this is done, at points remote from the power house neighboring pipe systems are still likely to form branches of a joint-circuit back to the negative terminal of the dynamo. Where the current passes from the tracks to the cable sheath or to a pipe line, no damage will be done to either of the latter, but the tracks themselves will be eaten away. Where the current leaves the sheath or pipe, however, and passes to the tracks or to other pipe systems, damage will be done to the sheath or pipe at the point or points where the current leaves them. When any interruption occurs to the track return circuit, as, for instance, when a bond is broken, or a rail is broken or temporarily removed, a large proportion of the return current is shunted around the break through ad- jacent pipe lines and the large current volume diverted through the latter within a very short time causes serious damage at the point where the current leaves the sheath or pipe in returning to the track rails at a point on the gen- erator side of the interruption. In order to determine where electrolysis is liable to occur, it is well to obtain or prepare a map showing the location of manholes and cables and the routes they take. Upon this map the electric railway lines may be traced with red ink. At all manholes measurements should be made between the rails and cables, water-pipes, gas-pipes, manhole frame (if metal), water in manhole, other cables and in fact all metal objects that are buried in the ground in that neighborhood which would in any way affect the cables. In making the measurements a voltmeter with a low reading scale should be used, attaching a wire to the positive terminal of the meter and another wire to the negative terminal. The free ends of the wires should be attached to strips of lead or to steel rods (old steel files with sharp points make good substitites) as it is found that when the ends of the copper wires are used a local action sometimes takes place which interferes with the true readings. If, when the wire attached to the positive terminal of the voltmeter is placed in contact with the rail, water-pipe, water in the manhole, manhole ELECTROLYSIS OF UNDERGROUND CABLE SHEATHS 469 frame, or other object, while the wire attached to the negative terminal of the voltmeter is placed in contact with the cable sheath; the voltmeter pointer is deflected to the right, the indication means that a current is flowing from the rail, pipe or manhole frame, etc., to the cable sheath. If the deflection is in the same direction as contact is made with each object, a record should be made to the effect that the cable is (negative) to the earth at that point, the rail, and to other pipe systems. If it is found when contact is made in any instance that the pointer deflects to the left the indication means that cur- rent is flowing from the cable sheath to the neighboring object and that the sheath is being slowly eaten away. The exact difference of potential may be learned by reversing the wires in the binding-posts of the voltmeter so that the pointer may move from its zero position at the extreme left of the scale, to a point on the right which indicates the existing voltage. If a reliable low-reading voltmeter is not at hand, and a Weston galvanom- eter is available, the latter may be used for electrolysis tests by connecting an external resistance of 5,000 ohms in series with the galvanometer. Then, with no shunt around the galvanometer movement coil, the scale will have a reading of o.i volt, in o.oi-volt divisions. Using the one- tenth shunt, the scale-reading will have a value of i volt in divisions of o.i volt. Using the one one-hundredth shunt the scale will have a value of 10 volts in divisions of i volt. CABLE TO CABLE, AND CABLE TO RAIL BONDING Undoubtedly there are instances of electrolytic corrosion of cable sheaths not attributable to stray railway currents, such, for instance, as occur where cables are laid in earth strewn with cinders, or where the character of the soil in which the cable is buried is such that galvanic action takes place between the cable sheath and neighboring metallic substances, or in cases where dur- ing the winter months frozen water-pipes are thawed out by heating them by passing currents of large volume through the frozen sections for a number of hours, 1 but inasmuch as the bulk of the trouble experienced is due to electric- railway return currents, it is good practice to take all possible precautions to insure a low resistance return path to the power station for these currents. In some instances it has been found advisable to run bare stranded copper- wire cables parallel to and bonded to the cable for the purpose of shunting stray currents which otherwise would flow through the sheath of the cable. Satisfactory operation of electric railroads requires that adequate bond- ing between abutting rails be maintained in order that the circuit resistance 1 Where this method of thawing water-pipes is practised it has been found that neigh- boring pipe systems are endangered to an extent dependent upon the proximity of such lines to the water-pipes being treated, upon the character of the sub-soil, and upon the electrical continuity of any intervening joints in the water-pipe between the points upon their surfaces where the thawing current is applied. 470 AMERICAN TELEGRAPH PRACTICE between the negative terminals of car motors and the negative terminal of the power generator will be as low as possible, and unless the condition of all bonds is constantly inspected, high-resistance contacts are liable to develop and remain undetected until considerable damage has been done to adjacent cable sheaths. So far as rail bonding is concerned it should be remembered that the conductivity of copper is about ten times that of the steel used in making the rails, the copper bond employed should, therefore, be of one- tenth the sectional area of the rail if the bond is to have the same current- carrying capacity as the rail. Where two or more cables terminate in, or pass through a manhole the various cables should be bonded together, preferably with a strip of lead 2 or 3 in. in width. Where lead is used in cable to cable bonding there is less likelihood of galvanic action than where copper bonds are used. Where cable sheaths are found to be positive to track rails, it is customary to bond the cable to the rails, and while there are certain objections to this practice and some risk incurred, the general experience is that the advantages outweigh the disadvantages. APPENDIX A REFERENCES TO PRINTING TELEGRAPH LITERATURE 1. THE BRETT PRINTING TELEGRAPH, " The Telegraph Manual," Shaffner, 1859, page 273. 2. THE BONELLI TYPO-TELEGRAPH, " Electricity and the Electric Tele- graph," Prescott, 1888, page 763. 3. THE BUCKINGHAM PRINTER, "American Telegraphy," Maver, page 436a. 4. THE BARCLAY PRINTING TELEGRAPH SYSTEM, Serial Article by William Finn, in Telegraph Age, N. Y., running from June 16, 1908, to March i, 1909. 5. THE BURRY PRINTER, Telegraph Age, N. Y., April i, 1903, page 169. 6. THE BAUDOT PRINTING TELEGRAPH SYSTEM, "The Hughes and Baudot Telegraphs," Crotch. Rentell & Co., London, 1908. 7. THE CARD WELL PRINTING TELEGRAPH SYSTEM, Telegraph Age, N. Y., June i, 1905, page 221. 8. THE "COMBINATION" TELEGRAPH PRINTER, "Electricity and the Elec- tric Telegraph," Prescott, 1888, page 608. 9. THE CREED TELEGRAPH PRINTER, Electrical Review, London, Sept. 25, 1908. Electrical Review, London, Dec. 4, 1908. Telegraph Age, New York, July i, 1907. 10. THE DEAN PRINTING TELEGRAPH SYSTEM, Telegraph Age, N. Y., Aug. 16, 1907, page 443. 11. THE ESSICK PRINTER, "American Telegraphy," Maver, page 431. 12. THE HOUSE PRINTING TELEGRAPH, "The Electromagnetic Telegraph," Lardner, 1853, page 117. "History, Theory, and Practice of the Electric Telegraph," Prescott, 1864, page in. "Electricity and the Electric Telegraph," Prescott, 1888, page 604. 