Nikola Tesla: Colorado Springs Notes, 1899-1900 Page 419

the loading of the secondary circuit, and this alters the mode of oscillation. Also, shunting the secondary with capacitance (as in the diagrams of 18 September and 19 September Figs. 2, 3 and 4) alters the spectrum of the oscillation in comparison with that yielded by an oscillator with two oscillatory circuits. Configurations such as those shown in Figs. 5 and 6 of September 19th can be considered as typical Tesla oscillators with a loosely coupled third circuit consisting of the extra coil and capacitive load. Then the greatest voltage at the free terminal of the extra coil is obtained when the natural resonant frequency of this circuit (together with the ball antenna) is the same as that of the strongest component in the spectrum of the oscillator.

September 22-23

Having investigated the tapering secondary Tesla started making a new, 15 m diameter cylindrical secondary. The criterion that the weight of copper in the primary and secondary should be the same follows from the requirement of equal losses in the two windings (losses in the copper). This way of calculating the gauge of the primary and secondary conductors is applied in designing LF transformers, but for HF transformers it only provides a rough guide, for a number of reasons, e.g.: the current ratio may differ considerably from the turns ratio, skin effect is not taken into account, etc.

September 25

As he often did earlier, before finalizing a set up Tesla measured the inductance of the primary and the regulation coil which he usually used as an added, adjustable primary inductance. The value he obtained for L_2p differs from that obtained earlier (see July 17th) by the same method.

September 26

By this method the frequency of an oscillator is found with a help of a resonant circuit of known parameters. When its resonant frequency is adjusted to coincide with the frequency of the oscillator, the voltage across its terminals, estimated by the strength of the spark across an “analyzing gap”, is a maximum. Tesla says that the excitation must be “convenient”. Since he introduced regulation of the excitation by means of the small gap b, it is clear that “convenient” excitation was obtained with loose coupling. Loose coupling between the primary and secondary circuits of a spark oscillator ensures that the two frequencies which such an oscillator normally produces are very close. Up to a certain degree of coupling, Tesla's oscillator produces a single frequency. According to Fleming and Dyke⁽³¹⁾, with an ordinary spark gap the maximum coupling coefficient for monochromatic oscillation is around 0.05 (certainly less than 0.1), while with a rotary break producing pulse excitation a coefficient of up to 0.2 gives good results. With higher coupling coefficients three components are obtained, even if the primary and secondary circuits by themselves have the same resonant frequency.

September 27

True to the principle that measurements should be checked by calculation, Tesla calculates the inductance of the same coil using the formula for a coil of infinite length,