LIBRARY OF THE UNIVERSITY OF CALIFORNIA. Deceived Accessions Nov} Class No. POLYPHASED ALTERNATING CURRENTS. (ILLUSTRATED}. BY E. HOSPITALIER. LONDON: H. ALABASTER, GATEHOUSE & Co., 22, PATERXOSTER Row, E.C. 3X VAN NOSTRAND COMPANY, NEW YORK. ^ x xvA . CONTENTS. PAGE INTRODUCTION . i I. TRIPHASED CURRENTS % . 6 i. GENERATION . . 10 ii. TRANSFORMATION . . - t .'..-. . . . 19 in. APPLICATIONS. . . . . . . -27 Electric Lighting . . . . 27 Motors -. . . . 31 iv. MULTIPLE TRANSFORMERS 53 II DIPHASED CURRENTS . . 59 i. GENERATORS . . --'... -59 ii. TRANSFORMERS . . . . . . .61 HI. MOTORS 65 iv. MULTIPLE TRANSFORMERS 67 APPENDIX A . . .75 APPENDIX B 77 POLYPHASED ALTERNATING CURRENTS. INTRODUCTION. THE question of polyphased alternate currents has become the topic of the day, and we may say that the application of these currents to the transport and distribution of electrical energy constituted the really interesting, original and novel feature of the Frankfort Electrical Exhibition. But whilst general attention was attracted by the transmission experiments between Lauffen and Frankfort, there is not sufficient import- ance attached to the question in its general bearings, i.e., to the employment of polyphased alternate currents as a general method of production, transformation, transport and distribution of electrical energy and its application to lighting, motive power, storage of energy, and electro-chemical operations. The study of poly- phased alternate currents includes all these questions, however, and it is with the object of treating them all methodically that we write this little work under the general heading of Polyphased Alternating Currents, the Lauffen -Frankfort experiments constituting merely one chapter (although certainly the most important one) of this special study. But, before entering upon our subject, we will show briefly how the increasing progress of electrical industry has led engineers to create a system which four years ago did not exist. From 1800 to 1830, the date of the discovery of induction by Faraday, the voltaic pile sufficed for electrical researches of a purely scientific character. From 1830 to 1850, during which period we can trace the gradual growth of electrical industry, then limited to telegraphy and electro-chemical operations, the pile was still sufficient for the applications for which it was required. From 1850 to 1870, we find attempts being made to substitute for the battery a more powerful and more economical generator of electrical energy, a continuous current generator. This problem was solved by Gramme in 1870. From 1870 to 1880 may be called the decade of arc electric lighting, and of the attempts made to obtain small sub-divided lights ; the incandescence lamp was the solution of this problem. The development and improvement of incandescence lighting occupied from 1880 to 1890, simultaneously with that of the distri- bution of electrical energy and its transport to great distances. In 1885 began the rivalry between continuous and alternating currents. Whilst continuous currents lend themselves readily to the transmission of power to a distance, they are ill adapted to transformation and distribution at a distance. Simple alternating currents, on the other hand, are well adapted for transport to a distance and transformation, but their application seems limited to thermic effects, such as electric lighting, electric welding, and chemical operations in which the currents act merely as a producer of heat. We are still in want of a perfect simple alternating current motor and a simple and practical method of transforming alternating currents into continuous currents, which would bring them into general use. While certain inventors were trying to solve these two problems, others, turning to account the magnetic properties of the revolving fields produced by two or more alternating currents with a suitable lag, saw in the methodical and systematic employment of these polyphased currents a means of realising a complete system of generation, transformation, distribution and general application of electrical energy. It is this system that we propose to study methodi- cally, examining in succession the generators, trans- formers, motors, and certain apparatus not named up to the present, made in view of transforming polyphased alternate currents into continuous ones. Researches are being made in different quarters, with the object of constructing a simple alternating current motor starting with a load and possessing the same properties as continuous current motors. Other inventors are en- deavouring to construct apparatus capable of practically transforming, under good conditions of efficiency , simple alternating currents into a continuous current. These problems are as yet only imperfectly solved, but if better solutions were arrived at, polyphased alternate currents would give place to simple alternate currents. Whilst waiting for this, we will consider the case as it stands at present, and devote this pamphlet to the general study of these currents, avoiding as far as possible, questions of precedence, which are continually cropping up in this particular case, and which, as matters now stand, are very difficult to decide. Before entering upon the description of polyphased FIG. 1. INTENSITY IN DIPHASED CURRENTS, A B, \VITH COMMON RETURN WIRE, c. alternate current generators, the first question that suggests itself relates to the choice of the number of currents to be used, when once we have admitted the necessity of having recourse to these multiple currents on account of their special properties. In order not to increase indefinitely the number of wires, it is evident that we should limit ourselves practically to the smallest number possible, and that thus our decision lies between 2 and 3. With two currents having a lag of a quarter of a period, we require four wires, or, at any rate, three, if one serves as a common return wire for the two currents. We have only to look at fig. i to see that the effective intensity passing through this common wire, c, is greater than that of the current passing through the other wires, A and B ; we must, therefore, give it a greater section FIG. 2. PRINCIPLE OF TRIPHASED ALTERNATE CURRENT GENERATOR which destroys symmetry. With three currents lagging by a third of a period, however, fig. 3 shows that the effective intensities are the same, and that their algebraic sum is nil at each moment. Thus there is symmetry, and consequently equality of section in the three conductors. We will, therefore, commence our subject with an examination of triphased currents with three conductors of equal section. I.-TRIPHASED CURRENTS. WE will imagine, for the sake of illustration, a Gramme ring (fig. 2), rotating at constant speed in a simple magnetic field, three elementary bobbins, a, b, c, being attached to this ring at 120 from one another. If we turn the armature these bobbins will become the seat of alternating electromotive forces which we will suppose to be strictly sinusoidal, which FIG. 3. DIAGRAM OF ELECTROMOTIVE FORCES AT EVERY INSTANT IN THE THREE BOBBINS OF A TRIPHASED ALTERNATE CURRENT GENERATOR. form, indeed, they assume, wath very little variation, in practice. Fig. 3 shows the value of these elect tomotive forces at each instant. These three bobbins may be connected with three outer wires in two different ways ; in a star or in a triangle (fig. 4). The star connection is termed by M. Dolivo-Dobrowolsky open circuit connection. and the triangle connection closed circuit connection. The expressions "star" and "triangle" seem, however, better suited to the case, and will be exclusively employed here. However this may be, either of these modes of coupling gives, in the outer circuits, three lagging currents, but the " star" mode has a certain advantage over the "triangle," in that it is possible to connect the common centre of the three circuits with a fourth or neutral wire, and to establish apparatus working as Star Coupling. Triangle Coupling. F IG . 4. COUPLING OF TRIPHASED ALTERNATE CURRENT GENERATORS simple resistances, incandescence lamps, for instance, between the neutral wire and each of the three other wires ; the corresponding bobbin of the generator then acts as an ordinary alternating current dynamo without either reacting or disturbing the working of the system, but the triangle mode of connection does not allow of this arrangement. In the foregoing illustration, we have supposed the generator to consist of a Gramme ring and three elementary bobbins, each occupying but a very small 8 fraction of the circumference. Practically, we can very easily realise an alternator with triphas.ed currents with an ordinary Gramme machine, by connecting three points 120 apart in the continuous winding with three insulated collecting rings, on which rest three brushes which are themselves connected with the outer conductors. We thus obtain a generator, the three bobbins of which are connected in a triangle. The inductors may be excited separately, or, which is often more convenient, be fed by a derivation taken on the collector, the machine thus producing at the same time continuous currents and triphased currents. (We will discuss later on the properties of an apparatus thus constituted.) But if we study a little more closely the electromotive forces developed at each moment in each third of the ring, we can see that they are not always utilised under the best possible conditions, and that thus we only obtain the algebraic sum of these electromotive forces, certain spirals tending to produce a current of one direction, and the others a current in an opposite direction. Thus, in practice, this arrangement has been given up, for it has the obvious disadvantage of increasing the internal resistance of the generator, and of reducing its specific power l and its efficiency. The drum winding has been preferred for the construction of large machines producing triphased alternate currents. 1 The specific power of a machine is the quotient of its power by its mass, and is measured in watts per kilogramme, or, in the English system, in watts per pound. Further, the possibility must be afforded of working motors of whatever size in a rational manner. The simple alternating current is at present inferior in this respect to the continuous current, whilst the so-called polyphase alternating current meets this demand in the most perfect manner. 