Bifilar pancake coil overunity experiment

[TinselKoala]

No i cannot, i have no induction cooker. Also using light bulbs is an improper way to measure power, it is very difficult to measure their brightness, and in these videos they don't even try to do that. If you did use heating elements instead or such, you could at least measure their temperature, and from that you could calculate power, though whether they are in the air or in the water, what is the temperature of the air or water, whether the wind blows or water moves, all should then be measured and included in the calculation. No matter how convincing something may look, without a proper measurements the experiments are worthless for research. I mean what you do may be great, all i'm saying is that you lack the necessary rigor.

Could you please explain how the measurements you took on your circuit represent "input power" and "output power" ? Your oscilloscope measures voltages. How do you get from voltages to power values? 

(I know how this is done, normally; I am asking you to explain precisely how you have done it in this case with this circuit as you showed, in plain English not multiple lines of code. As someone who obviously has some programming experience surely you know that one of the worst tasks a programmer can be faced with is debugging _someone else's_ uncommented code.)


This is the circuit you are using, right? Just for review purposes, what's the frequency and duty cycle of the 555 pulse train? Do you think a different signal generator would work the same here? And are you specifying the 2sc945 transistor as necessary or can other NPNs be used? I have a handful of 945s on hand but it's always nice to test alternatives. I'm also concerned about your "bifilar" coil of coaxial headphone cable. It's not really equivalent to the bifilar winding of Tesla's patent. If I were to construct and examine this circuit, I would have to construct a coil like yours, as all of my bifilar coils are true Tesla bifilars (a rarity). Fortunately I have some RG174/U laying around. Do you think that would make a suitable coil for this setup? Or would the true TBF be better? Regardless, if the claim is that there is something special about the particular coil that allows the measurements to result in OU... then other coil types should be compared to see if that is in fact the case.

[ayeaye]

Could you please explain how the measurements you took on your circuit represent "input power" and "output power" ? Your oscilloscope measures voltages. How do you get from voltages to power values?

In a pulsed circuit, each cycle has two parts, when the pulse is on and when the pulse is off. The coil and the resistor R3 there form an LR circuit. When the pulse is on, the transistor is open, and the current from the power source goes through that circuit. This is the input part, because the current from the outside power source goes through the coil and is consumed by the coil. When the pulse is off, the transistor is closed, and almost no current goes through R2.

(A very small current that goes through there (during the pulse off) goes to the opposite direction to the current through R3, that decreases it, so not considering that decreases the calculated power in R3, thus it can be disregarded and considered that no current goes through R2 during that part.)

Thus all the current that goes through the R3 in that part (pulse off) is the back-emf generated by the coil, that comes from no outside source. This is the output part. The power during the first part is the power used by the coil, and the power during the second part is the power generated by the coil, this is how the input and output powers are calculated.

Calculating the power used by the coil during the first part (the input power), is done by calculating the power used during the first part in the LR circuit, that is the current going through the R2, multiplied by the voltage on the LR circuit (on the R3), that is Ir2 * Vlr (Ir2 is Vr2 / R2). And then subtracting from that the power used during that part by R3, which is Vlr * Vlr / R3. Thus, all the equation is Vlr * Vr2 / R2 - Vlr * Vlr / R3. This was calculated by the Python script. I don't see how that calculation can be made without a Python or other code, but i don't know all the software that there may be for such calculations.

Calculating the power used by R3 during the second part (the output power) is easy. It's Vlr * Vlr / R3.

The circuit is that yes. Later (after the experiment) i thought that the circuit on the figure below would be better, which is the same with the R2 resistance reduced to 100 ohms. I estimated that a larger resistor there is not necessary, no damage will be caused to the transistor with R2 100 ohms. Having that resistor smaller enables more precise measurements, as the only thing that changes more in its voltage, is the upper part of the trace.

The frequency was 30.303 kHz. The 555 timer doesn't show the duty cycle, so it can only be seen from the oscilloscope traces and the graphs. As it's seen, the duty cycle was close to 22%. I found Arduino to be better, and i think that any signal generator would do, but i used 555 because it's cheap and more people can use it. I chose the c945 transistor because it's fast, but i think that any fast small NPN transistor would do.

I used coaxial cable in the coil, because i think that the coaxial cable has more capacitance than a pair of wires, as the other conductor is all around the first, and thus the area between the conductors is greater. So if there is no headphones cable (which i found that they in some places sell under the name "audio cable"), then it may be made of a coaxial cable. At that, my cable had a pair of coaxial cables, so i made of these the equivalent of two bifilar pancake coils, one on another, that i connected in series. The coil had 17 turns, two in series had the equivalent of 34 turns.

I think that all that matters is that it's a bifilar coil, with as great as possible capacitance, because i guess that it's this capacitance that may cause overunity. Thus it can be other coils, like bifilar pancake coils, the ones with more turns and more capacitance would be better i think, all i had was this headphones cable and i couldn't make a bigger coil. Also it can be other bifilar coil, such as a Rodin interference coil, that have considerable capacitance between the two bifilar windings. Maybe even a Bedini SG bifilar coil, but without the core, as i noticed that the coil having any core makes the back-emf narrower and seems to decrease the effect.

I think also that the possible overunity effect is greater when the frequency is closer to the resonant frequency. Because when the effect is caused by capacitance, then it depends on the self-oscillation ringing, which is relatively greater, the closer its frequency is to the duty cycle. I may be wrong in that, it's just what i thought. With my 555 timer i couldn't go over 30 kHz with the 20% duty cycle. What concerns the duty cycle, i think that it should be great enough for all duration of the back-emf, as cutting the back-emf i think is not good.

As i have only a small coil, i noticed that its self-oscillation frequency was somewhere near 1MHz, which is very high, i think a bigger coil with a lower self-oscillation frequency may be better.

[TinselKoala]

I've redrawn the circuit to aid in laying it out. Please check my redraw and let me know if I've made any errors.

[ayeaye]

I've redrawn the circuit to aid in laying it out. Please check my redraw and let me know if I've made any errors.

Yes that's correct.

I made R2 1k because i was afraid of possible high peaks from the coil, but there appeared to be none, so it can be made smaller.

The pinout of the c945 that i have was not as written on the datasheet, i found that out by testing. There apparently are these with two different pinouts. I don't know what is yours, mine was as on the figure below, that is, emitter is where the blue jumper wire goes, collector is where the brown wire goes, and base is in the middle.

[TinselKoala]

EBC is more usual for the TO-92 package, that is true. I haven't tested my handful yet. I think I have some from 2 different manufacturers.

I was going by this datasheet: