Page images

being in readiness, and the various switches closed, the alternator A is gradually speeded up until the voltmeter Vh indicates the required voltage of, say, 2,000 volts. This potential is then maintained constant for a definite period, usually from 20 to 30 minutes, or even longer, when the alternator is slowed down and finally stopped, the switches being left closed meanwhile. These are then

[graphic][ocr errors][subsumed][ocr errors][ocr errors][merged small]

opened, and the cable put to earth for a corresponding period, in order to dissipate any residual charge which might remain in it, and interfere with the subsequent galvanometer test. The ordinary insulation test is then repeated, and the result obtained should, if the insulation be perfect, correspond exactly with that obtained before the application of the alternating potential. If it be found to have fallen in value after being duly earthed, attention should first be given to the extremities, which, in view of possible surface leakage, should be prepared afresh in the manner previously described. Should this not have

even more.

the desired effect of raising the insulation under test to its previous value, the high voltage should be again applied, when, after a more or less protracted period, the insulation will probably give way as indicated by the flicking of the ammeter needle at B, and the sudden drop in voltage as denoted by the voltmeter Vh. The switch Sl should be immediately opened, and the alternator shut down, when the cable can be disconnected and the fault produced by the passage of the high voltage duly localised by one of the several methods to be described later.

As regards the testing voltage applicable in each particular case, there is no hard and fast rule determining this quantity, the general procedure being to subject the cable to an alternating voltage at least twice that at which it is intended to work in practice, for a time which, as before stated, varies with different authorities, and may be anything from ten minutes to an hour, or

This question of voltage and time limit is one which requires standardising, and it is to be hoped that some responsible body like the British Association or the Institution of Electrical Engineers will, at no very distant date, formulate a set of rules applicable to most of the cases met with in practice, and founded on past experience on the part of leading cable manufacturers and users, who, after all, are the individuals most competent to judge in such matters.

There is no doubt now that high voltages in electric light and power distribution are becoming so prevalent in this country that the ordinary megohmio results, as obtained by the direct deflection and other methods of insulation resistance measurement will go for nought unless accompanied by a corresponding guarantee of resistance to alternating currents at high voltages.

It may be imagined from a brief consideration of the facts that the actual power required for the application to an insulated cable of a test of this description is very small, owing to the fact that the secondary or high tension circuit is not completed; this, however, is not the case, on account of the self-induction of the transformer circuit, the output of the testing alternator being governed by the electrostatic capacity of the cable or wire under test. In the case of considerable lengths of fairly large cable immersed in water, the output required sometimes amounts to as much as 'ten or even twenty E.H.P. for a secondary voltage of from two to five thousand, and, in this connection, it is advisable to compute the power required before attempting to apply the test with what may prove to be insufficient plant for the purpose.

The aforementioned question of comparison between the direct deflection galvanometer tests before and after the stressing test, is a very important one, and should receive careful attention, inasmuch as an extremely slight drop in the insulation resistance, as recorded in terms of megohms per mile, indicates inherent weakness, and has often resulted in the ultimate breaking down of the insulation under a subsequent straining test, after an application of an hour's duration.

It is inadvisable to raise the testing voltage beyond a certain ill-defined limit, as it tends to strain the insulation unnecessarily, especially if, as is often the case in these days of cut prices, that insulation has been specially composed to suit a particular voltage ; it is far better, under the circumstances, to repeat the original dose or voltage, and maintain it, if necessary, for a longer period, rather than run the risk of permanently straining the insulation by subjecting it to a momentary current of considerably higher tension, a practice frequently resorted to in America as a time-saving device.

In connection with the high pressure testing of insulated cables and dielectrics generally, where a high pressure alternating current is required capable of gradual adjustment over a wide range, the attendant difficulties in the way of varying the speed of the primary testing alternator may be obviated by the adoption of a special type of regulating transformer designed for the purpose, and manufactured by Messrs. Cowans, Limited, of Salford, Manchester. These useful accessories are wound for pressures of from two to three thousand volts at a normal output of 20 horse-power, and are rendered capable of supplying any required voltage between zero and these limits by a simple rotary motion imparted by a convenient handle and worm-wheel attachment. A general view of the apparatus is shown in the accompanying illustration. If higher voltages be required, this apparatus

can also be used in conjunction with a second step-up transformer, preferably of the oil-insulated type.

The main principle of the apparatus consists in a movable shuttle" or H armature capable of rotary motion around its axis as a centre. This is encircled by a fixed, ring-shaped magnetic core; around this ring is wound part only of the secondary circuit, whilst the shuttle carries the remaining portion, together with the whole of the primary section. It will thus be seen that, by varying the relative position of the shuttle and fixed core through an angle of 180 degs., the secondary voltage can be raised from zero to maximum, or vice versa.

A secondary feature in the desigr of this transformer consists in short-circuited or shading" coils, which are wound on the shuttle at right angles to the active wind

[ocr errors]
[graphic][merged small][merged small]

ing, and serve to prevent an excessive drop of potential in the secondary winding, due to the magnetic flux set up between the two portions, fixed and movable, of the secondary circuit. By means of this latter device, the total drop in potential is reduced to 7 per cent., a very satisfactory and therefore negligible figure, especially in testing operations, where the voltage has only to be maintained for a limited period.

I am indebted to Messrs. Cowans, Limited, for the above details, which serve to explain the principle of a very useful invention.

If, as in the cases mentioned above, the cables under test be immersed in a tank, and, in fact, in many similar cases, it is more than probable that circumstances compel the placing or setting up of the testing instruments at some distance from the cable whose insulation resistance is to be determined. When this is so, a length of well-insulated lead must be run from the instruments to the extremity of the cable, and also a secondary lead, which need not necessarily be insulated, for the earth connection. Both purposes are admirably served by a length of concentric cable of small cross-sectional area, or a single insulated conductor armoured with a spiral lapping of iron or steel wire. The ends of such a lead require to be equally as carefully prepared as the actual extremities of the cable itself, and a preliminary test must be taken on the lead, its far extremity being left free, before it is connected to the cable, dl and the electrification readings being noted in exactly the same manner as if the cable itself were under test. The readings thus obtained, which should, if the lead be of good quality, and of no very great length, amount to but a few degrees, are deducted from the ultimate deflections obtained through lead and cable combined, the remainders being the true reading due to the cable alone.

Departing for the time being from the subject of insulation or high resistance measurement, let us proceed to examine the methods of dealing with the opposite extreme, or

(6) Low Resistance Measurements, which are similarly beyond the scope of the ordinary Wheatstone bridge method of resistance measurement.

The only practicable methods of dealing with resist

« PreviousContinue »