10 I. GENERATION. BY connecting together in a suitable manner the induced bars which arrive at each moment at the same relative position with regard to the inductors, we get a generator in which there are never, in any of the circuits, two opposite electromotive forces. Moreover, poly- phased current dynamos being always machines of great power, and turning at a low angular speed, multi- polar inductors should be employed. In Brown's generator at Lauffen, which may be taken as a type of a carefully studied machine, there are 32 poles and 96 induced bars, so that between two consecutive poles there are only three induced bars corresponding to each of the three circuits. As regards frequency, there does not seem to have been any attempt made to attain, with polyphased currents, the high frequencies attained with ordinary alternating currents (133 periods per second is the Westinghouse system). In fact, the number of periods is comprised between 30 and 40 per second. These low frequencies reduce the specific power of the generator, transformers and motors, but they allow of higher efficiencies being given by the motors, and this compensates, and more than compensates, for the reduction of specific power. The frequency would even be still lower, if it were not for the fear that this would involve periodic variations of luminous intensity which 11 would affect the incandescence lamps and, still more, the arc lamps. All the polyphased alternate current generators are low tension machines. The general opinion at present is, that it is better, when it is required to transport energy to a great distance, to construct machines of low electromotive force, strong, simple, easily con- structed and insulated, and to transform these currents of low tension in a suitable apparatus. We thus gain increased security by confining the high tensions to transformers which are inert and can be easily insulated, and which are rarely touched. Notwithstanding the employment of an intermediate apparatus, it is more economical to construct a low tension machine and a transformer than to construct a dynamo giving directly the high tension required for transport to a great dis- tance. The idea of producing low tensions and trans- forming them at the starting point, has been proposed several times for ordinary alternating currents, with a view to avoiding the dangers and the difficulties of con- struction of high tension machines ; but we do not think it has ever been practically applied before the Lauffen-Frankfort experiments. On account of the enormous intensities of the cur- rents produced by the low tension dynamos, the arma- tures are naturally left fixed, which facilitates the connections, and the inductors are made to revolve by supplying them with the exciting current by means of brushes and collecting rings communicating with the winding of the inductors. 12 Figs. 5 and 6 show a polyphase machine which, at 150 revolutions per minute, can convert 300 H.P. FIG. 5. The arrangement of the armatures is such as to gene- rate three alternating currents, the phases of which 13 are displaced towards each other by 120. The tension of each current amounts only to 50 volts, and the strength of the currents is 1,400 amperes each ; the armature is fixed, and the magnetic field is made to rotate. Corresponding to this great strength of current the winding of the armature has a large L. FIG. 6. section, when, as is here the case, we may dispense with the arrangement of several parallel circuits. The leads, which in this case are massive copper rods of 29 millimetres in diameter, are placed in correspond- ing holes of the iron of the armature, from which they are insulated by means of asbestos tubes. By this 14 arrangement the following very important advantages are secured : Foucault currents which would appear of extreme strength on the use of such massive conductors as commonly arranged are here absolutely non-existent. In fact, on experiments with bars up to 5 centimetres in diameter resting in a' perforated armature, it was quite impossible to detect even the smallest trace of waste of power by Foucault currents. The arrangement of the rods in holes permits also of an exceedingly simple construction of the armature, especially as regards insulation, which further, in con- sequence of the use of asbestos, admits of the produc- tion of a completely non-combustible armature. From a mechanical point of view, the winding is very well protected against injuries from without. Displace- ments and injuries to the wires can never occur in consequence of overloading and short-circuiting. Further, in consequence of the very favourable utili- sation of space, and the reduction of atmospheric resistance, the magnetising current is considerably less than with ordinary armatures. Perforated armatures for continuous currents were successfully constructed at Oerlikon as far back as 1885, but in the construction of the machines just described the advantages of this armature are evinced more strikingly than heretofore. Corresponding to the number of poles (32) of the magnetic field, there are 32 rods arranged for the armature circuit, which are all introduced serially by 15 means of suitable cross-pieces. The armature, in con- sequence of its three circuits, has 3 x 32 = 96 perfora- tions, and, of course, the same number of rods. The iron armature is held together by a cast-iron frame, which stands on projecting feet of the foundation plate, and can be displaced axially (see figure) for the sake of mounting or cleaning. The magnetic field of the machine has the peculiarity that all the poles, both positive and negative, are produced by a single exciting bobbin. This is effected by the following simple arrangement : The magnetising coil is wrapped upon a cast-iron ring, with two flanges resembling a pulley ; two steel rings, each with 16 polar projections, are applied to the ring on both sides, and screwed up accordingly. The position of the polar projections is selected so that they project between each other. Thus when the coil is excited there are formed alternately positive and negative poles, the one ring receiving only positive and the other negative poles. This arrangement admits of an ideal utilisation of the magnetic coils in consequence of a considerable reduction of the weight of copper along with a con- siderable reduction of excitation, as appears from the following data. The construction of this field is of surprising simpli- city and solidity, the entire magnetic field, with 32 poles, consists only of four pieces, which, in view of the fact that the field rotates, must not be undervalued. For conveying the exciting current, two rings fixed 16 on the nave are connected by wires with two discs se- cured to the frame of the machine, whereby these wires effect the conveyance of the current from the extreme clamps to the rings of the nave. The mag- netic field is placed movably upon a shaft of corre- sponding strength which rests in a strong double plummer block, the latter being screwed down to the foundation plate. The free, posterior end of the shaft is directly or indirectly connected with the point where power is given off or received. The total weight of copper in the magnetic field is about 300 kgs., i.e., only a fraction of the weight of copper of other machines of the same size and number of revolutions. When running without load, and at the normal angular velocity (150 revolutions per minute), 100 watts suffice to produce the tension of 50 volts in the armature, consequently only Jjth per cent, of the full duty of the machine, certainly a surprising result. Considering the reaction of the armature when the machine is fully loaded, the energy necessary to excite the machine is naturally greater, but it still does not exceed a fraction of one per cent. Experiments made have shown that at the full number of rotations and excitement to the normal tension the losses in friction at the bearings, resistance, hysteresis, etc., amount to 3,600 watts, i.e., about 1*6 to 1*7 per cent, of the full duty of the machine. The additional loss in the copper of the armature amounts to 3,500 watts, so that the entire commercial efficiency of the machine is about 96 per cent., a result which, to 17 our knowledge, has not been reached by any other machine of similar performance and speed. Corresponding to these slight losses the heating is quite insignificant, and the machine is therefore fully competent to bear without difficulty the overloading 32 bars in each Coil coupled in Series. The ends of each Bobbin, A A', B B', c c', are free, and permit the connections of the three, in star or in triangle. F IG . 7. WINDING OF BROWN'S TRIPHASED ALTERNATOR AT LAUFFEN. which often occurs in industrial operations, even if we do not take the asbestos insulation into account. The total weight of the machine without the bed- plate is only 9,000 kilogrammes. The Brown triphaser fulfils all the conditions that we have mentioned, and constitutes the most perfect c 18 type of generator of triphased alternate currents that we can mention. To complete this description, we append diagrams (fig. 7) showing the winding of the 96 induced bars in three circuits, each containing 32 bars in series. In order to facilitate the reading of the connections, we have substituted for the cylindrical winding a tron- conical winding, and we have placed the three series of connectors arranged at the two ends of the winding on circles of different diameters in order to avoid their superposition. The three windings give six free ends, which can be connected at will in a triangle or in a star, according to the applications intended. Before quitting the question of triphased generators, or triphasers, we wish to call the attention of professors desirous of showing their pupils the properties of these currents to a means of obtaining them without having a special machine constructed and without even adding collecting rings and connections to an ordinary Gramme machine. We have only to take an alternating current Gramme machine of the type known as self -exciting, with multi- polar inductors, and to establish suitable connections between the fixed elementary bobbins constituting the induced winding. The natural lagging of these bobbins resulting from the construction itself sup- plies us at once with polyphased currents capable of being used for the demonstration and repetition of the principal experiments, allied to the polyphased alter- nate currents. 