It is handy to have an additional coil for testing strong currents, and as this may be combined in one instrument at a trifle additional cost, we will get some line wire (No. 22) and wind six or eight turns of it around the coil outside the other wire; one end of this wire will be attached to an additional binding screw placed between the others, and the other end to left binding screw shown. The coil thus prepared may now be mounted in position. Pierce the board dial and the wood at its back with a hole large enough for the needle spindle to pass through from the back to the centre of the dial. See that the thick end of the inside needle hangs downwards, then place the coil in the position it is intended to occupy, and note how far the needle spindle protrudes on the face of the dial. If this is too long, nip off the end and file it up taper and smooth until it will work freely in a hole in the needle guard, with all parts in their proper places. This being satisfactory, secure the coil in its place by sealing wax, or, better still, by two thin straps of brass, held by screws at each end, placed across the coil. Now clean the free ends of the coil wires, insert them under the nuts of the binding screws, fix the indicating needle on the end of the spindle outside, and see that it hangs in a vertical position with the inside needle when the instrument is standing on a level surface. Secure it in this position, screw on the needle guard, fasten on the glass face, and the instrument will be complete.
§ 77. Provided thus with an efficient detector, the fitter may proceed to test his work. In cases of
new installations, take the wire off the carbon binding screw of the battery and attach it to one screw of the galvanometer (on the intensity coil side), next attach a piece of wire from the other binding screw of the galvanometer (the central one) so as to place the galvanometer in circuit. There should be no movement of the needle, and in proportion to the deflection of the needle, so will the loss or waste be. If loss is going on, every means must be used to remedy it. It is of the utmost importance to the effective working of the battery and bells that not the slightest leakage or local action should be allowed to remain. However slight such loss may be, it will eventually ruin the battery. Let damp places be sought out, and the wires removed from near them. Bad or injured coverings must also be looked for, such as may have been caused by roughly drawing the wires across angular walls, treading on them, or driving staples too tightly over them. Two or more staples may be touching, or two or more wires carelessly allowed to lie under one staple. The wire may have been bared in some places in passing over the sharp edges of the zinc tube. The backs of the pushes should be examined to see if too much wire has been bared, and is touching another wire at the back of the push-case itself. Or the same thing may be taking place at the junction with the relays or at the indicator cases. Should the defect not be at any of these places, the indicator should next be examined, and wire by wire detached (not cut) until the particular wire in which the loss is going on has been found. This wire should then be traced until the defect
has been discovered. In testing underground wires for a loss or break, it will be necessary first to uncouple the distant end, then to disconnect the other end from the instruments, and attach the wire going underground to the screw of the galvanometer. A piece of wire must then be taken from the other screw of the detector to the carbon end of the battery, and a second wire from the zinc end of the battery to the earth plate or other connection. Proceeding to that part of the wire where the injury is suspected, the wire is taken up, and a temporary earth connection having been made (water main, gas pipe, etc.), and by means of a sharp knife connected with this latter, the covering of the suspected wire penetrated through to the wire, so as to make a good connection between this suspected wire and the temporary earth plates. If, when this is done, the needle is deflected fully, the injury is farther away from the testing end, and other trials must be made farther on, until the spot is discovered. Wherever the covering of the wire has been pierced for testing, it must be carefully recovered, finished off with Prout's elastic glue, or gutta-percha, and made quite sound. The connections with the earth plates very frequently give trouble, the wires corrode or become detached from the iron pipes etc., and then the circuit is broken.
§ 78. When the fitter is called to localise defects which may have occurred in an installation which has been put up some time, before proceeding to work let him ask questions as to what kind of defect there is, and when and where it evinces itself. If all the
bells have broken down, and will not ring, either the battery or the main go and return wires are at fault. Let him proceed to the battery, examine the binding screws and connected wires for corrosion. If they are all right, let the batteries themselves be tested to see if they are giving the right amount of current. This should be done with the quantity coil of the detector. Should the battery be faulty, it will be well to renew the zincs and recharge the battery, if the porous cell be still in good condition; if not, new cells should be substituted for the old ones. Should the battery be all right, and still none of the bells ring, a break or bad contact, or short circuit in the main wires near the battery may be the cause of the mischief. If some bell rings continuously, there must be a short circuit in the push or pushes somewhere; the upper spring of one of the pushes may have got bent, or have otherwise caught in the lower spring. Pulls are very subject to this defect. By violent manipulations on the part of mischievous butcher or baker boys, the return spring may be broken, or so far weakened as not to return the pull into the "off" position. If, the batteries being in good order, any bell rings feebly, there is either leakage along its line, or else bad contact in the push or in the connections of the wires to and from the push. There should be platinum contacts at the ends of the push springs; if there are not, the springs may have worked dirty at the points of contact, hence the poor current and poor ringing. It is seldom that the bells themselves, unless, indeed, of the lowest quality,
give any serious trouble. Still the set screw may have shaken loose (which must then be adjusted and tightened up), or the platinum speck has got solder on its face and therefore got oxidised. This may be scraped carefully with a penknife until bright. Or, purposely or inadvertently, no platinum is on the speck at all, only the solder. A piece of platinum foil should be soldered on the spot, if this is so. Or again (and this only in very bad bells), the electro-magnets being of hard iron, may have retained a certain amount of permanent magnetism, and pull the armature into permanent contact with itself. This can be remedied by sticking a thin piece of paper (stamp paper will do) over the poles of the magnet, between them and the armature. In no case should the fitter cut or draw up out of tubes, etc., any wire or wires, without having first ascertained that the fault is in that wire; for, however carefully joints are made, it is rare that the jointed places are so thoroughly insulated as they were before the cutting and subsequent joining were undertaken. To avoid as much as possible cutting uselessly, let every binding screw be examined and tightened up, and every length of wire, which it is possible to get at, be tested for continuity before any "slashing" at the wires, or furious onslaughts on the indicator be consummated.
In conclusion, I beg to record my thanks for the very generous assistance which I have received in the compilation of the foregoing pages from the electrical firms of Messrs. Blakey Emmot, Binswanger, Gent, Judson, Jensen, and Thorpe.
[15] It must be borne in mind that the negative element is that to which the positive pole is attached, and vice versâ (see ss. 8 and 9).