Bifilar pancake coil overunity experiment

[ayeaye]

The oscilloscope's probe capacitance 10 pF has an impedance 1.2 k at 13 MHz, so it should not significantly change the measurement of current with an 1 ohm shunt resistor. Also when there is a much greater capacitance in the circuit, it should also not significantly change the phase shift. But in parallel with coils with a very low inductance, it certainly changes the resonance. How much it impacts the measurements, sure depends on the circuit.

My two cents, add a 10 pF capacitor everywhere where is an oscilloscope measurement. Or measure the actual capacitance of the oscilloscope's probe, and use a capacitor with exactly the same value. Then when actually measuring with oscilloscope, remove the capacitor. That way the oscilloscope shouldn't significantly alter the circuit, and the oscilloscope's probe shouldn't change the coil's resonant frequency.

Apart from not using as high frequencies as 13 MHz.

[ayeaye]

At 13 MHz, a capacitance 10 pF in parallel with a 1 ohm resistor causes phase shift approximately -0.05 degrees  https://keisan.casio.com/exec/system/1258032649 , thus measuring current with an oscilloscope and a 1 ohm shunt resistor shouldn't significantly alter the circuit.

I don't see that neither the resistors nor the oscilloscope measurement can significantly alter the Tinman's circuit, because the capacitance 10 pF is tiny, but i have not done the calculations. One thing, when measuring the voltage directly on the load resistor and not between two points and the function generator's ground, i see how to calculate the error, how to do it between the resistor and the function generator's ground, it's more difficult, and i have no clear idea how to calculate it. So for easier error calculation, it's better to measure the voltage directly on the load resistor with a floating scope, but certainly then has to be measured that there is no connection between the oscilloscope's and function generator's grounds.

So based on that, the Tinman's power measurements like should be reliable, except the statement by Tinman that the overunity disappeared when he used carbon resistors. Whatever resistors have no significant inductance in that circuit as seen above, so that shouldn't significantly alter the circuit, not to talk changing 3 times overunity to no overunity. So something is very weird. But nothing is never absolutely certain of course.