19 II. TRANSFORMATION. As polyphased alternate current generators are con- structed to produce currents of great intensity and low electromotive force, while the application of these currents, in the present state of electrical industry, is directed mainly to the transport of electrical energy to a great distance, it follows that these currents must be transformed twice between the point of production and the point of utilisation, the first time to increase their tension at ths expense of their intensity, and the second time to reduce this tension and render the currents harmless and capable of being manipu- lated. This is the work of the transformers. The polyphased alternate current transformers do not .differ essentially from ordinary alternate current transformers, and consist of closed magnetic circuits enclosed in two windings, one of thick wire and one of fine. In the apparatus supplied by the Allgemeine Elektricitats Gesellschaft and the Ateliers d'Oerlikon we find essentially the same principles, tne apparatus only differing in details of construction. Three cylindrical iron cores (fig. 8), formed of layers of thin soft sheet iron that has been well annealed, rest on an annular disc of sheet iron, and are covered with a second disc of sheet iron of the same thickness, 20 the two discs being formed of one sheet of iron, rolled like a ribbon on a bobbin. The three cylinders are each enclosed in two windings of copper wire. In the transformers placed at the generating station, the thick wires are connected with the dynamo, and the fine wires with the line of transport, as shown in fig. 9. FIG. 8. At the other extremity of the line (the arrival station) the fine wires of the transformer are in com- munication with the line of transport at a high potential, and the thick wires are connected with a distribution board, from which the conductors supply- ing the different utilisation apparatus radiate. We 21 22 can see that at the points of departure and arrival it is possible to connect the inductive and induced circuits either in a triangle or a star, as required. In the Lauffen- Frankfort experiments, for reasons which will become evident when the utilising apparatus is examined, the " star " connection has been preferred. The transformers at Lauffen and those at Frankfort have a co-efficient of transformation equal to 160, that is to say, a difference of potential of 50 volts esta- blished at the terminals of one of the inductive bobbins produces a difference of potential of 8,000 effective volts at the corresponding terminals of the induced circuit. This difference of potential is that which exists between the neutral point and each of the wires. If we trace the curves representing the differences of potential between each of the wires and the neutral point, and call e the difference of potential between the neutral point, o, and the outer extremity of the bobbin, a, we can easily see that the effective difference of potential, E, between any two wires is greater. The relation between E and e is E = e 2 sin 60 = 1*732 e. A difference of potential, e, of 8,000 volts between the neutral point and each of the wires corresponds, therefore, to a difference of potential, E, of 13,850 volts between any two wires. It is in consequence of the high tensions which are now used, and which will be exceeded eventually in subsequent experiments, that special precautions have 23 been taken in the construction both of the transformers and the line. As these precautions have been sum- Fir.. 10. ciently often described elsewhere, there is no need for us to enter further into them here. Fig. 10, which shows the details of construction of 21 the transformers made at the Oerlikon Works, under the direction of Mr. Brown, explains the essential principles of these transformers, which are constructed to transform, when fully charged, the 200 kilowatts which the dynamo is capable of producing. We see that the fine wire circuits are entirely enveloped by the thick wire circuits, and that the three extremities of these fine wires pass out of the transformer in three thick glass tubes, so as to avoid any accidental metallic contact between the cast iron casing containing the petroleum, and the high tension wires. This casing is, moreover, connected to earth, and with the neutral wire, and can be touched with impunity during the transmission. A tap placed at the bottom of the cast iron casing enables the oil to be drawn off from the transformer before it is taken to pieces for the purpose of examination. All the connections are external, so that the couplings can be effected without interfering with the apparatus. We can see by fig. 9 that in consequence of the con- nections established between the neutral points of the dynamo, and of the transformers which are connected in a star, the whole circuit is in metallic commu- nication with the earth. This mode of coupling, which is used exclusively for triphased alternate currents, increases the safety of the installation and reduces the danger at those points where the high tension wires enter the transformers, and along the wires themselves. The fatal accident which hap- pened at Lauffen, a few days before the closing of 25 the Exhibition, can only be attributed to the im- prudence of the victim and to the violation of the regulations forbidding both the public and the Ex- hibition staff to come within reach of the dangerous high tension points. The efficiency of triphased transformers has been very high, 92 per cent, at half load and 96 per cent, at full load. The transformers insulated by means of oil will not necessarily be employed in all transmissions of energy by polyphased currents. Dry transformers are also being constructed at 3,000, 4,000, 5,000, and even 10,000 volts; for instance, those of Mr. Ferranti, at Deptford, for simple alternate currents. The same can be done for polyphased alternate currents. The construction, therefore, can be simplified when the tensions do not exceed these figures, and the insulating liquid dispensed with. LINES. The high tensions of polyphased alternate currents compel us to have recourse to overhead lines specially insulated. The Lauffen-Frankfort experiments will afford us valuable information with regard to the quality of these lines, and we await the results of the experiments of the Commission with some impatience and curiosity, but it seems now to be universally acknowledged that the employment of large oil insu- lators with three cups involves a number of useless precautions. The insulators with a single cup, of the 26 Johnson and Phillips type, seem all that is required, and have not given rise 1 to any inconvenience. The expense of power in the line has been, in the Lauffen- Frankfort experiments : 3' i horse power for an output of 53 horse power. 8*0 ,, 93 25*2 149-8 the output being taken at the terminals of the secondary at the Frankfort transformers. We will not here discuss the transport of energy at a high potential by underground tri-cables ; it is to be feared that their high price would prevent their em- ployment, as they would add so considerably to the expense of the installation as to render all the econo- mical advantages of transport to a great distance of no account. As to canalisation at low tension, it is similar to that of continuous currents or ordinary alternate currents, and with the exception that three conductors are employed instead of two, differs in no way from it, so that we need not discuss it further. 27 III. APPLICATIONS. THE applications of the electric current may be classed into three essentially distinct groups, according to the nature of the transformation effected by the apparatus employed for its utilisation. These are : i. Thermic applications in which the electrical energy is transformed into heat ; this class includes electric lighting ; 2. Mechanical applications, and 3. Electro- chemical applications. But in these last applications simple ordinary alternating currents or polyphased currents cannot render great service in their natural form. They require to be first transformed into con- tinuous currents, and thus we- shall have to describe, in view of those applications of the future, which may require the employment of continuous currents, some apparatus which may effect this transformation very simply. ELECTRIC LIGHTING. In the Lauffen-Frankfort experiments the lamps used were exclusively incandescence lamps. It is probable that the feeding of arc lamps would involve certain difficulties, either through the self-induction of the regulating mechanism of these lamps, which tends to destroy the equality of the lagging of the three currents 28 or through the complex and ill-defined phenomena of which the arc is the seat, or, again, because the low frequency of the currents employed is detrimental to the steady lighting of objects to which a somewhat rapid movement is imparted ; this inconvenience is Incandescence lamps coupled on star-triphaser ; A, u, c, are groups of lamps on high voltage ; D, E, F, are low voltage lamps connected between the N w (neutral wire), and any of the other three wires. FIG. 11. familiar to us since the first experiments with the Jablochkoff candles. Up to the present only incan- descence lamps have been used for electric lighting to the exclusion of arc lamps. If we consider a triphased generator in which the " star " mode of coupling is used, we can see that it is 29 possible to arrange deviations from it in six different ways, three by connecting the lamps between any two wires : a and b, b and c, c and a, three by connecting them between the neutral wire, o, and each of the three other wires : a and 0, b and o, c and o (fig. n). The difference of potential between any two wires being, as we stated previously (page 22), 1*732 times as great as between the neutral point and any one of the wires, the lamps connected in parallel should have voltages in the same ratio, according to the points between which they are connected. From the point of view of the independence of the three circuits and the possibility of varying to a certain extent the number of lamps supplied in the three groups constituting the connection of each system, it is better to connect the lamps between the neutral wire and each of the three other wires. In the case in which the three groups give exactly the same intensity, we can easily see that no current passes into the neutral wire, the function of which is simply to compensate the inevitable inequalities of output that occur in practice. Each of the three cores of the three-branched transformer works for its own circuit, and the system, as far as incandescence lamps are concerned, may be compared to three independent transformers connected by a common return wire, but the phases of which lag by a third of a period. Thus, for each group of lamps in service, two wires will be sufficient, one connected to the neutral wire, and the other to the extremity of one of the three bobbins of the generator or transformer. 