588

ELECTRICAL

the transformer. The vertical ordinates in fig. 4 represent the observed P.D. of the two surfaces of this condenser-that is, the P.D. at the secondary terminals of the transformer.

It will be seen from fig. 4 that when the capacity is zero this P.D. was 2,500 volts. As the length of cable was increased continuously, this P.D. rose up to 8,500 volts, corresponding to a capacity of 2 microfarad, and then fell again to 2,500 volts, at a capacity of 45 microfarad. Hence, corresponding to a certain capacity-viz., 2 microfarad-the pressure is multiplied about 3 times, and there is a certain narrow range of capacity over which this exaltation of pressure takes place. At and beyond 45 microfarad capacity, the presence of the condenser reduces the normal pressure continuously. The pressure on the primary side of the transformer is also increased, as shown in the curve given in fig. 5, which represents a series of similar pressure observations taken on

REVIEW.

[MAY 8, 1891.

it. When transformers have been employed to test the insulation of armatures of high-pressure alternators, it has been noticed that under some circumstances large snapping sparks will jump off over air distances which could not possibly permit discharges at the actual P.D. of the secondary terminals of the transformer so used when unconnected with any such body having capacity.

In employing an alternator or transformer to test insulation between two opposed conductors, or in any way in which the opposed terminals gain capacity, it is necessary to be on the watch for this possible rise of pressure, and not to infer, because a transformer gives, say, 2,000 volts on open circuit between its poles when measured with an electrostatic voltmeter, that therefore this is the measure of the pressure which is being applied to a dielectric which forms the insulator of a condenser joined in across the poles of that transformer.

I am not at all sure that under some circumstances errors may not be committed in the mere use of an electrostatic voltmeter to measure P.D. between the poles of an alternator or a transformer, if that voltmeter happens to possess a certain capacity in relation to the inductance of the circuit in wbich it is joined in parallel.

10. It is easy to treat in the foregoing manner another case slightly more complicated.

Let the inductive line a b gf (fig. 7) be divided into three sections, a b and bg having each resistance R and inductance L, and the section

Curves showing variation in voltage in primaryl(low voltage) circuit by varying capacity in secondary (high voltage) circuit. The experiments have been made with No. 2 transformer, and current was obtained from W, machine, frequency being 100 . The exciting current was constant, and the alternator was running at 750 revolutions during all the experiments.

FIG. 5.

the primary side of the transformer. It will be seen that the same general effect takes place on the primary side. It will be seen how entirely these curves deduced from observation agree with the theory as illustrated by the curve in fig. 2. The curves in figs. 4 and 5 are generally similar, but, if superimposed, will be found to be not quite So. If the ratio of corresponding ordinates is set off so as to form a fresh curve (fig. 6), this curve will give the variation of change ratio

of the transformer corresponding to the various capacities, and we see that this change ratio is altered progressively from 1: 43 to 1: 57, as the capacity of the condenser on the secondary is changed from zero to 65 microfarad. It is evident, therefore, that in applying transformers to test cable, such as concentric cable, which possesses sensible capacity, we must be on our guard against this alteration of the change ratio of the transformer, (r else we may be misled as to the real pressure being applied to the cable.

9. These considerations point out that in applying either transformers or alternators to circuits which have capacity, we must have regard to the fact that for certain critical values of the inductance and the capacity for given frequencies, there may be a very considerable increase in the potential difference of the two sides of the condenser over and above that of the poles of the alternator or transformer when no condenser is present. Such effects have been noticed in practice of late. It has been frequently found that if a transformer which gives on its open secondary circuit a certain P.D. is applied to test the insulation of a length of concentric cable by connecting its terminals to the two conductors, the far ends of the cable being insulated, under some circumstances an incicased P.D. is observed between the free ends of the cable over and above that at the terminals of the transformer when the cable is not connected to

Bb + a

(92)

= I {(A a is 1) sin p t + (Ba + Ab + b) cospt) (23) from which we deduce that the ratio of the maximum currents, Land 11, in the first and second sections of the line is given by the equation, I = I√(A + 12 + B2) (a2 + b2) + aba - B-2aA+1+2 Bb + 1 (24) and in like manner the ratio of the maximum values of the potentials Vo and V1 can be obtained.

It is possible also to show that for certain critical values of L, C, and R, the potentials Vo, V1, and v, at the terminals of the alternator and at those of the condensers, progressively increase, as well as the currents in the section of the line. Cases of this kind, beyond a single condenser, are best treated as Mr. Blakesley has done by geometrical methods.

11. In the limited case of a condenser charged by an alternating electromotive force through an inductive circuit, we may arrive at a direct analytical solution of the relation between the coefficients of induction and capacity of the system and the mean square of the charging current, and of the potential difference at the poles of the condenser or at those of the alternator, as follows:-As before, let vo and vo stand for the maximum and instantaneous values of the potential difference at the poles of the alternator; let vl and be the same quantities at the terminals of the condenser; and let I and i be the same for the current flowing into the condenser having a capacity c whilst the line connecting the condenser with the alternator has an inductance L and a resistance R. We have then, by, fundamental equations, the following relations :

cos p/ +

sin pt;

589

practice, and the one which has to be compensated for in some way or other. As the difference between the two kinds has been fully explained by Mr. W. H. Preece, F.R.S., Prof. -Ayrton, F.R.S. (the latter has also given a formula for calculating the value of each), and others, we don't intend to enlarge further on the subject. Suffice it to point out here that a current due to electro-magnetic induction has the same direction throughout the secondary wire, whereas electrostatic induction causes a current of opposite direction at each end of the secondary wire.

B

In figs. 1 and 2, where A is the primary and в the secondary

which gives L in terms of observable quantities.

The current at any instant flowing across the dielectric of the cable may be spoken of as the condenser current. It is always in magnitude equal to the product of the capacity of the condenser and the time rate of change of the mean potential difference, v, of its surfaces. It follows that, in the case of simple periodic currents, the square root of the mean square of this current, I, which is what we measure with a dynamometer, is equal to the product of the square root of the mean square of the condenser plate potential difference, V1, and the quantity, c p, where c is the capacity of the condenser, and p, as usual, is 2 times the frequency, or I = cp V1. This follows at once from equations (19) and (31). The value of the potential difference at the terminals of the condenser, as measured, say, by an electrostatic voltmeter, is therefore determined by the magnitude of the condenser current flowing into it, as measured by a dynamometer when the frequency of the current is known.

(To be continued.)

REMARKS ON INDUCTION IN MULTIPLE CORE CABLES.

By P. CHR. DRESING and K. GULSTAD, of the Great Northern Telegraph Company.

HAVING been engaged for some time in investigating the inductive effect between the cores in a two and three-core cable, with a view of finding some means of diminishing or neutralising the disturbance from core to core, we have had occasion to carry out a number of experiments in connection therewith, which, we hope, will prove of interest to the readers of this paper.

We also think that the result of our work is such that it may, to some extent, throw light on the disputed question, whether one mainly has to do battle with electrostatic or electro-magnetic (dynamic) induction between wires suspended on the same poles or cores in the same cable.

Several experimenters have been at work from time to time, and the subject has been ventilated from several points of view; but it would appear that the electro-magnetic effect has been considered the principal one, and until lately, when Mr. J. J. Carty published his paper, "A New Work of Telephone Induction " (the REVIEW, December, 1889), and "Inductive Disturbances in Telephone Circuits (the REVIEW, March, 1891), very little was heard of the electrostatic induction between wire and wire. All our experiments which relate more especially to cores and cables prove beyond doubt that the electrostatic effect is by far the most important; we may say that it is in reality the one felt in

wire, the arrows in I show the dynamic, and in 2 the static effect when currents are sent from the battery at A. It is also evident that in B, fig. 2, there is a point which is neutral, and where no current is felt.

As a matter of fact, all the phenomena as represented above will not appear in multiple core cables of considerable length, simply because the inductive capacity of the secondary core will destroy the relatively weak inductive impulse long before it reaches the distant end, and also because the primary current is weak at that end compared to its first rush into the core at the near end. The consequence of this is that the disturbance in the neighbouring wire, caused by induction, even when very delicate instruments are employed, is of no importance whatever when receiving at the same end of both cores simultaneously; but the very moment that A sends while B receives, great interference at once occurs between the cores, and the signals on B's instrument become distorted, until proper means of counteracting and neutralising the induction are introduced. We shall treat of these later on.

In order to determine whether this induction is dynamic, and due to the powerful charges and discharges of the current in the primary wire, we made the following experiment

with an artificial two-core cable, as shown in fig. 3, where A and в represent the conductors prepared of strips of tinfoil; a and bare sheets of tinfoil, separated from A and B by paraffin paper.

By making or breaking the current in A, no sound whatever could be heard in the telephone, T; the weak dynamic effect could not be detected, and no other induction is possible in this case.

The arrangement was altered, and connections were made as shown in fig. 4, viz., A and B got the same return wire. The telephone repeated now every manipulation of the battery key. The reason for this might be explained by saying that a minute part of the current found its way through B

instead of through the return wire. This latter had, however, no appreciable resistance, and a very sensitive mirror galvanometer gave no deflection. But to make quite sure, and at the same time to prove that the induction was electrostatic in its nature, we connected a and b to the return wire by means of a copper wire (the dotted line). The sound emitted from the telephone ceased immediately, because the static induction between A and B was screened by the sheets, a and b. A and B were next insulated, as shown in fig. 5, which in reality means that the artificial cable was replaced

by a condenser. In this way the inductive action was considerably increased, and a galvanometer could now easily take the place of the telephone.

From the above we conclude that the principal induction felt in multiple core cables must be electrostatic, and if it be said that the distance between the cores in a real cable is much greater than in an artificial one, and that the conditions are different, we need only refer to the induction between wire and wire in telephone cables, where each individual core is surrounded by tin foil, electrically connected throughout and with the earth. How could one possibly screen effectively the induction between such two conductors, if it were a question of dynamic induction, which could only be remedied by passing the return wire between the inducing (primary) and the induced (secondary) wire; and no one

would surely assert that the tin foil acts as a return wire. The same view applies to submarine cables with three, or more, cores, and as it is a fact easily demonstrated, that the induction between two of the cores is considerably reduced when the third is put to earth, we consider that this would be quite impossible if it were a dynamic effect.

We now come to the question as to how it is possible that electrostatic induction can take place between cores submerged in water and surrounded by the sea.

The following experiment was made (fig. 6):-a and b were connected by c with a return wire, z; x and y are short pieces of wire connecting a and b with z. In в a telephone

[MAY 8, 1891.

was placed, the battery had an E.M.F. of about 90 volts, and the condenser a capacity of about 17 microfarad for one of the conductors, A or B. On closing or opening the battery no sound was heard in the telephone, neither was the galvanometer affected, although it was very sensitive and capable of showing induction representing about both of the capacity of A or B. The screening effect of a and b was very perfect indeed under these circumstances, but as soon as x, y, and: -more especially z-had any appreciable resistance, even a fraction of an ohm, the telephone emitted a sound whenever the battery was applied; and when z was equal to some hundred ohms the galvanometer was deflected also. From this it appears proved that whenever there is any, even ever so small, resistance in the earth connection, induction can be traced from core to core.

To explain this, we altered the connections in fig. 6 a little by removing the resistances x and y, z being again the resistance of the return wire as before, and for simplicity's sake, we made b two sheets of tinfoil (fig. 7). By manipulating the battery key the following now occurs :-The inductance in z will for some very short time retard the charging or discharging current through it, and the consequence will be that B is charged momentarily through the tinfoil sheets, a and b (vide "Static Induction," by K. Gulstad, C.E.). This induction does in fact appear in the interval that elapses from the moment of inductive action between the cores, until this effect can be equalised through the earth, or perhaps better between the iron wires and the outside of the

cores.

It is very instructive to observe this induction on a mirror galvanometer. The spot of light moves at first sluggishly, comparatively speaking, for the charging current, and being very suddenly stopped when the discharge passes through the resistance gets a blurring appearance. What is it that causes the resistance that gives rise to this inductive effect; where is the resistance situated, and how great is it? These are questions which we cannot answer satisfactorily, as we have never had opportunity of testing a multiple core cable since we commenced these investigations, but we hope to be able to offer our readers an explanation that seems feasible.

Our view is this, that the resistance in the earth connection is partly of an electrolytic nature, and that it consequently is situated between the cores and the iron wires. Different circumstances seem to point to this, more especially the fact that the induction varies a good deal in strength and appearance, according as the copper or the zinc pole of the battery be connected with a in the case when z is representing water. From this it would also appear that a charging current, like any other current, must overcome a certain inductance, caused by momentary polarisation; we may here refer to Sir William Thomson's experiments with semi-conductors, amongst which water in certain respects must be included, and where he points out that these semi-conductors after a little time only are able to act as conductors for static electricity. It may also be of interest to quote Mr. Latimer Clark's words in reference to the effect of mutual induction between cores in submarine cables-vide, the Society of Telegraph Engineers' Journal, Vol. iv., 1875 :

"It is true that in the Wexford cable, with instruments of the most sensitive kind, I have witnessed a sensible induction from wire to wire. That current, I imagine, was caused in the manner I have indicated that the sudden positive electrification of one of the wires threw off a small quantity of positive electricity from its exterior, and that electricity, before it could distribute itself from the cable into the sea would momentarily raise the tension of the whole core of the cable, and cause an inductive discharge."

In another issue of the REVIEW we hope to explain other causes for the static induction between cores in multiple core cables, and also to describe the means by which the problem of overcoming these disturbances has been solved.

Jerry Builders and Lights.-A contemporary, i referring to a certain exhibition in its last issue, states tha the Town and County Electric Light Company showe "primary batteries and incandescent lamps for lighting private houses that will last three months without attention This is distinctly a libel on jerry builders, whose honse generally last longer than three months.

MAY 8, 1891.]

THE DEPTFORD CENTRAL STATION.

THERE are probably few undertakings exciting greater interest at the present moment than the Deptford central station. This is reasonable, for since electric light was first used it is the largest and most important scheme that has been proposed. Mr. Ferranti's plan of generating and transmitting high pressure currents is too well-known to need recapitulation here. It must be said, however, that the work has been executed with a boldness that is startling, and we cannot withhold praise from the manner in which almost insurmountable obstacles have been overcome. But Mr. Ferranti had a "free" hand, and was backed up by a fearless and confident body of directors, whose downright honesty of purpose, despite our adverse criticism, we have never impeached. If success was ever deserved it is in the case of Mr. Ferranti and the London Electric Supply Corporation.

Deptford central station is conveniently situate at the riverside near Greenwich Hospital. It is the generating station of Mr. Ferranti's system. Here the large currents are generated and transmitted to the four distributing stations in London, of which the Grosvenor and Trafalgar only are at present in working order. The Pimlico and Blackfriars are in course of construction.

In this article it is intended, with the aid of illustrations, to give some slight notion of what the Deptford station is like. It must be understood, however, that it is by no means in a finished state, for, as will be noted elsewhere, the great 10,000 H.P. dynamos are not yet completed, and probably months will elapse before they are ready. Our readers, however, will be able from the following particulars to appreciate the greatness of the concern.

The station is a large building of 210 feet in length and 195 feet in breadth, the height of the main building is 100 feet, and chimney shafts 150. The overhead travellers, the huge planing machines and lathes, give one more the impression of a big workshop than a generating station. This is occasioned by the fact that the whole of the plant is made on the premises, a matter with which we will deal later.

The two large dynamos shown in the illustration are of 1,500 H.P. and generate current at 10,000 volts. The diameter of each armature is 13 feet, and the height of the machine from the bed plate 14 feet 6 inches. Each machine possesses 48 armature bobbins, which are copper strips wound on a core of insulated brass strips, and connected in pairs. New ones can be replaced on the armatures in a few minutes, a characteristic which is found in all the Ferranti dynamos. On the faces of the field magnets are fixed two ebonite caps fastened one inside the other with Ferranti compound. Each of these caps is tested to 20,000 volts, they thus prevent sparking from the armature to the frame of the machine.

Special attention has been paid to the effective lubrication of all parts of the machinery. At the end of the shaft of each 1,500 H.P. dynamo is a double set of pumping gear, which, drawing the oil from tanks underneath forces it into the bottom of the bearings, the oil afterwards passing through a filter back into the tank. A hollow chamber is fixed round the bearings of the dynamos through which cold water circulates freely.

By means of special machinery the field-magnet frames of the dynamo can be drawn back for cleaning in less than ten minutes. The exciting machines (Allen & Kapp dynamos) are placed at the rear of the large dynamos; they are of slow speed, 200 revolutions, 50 volts, and 400 ampères each. By means of special switches they can be connected to any of the dynamos, or, run in parallel, amperemeters and voltmeters on each machine indicating the exact state of each exciter.

It may be a fit period to mention that all machines are now run in parallel. Formerly when changing the load at the Grosvenor from one dynamo to the other the lights blinked, which caused customers to complain. Since, however, running in parallel, the load is transferred without the least flicker.

We give a front view of the engines driving the large dynamos. They are of 1,500 H.P., and, as will be seen, are of the vertical marine Corliss type, made by Hicks Hargreaves & Co., Bolton. The driving power is transmitted by 40 5-inch ropes on a flywheel of 22 feet diameter, weighing 60 tons. The stroke of the engines is 4 feet. The low pressure

591

cylinder measures 56 inches, the high pressure measuring 28. A special valve gear is employed on these engines; and by means of a wire rope the man in charge of the switchboard can open or close the valve at his pleasure. Swivel bearings are employed on all the dynamos and engines.

Although the corporation intended that the Grosvenor station was ultimately to be a distributing centre, current was generated there for some time. Owing, however, to the Courts issuing an injunction, the generating plant was stopped and removed to Deptford. This is shown in one of the illustrations, and is driven by two new 700 H.P. engines. The two dynamos of 625 H.P. are of slow speed, and each gives a pressure of 2,500 volts, which is converted by four 150 H.P. transformers to 10,000 volts. A Siemens exciter is used to each dynamo, a small Kapp exciter being available in case of accident. As in the other dynamos a special opening gear is attached-in this case worked by hand-one man dividing the parts in five minutes.

The engines are fitted with Corliss valves, with high and low pressure cylinders built in tandem. The diameter of the flywheel, weighing 35 tons, is 24 feet, 17 ropes driving the dynamos. As in all other parts of the machinery, swivel bearings and brasses are fitted, and all bearings are lined with white metal.

The switchboard, indistinctly seen in the illustration, is placed upon a specially insulated platform. Here the current from the four dynamos is brought and passes through amperemeters, and then, by a switch with a three ft. break, on to the omnibus mains, which are connected with the London mains. Each of the 1,500 H.P. dynamos is connected in the same fashion, so that all may be at once placed in parallel. A separate ampèremeter on each London main indicates the amount of current passing through it.

The resistances for regulating the voltage are placed between the foundations of the engines and are worked by ebonite tubes running in slides. The fuses were originally placed in porcelain tubes, but these are being replaced by specially tempered glass, which are of greater convenience, as the exact state of the fuse is then seen.

The boiler house contains twenty-four 500-horse-power tubular boilers divided into four batteries of six boilers each. It is intended to place on the top an additional twenty-four, divided in a similar manner. This space at present is occupied by a tank which holds 800,000 gallons of water. Underneath the boilers is placed a forced draught engine to facilitate making steam rapidly in the case of a foggy day.

The most novel and curious of the illustrations is an armature ring for one of the 10,000 horse-power dynamos. Lying on the ground it measures thirty-five feet in diameter, it is made of cast iron in eleven pieces, and is to be fastened to the dynamo shaft by cast iron arms or spokes, each of which will be in turn secured to the shaft by a double milled steel ring shrunk on, whilst as an additional security 22 solid steel bolts, 6 inches diameter, each weighing when finished 12 cwt., passing through the outside of the armature ring, screw direct into a steel ring mounted round the centre of the dynamo shaft.

The armature and shaft when completed will weigh 225 tons, and the field-magnets 350 tons more. This is exclusive of the massive bed plate on which both the engines and field magnets will stand. One 5,000 H.P. engine is to be fixed at each end of the dynamo shaft; the armature thus being in the centre, and taking the place of the ordinary flywheel, which is entirely dispensed with; the bobbin holders are fixed in the same way as on the 1,250 machines, 132 coils being used. It may be mentioned here that five dynamos of this type, each supplying 200,000 lights, are to be built.

To give a slight idea of the extraordinary size of the engines to be employed for driving, the measurement from the ground to the top of the high pressure cylinder is to be 48 feet. The over-all dimension of the dynamo is 45 feet, and of this 16 feet is below the floor level. Yet, in spite of the colossal proportions, it is said the dynamos will be so easy of manipulation that they can be drawn apart for cleaning in less than five minutes.

To the left of the armature ring may be seen one of the finished shafts, which before turning weighed 30 tons, it now weighs about 20 tons, the diameter of the centre part measuring three feet; and there is a 12-inch hole from end to end of the shaft. It may here be mentioned that the steel

D