13. THE HUGHES PRINTING TELEGRAPH, "History, Theory and Practice of the Electric Telegraph," Prescott, 1864, page 139. "Electricity and the Electric Telegraph," Prescott, 1888, page 608. 14. THE HUGHES TYPE-PRINTING TELEGRAPH SYSTEM, "Telegraphy," Herbert, 1906, page 370. "Hughes Type-printing Telegraph System," Wyman & Sons, London, 1906. 15. THE MORKRUM PRINTING TELEGRAPH, Telegraph Age, N. Y., June. 16, 1912. 16. THE MURRAY PRINTING TELEGRAPH SYSTEM, "Telegraphy," Herbert, 1906, page 826. 471 472 AMERICAN TELEGRAPH PRACTICE 17. THE PHELPS TYPE-PRINTING TELEGRAPH, " Electricity and the Electric Telegraph," Prescott, 1888, page 736. 18. THE PHELPS MOTOR PRINTER, "American Telegraphy," Maver, page 4i9b. 19. THE ROWLAND PRINTING TELEGRAPH SYSTEM, Proceedings of the American Institute of Electrical Engineers, Vol. XXVI, 1907, page 507. 20. SIEMENS TYPE-PRINTING TELEGRAPH. " Electricity and the Electric Telegraph," Prescott, 1888, page 734. 21. THE WRIGHT PRINTER, Telegraph Age, N. Y., May 16, 1910, page 348. APPENDIX B SPECIFICATIONS FOR THE CONSTRUCTION OF HIGH-TENSION POWER TRANSMISSION LINES ABOVE TELEGRAPH WIRES 1 GENERAL (a) These specifications apply to constant-potential power transmission lines of over 5,000 volts. (b) These specifications prescribe a certain minimum standard of con- struction for the high-tension line which is required in order to provide a reasonable degree of security against the failure of any portion of the high- tension construction that might allow the high-tension wires to come into contact with the telegraph wires. (c) It is not the purpose of these specifications to restrict the high-tension construction narrowly in details, but to stipulate the fundamental principles which must be followed in order to attain reasonable safety. (d) Each portion of the high-tension line shall have sufficient strength to resist the maximum mechanical stresses to which it may be subjected, due allowance being made for a factor of safety suited to the degree of uniformity of the material, the character of the material with respect to deterioration and the nature of the stress, as hereinafter specified. (e) Obviously the maximum mechanical loads upon the high-tension con- struction will usually occur when the wires are coated with ice and subjected to the maximum wind velocity at right angles to the line at the minimum temperature. (f) The maximum stresses in the high-tension construction shall be com- puted on the basis of a wind pressure of 20 Ib. per square foot of plane area, or 12 Ib. per square foot of projected area for cylindrical surfaces. These values are based upon a maximum actual wind velocity of 70 miles per hour and are to be used in connection with the following coincident conditions : (1) Maximum coating of ice, 1/2 in. in thickness. (2) Minimum temperature, zero degrees Fahrenheit. NOTE. In a few sections, in southern portions of the country, minimum temperatures of zero degrees and ice formation are not encountered. For transmission lines constructed in such regions the above requirements may be suitably modified to accord with local climatic conditions. In no case shall the minimum temperature be taken above 30 F. 1 From Standard Specifications. 473 474 AMERICAN TELEGRAPH PRACTICE (g) The general types of construction which shall be employed, the factors of safety to be observed, and the minimum sizes and strengths of materials, shall be as specified below. (h) Where galvanizing of iron or steel is required, it shall conform to the requirements of the appended specifications for galvanizing for iron and steel. TOWERS AND POLES (1) Material. The poles supporting the high-tension conductors where these are above the telegraph line shall preferably be of steel. Reinforced concrete or wood poles may be employed under suitable restrictions as herein- after specified. (j) Factors of Safety. (i) Poles shall have the following minimum fact- ors of safety according to the nature of the materials employed: Steel 3 Reinforced concrete 4 Completely creosoted wood 5 Other wood 6 (2) The poles, at the terminals of the portion of the high-tension line cov- ered by these specifications, shall be of such strength as not to break under the maximum load conditions, if any, or all, of the conductors in the spans outside this portion should break. (k) Wood or Reinforced Concrete Poles. (i) Wood poles shall not be used where inflammable materials, such as structures, are situated within a distance sufficient to cause an appreciable fire hazard to the pole. (2) If wood poles are employed, surrounding underbrush and grass must be removed for a sufficient distance to avoid fire hazard. (3) Wood or reinforced concrete poles must be provided with a grounded copper wire or an approved equivalent metal strip, placed at the side of the pole and extended to the top of the pole and over the top of the pole. In the case of wood poles this grounded conductor shall be extended down the opposite side of the pole to the top of the lowest cross-arm, Fig. 419. This grounded conductor shall be of sufficient conductivity to carry safely the maximum short-circuit current. This grounded conductor provides lightning protection and, in the case of wooden poles, serves to prevent arcing and setting fire to the pole in case a high-tension wire becomes detached from its insulator and rests against the side of the pole. (1) Guys. (i) Where guys can be placed, the total strength of the guyed structure shall be sufficient to sustain the maximum stress with factors of safety not less than those specified in section (j). (2) All guys shall be anchor guys, guys to anchored stubs or rock guys. (3) Methods of anchoring, locations for anchors, and depth and character of setting shall be such as to render effective the full strength of the guy. APPENDIX B 475 (4) Guys shall be of galvanized steel strand not smaller than five-six- teenths in. in diameter. (5) Strain insulators are not required, but if these should be placed in guys, each strain insulator shall have a breaking strength not less than that of the guy in which it is placed. Every guy which passes over or under any electric wires, other than these carried upon the guyed pole shall be so placed and maintained as to provide at all times a clearance of not less than 2 ft. between the guy and such electric wire. (m) Minimum Size Wood Poles. No wood pole, whether guyed or not, shall be less than 8 in. in diameter at the top. a A a ft ft Not less than ^Galvanized Strand -y ^-Galvanized / IronStr/ps * ft FIG. 419. (n) Replacement of Wood Poles. Wood poles shall be periodically inspected and shall be replaced before their strength falls below two-thirds of their initial strength. (o) Structural Steel Poles or Towers. (i) All structural steel shall conform to specifications for open-hearth railway bridge or medium steel adopted by the Association of American Steel Manufacturers. (2) All steel poles and towers shall either be galvanized or thoroughly painted with not less than three coats of an approved metal preservative. 476 AMERICAN TELEGRAPH PRACTICE Painting shall consist of at least one shop coat and two field coats, preferably all of different shades of color. (3) Steel poles and towers shall be thoroughly grounded in a manner satisfactory to the telegraph company. (p) Unit Strength of Materials. The fiber stresses to be employed in computing the strengths of poles shall not be more than as follows: Working fiber stress ( medium 20,000 Ib. \ railway bridge 18,500 Ib. Cedar 600 Ib. Chestnut 800 Ib. Creosoted yellow pine 1,200 Ib. The working-fiber stresses given above include allowances for factors of safety in accordance with the preceding requirements. (q) Setting Poles. (i) Great care shall be taken in setting poles at high-tension crossings to secure firm foundations. (2) Exposure to washouts shall be avoided. (3) Poles shall not be set on sloping banks when other location is practi- cable. Where poles are necessarily set on sloping banks they shall be well reinforced by cribbing. (4) In sandy or swampy soil concrete foundations shall be provided for wood poles. Each foundation shall contain not less than two cubic yards of concrete. (5) Concrete shall not be leaner than one part of cement to two and one-half parts of sand, to five parts of broken stone. An equivalent gravel concrete may be used. Cement shall be Portland cement conforming to the standard specifications of The American Society for Testing Materials. Sand shall be clean and sharp. All concrete shall be mixed and placed thoroughly wet. WIRES (r) Spans Covered by these Specifications, (i) Crossings. The con- struction herein specified applies to the cross-over span. Where the distance from the topmost high-tension wire at either pole of the cross-over span to the nearest wire on the telegraph line is less than one and one-half times the height of the topmost high-tension wire above the ground at the high-tension pole, the requirements specified for the cross-over span shall be considered as applying also to the next high-tension span adjacent to that pole, Fig. 420. (2) Parallel Lines. Where the high-tension line must necessarily be constructed higher than and parallel to the telegraph line, and separated from the latter by a distance less than the height of the high-tension poles, the construction shall conform to the requirements for the cross-over span APPENDIX B 477 as herinafter specified. The requirements shall also apply to each span next adjacent to the portion above the telegraph line, unless the distance from the nearest telegraph wire to the topmost high-tension wire on the high-tension poles at the end of the over-built section, is greater than one and one-half times the height of the topmost high-tension wire from the ground, Fig. 421. (s) Factors of Safety. The length of the cross-over span and the sag of the wire shall be so proportioned, with reference to the kind and size of wire \ v q \ a a e e gf~'' 1 ^ , h -% w^ -^f ~ / '// ' ^ *uu : Note: * ( If "a "is less than 1^ times "h " the Requirements for the Cross- over Span shall apply also to the adjacent Span. FIG. 420. and method of suspension, that a factor of safety of at least 2, and no stresses beyond the elastic limit of the material will be obtained under the maximum conditions specified in clause (f). (t) High-tension Conductors. (i) Stranded wire shall be used for the high-tension conductors in the cross-over span and other spans covered by the requirements of these specifications. Each shall consist of not less than seven component wires. (2) The minimum sizes of conductors shall be Copper Aluminum. . Not less than No. o B. & S. gage. Not less than No. oo B. & S. gage. (3) There shall be no joints in the conductors in the spans requiring special construction. (u) Precautions against Injury to Wires from Arcing, (i) Separation. The minimum separation between wires on centers shall be as follows: 478 AMERICAN TELEGRAPH PRACTICE M APPENDIX B 479 Voltage between wires Minimum separation on centers Under 1 2 tjoo 2 ft. 1 2 500 to 10 ooo 2^ ft. 20 ooo to 20 ooo 3! ft. ?o ooo to 30 ooo 4 ft. 40 ooo to 50 ooo 5 ft. 60 ooo and. over 6 ft. (2) At Insulators. (i) At the poles forming the termini of the spans covered by these specifications, each conductor shall be so protected at the point of attachment to the insulator, that if the insulator breaks down elec- trically, the resulting arc will not burn the conductor. This may be accom- plished by providing between the conductor and the insulator a metal cap which will interpose at least 1/2 in. of metal between the line conductor and the head of the insulator. Also the conductor shall be protected from an arc for a distance of not less than 24 in. on each side of the center of the insulator head by a serving of wire or a sheet metal envelope not less than No. 6 B. & S. gage in thickness. (2) The wire serving or sheet metal envelope shall be of the same kind of metal as the line conductor which it protects. (v) Minimum Clearance above Telegraph Wires. The high-tension construction shall be such that at a temperature of 130 above the minimum temperature (clause F.), the lowest high-tension wire shall clear the highest telegraph wire or cable by not Jess than 8 ft. Where practicable, no telegraph pole shall be closer than 1 5 ft. horizontally to the nearest high-tension wire. The telegraph crossarms may be spaced 15 in. on centers at crossings in order to allow high-tension poles of minimum height to be used. (w) Unit Strength and Elasticity of Materials. (i) The tensile strengths to be employed in computing the wires shall not be more than as follows: Working strength, pounds per square inch Hard-drawn copper stranded conductor Hard-drawn aluminum stranded conductor 30,000 12 OOO (2) The modulus of elasticity may be taken as follows: 480 AMERICAN TELEGRAPH PRACTICE Modulus of elasticity Hard-drawn copper stranded conductor Hard-drawn aluminum stranded conductor Steel strand 12,000,000 7,500,000 22 OOO OOO (x) Coefficient of Linear Expansion. The coefficient of linear expansion per Fahrenheit degree may be taken as follows: Copper o . 0000096 Aluminum o . 0000130 Steel o . 0000064 (y) Method of Attachment. In all spans covered by these specifications, the high-tension conductors shall be attached to the insulators on each side of the span by mechanical clamps or approved ties. Ties such as are ordinarily employed for signaling wire shall not be used. The clamps or ties shall have sufficient grip and shall be set up sufficiently tight so as to hold the conductors up to stresses equal to the working strengths of the conductors and shall be of such a design as not to injure the wire. If ties are used, the tie wires shall be attached to the line conductor at a distance from the head of the insulator not less than one- tenth of the distance specified in section (u), and in no case less than 4 in. CROSSARMS (z) Material. The crossarms supporting the wires or strands shall be of steel or creosoted wood. (aa) Factors of Safety. Crossarms shall have the following minimum factors of safety according to the nature of the material employed: Steel 3 Creosoted wood 5 (bb) Loads on Crossarms. The crossarm and its attachment shall have sufficient strength to provide against breaking, in the case of the breaking of any or all of the wires in the span adjacent to the cross-over span. (cc) Steel Crossarms. Steel crossarms should preferably be used. Steel crossarms shall be thoroughly grounded. All portions of the ground connec- tion shall have sufficient conductivity to carry safely the maximum short- circuit current. (dd) Wood Crossarms. (i) If wood crossarms are employed they shall be treated with creosote or dead oil of coal tar in accordance with approved specifications. APPENDIX B 481 (2) Wood crossarms shall be provided with grounded galvanized iron plates or grounded copper wires on their upper surfaces. Plates shall not be less than 1/4 of an inch in thickness, and of a cross-sectional area not less than that of the ground wire. If copper wires are employed they shall be of suffi- cient conductivity to carry safely the short-circuit current. Ground wires or plates shall be firmly attached to the crossarms. (ee) Protection of Metal from Corrosion. All portions of steel crossarms and their fittings, and the center bolts, braces, ground plates and other fittings of wood crossarms shall be thoroughly galvanized. (ff ) Protection Against Line Conductors Falling Clear of Crossarms. At spans where these specifications apply, angles in the route of the high-tension line shall be avoided wherever practicable. At these spans, if mechanical clamps are not employed, the outer high-tension line conductors shall in all cases be attached so as to pull against the insulators. PINS (gg) Material. Steel pins shall be used. (hh) Strength of Pins. Pins shall be sufficiently strong to provide a factor of safety of 3 against stresses produced by the maximum wind pres- sure on the wires loaded with ice and also against stresses produced by the breaking of the wire in the span adjacent to the crossing span. (ii) Grounding of Pins. Pins shall be thoroughly grounded. INSULATORS (jj) Material. Porcelain insulators shall be used for supporting the high- tension conductors. (kk) Mechanical Strength. The insulators shall be sufficiently strong so that, when mounted, they shall be able to withstand without injury twice the maximum mechanical stress to which they will be subjected with the line conductors attached as herein specified. (11) Dielectric Strength. Where tested under approved methods each insulator shall be capable of resisting three times the normal voltage when tested dry and twice the normal voltage under spray test. (mm) Disk Insulators. Where suspension insulators are used, each individual disk shall be provided with interlinked attachments so that, in case the porcelain should be shattered, the conductor would remain mechanic- ally attached to the crossarm. The support next adjacent to the crossarm shall be thoroughly grounded. LIGHTNING PROTECTION (nn) Each pole and tower, in the portion of the high-tension line covered by these specifications, shall be provided with a grounded lightning-protec- tive device extending above the top of the pole or tower and not less than 3 ft. above the highest conductor. 31 482 AMERICAN TELEGRAPH PRACTICE CONSTANTS, UNIT STRESSES AND FORMULAE TO BE USED IN COMPUTING STRENGTH OF TRANSMISSION LINES UNIT STRESSES POLES AND TOWERS Allowable working fiber stress Steel. . medium railway bridge. Cedar Chestnut Creosoted yellow pine . . . 20,000 Ib. per square inch. 18,500 Ib. per square inch. 600 Ib. per square inch. 800 Ib. per square inch. 1,200 Ib. per square inch. WIRE AND STRAND Allowable tensile strength Stranded copper. . . . Stranded aluminum. Steel strand. . 30,000 Ib. per square inch. 12,000 Ib. per square inch. (According to the character of the material, a factor of 3 being used). Modulus of elasticity Stranded copper. . . . Stranded aluminum. Steel strand. . 12,000,000 7,500,000 22,000,000 Coefficient of linear expansion per degree Fahr. CoDoer o . 0000096 Aluminum . . . o. 0000130 Steel o . 0000064 APPENDIX B 483 CROSSARMS Allowable working fiber stresses Steel. medium railway bridge. Creosoted yellow pine 20,000 Ib. per square inch, 18,500 Ib. per square inch. 1,200 Ib. per square inch. PINS Allowable working fiber stress Steel. 20,000 Ib. per square inch. Wind Pressure. P= pressure in pounds per square foot. 7 = actual velocity of wind in miles per hour. For plane surfaces P =o.oo4F 2 . For cylindrical surfaces P =0.002572 (P = pressure per square foot of projected area.) For velocity of 70 miles per hour P =20 Ib. for plane surfaces. P =12 Ib. for cylindrical surfaces. Sleet and Ice. Weight of ice per cubic foot, 58 Ib. Weight of block of ice i ft. long and rsq. in. section, 0.403 Ib. Poles. A pole is essentially a beam fixed at one end. The ordinary beam formulae apply. The strength of a pole is given by the formula M+ y where M = moment of the forces about the ground line (or other point at which the strength is being considered) . p = maximum fiber stress. / = moment of inertia of section of pole. y = distance from center to most strained fiber. For a pole of circular cross-section 484 AMERICAN TELEGRAPH PRACTICE where D = the diameter of the pole in inches and the moment arms of the forces are expressed in feet. p is the maximum ultimate fiber stress or the allowable working fiber stress according as the ultimate strength or safe working strength of the pole is desired. Forces Acting on a Pole Transversely. Wing pressure on pole. Wing pressure on conductors. The approximate moment at the ground due to wind pressure on the pole would be ^ 72 P = wind pressure per square foot of projected area. H = height of pole above ground in feet. DI = diameter of pole at ground. D 2 = diameter of pole at top. The moment at the ground due to wind pressure on the wires would be 24 L = height of wires above ground in feet. n = number of wires. Di = diameter of conductor loaded with ice. Si = and Sz = lengths of adjacent spans in feet. The total moment is the sum of M P , Mci, Mc 2 , etc. Conductors. A metallic conductor is elastic and also expands and con- tracts with changes in temperature. When a wire in a span is cooled it contracts, making the sag less and increasing the tension in the wire. The elongation of the wire due to the increased tension tends to increase the sag and to diminish the tension. When a wire is loaded by sleet or by wind pressure the increased tension in the wire causes it to stretch and the sag to increase. The increase in the sag tends to reduce the tension. The formulae for computing these various changes are as follows: Relation between temperature and sag 81 fJ 2 j 2\ , T T P v d, where sag and span are expressed in feet. where sag is in inches and span in feet. APPENDIX B 485 Symbols. a = temperature variation for small changes in sag. /o = initial temperature. F = length of span. do = sag at temperature to. di =sag at temperature t\. c = coefficient of linear expansion per degree F. e = modulus of elasticity. p =load per foot of wire. s = cross-section of wire in square inches. By assuming small changes in the sag, successive values of a may be found from which a curve showing the variation of sag with temperature may be made. Relation between tension and sag. pY 2 T= --g-j- (sag and span in feet) Length of wire. (8 d 2 \ iH TFs) (span and sag in feet) O Elongation due to change in tension. TL ~ es Example. Poles: Length of pole, 40 ft. Height of pole above ground, 34 ft. Length of adjacent spans, 100 ft. and 120 ft. Wires: One No. o wire on top of pole. Two No. o wires on crossarm 3 ft. below top. Two No. 12 telegraph wires on crossarm 7 ft. below lower power wires. To find dimensions of cedar pole to give factor of safety of 6 with a wind velocity of 70 miles in a direction at right angles to the line and 1/2 in. thickness of ice on each wire. Wind pressure on upper wire. Diameter of No. o wire = 0.37. Diameter of No. o wire covered with 1/2 in. ice = 1.37. ^PLnDj (5 t +5a) 24 _ 12.3X34X1X1.37(100+120) _ MCl ~ ~2^T -5,253- 486 AMERICAN TELEGRAPH PRACTICE Wind pressure on two middle wires: 12.3X31X2X1.37(100+120) Mc 2 = - ~ = 9>579- Wind pressure on telegraph wires: Diameter of No. 12 wire = 0.104 in. Diameter of No. 12 wire covered with 1/2 in. ice = 1.104 m - 12.3X24X2X1.104(100+120) Mc s = ^- - = 5,975- Wind pressure on pole assuming diameters at butt and top to be 17 in. and 8 in. , 6,500 I2 -3X34 2 X33- (If the result gives dimensions of poles much different from the values assumed a second approximation should be made.) Total moment = 27,307 For cedar p = 600 6007TZV 4.91 Z> 1 3 = 27,307 1)1=17.1 in. Circumference at ground line = 54* in. Example. Conductors: To find sag of wire at 60 F. such that the wire will have a factor of safety of 2 at o F. with ice 1/2 in. thick all around the wire and wind blowing at right angles to the line at a velocity of 70 miles an hour. Size of wire No. o. Wire of stranded copper. Length of span 200 ft. Diameter of wire = 0.370 in. Cross-section of copper = 0.083 sc l- m - Diameter of wire + 1/2 in. ice = 1.370 in. Cross-section of wire +1/2 in. ice = 1.47 sq. in. Cross-section of ice = 1.39 sq. in. Weight of wire per foot = 0.323 Ib. APPENDIX B 487 Weight of ice per foot = 0.403X1.39 = 0.560 Ib. Weight of ice and wire per foot = 0.560+0.323=0.883 Ib. 1.370 Wind pressure per foot = ---- X 12. 3 = 1.40 Ib. Resultant pressure per foot = V i.4o 2 + o.883 2 = Breaking weight of No. o wire = 4,980 Ib. With factor of safety of 2, the allowable tension in the wire = 2,490 Ib. At o the wires weighted with wind and ice d = sr 1.65X40000 8X2490 ^ = 3.3 ft. With sag of 3.3 ft. the length of wire 200(1 + 3 Contraction due to removal of wind and ice : TL E = es 200.145 1 2000000 X 0.083 = 0.000201 T (T in this case being the differ ence in tension.) Tension Length Sag 2,490 200. 145 3-3 2,39 200. 125 3-o6 2,290 200. 105 2.80 2,190 200.085 2.52 2,090 200.065 2.21 1,990 200.045 1.84 1,890 200.025 i-37 1,790 200.004 0-55 488 AMERICAN TELEGRAPH PRACTICE The sag for each of the above lengths is determined from the formula; = 8.65 \/L 200 Increase in tension due to contraction of wires: wY 2 "If 0.323 X 40000 ST 1615 T Tension Sag * 1,615 I.O 1,700 i, 800 1,900 o-95 0.9 0.85 Plotting these two relations of tension and sag with sags as abscissae and tensions as ordinates, the intersection of the two curves shows the sag and tension at equilibrium. From the curve: Sag = 0.89 ft. = 10.7 in. Tension = 1,820 Ib. Which represents the conditions in the wire at o F. without wind or sleet. To find the sags at other temperatures: 4 1 2 ec s d n d* 54 54 cY 2 54X0.0000096X40000 = 0.0482 2 ec s APPENDIX B 489 Assume variations of i in. in the sag, ^0=10.7 in. ^1 = 11.7 in. a = 0.0482 (22.4) + 202oX = 1.08+16.20 = 17.30 I7-30 ^0=11 7 in. = 14.8 32.1 12.7 12.9 45-o 13-7 11.4 5 6 -4 14-7 10.2 66.6 15-7 9-3 75-9 16.7 8-5 84.4 17.7 7-9 92-3 18.7 7-3 99.6 19.7 6.Q 106.5 20.7 6.6 113.1 21.7 6.2 H9-3 22.7 6.0 125.3 Plotting with sags as abscissae and temperatures as ordinates, the sag at any temperature can be read from the curve; for example, the sag at 60 should be 15 in. SAG OF WIRES Tables showing the sag at ordinary ranges of temperature for various sizes of stranded copper and aluminum wire to give a factor of safety of 2 at o F. under conditions of 70 miles per hour wind velocity and 1/2 in. thickness of sleet on the wires. NO. o STRANDED COPPER Spans in feet IOO Temp. or less 125 150 200 250 300 400 500 600 ' Sags in. in. in. in. in. in. ft. ft. ft. 2 4 6 ii 23 40 8 i<5| 29 30 2 4 7 13 27 46 9 17 29^ 60 3 5 9 15 32 52 10 IT* 30^ 90 3 7 ii 19 37 59 I0| ii 3i 120 4 9 14 23 42 67 i J 9 31* 490 AMERICAN TELEGRAPH PRACTICE NO. 2/0 STRANDED COPPER Spans in feet Temp. IOO or less 125 150 200 250 300 400 500 600 Sags in. in. in. in. in. in. ft. ft. ft. 2 3 5 10 18 30 6 "i IP! 30 2 4 6 ii 22 37 7 Mi 201 60 3 5 7 13 26 44 7^ 13* aii 90 3 6 9 16 31 5i 8i 14 22 120 4 7 ii 19 36 58 9 15 23 NO. 3/0 STRANDED COPPER Spans in feet Temp. IOO or less 125 150 200 250 300 400 500 600 Sags in. in. in. in. in. in. in. ft. ft. 2 3 4 9 15 24 56 9 isi 30 2 4 5 10 18 30 64 gi i6 60 3 5 6 12 22 36 73 iof 17^ 90 3 6 8 14 27 43 82 II ifti 120 4 7 10 17 32 50 9 2 12 19 NO. 4/0 STRANDED COPPER Spans in feet Temp. IOO or less 125 150 200 250 300 400 500 600 Sags in. in. in. in. in. in. in. ft. ft. 2 3 4 9 12 17 39 6 I2| 30 2 4 5 10 15 21 44 7 13! 60 3 5 6 12 18 26 So 8 14* 90 3 6 8 14 22 31 58 8 15* 120 4 7 10 17 26 37 66 9^ x6i APPENDIX B NO. 2/0 STRANDED ALUMINUM 491 Spans in feet Temp. or less 125 150 200 250 300 400 500 Sags in. in. in. ft. ft. ft. ft. ft. 2 13 25 4* 6| n| 23! 38 30 3 15 3i 5 7 12 24 38* 60 6 iQ 37 si 7i Mi *4i 39 90 ii 25 42 6 8 13 25 39^ 120 17 32 47 6i 8| tii 25^ 40 NO. 3/0 STRANDED ALUMINUM Spans in feet Temp. 100 or less 125 ISO 200 250 300 400 500 600 Sags in. in. in in. ft. ft. ft. ft. ft. 2 8 15 33 5 81 19 32 45 30 3 10 22 42 Si 9 i9i 3*i 4Si 60 6 15 28 5i 61 9i 20 33 46 90 ii 21 34 58 7 10 20^ 33i 47 120 17 28 40 64 7i ii 21 34 47i NO. 4/0 STRANDED ALUMINUM Spans in feet Temp. or less 125 150 200 250 300 400 500 600 Sags in. in. in. in. in. ft. ft. ft. ft. 2 5 10 20 40 6 15 26 37i 30 3 7 18 30 So 7 16 26^ 38 60 6 ii 25 40 59 7i i6i 27^ 39 90 ii 18 32 4 8 67 8| ?7i 28 39i 120 17 27 30 56 75 9 18 28 40 APPENDIX C TELEGRAPH ' Morse Continental A . CHARACTERS =1 Morse T Continental B . - ... ... u C ... . V D W E ... "X F . Y .... G- 7 TT & I 1 ... ..... J ... K 2 L 3 , M 4 .... N .... . . 5 . ., 6 P . 7 Q 8 9 . s . . ^ , . . o Short Numerals Generally Us 1 |3 15 ed By Continental Operators \i . IQ r .0 Morse Continental 1 Phillips : Colon i m m * Colon Dash . - * * * * Fraction Line . _ - Dash - Hyphen . . ' Apostrophe . __ ^_ . . . . . Pound Sterling - _ - / Shilling . . o * * * * d. Pence - - __ _ * * $ Dollars jjj Cents ~* * * : " Colon Followed by Quotation [ ] Brackets . ' _ _ Quotation within a Quotation ,, _,._.., . .,_. ... End of Quotation ... . ^ . . . End of Quotation within Quotation p prr p n t, __ Capitalized Letter * * Ttalirs or T T nderline , __ . . . . . * 492 APPENDIX D USEFUL TABLES COIL-WINDINGS, RESISTANCE, AND OPERATING CURRENT OF TELEGRAPH INSTRUMENTS Instrument Resistance, ohms Turns of wire Gage of wire, B. &S. Normal operating current, mil- amperes Single line relay Single line relay 75 150 2,350 per spool 3,600 per spool. . . . 29, single silk 30, single silk 80 40 Single line relay Sounder 250 10 3,900 per spool. . . . i, 080 per spool 32, single silk 24 cotton 25 Sounder Polar relay . . . 150 IOO 3,500 per spool. . . . i, 600 per section 33, single silk 29 single silk 50 20 Polar relay Polar relay Polar relay 2OO 300 ,OO 1,400 per section. . . i, 800 per section. . . 2 500 per section 32, single silk 33, enameled 34 single silk 20 20 20 Neutral relay Neutral relay 60 1,400 per section. . . i 600 per section 30, single silk 33 single silk 60 60 Transmitter Transmitter 2O 150 1,240 per spool 3,600 per spool 26, single silk 30 single silk 200 CQ WIRE GAGES BROWN. & SHARPE'S GAGE The B. & S. Gage is standard for copper wire and is understood to apply to all cases where size of copper wire is mentioned in any wire gage number. By referring to table it will be seen that in the B. & S. Gage, to all practical purposes, the area in circular mils is doubled for every third size heavier, by gage number, and halved for every third size lighter, by gage number. Every tenth size heavier by gage number has ten times the area in circular mills. Every 10 B. & S. Gage wire has an area of approximately 10,000 circular mils, and from this base the other sizes can be figured, if a table should not be at hand. 493 494 AMERICAN TELEGRAPH PRACTICE WIRE GAGES Iron wire Mile ohm at 60 Fahr. is 4500 Ibs. 100% pure. H. D. Copper wire " " " " " " 859 " " " 1 S IRON g fi ? IRON C 13 H. D. COPPER 97-95% Conductivity d) CO r M w .2 rt i ii Weight, Resist- 5 | SS Weight, Resist- S rt 8 Weight, Resist- g O a Lbs. ance, 60 < Q Lbs. ance, 60 S O Q Lbs. ance, 60 PQ Per Mile F. Ohms Per Mile F. Ohms Per Mile F. Ohms 2*8 825 3 258 932 4.99 3 229 729 6.38 3 229 838 .04 4 238 787 5.97 4 204 578 8.05 4 204 665 .32 5 220 673 6.98 5 182 460 IO.II 5 182 529 .65 6 203 573 8.20 6 162 364 12.79 6 162 419 .09 7 180 450 10.44 7 144 288 16.16 7 i-U 331 .65 8 165 378 12.43 8 128 228 20.41 8 128 262 3-35 9 148 305 15.41 9 114 181 25.71 9 114 208 4.22 10 134 250 18.80 10 102 145 32 . 10 10 102 166 5.28 ii I2O 200 23.50 il 91 "5 40.47 ii 91 132 6.65 12 109 165 28.48 12 81 91 51.15 12 81 105 8.36 13 95 125 37.6o 13 72 72 64.65 13 72 83 10.55 14 83 95 49 47 14 64 14 64 65 13 29 15 72 72 65.27 IJ 57 1 5 57 5 2 16 78 16 16 18 18 26 42 58 20 32 16 53.63 CLASSIFICATION OF GAGES In addition to the confusion caused by a multiplicity of wire gages, several of them are known by various names. For example: Brown & Sharpe (B. & S.) = American Wire Gage (A. W. G.). New British Standard (N. B. S.) = British Imperial, English Legal Stand- ard and Standard Wire Gage and is variously abbreviated by S. W. G. and I. W. G. Birmingham Gage (B. W. G.) = Stubs', Old English Standard and Iron Wire Gage. Roebling = Washburn Moen, American Steel and Wire Co.'s Iron Wire Gage. London = Old English (not Old English Standard). As a further complication: Birmingham or Stubs' Iron Wire Gage is not the same as Stubs' Steel Wire Gage. GENERAL USES OF VARIOUS GAGES B. & S. G. All forms of round wires used for electrical conductors. Sheet copper, brass and German silver. APPENDIX D 495 U. S. S. G. Sheet iron and steel. Legalized by act of Congress, March 3, 1893. B. W. G. Galvanized iron wire. Norway iron wire. American Screw Co.'s Wire Gage. Numbered sizes of machine and wood screws, particularly up to No. 14 (0.2421 in.). Stubs' Steel Wire Gage. Drill rod. Roebling & Trenton. Iron and steel wire. Telephone and telegraph wire. N. B. S. Hard drawn copper. Telephone and telegraph wire. London Gage. Brass wire. EQUIVALENTS OF WIRES B. & S. GAGE = 16-9 = 32-12 = 64-15 = 16-10 = 32-13 = 64-16 = 16-11 = 32-14 = 64-17 16-12 = 32-15 16-13 = 32-16 = 16-14 = 32-17 16-15 = 32-18 = 16-16 ' 16-17 16-18 0000 = 2-0 = 4-3 = 8-6 = ooo = 2-1 = 4-4 = 8-7 = oo = 2-2 = 4-5 = 8-8 = o = 2-3 = 4-6 = 8-9 = I = 2-4 = 4-7 = 8-10 = 2 = 2-5 = 4-8 = 8-1 1 = 3 = 2-6 = 4-9 = 8-12 = 4 = 2-7 = 4-10 = 8-13 = 5 = 2-8 = 4-1 1 = 8-14 = 6 = 2-9 = 4-12 = 8-15 = 7 = 2-10 = 4-13 = 8-1 6 8 = 2-1 1 = 4-14 = 8-17 9 = 2-12 = 4-15 = 8-18 10 = 2-13 = 416 i 1 = 2-14 = 4-17 .... 12 = 2-15 = 4-18 13 = 2-16 = 4-19 14 = 2-17 .... 15 = 2-18 16 = 2-19 . . . .' .... 496 AMERICAN TELEGRAPH PRACTICE CURRENT REQUIRED TO FUSE WIRES OF COPPER, GERMAN SILVER AND IRON B. &S. gage Copper, amperes German silver, amperes Iron, amperes B.&S. gage Copper, amperes German silver, amperes Iron, amperes 10 333- 169. IOI . 26 20. 6 10. 6 6.22 ii 284. 146. 86. 27 17.7 9.1 5.36 12 235- 120.7 71.2 28 14-7 7-5 4-45 13 200. 102.6 63- 29 12.5 6.41 3-79 14 166. 85.2 50.2 30 10.25 5-26 3-n IS 139- 71.2 42.1 3i 8-75 4-49 2.65 16 117. 60.0 35-5 32 7.26 3-73 2. 2 17 99. 50-4 32.6 33 6. 19 3-i8 1.88 18 82.8 42.5 25-1 34 5-12 2.64 i-55 iQ 66.7 34.2 20.2 35 4-37 2.24 i-33 20 58.3 29.9 17.7 36 3-62 1.86 1.09 21 49-3 25-3 14.9 37 3-o8 1-58 93 22 41.2 21. I I2.S 38 2-55 i-3i 77 23 34-5 17.7 IO-9 39 2. 2O 1-13 .67 24 28.9 I 4 .8 8.76 40 1.86 95 .56 25 24.6 12.6 7.46 APPENDIX D 497 THERMOMETER SCALES Centigrade Fahrenheit Centigrade Fahrenheit Centigrade Fahrenheit 100 212.0 1 66 150.8 32 89.6 99 2IO. 2 65 149.0 3i 87.8 98 208.4 64 147.2 30 86.0 97 206.6 63 145-4 29 84.2 96 204.8 62 143-6 28 82.4 95 203.0 61 141.8 27 80.6 94 201 . 2 60 140.0 26 78.8 93 199.4 59 138.2 25 77.0 92 197.6 58 136.4 24 75-2 9i 195.8 57 134-6 23 73-4 90 194.0 56 132.8 22 71.6 . 89 192. 2 55 131.0 21 69.8 88 190.4 54 129. 2 2O 68.0 87 188.6 53 127.4 19 66.2 86 186.8 52 125.6 18 64.4 85 185.0 5i 123.8 17 62.6 84 183.2 50 , 122. O l6 60.8 83 181.4 49 120.2 15 59-o 82 179.6 48 118.4 14 57-2 81 177-8 47 II6.6 13 55-4 80 176. o . 46 II4.8 12 53-6 79 174.2 45 II3.0 ii 5i-8 78 172.4 44 III . 2 10 50.0 77 170.6 43 109.4 9 48.2 76 168.8 42 107.6 8 46.4 75 167.0 4i 105.8 7 44-6 74 165. 2 40 I04.O 6 42.8 73 163.4 39 IO2. 2 5 41.0 72 161.6 38 ICO.4 4 39-2 7i 159.8 37 98.6 3 37-4 70 158.0 36 96.8 2 35-6 69 156.2 35 95-o I 33-8 68 154-4 34 93-2 32.0 67 152.6 33 91.4 Seventy-five degrees Fahrenheit, or 23.8 C. is the standard temperature for measuring electrical resistances in submarine cable tests. Sixty degrees Fahrenheit, or 15.5 C., is the standard temperature for measuring the electrical resistance of wire for general telegraphic purposes. 32 INDEX Absolute units, 5 Action of gravity cell, 13 of condenser as static compensator, 252 "Added" resistance of Field quadruplex, 298 Aerial cable twisted pair rubber insulated, 453 open lines, speed of signaling over, 209 Alphabets, 492 elements of, 207 Alternating-current generator, phantoplex, 396 motors, 32 source of power, 31, 176 Amalgamation of zinc, 14 Ammeters, 159 Ampere, 9 Ampere- turns, 26, 160 Annunciator board connections, 358 branch office, 356 differential, 357 Needham, 360 Anode, n Armature, 29 closed-coil, 26, 29 drum-wound, 29 dynamo tor, 37 lap-wound, 30 open-coil, 29 ring-wound, 29 suspension of relay, 215 wave-wound, 30 Arresters, lightning, 119 Artificial line, 252 rheostat, 261, 327 Atkinson repeater, 227 Auto-starter, for a.-c. motor, 43 Automatic starter, motor, 42 telegraphy, 402 Postal system, 415 transmitter, 406 Auxiliary power switchboard, 68 B-side call bells, 366 "kick," 309 Balance, capacity or static, 333 Balancing duplex, 331 quadruplex, 331 rules Postal Tel. Co., 336 W. U. Tel. Co., 337 Ballistic galvanometer, 155 Barclay direct-point repeater, 386 Battery, arrangement, 3-wire system, 62 at one end of line only, 108 at both ends of line, 108 circuit arrangements, 71 closed circuit types,, n double fluid cells, 1 1 duplex, 266 intermediate, 43, 347 internal resistance of, 166 open circuit types, n primary, 10 required to operate single Morse lines, 112 single fluid cells, 1 1 switching systems, 58 testing, 159 Baume scale, 15 Berry pole-changer, 286 Blavier test, 177 Bonding cable sheaths, 469 Boosters, 64 Branch office annunciators, 356 combination single and duplex set, 390 connected with main office duplex over one wire, 390 control of direct-point repeater, 387 of quadruplex repeater, 388 definition of, 130 instrument arrangement, 352 wiring, 352 Bridge balance, 332 duplex, 267 Wheatstone, 160 Bridging telephone set, 433 Bridle wire, 466 B.P.O., quadruplex, 329 Brown & Sharpe's wire gage, 493 Brushes, dynamo, 30 Bug-trap neutral side of quadruplex, 309 Bunnell key, 117 499 500 INDEX Bus-bars, 68, 265 Cable, aerial, rubber insulated, specifica- tions for, 462 aerial and underground, 455 office, specifications for, 463 sheath bonding, 469 testing, 197 Cadmium cell, 21 Call box, Gill selector, 376 Calorie, 7 Capacity balance, 333 condenser, 164 electrostatic, 89, 202 inductive, 91 unit of, 8 Carhart-Clark cell, 21 Cartridge fuse, 41 Cathode, n Catlin self-adjusting repeater, 244 Cell, n bichromate, 18 Carhart-Clark, 21 Clark, 21 dry, 20 Edison-Lalande, 12, 19 Fuller, 12, 1 8 gravity, 12, 15 Lalande, 12, 19 Leclanche, 12, 17 standard, 21 Weston, 22 C. G. S., system of units, 7 Charge on conductors, 89, 202, 334 Chemical electricity, i symbols, 12 Choke coil, 121 Circuit calculations, 72 divided, 70 efficiency, 212 grounded, 70 joint, 85 metallic, 70 shunt, 83, 86 Closed-circuit cells, n Closed-coil armatures, 26 Codes, telegraph, 492 Co-efficient of temperature, 8c Coil windings of telegraph instruments, 493 Collector rings, dynamo, 25 Combination repeater, 380 sets for single or duplex working, 390 Common battery feeding several lines, 113 Commutator, 30 Composite telephone and telegraph circuit, 442 Compound dynamo, 27 Concentrated circuit annunciators, 359 Condenser, 162 bug- trap, 314 discharge, timing, 335 in connection with artificial line, 253 method of measuring inductance, 103 reading, 270 signaling, 269 testing, 342 Conductance, 10 leakage, 203 Conductivity, 10 Mattheissen's standard of, 79 measurements, 190 specific, 77 Conductors, 2, 3, 76, 81 Constant of galvanometer, 156 Continental alphabet, 492 Continuity preserving transmitter, 250 tests of line wires, 196 Conversion factors, 77 Copper wire, conductivity of, 494 diameter mils, 494 resistance of, 494 specifications, 449 weight per mile, 494 Core, 100 Coulomb, 9 Counter e.m.f., inductive, 103 Cross-bar switchboards, 134 Cross-connecting frame, 143 Cross-fire, 209 Cross, location of, 173 Current, 4, 9 of charge on line wires, 334 proportions in quadruplex circuits, '299 ratio in quadruplex circuits, 299 rectifiers, 54 regulation, dynamo, 45 required to fuse wire of various gages, 496 unidirectional, 25 values in telegraph relays, etc., 436 d'Arsonval galvanometer, 154 Davis-Eaves quadruplex, 304 Decrement quadruplex, 314 INDEX 501 Depolarizer, battery, 14 Derived mechanical units, 5 Diehl bug- trap, 313 Difference of balance pole-changer key closed and open, 339 of potential, 4 Differential annunciator, 357 bug- trap, 314 galvanometer, 155 neutral relay, 289 relay, 251 winding of polar relay, 259 of repeater relay, 224 d'Humy tape reperforator, 416 self-adjusting repeater, 243 Diplex, 290 Direct-point repeater, 383 branch office control, of, 387 Distributing frames, 151 Distribution of current in divided circuits, 114 Disturbances induced in telegraph lines from a.