80 As a scientific curiosity, M. Dolivo-Dobrowolsky showed us at the Frankfort Exhibition an incandes- cence lamp at work with three filaments arranged as shown in fig. 12. To any one unacquainted with the pro- perties of polyphased alternate currents, it is somewhat puzzling to see a lamp with three filaments of equal section and equal length arranged in derivation on the 6 c Coupling of a three-filament lamp on the three wires, a, b, c. FIG. 12. three wires connected by their other extremities and traversed by equal currents. In order to understand this phenomenon, we have only to remember that if the effective intensities are pretty equal in the three filaments, they are unequal at each instant, and that each of the wires serves as a common return wire for the current passing at the same moment through the two others. It is not probable that an important 81 practical application will ever be made of such lamps, but it was very interesting to note their curious properties. MOTORS. The chief interest attached to polyphased alternate currents lies in the production of motive power. Although with these currents it is possible to obtain motors that are either synchronous or asynchronous, they are used chiefly as asynchronous motors, and we shall only discuss motors of this kind. An asynchronous motor, worked by polyphased currents, is characterised by two essential parts : an armature and an inductor moving relatively to one another. The inductor is connected directly with the source of electricity, dynamo or transformer, and the armature constitutes a circuit closed electrically upon itself, and not connected electrically with the inductor. The relative positions of these two parts differ according to the power of the motors. In motors of low power, the inductor is fixed and the armature is movable ; in high power motors, on the contrary, the armature is fixed and the inductor is movable. We may here remark that with polyphased currents a great confusion of terms still prevails with regard to the name to be given to the two essential parts of the motor. For our part, we keep to the expressions used to describe continuous current motors, and apply the name of inductor to that part of the electrical 32 circuit producing the field, and that of armature to that part that moves as to the field and so becomes the seat of induced currents. Low Power Motors. In low power motors, as we have just said, the inductors are fixed and the armature is movable. This inductor consists of a series of bobbins, or to speak more generally, of windings traversed by triphased alternate currents. These currents develop in the Gramme ring producing a rotating field with triphased alternate currents. FIG. 13. interior of the cylindrical space left by the winding a revolving magnetic field. If the winding is simple or bi-polar, we get a single magnetic field, the angular speed of which corresponds to one revolution per period. If the winding is combined to produce four, six, or eight poles, the angular speed of rotation of the field is only half, a third, or a quarter of a revolution per period. The frequency or number of periods varies at present between 30 and 40 per second. S3 This revolving field may be realised theoretically by more or less complex windings. One of the simplest methods of realisation is to take a Gramme ring wound round, but without a collector (fig. 13). By attaching three points of the winding 120 apart to a triphased alternator, we get in the interior of this ring Winding of Brown's 15 kilowatt triphased alternate current motor.. FIG. 14. a revolving magnetic field ; the three elementary bobbins constituted by the three-thirds of the ring are then coupled by the triangle method. But more often, and this is how Mr. Brown's motors are constructed, the drum winding is used, the copper strands forming this winding being connected together by connectors arranged round the periphery, and leaving the cy- 34 lindrical space in which the armature is placed perfectly free and accessible. As regards the constancy of the revolving magnetic field produced by a winding constituted in this way, it is difficult to say how far we can depend upon it, on account of the hysteresis, the reactions of the armature, and the more or less sinusoidal form of the current passing through the three bobbins. We can demon- strate, however, and we recommend young students of electricity to go over the trigonometrical calculations which are very simple, of this demonstration, that, in the case of three similar bobbins, lagging by a third of a period, traversed by equal sinusoidal alternate currents, also lagging by a third of a period, placed in a homogeneous medium of constant magnetic per- meability, the magnetic field is constant. The same may be said, and has often been proved, in the case of two alternate sinusoidal currents lagging by a quarter of a period, passing through two bobbins placed at right angles. We may also quote as a type of inductive winding that of the Brown motor of 15 kilowatts. From figs. 15 and 16 it will be seen that the armature is stationary, and in the general arrangement quite similar to the above-mentioned 300 H.P. generator. In consequence of the smaller size, the inductor contains more than three wires per pole. The three circuits are connected up like a Thomson - Houston armature, and the winding is so arranged that four rotating poles are produced, with 40 C5 Lateral view and vertical section cf Crown motor. FIG. 15. End view of Brown motor. FIG. 16. 36 cycles per second : the motor makes about 1,200 revolutions per minute. The normal tension for the motor is 80 volts, a reduction to 50 or an increase to over 100, will not make a practical difference in the speed. Of course, in the first case, the heating of the armature wire is a more considerable one, in the second, the iron heating is increased. The magnetic field is formed of a laminated ring with holes, in which are placed insulated copper bars; the free ends on both sides are connected by copper rings : this pro- duces the whole field, and it is not easy to imagine a more simple construction. Before giving results of the trials, some general data of the motor may be of interest. The armature has 90 wires of about 40 mm 2 , section ; the mode of winding and connecting up is shown diagrammatically in the drawing. The copper weight is 20 kg., the iron weight about 100 kg. ; the breadth of the armature is 200 mm., the outer diameter about 500 mm. The rotating magnet carries 54 copper bars with a section of 100 mm 2 . The weight of the copper is 15 kg., that of the iron is 70 kg. Recent trials in Oerlikon with this motor showed that it can easily supply 20 H.P., the heating even with a con- tinuous run is under normal, as no part of the motor showed a higher temperature than 20-25 C. above the temperature of the test room. It may be of interest to mention that the rotating armature shows practically no heating. The difference of speed be- tween no load and full load is only about 3 per cent. The torque of the motor at starting is a very con- K r T\V siderable one. It could easily put in motion a 25 H.P. ventilator. The same is the case with an extra excited dynamo, which was working on a 20 H.P. absorbing resistance. Over 200 kg. torque could be obtained on the 20 cm. pulley. The motor can carry also a con- siderable overload and runs without the least noise. The starting of the motor is similar to a continuous current motor controlled by a resistance in series with the armature circuit. Regarding the commercial efficiency no accurate tests have been carried through. The losses are about the following : In the armature, copper and iron, they amount to about 4 per cent., in the magnet copper and iron, including friction, to 4 per cent. From this results an efficiency of 92 per cent. A commercial efficiency of 90 per cent, can be taken as absolutely sure. That the losses are very small is further clearly shown by the very moderate heating, in spite of the small size of the motor and the trifling ventilation. A further advantage is that the commercial efficiency remains a high one, even if the motor works considerably underloaded. The weight of the motor is 420 kg., that is to say, a trifle over 20 kg. per H.P. For locomotive purposes, as in the arrangement devised by M. Heilmann, it would not affect the solidity of the design to reduce the weight to 300 kg., the result of which would be only 15 kg. per H.P. This result must be looked at as a very good one, especially considering the low current density the motor is worked with. As the bearings are automatically oiled and contain large oil reser- 38 voirs, this motor can run without any control for months. The theoretical winding represented in the figure shows 54 bars only (not go), or 18 for each of the three bobbins. These three bobbins are besides divided into two halves connected in tension, as shown in the diagram (fig. 17). By this mode of coupling the 9 free ends of wire are reduced to three, which communicate directly with the triphased Connections of the six windings in the inductor of Brown s 15 kilowatt motor FIG. 17. generator. In order to reverse the direction of the rotation of the field, it is sufficient to reverse the current in two of the three windings ; we thus change the direction of rotation of the field, 'and, consequently, that of the motor : this is very easily effected by means of a reversing commutator. The armature consists, in small motors, of a series of longitudinal bars parallel to the axis, and connected in derivation. A revolving field motor could be obtained by placing in the space formed by the in- 39 ductors a Ferraris copper cylinder or one of solid iron, but as the revolving field is not homogeneous, the Foucault (eddy) currents developed in these masses of copper or of iron would not be directed so as to produce a very appreciable motive couple, and thus the efficiency and specific power of the motor would be reduced. In order to obtain high efficiency and high specific power, we must modify the armature and construct it as shown in fig. 18. A cylinder formed of discs of thin sheet iron insulated electrically, is pierced Armature of the Brown 15 kilowatt motor. FIG. 18. in its periphery with holes parallel to the axis, in which are inserted bars of copper, also insulated, the extremities of which, on each side of the cylinder, are soldered into two rings of copper, and form a winding in derivation. The iron discs serve to reinforce the magnetic field, without producing in it any Foucault currents ; the induced currents are generated in the bars of copper and are thus developed in the direction corresponding to the maximum torque between the revolving magnetic field and the induced wires. 40 Theory. The theory of the motor so constituted does not differ essentially from Messrs. Hutin and Leblanc's revolving field motor. The formulae relating to this motor show that the angular speed of the motor when turning freely tends to equal that of the re- volving field, the induced circuit then remaining motionless with regard to the field. When the charge is increased the speed is diminished, but remains very little different to that of the field, if the electrical re- sistance of the induced circuit and its self induction are themselves very small. On account of the low relative speed of the field and the armature, each of the wires constituting the winding is the seat of a periodic electromotive force of low frequency, the period of which, when turning freely, may last several seconds, and which, when fully charged, generally remains less than one period per second. As the different wires constituting the winding, lag in the magnetic field, and, when the permanent rate of working is established, are the seat of alternate currents of the same period, it follows that the motive torque produced by the sum of the torques exercised by each wire remains constant, and that the motor turns with an angular speed as uniform as that of a continuous current motor bearing elementary bobbins equal in number to the induced wires. 41 Owing to the low resistance of these induced wires and to their feeble impedance, on account of the low frequency realised, the induced currents are of great intensity and produce a powerful torque, both at starting and when working normally. This powerful torque, at starting, constitutes an important and interesting characteristic of revolving field motors, whatever may be the method adopted for producing this revolving field. But the employment of poly- phased alternate currents has another advantage, of no less importance, especially for motors of low power. From the very fact that the inductive system is con- nected with the three wires coming from the trans- former or the generator without any other break than the interrupter, and that the armature forms a distinct circuit, absolutely closed upon itself, the motor re- quires no commutator, no rubber, no sliding contact, and its maintenance is reduced to simply filling the oil cups of the two bearings which support its axle. It is difficult to conceive an apparatus that would be more simple, less likely to get out of order, or better suited to the distribution of electrical energy when it is required to feed motors of low power. The only inconvenience attached to the employment of these motors is the third wire, but this would not prove a serious drawback in the case of fixed installations, considering that with continuous currents distribu- tions are carried on requiring three, and even five wires. 42 High Power Motors. We have already said that for motors of great power and the only type that we can mention as having entered into the domain of industry, is the 100 horse-power type constructed by M. Dolivo- Dobrowolsky the respective positions of the armature and the inductor are changed ; the inductor receiving the current from the dynamo or transformer is movable, whereas the armature closed upon itself is fixed. This arrangement tends naturally to complicate the con- struction, since the current has to be brought to the movable inductor by brushes and collecting rings. This is the reason for this construction. We have stated and proved that the inductor is traversed by currents of high relative frequency, whereas the arma- ture, on the contrary, is the seat of currents of very low frequency. It follows that the inductor is sub- jected to very frequent reversals of magnetisation, whilst the armature describes a much smaller number of magnetic cycles in the same time. From the point of view of losses by hysteresis, therefore, it is better to give the smallest possible mass to the system subjected to frequent magnetic reversals, and without importance to have large magnetic masses in those parts of the motor that are subjected but rarely to reversals of magnetisation. These conditions are obtained in motors of great power by putting the inductor inside 43 and the armature outside. The losses from hysteresis are thus reduced to a minimum. In motors of low power the efficiency is a secondary consideration, and is to a certain extent sacrificed to simplicity by leaving the inductor movable, which does away with brushes and collecting rings. Fig. 19 shows the outward ap- pearance of M. Dolivo-Dobrowolsky's 100 nominal horse-power motor, which worked the artificial cascade at the Frankfort Exhibition. We find here a movable inductor receiving the current through six collecting rings, although this number will be reduced to three in practice. As constructed at present the six rings com- municate with the six free ends of the three windings, and thus enable the three bobbins, or to speak more correctly, the three windings, to be connected either in a triangle or a star, as required, so that we can deter- mine by experiment which coupling is most favourable to the good working of the motor under the different conditions of its practical employment. This winding is multipolar and produces on the periphery of the inductor four revolving poles instead of two, so that when working normally and freely the angular speed is only about half the frequency of the generator or transformer. If, for instance, the frequency of the alternating current is 20 cycles or periods per second, which corresponds to 1,200 periods per minute, the angular speed of the motor will remain always less than 600 revolutions per minute, this speed of 600 revolutions per minute being, moreover, the maximum limit compatible with free working. Thus the re- 44 45 volving field motors can never run too fast, which is a great advantage in certain applications. When reduced to its absolutely essential elements, the triphased current motor of M. Dolivo-Dobrowolsky consists of two concentric rings of soft iron, the outer ring acting as a fixed armature, and the inner as a movable inductor, as in the triphased alternate gene- rator of Mr. Brown described before. The two rings are pierced with holes parallel to the axis, very near to the periphery, in which are inserted the inductive and induced conductors, the connections between the con- ductors being made by means of connectors arranged respectively on the surfaces of both parts. The conductors forming the inductor are coupled together so as to produce a revolving magnetic field with four poles, the fixed induced conductors are also coupled together in three circuits in a star, and so arranged that a resistance can be introduced into them at the moment of starting, a precaution which seems only necessary for motors in which the power exceeds 15 kilowatts, for Mr. Brown's motor is not provided with this arrangement, which complicates the con- struction of the armature, but looks indispensable for starting. Different Modes of Coupling for the production of a Revolving Magnetic Field. We have supposed up to now that the three elementary bobbins producing a magnetic field were connected in a star. In a paper sent by M. Dolivo- 46 Dobrowolsky to the International Congress of Elec- tricians at Frankfort, the writer indicates two other arrangements for the same purpose, which are shown by figs. 20 and 21. The first (fig. 20), which can be Star coupling of three bobbins for the production of a rotating field. FIG. 20. understood at a glance from the diagram, is simply the connection of the three bobbins in triangle. We can see that, by this arrangement, the intensities i } , i, z,, of the currents passing through the three bobbins lag 60 47 behind those passing into the conductors, i lf I 2 , I 3 . The coupling shown in the diagram (fig. 21) comprises six bobbins, and, in accordance with the remark we have just made, gives, by means of three currents lagging by Production of six currents lagging of th of pe.ioj with three currents lagging cf a Jrd (Dolivo-Dobrowoliky coupling). FIG. 21. a third of a period, six currents lagging one behind the other by a sixth of a period. By arranging suitable proportions between the number of spirals and the thicknesses of the wire on the two series of bobbins, 48 I,, I 2 , I 3 , on the one hand, and i 1} i. 2 , * 3 , on the other hand, we obtain six equal magneto-motive forces lagging by a, sixth of a period, which combine to produce a constant revolving magnetic field. But as it is easy to show that three currents, or even two, only, are sufficient to produce a constant revolving magnetic field, the complex arrangement indicated by M. Do- browolsky seems to us, as far as we understand at present, needlessly to complicate the winding of the inductive circuit without any important advantage being gained by it. In short, if for motors of great power, and for considerations relating to the frequency of the magnetic cycles to which the armature and the inductor are subjected, it is expedient to place in the centre the inductor, or the part which receives the current, with a view to reducing its volume as much as possible, three collecting rings will be sufficient in any case lo convey the current instead of six, and a simpler winding can be used in a star or a triangle to ensure the constancy of the magnetic field. We can also place the inductor inside without making it turn, and thus the collecting rings can be entirely dispensed with. Instead of variable resistances being introduced into the induced circuit at the moment of starting, they would be introduced into the inductive circuit, which would enable the revolving induced circuit to be left