Fig. 1.
C is the core, consisting of a bundle of soft iron wires as fine as can be obtained. The greater the subdivision of the core the quicker will it respond to the magnetizing current in the primary coil, and lose its magnetism when the current ceases. It has another advantage, in that the disadvantageous eddy, or Foucault currents, are lessened, which fact, however, is of not enough importance to need extended consideration.
Many coil-makers saturate the core with paraffin or shellac, which is of slight benefit. This core is wrapped in an insulating layer of paraffined paper or enclosed in a rubber shell, there not being any great necessity to use more than ordinary insulation between the core and the primary coil.
In the majority of induction coils or "transformers" used in the alternating current system of electric lighting, the iron cores form a closed magnetic circuit. A closed magnetic circuit in a Ruhmkorff coil could be obtained by extending the iron core at each end and then bending and securing the ends together, forming, as it were, a ring partly inside and partly outside the coil. But although the inductive effects would be heightened and less battery power required, the slowness of the circuit to demagnetize would alone be detrimental to rapid oscillations of current.
There would also be a loss from a greater hysteresis (energy lost in the magnetization and demagnetization of iron). A core magnetizes quicker than it demagnetizes, and the latter is rarely complete; a certain amount of residual magnetism remains, hysteresis being strictly due to this retention of energy (Sprague). Hysteresis shows itself in heat, but must not be confounded with Foucault or eddy currents. The latter are corrected by subdividing the metal, but the former depends upon the quality of the metal, and increases with its length.
Moreover, a coil with a closed magnetic circuit requires an independent contact breaker.
In most of the alternating currents used in lighting their rapidity of alternation is but one hundred and twenty-five periods per second. As in the simple electromagnet, the proportions of diameter and length of the primary coil and core will determine its rapidity of action. A short fat coil and core will act much quicker than a long thin one. But on a short fat coil the outside turns would be too far removed from the intensest part of the primary field. A good proportion of core length is given in the following table:
| Spark Length of Coil. | Iron Core. |
|---|---|
| ¼ | 4″ × ½″ |
| ½ | 5″ × 10∕16″ |
| 1 | 7″ × ¾″ |
| 2 | 9″ × 1″ |
| 6 | 12″ × 1⅛″ |
| 12 | 19″ × 1½″ |
The primary coil P consists of two or not more than three layers of insulated copper wire of large diameter, being required to carry a heavy current in a 2-inch spark coil, probably from 8 to 10 amperes. In designing the primary coil a great advantage arises from using comparatively few turns but of large wire. Each turn of wire in the primary has a choking effect upon its neighbor by what is termed self-induction.
As the primary coil and core may be considered as an electro magnet, it may not be out of place to notice the rule governing such. Magnetization of an iron core is mainly dependent upon the ampere turns of the coil surrounding it—that is, one ampere carried around the core for one hundred turns (100 ampere-turns) would equal in effect ten amperes flowing through ten turns. Practically speaking, there would be certain variations to the rule, for one difficulty would arise in that the smaller wire used in conveying the smaller current would fit more compactly and allow more turns to be nearer the core, the active effect of the turns always decreasing with their distance from the core. And although a large current and few turns would not have so much self-induction, there would be trouble at the contact breaker, owing to the large current it would have to control.