-c. lines, 424 Divided circuits, 70, 114 Double-fluid cell, 1 1 Dry-cell, 20 used to operate open-circuit telegraph system, no Ducts in operating-room floor, 148 Duplex, 249 balancing, 331 battery/ 266 bridge, 267 city line, 276 double current, 254 high efficiency, 273 high potential leak, 272 polar, 255 Postal system, 275 repeater, 383 short line, 278 single current, 249 Stearns, 249 Western Union, 271 Dynamo, 23 commutator, 25 current regulation, 45 feeding several lines, 115 field-magnet winding, 24, 25 magnetic circuit, 27 quadruplex, 295 Dynamo tor, 36 Dynamotor switchboard wiring, 59 Earth connections, 128 currents, 167 potentials, 167 Edison-Lalande cell, 12 Nickel Iron storage cell, 53 Effects of temperature upon resistance of wires, 80 Electric charge on line conductors, 334 Electrical measuring instruments, 153 Electrolysis, 467 Electrolyte, storage battery, 50 specific gravity of, 52 Electrolytic rectifiers, 55 Electromagnets, 26, 96 Electromagnetic induction, 92 units, 5 Electromagnetism, 96 Electromotive force, 4, 9 Electron theory, VII Electrophorus, i Electropoin fluid, 19 Electrostatic capacity, 202 of conductors, 89, 91 flux, 202 induction, 92 from passing clouds, 120 units, 5 Energy, electric, 10 kinetic, 3 potential, 3 Erg, 99 Escapes, location of, 177 Exploring coils, 198 Extra current of self-induction, 279 Fall of potential, 87 Fault-finders, 197 Fault location in cable conductors, 176 in quadruplex apparatus, 341 Field excitation of dynamos, 26 key quadruplex, 295 rotating magnetic, 32 Figure of merit of relays, 215 Fisher loop test, 177 Flux, electrostatic, 202 magnetic, 8, 100 Frictional electricity, i Force, electromotive, 4 magnetomotive, 8 Freir relay, 315 502 INDEX Fuller cell, 12 Fundamental units, 5 Fuse, enclosed, 40 wire, 41 Fuses, 126 in motor circuits, 40, 41 Fusing current, wire of various sizes, 496 Galvanized iron wire, specifications, 450 Galvanometer, 153 d'Arsonval, 155 Ballistic, 155 differential, 155 in quadruplex circuit, 329 shunts, 156 used as low-reading voltmeter, 469 Gerritt Smith quadruplex arrangement, 312 Ghegan repeater, 229 Gilbert, 99 Gill selector, 369 call box, 370 Gravity battery calculations, 75 quadruplex, 292 cell, 12 Ground coil in quad, circuit, 302 contacts on lines, location of, 172 wires, 128 Grounded circuit, 70 line telephone circuit, 432 telephone circuit connected with metal- lic circuit, 434 Half-deflection method of measuring resist- ance, 157 Half- repeater, 375 Milliken, 381 Hard drawn copper wire, 449 Heat, effect of upon resistance of wires, 80 Helmholtz's law, 94 Henry, 10 High potential leak duplex, 272 tension line crossings above telegraph lines, 473 Holding coils of neutral relays, 317 Horse-power, 7, 10 Horton repeater, 238 Hot-wire meters, 158 House circuit repeater, 382 Howler telephone signal, 443 Hydraulic analogy of electrical action, 3 Hydrometer, 15, 1 6 Hysteresis, 100 Impedance, 95 of receiving instruments, 212 of retardation coil, 435 coil simplex, 437 coil, W. U. quad., 321 Increment key, B.P.O., quad., 329 Inductance, 10 a factor of "time-constant," 102 in electric circuit, 205 measurement of, 103 of polar relays, 214 Induction motor, 3 2 Inductive capacity, 8 of conductors, 91 disturbances from a.-c. lines, 424 reactance, 435 Inductorium, 317 Insulation resistance of condensers, 165 of line wires, 184 Insulators, 3, 431 transposition, 431 Iron wire, diameter mils, 494 mechanical and electrical requirements, 4Si resistance of, 494 specifications, 450 weight per mile, 494 Intermediate battery, 42 offices on single lines, 1 08 Morse loop connected into duplexed line, 392 test office, 130 Internal resistance of battery, 73 of quadruplex apparatus, 298 J-hooks, 431 Johnson coil, 282 Joint resistance of circuits, 82 Jones quadruplex, 295 Joule, 7 Kelvin's method of measuring resistance of galvanometer, 157 Key, Bunnell, 117 Keyboard perforator, 406 Kilovolt, 9 Kinetic energy, 3 Kleinschmidt perforator, 407 KR law, 205 Lag of magnetization behind current, 101 Law of shunts, 84 INDEX 503 Lead covered aerial and underground satu- rated paper cable, 455 Leak resistance of Field quad., 298 Leakage conductance, 203 Leclanche cell, 12 Leg-board connections, direct-point re- peater, 384 duplex repeater, branch office control, . 389 Leg-boards, 344 Legs to branch offices, 353 Life of gravity battery, 1 7 Line capacity too high to be balanced with total capacity of condensers, 340 resistance box, W. U., quad., 326 Lines of force, 97 Lightning arresters, 121 location of. 1 24 disturbances, 119 Local circuits, single lines, 108 multiplex sets, 265 connections, W. U. quad., 351 Lodestone, i Long-end current, 295 Loop-boards, 344 Loops to branch offices, 352 Loops witch, Western Union, 350 Loop tests, 173 Loss in transmission efficiency, cables, 206 Magnet, electro, 104 permanent, 2 Magnetic field, 205 flux, 8, 100 induction, 8 leakage, 104 moment, 8 reluctance, 8 saturation, 98 units, 5 Magnetomotive force, 27, 28, 99 Main-line call bells, 366 switchboards, 130 Make spark, 284 Mallet perforator, 403 Mattheissen's standard of conductivity, 79 Measuring distant quad, battery, 342 Mechanical units, 5 Megohm, 9 Metallic circuit, 70 composite, 443 quadruplex, 308 Microfarad, 10 Mile-ohm, 77 Milliammeter in quad, circuit, 327 Milliken repeater, 236 half-repeater, 381 Mirror galvanometer, 154 Morse alphabet, 492 Motors, alternating current, 32 compound, 31 direct current, 30 series, 31 shunt, 31 three-phase, 44 two-phase, 44 Motor current regulation, 37 -dynamo, 35 -generator, 34 -starters, a. c., 43 d. c., 38, 41 solenoid, 42 Multiple arrangement of cells, 72 connection of relay windings, 217 -series arrangement of cells, 73 Multipliers for voltmeters, 158 Murray loop test, 171 Needham annunciator, 360 Negative pole to line on closed key, 340 Neilson repeater, 232 Neutral relay, 289 with holding coil, 317 with short cores, 317 Western Union, 320 side of quad, signaling systems, 366 O'Donohue shunt repeater, 391 Office cable specifications, 463 wire specifications, 465 Ohm's law, 5 "Ohmic" balance, 331 Open circuit cells, 1 1 system of telegraphy, no Oscillatory discharges, 121 Overcompounding dynamo winding, 28 Overload motor-starter, 41 Paper insulated aerial cable, 455 Parallel arrangement of cells, 74 Perforated tape, 410 Perforator, key-board, 406 mallet, 403 Periods of reversal, 317 504 INDEX Permanent magnet, 2 state, 91 Permeability, 9, 98 Phantom telephone circuit, 441 Phantoplex, 395 quad., 398 transformer, 400 Pig-tail cable connections, 137 Pin- jacks, 138 connections, loop-board, 349 switchboard, 140 Platinum contact points, 280 Plugs, double conductor, 135 Polar duplex, 255 operation, 262 repeater, 385 phantoplex-quad., 398 relay, 257 inductance of, 214 windings, 214 Polarity of magnet, 104 of solenoid, 97 Polarization of cells, 14 Pole-changer, 255 of Wheatstone automatic, 409 Porous cup, 17, 1 8 Postal automatic, 415 receiver and transmitter circuits, 421 tape take-up gear, 420 direct-point repeater, 384 dynamo arrangement, 59 gravity battery quadruplex, 294 loopswitch, 348 quadruplex, local circuits, 344 repeater instrument rack, 394 rules for balancing, 337 spark control, 283 Potential difference, 4 energy, 3 fall of, 87 leads, 60, 67 Pothead