Nikola Tesla Books
of capacity of the elevated sphere with the rise from the ground would be greater than before found. The absolute rate of increase can be approximately estimated from the period of vibration. As before found the special coil with 400 turns had a self-induction of 44,772,000 cm. Now, however, with 404 turns this would be increased about 1% so that the self-induction would now be 44,772,000447,720 } 45,219,720 cm. But to this should still be added the self-induction of one turn of the secondary and wire leading to the ball and also the wire leading from the bottom of the special coil to the first turn of the secondary. The total length of these three wires is 240 feet and this is about 12% of the total length of the wire in the special coil which is 2854 feet. But inasmuch as the one secondary turn was very close to the primary and inasmuch as the other two wires were not coiled up, the self-induction of these wires was comparatively small, estimated a little over 200,000 cm so that the total self-induction was with fair approximation: 44,500,000 cm, or about 0,0445 henry (calculated). Now, with a ball of 38.1 cm capacity the period of secondary would have to be:
$! {T_{s} = {{2 \pi \over 1000} \sqrt{{38.1 \over {9 \times 10^{5}}} 0.0445}} = {{2 \pi \over {3 \times 10^{7}}} \sqrt{1695.45}} = {86.19 \over 10^{7{}}}} $! and
n = 116,000. This would be ignoring the internal capacity of the coil itself and its effect in slowing down the vibration. Now this capacity can be approximately estimated as well as the absolute increase of the capacity of the sphere from the primary vibration.
Taking the figures with the ball at a height of 33.66 feet from the ground, the primary vibration was:
$! {T_{p}' = {{2 \pi \over 1000} \sqrt{0.15264 \times {118,600 \over 10^{9}}}} = {26.7 \over 10^{6}}} $! or $! {267 \over 10^{7}} $!
This vibration is evidently slower than one of period Ts and it will be easily seen that
$! {{T_{s} \over T_{p}'} = {86.19 \over 267} = {\sqrt{{38.1 \over C + 38.1}}}} $!
where C is the internal capacity of the coil. Assuming for the present this capacity as not being distributed along the coil but in one place, we get
$! {C + 38.1 = 38.1 \times \left({267 \over 86.19}\right)^{2}} $! and $! {C = {\left[{\left({267 \over 86.19}\right)^{2} - 1}\right]} \times 38.1} $! cm.
Following this up we get for C value C = [(3.098)2 - 1] x 38.1 = (11.597 - 1) x 38.1 = 10.597 x 38.1 = 403.75 cm.
This is not the actual capacity of the coil but that ideal capacity by which the sphere should be increased to give the vibration of the primary.
We can now write
$! {T_{p}' = {{2 \pi \over 10^{3}} \sqrt{0.0445 \times {(403.75 + c) \over {9 \times 10^{5}}}}}} $! . . . . . . . . . . 1)
216
October 9
He continues to improve the method for variable capacitance measurement of a sphere by varying the height. Last measurements he performed on Oct. 5, but he did not provide the calculated values.
After that he was improving the apparatus as a whole, and changed the "special coil" connection method. "Weaker Link", which was a main characteristic of the change, indicated immediately its good characteristics. Less excited "special coil" was easier to be adjusted in resonance, because there were no current streamers. The effect of parasitic capacitance was reduced to mainly distributed capacitance of the "special coil". Prior to the sphere capacitance variation measurement related to height variation, he determines the "special coil" distributed capacitance. He assumes that the circuit with a sphere resonates at frequency Ï0, which is defined by oscillator primary circuit so that it could be written: Lp1Cp = Lsc(c + C), where Lp1 and Cp are total inductance (with the regulating coil included) and primary oscillator circuit capacitance, respectively. Lsc is ''special coil'' inductance (with inductances connections included). C is distributed or parasitic capacitance of the "special coil", and c is the sphere capacitance. At some other height only the sphere capacitance is changed in the circuit of the "special coil". In order to achieve the resonance between the system and the signal from the oscillator, Tesla varies the inductance in the primary circuit. When the generator frequency is equal to the resonant frequency of "free coil" circuit, according to Tesla, the following equation could be written: Lp2Cp = Lsc(c' + C).
Division of this equation by the previous one results in: c' = Lp2(c + C) - C, which corresponds to Tesla's equation from which c' is determined. Due to a calculation error when C was calculated, Tesla's various results for the sphere capacitance are higher by approximately 10%, but this does not essentially change the main conclusions. When calculating the coil distributed capacitance* he uses the relation LpCp = Lsc(c + C) for the sphere at a height, where he considers that its capacitance is close to the theoretical value for the remote sphere.
When commenting on results, he mentioned that this assumption could be the cause of the error.
* Under the term coil distributed capacitance Tesla considers the total capacitance among coil turns. Here he uses another definition "internal capacitance", which is similar to the one which is normally used even today.