Some means of charging the condenser which produces the oscillatory discharge is necessary. An induction coil is the most practical for the amateur.

The induction coil consists of a primary coil of wire wound around a central iron core and surrounded by a secondary coil consisting of many thousand turns of carefully insulated wire. The primary coil is connected to a source of direct current which also includes an interrupter to "make" and "break" the current in rapid succession. Every "make" of the circuit and consequent magnetization of the core induces a momentary inverse current in the secondary, and every "break" and corresponding demagnetization a momentary direct current. Normally, the induced currents would be equal, but by means of a condenser shunted across the interrupter the circuit when "made" requires considerable time for the current and magnetization of the core to reach a maximum value, while when broken the demagnetization and current drop are nearly instantaneous. The value of the induced electromotive force in a circuit varies as the speed at which the magnetic lines of force cut the circuit, and so the induced e.m.f. at "break" is thus rendered high enough to leap across a gap in the shape of sparks.

The formulas connected with induction coils depend upon conditions which are never met in actual practice and cannot be relied upon. To construct a coil of a given size, it is necessary to use dimensions obtained empirically. Therefore it is well for the amateur to stick closely to lines and hints which are given here or which appear in some up to date book on induction coil building.

For a long time the induction coil was an expensive, inefficient instrument, until wireless telegraphy demanded of it more rigid and efficient design and construction. It was the aim of manufacturers to produce the longest possible spark length with a minimum amount of secondary wire. As a result of this demand, wireless coils are now made with a core of large diameter and give heavier and thicker sparks. The secondary in this case is short and uses wire of large cross section in order to reduce the resistance and minimize the heating.

No one part of an induction coil may be developed to its maximum efficiency without seriously influencing and lowering the efficiency of the other parts. The following suggestions regarding the construction are given that they may prove a useful guide to the amateur coil builder. The parts will be considered in their natural order of construction.

Core.—Some experimenters not quite familiar with the principles of magnetism reason that if an induction coil were provided with a closed core as the transformer, the efficiency of the coil would be materially increased. But this is not so, for the magnetization and demagnetization of the iron cannot take place rapidly enough in a closed core when an interrupted direct current is employed in place of an alternating current.

The core of an induction coil is therefore always straight. For the same reason, it is never solid but is made up of a bundle of soft iron wires in order that rapid changes in magnetism may take place. The wires are always of as high a permeability¹ as possible so as to create a strong magnetic field. Swedish or Russian iron of a good quality is the best, as its hysteresis² losses are small. The smaller the diameter of the wire the less will be the eddy current losses and heating, but the permeability is also rendered less and the core will not be so effective, as the amount of iron is thereby decreased and the oxidized surface increased. No. 22 B. S. gauge wire is the best size for the average core.

Wires of a good quality may be purchased already cut to various lengths. To buy them in this form will save a great deal of the labor required in building a core. If the wires are not quite straight they may be straightened by rolling them, one at a time, between two boards. It is best to reanneal the wires in the following manner. Place them in an iron pipe and plug the ends of the pipe with clay. Then lay it in a coal fire until the whole mass attains a red heat. The fire is then allowed to die out gradually with the pipe and wires remaining in the ashes until cool. When cool remove them from the pipe and rub each one with emery paper until bright. After this cleaning, the wires are dipped in hot water and dried. They are then dipped in a good quality of varnish and allowed to dry again.

The varnish serves to interpose resistance to the eddy currents generated in the core and renders the losses due to this cause much less. A strong paper tube having an internal diameter equal to the diameter of the finished core is made by rolling the paper on a form and cementing with shellac. When perfectly dry. the tube is removed and the wires tightly packed in it. The following table gives the core dimensions for practical coils of different sizes.

Primary Winding.—The ratio of the number of primary turns of an induction coil to the number of secondary turns bears no relation to the ratio of the primary and the secondary currents. It has been found in practice that two layers of wire wound tightly on the core constitute the best primary. The primary should always be thoroughly shellacked or covered with insulating varnish. Since there is almost no ventilation in the primary the wire must be large enough to avoid all heating. A table containing the various sizes of primary wires is given below.

In large coils, the inductance of the primary causes a "kick back" and sparks are liable to pass between the adjacent turns. For this reason, it is always well to use double cotton covered wire and to further thoroughly insulate it by soaking the primary and core in a pan of melted paraffin and allowing them to harden therein. Afterwards the pan is slightly warmed to loosen the cake of paraffin and the excess of wax removed by scraping with a blunt instrument so as not to injure the wires. Paraffin contracts upon hardening, and the proper method to impregnate a porous substance is to allow it to soak and become set in it upon cooling.

A good method of reducing the "kick back" and also the size of the condenser shunted across the interrupter is to form the primary of a number of turns of smaller wire in parallel, the effect being to give a conductivity equal to a single wire of large diameter and at the same time to make a more compact winding of the primary on the core. This method of winding is very desirable in large coils, as it reduces the cross section of the primary and allows the secondary to be placed nearer the core, where the magnetic field is the strongest.

The primary winding ought to occupy nearly the whole length of the core, since there is no gain in carrying the end of the core very far beyond the end of the primary, for most of the magnetic lines of force bend at the end of the primary and return without passing through the extreme ends of the core.

Insulating Tube.—The successful operation of an induction coil without breaking down when under strain depends largely upon the insulating tube which separates the primary and secondary. Hard rubber tubes are perhaps the best. A tube may be easily built up of several layers of 1/2-inch sheet hard rubber by steaming it so as to soften it and then wrapping it around a form. The tube should fit the primary tightly and be about one inch shorter than the core. After the tube is in place it is poured full of beeswax and rosin in order to fill all interstices and prevent sparks due to the condenser effect of the windings from jumping from the inside of the tube to the primary.

Secondary.—A coil used as a wireless telegraph transmitter must have wire of large cross section in its secondary so as to obtain a heavy disruptive discharge. Numbers 34 and 32 B. S. are generally used for small coils and numbers 30 and 28 B. S. for large coils. Silk covered wire is the usual practice, but enameled wire is coming into use. Cotton covered wire takes up too much space and has poorer insulating qualities.

Enameled wire is insulated by a coating of cellulose acetate, which has a dielectric strength of about twice that of cotton and takes up much less room than silk. There is, then, with enameled wire a great saving in space, and a greater number of turns may be placed on the secondary without increasing its mean distance from the core. The following table shows the comparative diameters of silk and enamel covered wires suitable in size for use on the secondaries of induction coils.

In winding enameled wire it must be taken into consideration that the insulation of enameled wire is rigid and has no give. Consequently, to allow for expansion, enameled wire must be more loosely wound than fiber or silk covered wire. The occasional insertion of a layer of paper in winding will give room for expansion and at the same time not add greatly to the diameter.

The length of the secondary is generally not much more than one-half the length of the core. Coils giving sparks up to 2 inches in length may be wound in two sections or in layer windings, but the layer winding is not recommended for coils giving sparks over one inch. It is best in a coil of this kind to insert an occasional layer of paper. The paper should be well shellacked or paraffined and be of a good grade of linen. It should project about one-quarter of an inch from the ends of the secondary as shown by the sectional drawing in Fig. 28.

This insertion of paper increases the insulation and renders the liability of sparks jumping from layer to layer much less, as is the case when the layers are very long.

The secondaries of large coils are made up of "pies" or "pancakes" from one-eighth to three-eighths of an inch in thickness. The "pies" are separated from each other by a triple thickness of blotting paper which has been thoroughly dried and then soaked in melted paraffin. In cutting the blotting paper, much labor may be saved if a metal template of the required size is first cut from sheet brass and then laid on the blotting paper, which is cut by scoring around the edge of the template with a sharp knife.

The "pies" are wound in a bobbin or form such as is shown in Fig. 29.

The disks or flanges are made of sheet brass and mounted on an arbor so that the form may be placed in a lathe or some other contrivance for revolving it. The core is beveled in order to facilitate the removal of a completed "pie" from the winder. The flanges of the winder are clamped against the core by two nuts placed on either side. The "pie" is removed by unscrewing one of the nuts and removing one of the flanges.

In winding silk covered wire it is first passed through a mixture of beeswax and rosin or a bath of melted paraffin. The excess of wax is removed by passing the wire through a slit made in a pad of paper or by rubbing against a piece of felt. Fig. 30 shows such a contrivance.

The wire passes from the reel over an ordinary spool down into the pan of paraffin, out of the paraffin, over another spool, and rubs against a piece of felt to remove the surplus paraffin. The spools are mounted with a screw and a washer so that they will turn without friction.

The wire is guided, when winding, by the fingers. If it is wrapped with a piece of felt and held between the thumb and forefinger it will run without friction and not cut the fingers. It is necessary that the wire should be closely watched for kinks, etc. which would cause the wire to break. Oftentimes the wire is broken but is held together by the insulation. Therefore each "pie" should be tested for continuity when completed. This is best accomplished by means of a galvanometer and battery. All imperfect "pies" should be rejected, as one of them would cause serious trouble if embodied in the coil. In soldering the secondary wires, acid must not be used as it soon corrodes the fine wires. Rosin is the best flux for this purpose. When building a small coil with a "layer" winding it is absolutely necessary that the wire should be wound on in smooth even layers. In a built-up secondary having "pies" not greater than 1/4 inch in thickness such great care is not necessary.

Fig. 31 shows the methods of connecting up the pies or pancakes. In A, the inside of one section is connected to the outside of the next, and so on. The maximum voltage which can exist between the adjacent sections in this case is equal to the e.m.f. generated by one "pie" and is equal throughout. In B, the coils are connected alternately inside and out. The voltage ranges from zero at the points where they are connected, to a value equal to twice the e.m.f. developed by one section. It would seem that there would be a saving in insulation space of one-half in the first case, but it is not so since the connecting wire passes between the "pies" and therefore the insulation must be twice as thick or exactly equal to that in the second case. The latter method (Fig. 31 B) is the best and most convenient. When the "pies" are connected in this manner the current must flow through alternate sections in opposite directions. To accomplish this it is not necessary to wind every alternate coil in an opposite direction, but merely to turn them around and connect them with the direction of their windings reversed as shown by the arrows and the bevels in Fig. 31. The connections between the sections must be very carefully soldered.

After the secondary is assembled the coil should be placed in a tight receptacle or tank containing melted paraffin. The tank is then connected to an air pump or aspirator and the air exhausted. The diminution of pressure causes any air bubbles in the windings to expand and be pumped out. After standing a while, the pressure of the atmosphere is readmitted and the place of the bubbles will be occupied by paraffin which has been forced in under pressure.

Mounting.—A coil for wireless work is best mounted as shown in Fig. 32 and used with an independent interrupter. The coil may then be placed under the operating table or on the wall, out of harm's way, and the interrupter on the table, where it is handy to the adjustment of the operator.

The case is simply a rectangular hardwood box large enough to contain the completed coil. Two binding posts mounted on the side of the box connect with the primary winding and two on the top of the box lead to the secondary terminals. The box is filled with boiled oil or melted paraffin and sealed up by screwing on the lid. If desirable, the secondary binding posts may be mounted on the top of a short piece of hard rubber rod as illustrated in the drawing.