Bifilar pancake coil overunity experiment

[ayeaye]

Honestly, what i first thought after seeing the TinselKoala's video, was the following. First impressions may be wrong.

I did experiments with moderate frequencies, 20 kHz, something. And then i thought, they want me to abandon it all, and replace it with that instead. With frequencies several megahertz, just because exactly there supposed to be some indications of overunity, in spite that no one can never be sure of course.

And yes i'm well aware of all the difficulties of high frequency circuits and their measurements. Like when one uses ordinary resistors at several megahertz, one can get some outlandish results. Because they are inductive at these frequencies, and cannot really measure instantaneous current. And then these indications of overunity too turned out to be wrong, perfect unity with carbon resistors. So a single reason why to go there at all, like dropped. Why to go there at all, why to start from where it's the most complicated? Was it just an additional waste of time? Goes like against common sense.

[F6FLT]

How to measure RF voltages relative to the ground? Here is the simplest and more robust method (but limited to sine currents > 0.5V):

[ayeaye]

Capacitor? Yes, high capacity capacitors can be used instead of resistors, to measure high frequency voltages. The instantaneous currents can then be calculated from the change of voltage seen in the oscilloscope trace. No need for diode, it's about observing the instantaneous voltage on the capacitor all the time, not to gather over a long time, and then measure.

About measuring an instantaneous current in that way. The oscilloscope math is not capable of such calculations, except the most expensive ones perhaps. This is one reason why i took so much effort to show how to do calculations with a Python script, from the oscilloscope's waveform data. That one would not be limited by the oscilloscope's math capabilities, like one wouldn't have to add more circuit elements that all may cause an additional error, just to be able to calculate power.

It's possible, and can overcome the problem of inductive resistors, but better to avoid the multiple megahertz circuits altogether. Most oscilloscope probes are also said to be 100 MHz, but i doubt that they work well over 10 MHz. When the oscilloscope is 100 MHz, it supposed to accurately measure a 100 MHz sine, all the probes and resistors in it should be the ones that are proper for that. But that's sure at the limit of accuracy.

[F6FLT]

Most of ordinary resistors have a negligible impedance even at 30 MHz: inductance < 0.1 - 1 µH, capacitance < 0.5 pF (except wirewound resistors). Here they are not the problem.
Osciloscope probes are a real problem due to their high capacitance. They disturb the measured circuits. They can drastically change a resonance frequency.

[ayeaye]

Osciloscope probes are a real problem due to their high capacitance. They disturb the measured circuits. They can drastically change a resonance frequency.

Yes there may be up to 10 pF capacitance between the oscilloscope's probe and oscilloscope's ground. One should consider how much that can change the circuit at the given frequency. As i said, if there is no other way to be certain, then one way is to make a simple circuit, measure it, and then simulate the same circuit in LTspice. At that a simple circuit that certainly has no overunity, like a simple transformer instead of a bifilar coil, or such. Then see how great the difference is.

It's weird how power measurement is so great problem in electric devices. Electric devices can really be the most accurately and precisely measured, more than another phenomena.