wire, 466 Power, unit of, 7 Power-board, auxiliary, 68 telegraph, 58, 65 Primary battery, n Printing telegraphs, literature, 471 systems, 422 Proportion of currents in quad, circuit, 299 Quadruplex, 287 Quadruplex battery measurements, 342 3-wire system, 62 B. P.O., 329 decrement, 314 Davis-Eaves, 304 fault location, 341 Field key system, 295 Gerritt Smith arrangement, 312 Jones system, 295 local circuits, Postal system, 344 Western Union, 328 with no-volt battery, 347 management, 339 metallic circuit, 308 Postal system, 3.04 repeater, 388 signaling bells, 361 single dynamo system, 306 theory, gravity battery system, 288 Western Union, 318 operation of, 323 Quantity of electricity, 9 Ratio of currents in Field quad., 299 Reactance of retardation coil, 435 Reading condenser, 270 sounder, quadruplex, 310 Recorder, Wheatstone, 403 Rectifiers, electrolytic, 55 mercury-arc, 54 Relay armature suspension, 215 characteristics, 214 differential, 251 neutral, 289 winding, 259 Freir neutral, 315 Morse, 109 Neutral, W. U., 320 phantoplex, 396 polar, 257 repeater, 225 single line, connections of, 117 windings, 214 Releasing current, 203 Reluctance, 8 of magnetic circuit, 100 Reluctivity, 9 Remanence, 100 Repeater adjustments and management, 246 Barclay direct-point, 386 combination half-set and single line, 380 direct-point, 383 INDEX 505 Repeater, half-set, 375 half-Weiny, 377 house circuit, 382 instrument arrangement on rack, Postal, 394 table, W. U., 393 MiUiken half-set, 381 * O'Donohue shunt, 391 quadruplex, 383 station, definition of, 130 self-adjusting, 243 single-line, 219 three-wire, 240 Repeating coil method of tying telephone lines together, 434 telephone, 447 Repeating sounder bug- trap, 311 Reperforator, bearing adjustment, 419 tape, 415 Reserve power, 32 Residual magnetism, 100 Resistance, 4, 9 added, of Field quad., 298 affected by heat, 80 joint, 82 leak, of Field quad., 298 measurements of line wires, 175 of earth contacts, 169 of lines, 76 specific, 77 Resistivity, 10 Retardation, 202 coil, impedance of, 435 method of tying telephone lines together, 434 coils, telephone, 448 Reversals of current, 291 Reversing key, B. P. O., quad., 329 Rheostat, artificial line, 261 motor starting, 38 Rotating magnetic field, 32 Rotor, 32 Saturated paper cable, 455 Screening telegraph relays from inductive disturbances, 427 Second side of quad, signaling systems, 366 Selector connected into duplexed line, 371 into single line, 372 signaling, 368 Self-induction, in line wires, 205 Self-induction of relay balanced with shunted condenser, 217 Semi-automatic transmitters, 208 Series-multiple arrangement of cells, 73 Series telephone set, 433 Series-wound generator, 26 Several lines worked from a single battery, in Short-end current, 295 Shunt circuit, 83 Shunted condenser, 217 Shunt-field winding of dynamo, 28 Shunt, galvanometer, 156 repeater, 391 Shunts, law of, 84 Signaling condenser, 269 systems for quad, circuits, 361 Simplex telephone and telegraph circuit, 437 Simultaneous telephony and telegraphy, 434 Single dynamo quad., 306 Single-field dynamotor, 36 Single fluid cell, 1 1 Single line repeater, 219 Atkinson, 227 Ghegan, 229 Horton, 238 Milliken, 236 Neilson, 232 Toye, 231 Weiny-Phillips, 222 Single Morse circuit, 106 phase series motor, 32 Skirrow switchboard, 139 Solenoid, 96 Sounder circuit B-side of quad., 310 single line, 109 Spark at contact points, 279 Specific conductivity, 77 gravity of electrolyte, 15 of storage battery electrolyte, 52 inductive capacity, 8 resistance of conductor, 77 Specifications for iron and copper wire, 449 Speed of signaling, 201 related to receiving end impedance, 213 over aerial lines, 209 through cables, 204 Split-plugs, 135 Spring jacks, 136 jack connections of loop board, 349 Squirrel cage motor connections, 45 Standard cell, 21 506 INDEX Starting rheostat, 38 Static balance, 333 Stator, 32 Stearns duplex, 249 Storage battery discharging circuit, 48 Edison, 53 for telegraph service, 47 initial charge, 50 low cell indication and treatment, 51 multiple type, 47 obtaining additional life, 53 supplying current for several lines, 115 Stranded galvanized steel wite, 452 Strap and disk switchboards, 131 Strength of received signals, 211 Submarine cable specifications, 456 Sulphate of copper solution, 14 Superimposed phantoplex circuit, 398 Susceptibility, 9 magnetic, 96 Switch-blocks, 140 Switchboards, cross-bar, 134 main line, 130 pin-jack, 140 strap and disk, 131 Symbols, 5 Synchronous motors, 32 Table of ratings, storage battery, 50, 53 Tape, perforated, 405, 410 take-up gear, Postal automatic, 420 Telafault test set, 197 Telephony, 432 bridging telephone set, 433 composite circuit. 442 condenser capacities, 445 terminal connections, 445 with intermediate telegraph station, 444 with no intermediate telephones, 444 grounded line composite, 442 connected with metallic circuit, 434 telephone circuit, 432 howler signal, 443 intermediate station on composite line, 444 telegraph station connected into simplex, 439 station on simplex circuit, 438 telephone station on phantom circuit, 440 and telegraph on simplex, 439 Telephony metallic circuit composite, 443 metallic telephone line, 433 phantom circuit transpositions, 440 simplex circuit, 441 telephone circuit, 439 repeating coils, 447 coil method of tying lines together, 434 retardation coils, 448 series telephone set, 433 simplex, bridged impedance coil type, 437 transposition of telephone lines, 428 Temperature co-efficient, 80 Terminal office, definition of, 130 switchboard, 131 resistance of quad., 301 Tests with telephone receiver, 195 Theory of electricity, VII Thermal electricity, i Thermometer scales, 496 Three-phase motor, 34, 44 Three wire system, 62 Time-constant, 101 of relays, 216 Timing condenser discharge, 335 Toye repeater, 231 Transfer jacks, 149 Transformer, phantoplex, 400 Transmission efficiency equivalents, 437 losses, 206 Transposition insulators, 431 of wires, 428 Transmitter, automatic, 406 duplex, 250 quadruplex, 293 repeater, 224 semi-automatic, 208 Twisted pair paper cable, aerial, 453 Two-phase motor, 44 Underground cable, paper insulated, speci- fications, 455 sheaths, electrolysis of, 467 Underload rheostats, 41 Unidirectional currents, 25 Unit strength of pole, 8 Units, 5 Vacuum gap arrester, 123 Variable state, 91 INDEX 507 Varley loop test, 173 Voltaic cell, 1 2 Voltmeter, 157 tests, 182 Walking-beam pole-changer, 256 Way-office, definition of, 130 Wedges for use with spring jacks, 136 Weiny-Phillips half-repeater, 377 repeater, 222 Western Union 5-U retardation coil, 448 bridge duplex, 271 direct-point repeater, 386 distributing frame, 151 dynamo arrangement, 61 instrument arrangement on repeater tables, 393 loopswitch, 350 main-line switchboard, 150 proportional test set, 192 quadruplex, 318 local connections, 328 rules for balancing, 337 signaling system for multiplex sets, 363 spark control, 283 Weston cell, 21 Wheatstone automatic, 402 Wheats tone bridge, 160 measurements, 170 system, mallet perforator, 403 recorder, 403 transmitter, 406 Windings of telegraph instruments, 493 Wire, annunciator, 466 bridle, 466 call circuit, 466 gages, 493 gages, classification, 495 hard drawn copper, specifications, 449 iron, specifications, 450 office, specifications, 465 outside, twisted pair, 466 pothead, 466 Wires, telegraph, crossing under high-ten- sion lines, 473 Wireless trouble finders, 200 Wiring of telegraph power switchboards, 65 Work, unit of, 7, 99 Zinc, 12 amalgamation of, 14 202 Main Librar All BOOKS MAY BE RECALLED AFTER 7 DAYS Renewol, and Recharge, may b. made 4 day, prior o the due do,, BOOK* mnu I*A D_ I L Books may be Renewed by calling AS STAMPED BELOW UNIVERSITY OF CALIFORNIA BERKELEY FORM NO. DD6 BERKELEY, CA 94720 U.C.BERKELEY LIBRARIES VQ 04192 263701 UNIVERSITY OF CALIFORNIA UBRARY