closed upon itself as in Mr. Brown's 15-kilowatt motor. Notwithstanding their simplicity, the triphased alter- nate current motors of great power are still capable of 49 being improved and simplified, combining the require- ments of theory with the necessities of practice, and thus extending the applications for which they can be employed. Efficiency. The results of the experiments of the Lauffen-Frank- fort transmission on the motors have not been as yet published, and it is difficult to have any opinion on the efficiency of these motors. But efficiency tests have been made by M. Dolivo-Dobrowolsky with a motor of normally 2 H.P., running with 70 volts, at periods of 35 per second. The weight of this machine was 165 kilogrammes, and it was found that it did not get unduly warm when running for long periods at 2 H.P. For a short time it could be loaded to 4 H.P. on the shaft, without very seriously reducing its efficiency. The variation of speed between running empty and with a load of 2 H.P. was only 6 per cent., and by overloading the motor 100 per cent., the speed was still appreciably high, showing its superiority over alter- nating current motors, which stop altogether when overloaded. When running empty, the motor consumed 205 watts, and the useful mechanical work when applied, increased almost in exact proportion with the electrical energy used, until it reached 3 H.P., when it fell somewhat. A series of tests demonstrated that the efficiency of this motor at half load, or i H.P., was 75 per cent. ; at its normal load, 2 H.P., 80 per cent., and it reached the maximum, when giving 2*3 H.P., when E yu it was 81*4 per cent. The speeds, unfortunately, are - - - 1 Synchronous Triphased Motors. The rotating field motors, which we have described, being self-starting on heavy loads, and being able to work at any speed between zero and a maximum, limited by the frequency and number of poles, will be always preferred to synchronous motors. The best means to realise a synchronous motor is to take a triphaser like the Brown generator, but not excited. This generator connected to the three wires carrying a triphased alternate current can work as a motor, in which case it runs synchronously with the generator, but unlike a common alternating machine, it can start itself. The field must not be excited until synchronism has been obtained. In case the generator and the motor are set in action simul- taneously, the excitation of the magnet may be effected at the beginning, and in this case the initial torque is always very great, but it is only excep- tionally that we can find circumstances in which these conditions are fulfilled. The use of synchronous polyphased motors is, therefore, very restrained, and needs not more indications than their practical possibility. 1 The publication of this little pamphlet has been too long delayed in the expectation of the report ot the Frankfort committee at the 1891 Electrical Exposition. Some information has been published on the trans- mission itself, and we give these particulars in Appendix A, but nothing h.is transpired yet as to the efficiency of the Dolivo motor. The field is open to ?ny suppositions. 51 APPLICATIONS OF TRIPHASED MOTORS. When we consider the facility with which the triphased current motors are started, and the ease with which these currents are transformed, we can foresee that they will be applied successfully to the transport of great motive power to long distances, notwithstanding the obvious inconvenience of em- ploying three wires instead of the two required for continuous currents and ordinary alternate currents. The possibility of localising the dangers of the high tensions in the line and the transformers, and of thus rendering the generators and receivers absolutely harmless, are sufficient to justify this preference, not to mention the facilities for sub-dividing the me- chanical power afforded by the system of distribution by triphased currents. Transports of motive power into mines, where fire-damp is so dangerous, can also be easily effected by means of polyphased motors, as they have no friction contact, and no part likely to give off sparks. Sparks will only be produced at the interrupter when the system is started and when it is stopped, but it will always be possible to protect this part and render the sparks absolutely harmless. Also, for the railway system recently proposed by M. Heilmann, the triphased alternate current motor without collectors or brushes constitutes an almost ideal method of transmitting motive power between the fixed machine placed on the locomotive and the axles to which it imparts the rotatory movement, 52 since these motors suppress by their very con- struction all friction contacts, and consequently all maintenance of the non-accessible parts. Lastly, in installations worked by motors that are not easy of access, and which cannot, therefore, receive frequent attention certain ventilators of large buildings, for instance it will be advantageous to employ poly- phased motors, which will only require to be visited periodically, in order that the oil-cups may be filled up. It might even be expedient in this case to use a distribution of electrical energy by continuous currents, and to transform part of them into poly- phased currents. We are thus led, naturally, to the study of the apparatus by means of which this trans- formation, or the reverse, can be effected. 53 IV. MULTIPLE TRANSFORMERS. AFTER discussing the application of polyphased alter- nate currents to the production of lighting by incandes- cence and of motive power, we are naturally led on to the study of theprocesses by which these currents can be trans- formed into continuous currents to meet the requirements of electro-chemical operations, the charging of accumu- lators, and other applications to which we shall have occasion to refer. These transforming apparatus, to which we give the provisional name of multiple transformers, are now invented and practically realised. The Frankfort Exhibition has made us acquainted with two very distinct types, one for triphased currents, devised by M. Dolivo-Dobrowolsky, and the other for diphased currents, by M. Schuckert. We will in this- article discuss the triphased multiple transformers ; the corresponding apparatus for diphased alternate currents- will be described when we come to diphased currents- (page 67). In order to thoroughly understand the principle of the multiple transformers, to which, properly speaking, we should give the name of Continuous Current and Polyphased A Iternate Current Generator-Motors, we have merely to glance at fig. 22, which shows the apparatus reduced to its essential elements: A 54 Gramme ring provided with collector and brushes, excited either in derivation or by a separate source. To three points of the winding, 120 apart, are connected three derivation wires communicating with three insulated collecting rings, on which rub three brushes connected with three wires, a, b, c. A pulley is fixed upon the axle bearing the ring and the collector. We will represent by A the pulley, by B the collector and its Generator motor for continuous and triphased alternate current (Diagram showing the principle). FIG. 22. brushes, and by c the three collecting rings, a, b, and c, and their brushes. An apparatus constructed in this manner is capable of six essentially distinct functions, and this justifies the length of the name which we have coined in order to define it. i. By making the pulley, A, turn mechanically, we get at B a continuous current. Thus the system works 55 as a continuous current dynamo, and is really nothing else. 2. By supplying a continuous current at B the ring is made to turn, and mechanical power is obtained at the pulley, A. In this second case we have a continuous current motor. These two functions are well known, and we only mention them with the idea of giving a complete list of all the functions of which the apparatus is capable. 3. By supplying mechanical power to the pulley, A, it turns, is excited, and is, as we have seen, capable of producing continuous current at B. But if, instead of collecting the current at B, we collect it at c, on the three brushes, a, b, c, we shall obtain triphased alternate currents, and we shall thus get a triphased alternate current generator, which is self-excited by a continuous current which it produces itself. This is a very conve- nient method of constructing, at little expense, a triphased alternate current generator. Any shunt dynamo lends itself readily to this transformation by the addition of three rings on the extremity of the axle opposite to the collector, and connecting these rings to three points of the armature windings selected at 120 apart. 4. By supplying triphased alternate currents to, the three rings, a, b, c, when once synchronism is established between the angular speed of the axis of rotation and the frequency of the alternating currents, we get a synchro- nous alternate current motor, and collect mechanical power 56 at the pulley, A. The apparatus thus works as a synchronous alternate current motor. It must be observed that the working of this synchronous motor differs essentially from that of the motors with a revolving magnetic field which we have described in a former article. These motors can, in the first place, be only started when free without any load, and when not excited. The initial motive torque is produced by the reactions of the armature on the inductors. When synchronism is obtained the machine can be excited with its own current, and under these conditions, this synchronism is kept up, even when the machine is loaded, but it is indispensable to the working. The revolving field motors have the property of slackening their speed with the charge, and are even capable, if the resistance of the induced circuit is made to vary, of turning at any speed whatever. The alternate current motors, such as we have described, either turn syn- chronically or stop. We therefore mention this property of the multiple transformer without, however, attaching any practical importance to this fourth special function. 5. On continuous current being supplied to the brushes, B, the apparatus turns, is excited, and triphased alternate currents are collected at c. Thus we have a transformer of continuous currents into alternate triphased currents. Of course any winding whatever of a continu- ous current dynamo, furnished with the three collecting rings can be made to act in the same way. M. Dolivo- 57 Dobrowolsky, at the Frankfort Exhibition, used the drum-winding, taking the continuous current from accumulators and transforming it intotriphased alternate currents working his little revolving field motors during the stoppages of the Lauffen transmission. In a paper to the Societe Internationale des Electriciens, the writer of this pamphlet used, for the same purpose, a small series continuous current dynamo, constructed by M. Hillairet, with a Gramme ring, furnished with three collecting rings for repeating the principal experiments relating to revolving fields and polyphased alternate currents. This mode of transformation may be applied to certain special cases when, for instance, it is required to work from a great distance an electric motor that is not easy of access, that cannot easily be examined and attended to, and the sparks of which issuing from the collector entail a certain amount of trouble or inconveni- ence. The continuous current, when transformed into polyphased currents, can in such a case serve to work a revolving field motor without either collector or brushes, and above all, without sparks. 6. The most important of the transformations which can be effected by means of the apparatus is unques- tionably that of which we have now to speak. By send- ing into c triphased alternate currents, the apparatus is made to turn synchronically with the frequency of the alternating current, and continuous current is collected at B. The apparatus is a transformer of triphased al- 58 ternate currents into continuous current. The mode of transformation is similar, but of a contrary nature, to that which we described when speaking of the fourth function of the multiple transformer. We can see that by connecting utilising apparatus at B, these latter will be traversed by continuous currents, and may be either motors, accumulators, electrolytic baths, etc., as re- quired. We have had occasion to see an application of this kind at the Frankfort Exhibition, namely, the working of many little motors, made by M. Schuckert with his diphased alternate currents. From what we have just said, we see that it is possible to conceive a great number of continuous current and polyphased alternate current motor generators, by suppressing those parts that are useless for certain special applications, and by giving the essential organs known arrangements combined in different ways. We have been careful to describe the principle only of these curious transformations, not entering into questions of priority which our present information will not enable us to decide, or even to study, with any degree of completeness. We shall now describe the principle of apparatus for producing, transforming and utilising diphased alternate currents. 59 H.-DIPHASED CURRENTS. ALL the apparatus that we have mentioned up to the present work with three wires and three currents, under conditions of perfect symmetry, as we showed at the beginning of this series of articles. We have now to speak of the properties of two-phased alternate currents, the study of which will be greatly facilitated by what we have already found out with regard to triphased currents, and we shall follow the same order in discuss- ing them. GENERATORS. The first generator of diphased alternate currents was clearly described in Nikola Tesla's American patent, No. 381,968, dated May ist, 1888 (application filed October I2th, 1887), as a generator capable of working a synchronic revolving field motor. This generator consisted, essentially, of two bobbins connected at right angles on the same cylindrical armature in the magnetic field produced by suitable inductors. It was a generator with four wires and three currents, but the number of wires can easily be reduced to three by rendering common any two of the wires belonging, respectively, to each circuit. The diphased alternate current generators shown by M. Schuckert at Frankfort in the transport of motive power established between 60 the Palm Garden and the exhibition differs in principle from M. Tesla's in the crossing of the circuits, which altogether prevents the employment of a common return wire, and necessitates the use of four wires proceeding from the generator. Fig. 23 shows the winding of this generator, consisting of a simple Gramme ring revolving with a uniform angular speed in the magnetic field produced by inductors excited by a separate current. FIG. 23. DIAGRAM SHOWING THE CONNECTIONS OF A SCHUCKERT ALTERNATE DIPHASED GENERATOR. On the winding of this ring four collections of cur- rent are taken, the two points diametrically opposite a, a' communicate with one circuit, and the other two diametrically opposite points, b b', communicate with the second circuit. As the ring is movable, and the inductors fixed, the permanent connections between the four points of the generator, and the four con- ductors, are established by means of four collecting 61 rings and four brushes, which are not shown, in order not to complicate the diagram. It would, moreover, be easy to reverse the arrangement by leaving the induced ring fixed, and making the inductors turn. In this case we should get permanent communication between the generator and the outer conductors ; two rings and two brushes are sufficient to bring the exciting current to the movable conductors, as in Brown's triphased generator, which has already been described. On examining this generator more closely, we see that if the circuits traversed by the two alternate currents are clearly separated on leaving the armature, they are common in the armature itself; each quarter of the ring is traversed, at each instant, by the algebraic sum of the two outer currents taken with different signs, according to their connections. Many other combinations might be employed in order to realise a diphased alternator ; but we merely mention those that have been practically realised and applied to some practical use. TRANSFORMERS. We have just seen that the diphased alternate current generators generally contain four wires. For reasons of economy on the one hand, and reduction of danger on the other, these generators are made of low-tension ; it is therefore necessary to transform these currents when they are to be transported to a great distance. The transformer effects this transformation, and, more- over, enables the four wires to be reduced to three along 62 the line of transport. Fig. 2\ shows how this result is easily obtained by a suitable combination of the induc- tive and induced windings. In fact, a double transformation is generally resorted to, so as to reduce upon arrival the high potential of the line, and to work at low tension on the utilising apparatus. The Frankfort installation is a typical instance of this double transformation which enables us to have again, upon arrival, the two lagging circuits, 100 volts. Transformer. 2 000 volts. Transformer. 100 volts. Diagram showing the double transformation of diphased alternating currents (first from 100 volts to 2,000, second from 2,000 volts to 100). The two circuits are perfectly separated on departure and arrival, but have a common return along the line of high tension trans- mission. FIG. 24. perfectly distinct and independent, with four wires instead of three. The diphased transformer employed by M. Schuckert to transform the 100 volts furnished by the dynamo into 2,000 volts, and again to re-transform upon arrival the 2,000 volts into 100 volts, is shown as a whole in fig. 25, and the details of its construction in fig. 26. This multiple transformer belongs to the class of apparatus with a closed magnetic circuit, and consists, in prin- ciple, of two annular discs of soft iron. The thicker of these two discs is provided with radiating slots of FIG. 26. 64 rectangular section, into which fit the elementary in- ductive and induced bobbins, which are triangular in form. These bobbins contain windings, the numbers of which are in the same ratio as the co-efficient of transformation, and each of them forms a complete group. In the model shown there are twelve radiating slots, into which can be fitted six pairs of bobbins, three pairs for each of the two currents. We can, moreover, modify within considerable limits the ratio of transformation by modifying the coupling of the different bobbins, according to whether they are con- nected in series or in quantity. The magnetic circuit is closed by adjusting on the first disc furnished with its bobbins the second disc forming a sort of lid for the bobbins. To avoid all danger, the two discs are con- nected to earth. This arrangement of the transformers is one of the most simple, and has the advantage of being applicable, if the bobbins are suitably coupled, to ordinary alter- nating currents as well as to diphased and triphased alternating currents. APPLICATIONS. As with triphased currents, it is possible to apply diphased alternate currents to lighting and to motive power, and to transform them into continuous currents for those applications in which electrical energy can only be utilised under the form of continuous currents. Space does not permit of our entering further into the application of these currents to lighting ; the incande- 65 scence lamps being connected in derivation on each of the two distinct circuits. We will merely say a few words about motors, and the apparatus employed for transforming these currents into continuous currents. MOTORS. The diphased current motors really preceded the triphased current motors. They are synchronic or asynchronic. Although we cannot speak positively on this subject, it seems that the asynchronic motors were invented independently, and about the same time in Italy by Prof. Ferraris, and in America by M. Nikola Tesla. In both cases the researches and experiments date as far back as 1886, but M. Ferraris's motor was not made public until March, 1888, in the Atti delta R. Accademia delle S dense di Torino ', and the patents granted to M. Tesla, which were applied for in October and November, 1887, were not published until May ist, 1888. Fig. 27 shows the principle of Prof. Ferraris's two- current motor; it is reproduced from the original paper. We see here two rectangular circuits, traversed by two currents, with a lag of a quarter of a period, producing the rotation of a copper cylinder placed in their midst. The theory of the working is the same as that of the three-phased synchronic motor, so that we need not discuss it further. We may mention, how- ever, that Prof. Ferraris, at that time, did not consider that a revolving field motor had any industrial import- ance, even although he was studying its dimensions 66 in order to increase its power and its efficiency (page 10 of the report), whereas M. Tesla took out his patents quite aware of the industrial importance to which revolving field motors might attain. In fact, they have been applied in America to certain purposes which our American confreres have vaguely mentioned ; so vaguely, indeed, that it is difficult to say to what extent these applications have met with industrial success. We may say that up to the present there FIG. 27. PRINCIPLE OF FERRARIS'S REVOLVING FIELD MOTOR. does not exist, in Europe, a real industrial asynchronic motor with diphased currents. We do not know how many are in existence in America, and should be glad to receive information on this point. The same remark might be applied to synchronic motors with diphased alternating currents, for the only case that we can mention . is that of the transport of power established at Frankfort, and the motor worked at the Exhibition constituted rather a multiple trans- former than a motor properly so called. MULTIPLE TRANSFORMERS. All that we have said with regard to triphased current multiple transformers applies equally well to diphased current multiple transformers. In fact, M. Schuckert exhibited at Frankfort a generator-motor with continuous currents and diphased alternating currents similar in principle to that which we described in our last article. This apparatus is shown in fig. 28. It consists of a Gramme ring furnished with its collector, and moving in the field produced by inductors arranged along the two surfaces of this flat ring. Four points 90 apart are taken in the winding and connected with four metallic rings insulated from one another, and against which press four rubbers. This apparatus, constituted in this manner, may perform six distinct functions : i, it may act as a con- tinuous current generator ; 2, as a self-exciting diphased alternate current generator ; 3, as a continuous current motor ; 4, as a diphased alternate current motor ; 5, as a transformer of continuous currents into diphased alternate currents ; 6, as a transformer of diphased alternate currents into continuous currents. It was this sixth and last function that was performed by the apparatus exhibited at Frankfort by M. Schuckert. The high tension alternating currents (2,000 volts) issuing from the first transformer placed in the Palm Garden arrived at the Exhibition by three wires, and were received into a second transformer where they 68 69 were brought back to the potential of 100 volts, and separated into two distinct circuits with four wires. These four wires communicated with the four rings of the apparatus, which rings are shown at the left of fig. 28. The diphased alternate currents passing through the ring tend to produce in it a revolving magnetic field. This field reacts on the inductors, the circuit of which is at this moment open. This produces a motive couple which tends to turn the armature in the same direc- tion as the field and synchronically with it. When synchronism is established all we can do is to close the circuit of excitation on the brushes of the continuous current collector, in order to excite these inductors permanently and ensure this synchronic rotation be- tween the generator motor and the frequency of the diphased currents which feed it. The continuous current canalisation is branched on these same brushes, so that from the moment when synchronism is attained and the circuit of excitation closed, we get continuous current supplied by the transformation of diphased alternate currents. This continuous current served at Frankfort to feed con- tinuous current motors of low power working different implements in the gallery constituting the workshop, so that the electrical energy produced at low tension in diphased alternate currents on four wires was trans- ported at high tension in diphased alternate currents on three wires, transformed into low tension alternating currents on four wires, and, lastly, re-transformed into 70 continuous currents on two wires. The experiments of the trial committee (not yet published) will show at what cost all these transformations of energy were effected, and, considering the contradictory opinions expressed as to polyphased alternate currents, the sooner the better. CONCLUSIONS. After this rapid sketch of the properties of polyphased alternate currents, the reader will be convinced that they have many curious and interesting properties, and that as they constitute, as has sometimes been said, a third form of electrical energy, these special properties are worthy of being more fully discussed -than in this pamphlet, in which we have given a brief survey of the question. We shall, moreover, soon be enlightened as to the industrial and commer- cial value of this new branch of electrical engineering by the experiments of the Frankfort Committee and the results of the transmissions of motive power now being established with polyphased currents in different parts of Europe. We publish in Appendix B the description of the first distribution of electrical energy by triphased currents, made by M. Oscar von Miller at Heilbronn, at a distance of seven miles from Lauffen, the generating station. Some valuable information is given by this first application. The present importance of these Currents may be greatly diminished by the discovery of a simple alter- nating current motor starting with a load, or by that of 71 an apparatus capable of transforming simple alternating currents economically into continuous currents. It is also possible, and even probable, that polyphased currents will constitute merely a transitory system of transmitting electrical energy to a distance, as the Jablochkoff candle constituted a transitory system between the arc lamp and the incandescence lamp. However this may be, polyphased currents will mark a new era in electrical industry ; if their life is to be short, it will at least not have passed without honour and interest. APPENDICES. APPENDIX A. THE LAUFFEN-FRANKFORT TRANS- MISSION. THE following information with regard to the results obtained by the testing committee was furnished by the Council of the Frankfort Exhibition. The work of the committee consisted in testing three sets of plant. The first was that put down by the Allgemeine Elektricitats Gesellschaft, of Berlin, in conjunction with the Maschinenfabrik Oerlikon for transmitting 300 H.P. from the Wurtemburg Portland Cement Works at Lauffen, on the Neckar, to the Exhibition, a distance of 109 miles. The measurements were made by Prof. Dietrich, Dr. Feussner, Dr. Heim, Dr. Kopp, Messrs. Nizzola and Schmoller, Profs. Stenger, Teich- mann, Voit and Weber. Table I. shows the results obtained. It is to be noted that column 12 gives the efficiency expressed as a percentage ratio between the energy consumed by the lamps at Frankfort and that delivered by the turbines at Lauffen. The measure- ments taken at Lauffen, which were carried out under Prof. Killter and Councillor Lindley, with the view of obtaining the efficiency and other interesting points with pressures of 25,000 and 30,000 volts, will be pub- lished later on with the report on the Exhibition. 76 -ransuoo ptre sarnqxri; Surumsuoo POB oniBn^p IQTJUOJSUBI^ XjBpu januojsirexj A'jBpaooas 01 pat[ddns &i9ug[ uo!)onpuoo m BSOI U9UUOJ8UBJ^ AaBtnud jo ^nd;no Xq paii -euiqjn; aiiddns 'J- f CDCOOOCOOOCOCOCOCOi>t>COCOCOX>t>t* W W' ^^ ^r^ ^ ^ \f$ 0j ^Q ^Q *^ TJH which also contains the necessary safety cut- outs. By means of this superposed arrangement, it is easy to get at every part of the transformers and their connections to the transmission leads and safety devices. The Government of Wurtemburg has been very exacting in the matter of safety appliances, espe- cially those placed at the village of Southeim, which lies half-way between Lauffen and Heilbronn. At Southeim is a 20 kilowatt transformer station for local consumption, containing the transformer in a locked- up room, and a switchboard with instruments and G 82 Transformer. Transformer. Cable connection and safety device. Primary leads. Secondary leads. FIG. 3. 83 safety devices, as, also, a telephone. The telephone circuit runs on the same poles as the transmission leads, and is only about 28 inches below them. It serves the double purpose of telephonic communica- tion and as pilot wires for taking measurements. The telephones are said to work admirably in fair weather, and that only during very strong winds can any dis- turbance be observed. Messrs. Reiner, of Munich, have provided transmitters and receivers of a special design to work in proximity of the " Drebstrom " cir- cuits with the least amount of inconvenience to the users of the telephone line. During a storm, when the overhead wires assume varying positions relatively to each other, sufficient induction is set up to make telephoning difficult. Animated discussions have recently taken place in Germany between the postal authorities and electric light engineers on the subject of interference with telephones and telegraphs. In fact, special laws are being passed to protect the Government from disturbance created by the increas- ing use of heavy currents for purposes of transmission of electric energy. To judge by the fierce arguments at the meetings of the German electrical societies one is bound to come to the conclusion that there are two opposing parties. One, a conservative party consist- ing of telegraph and telephone engineers backed by Government officials, who are called " Schwachstrom " (weak current) engineers, and the progressive party, composed of " Starkstrom " (strong current) en- gineers. 84 The consumers at Heilbronn are supplied with electricity meters devised for three-phase circuits by Dr. Aron. A special kind of arc lamp, of which either two or three run in series, has been constructed by Messrs. Korting and Mathiesen. Motors of from 2 to 15 H.P. capacity are being made by the Allge- meine Elektricitats Gesellschaft. These motors are self starting, run at constant speed with varying loads have no commutators, and are reversible. With the exception of charging accumulators, this multiphase system has the combined advantages of the alternat- ing and the continuous current systems without the disadvantages of either. re n.