Bifilar pancake coil overunity experiment

[tinman]

I agree with you Brad. There is a ground problem if the scope and FG are not floating and share a common ground, which is likely.

Another problem is the  transformer itself. If it is made with two wires wound together, there is a strong capacitive coupling and the so-called "isolation transformer" is in fact a near all-pass circuit.
Even if it is made of two well separated coils, for instance coils diametrically separated on opposite sides of a tore core, the primary/secondary capacity is still on the order of pF to tens of pF, which is too much at Mhz frequencies.

Moreover the transformer adds possible concerns, for example making the FG appearing with a complex impedance.

I therefore propose the following simplified version, see picture.

No isolation transformer.
Only one ground.
The output power is easy and direct to obtain.
The input power is a little more difficult because the input current measurement is now made on the hot side of the coil. Nevertheless, I think there will be far less experimental bias than before.

If a COP>1 is always obtained, then it becomes very interesting. I hope Parzman and Itsu will try again with this new configuration.

Yes,that test circuit will give you accurate results.
I did miss the capacitive coupling problem on the isolation transformer though--good pickup there.


Brad

[Void]

I therefore propose the following simplified version, see picture.
No isolation transformer.

Hi F6FLT. Trying to measure current accurately across a CSR when one end of the transmit coil is
not connected to anything I have found from experience has questionable reliability as well. You need the isolation transformer
to improve measurements, IMO, but significant capacitive coupling between coil windings on all transformers is always
there at higher frequencies, so that always impacts measurements. The two pancake coils are coupling a fair bit
by capacitive coupling in this arrangement as well, IMO.

This is why I have been pointing out that I think it is a very good idea to look for ways to double check measurements
on these types of circuits (should do this on all measurements in 'OU' circuit testing really), and I don't really ever trust in
the accuracy of these type of measurements especially when COP > 1 is being indicated. Chances are measurements are not telling the truth.

At any rate, Brad has emphasized what I had pointed out as well, that the phase measurements on scopes on
these type of circuits at higher frequencies can be unreliable. I would suggest people try moving their scope probe leads
all around after connecting up their scope probes when the circuit is being driven/powered, and if you see the phase shift
amount on the scope between waveforms changing, then you obviously can't trust the phase shift you are measuring.
Brad, did you try moving the scope probe leads around when you were measuring the phase shift > 90 degrees?

If someone thinks they are measuring a COP of about 2 or higher, then they should have enough 'excess energy' in the circuit to
self loop. If it fails miserably when you try to self loop the circuit, then you are probably looking at measurement error. 
Self looping is the great equalizer. An OU experimenter's essential reality check.  Omit this at your own peril.

[ayeaye]

Trying to measure current accurately across a CSR when one end of the transmit coil is not connected to anything I have found from experience has questionable reliability as well.

No, one end of the input winding not connected does not decrease reliability. Coil is what we measure. It sure has a strong capacitive coupling. When it is all capacitive, then there is no voltage on R1 at all, all the bifilar coil acts just as a capacitor. Sure no overunity in that case. This is why it is interesting and important to measure voltage separately on R1. When some current goes through the output winding, then the current in the circuit is caused both by the capacitive coupling and that current.

"You need the isolation transformer to improve measurements,"

No, isolation transformer has nothing to do with the capacitive coupling in the bifilar coil. As F6FLT pointed out, when both the oscilloscope and the function generator have a common ground, there is no need for isolation transformer, and removing that transformer greatly improves the measurement accuracy, as one more part of the circuit is removed that can be a source of errors.

Now the possible cause of overunity, the way i guess. The current induced in the secondary winding fills the capacitance, the current doesn't all go through all the winding, thus might not have a full Lenz effect, thus more current can be induced that fills the capacitance, than provided by the Faraday's law.

"Self looping is the great equalizer. An OU experimenter's essential reality check.  Omit this at your own peril."

No, not at all, quite the opposite. This self-looping will be just adding more components to the circuit that all are possible causes of errors. And self-looping doesn't really prove overunity, self-looping can occur also for other reasons than overunity. Self-looping may not last long, for how long should it last to show overunity? This just makes the experiment ambiguous instead of doing what should be done, measuring overunity, that is the input and output power of the part of the circuit that is considered to possibly have overunity. Self-looping is a very different task, it is engineering and not research. It can be done after the research shows a result of overunity that may be enough for self-running. But it's engineering not research.

I don't see no reason why the input and output power of that circuit cannot be accurately measured, having a good and well calibrated oscilloscope. The probes may give more error with higher frequency yes, thus swapping probes may slightly change the results and it may be done to estimate error. But some 1.5 MHz is not that great frequency for a good oscilloscope, including its probes, and it can be measured with sufficient accuracy.

That the measurements are complicated and must be carefully made, there is no doubt though.

[Void]

I don't see no reason why the input and output power of that circuit cannot be accurately measured, having a good and well calibrated oscilloscope.

I have mentioned some reasons why measurements on these types of circuits can be very tricky,
and can easily mislead a person if they are not very careful. Taking the attitude of being willing to question and
closely examine all measurements, especially when something out of the ordinary is measured, and in such a case try
to find ways to double check measurement results, is a very good idea IMO. This is my point of view based on lots of
experience doing measurements in these types of circuits. Good luck to all with your measurements. May you be fooled
no more than half of the time.

[itsu]

I was using a battery operated (so ground isolated) FG all the time as mentioned before.
I used also an isolation transformer to be compatible with TK his setup.


Anyway, latest setup from F6FLT:

Still using the 2th bifilar coil.

Math function cannot be used in this setup (1 math function at a time only for my scope).

Screenshot 1 shows the values:

CH1 = 1.786V
CH2 = 1.750V
CH3 = 868.8mV
CH4 = 44.4mA   (i added the current probe value (right before R2 in the diagram) to show the difficulty to correctly
                           measure the voltage (current) across R2 (CH1 - CH2) in this setup)

Difference between CH1 and CH2 is 36mV, but this is very debatable, see lateron.

Calculating:

Pin = CH1 * (CH1-CH2)/R2
Pin = 1.786 * (1.786-1.75)/1
Pin = 64.29mW

Pout = CH3²/R1
Pout = 0.8688²/20
Pout = 37.74mW

COP = 37.74/64.29 = 0.58

Using the current probe value of 44.4mA, then we get:
Pin   = 79.2mW
Pout = 37.74mW
COP  = 0.47


But.......

note the phase difference between CH1 yellow and CH2 blue, its 0, meaning pure resistive.
note the phase difference between CH4 green and CH2 blue, its -52° (plus 10 = -62°) which should be 0 as both are measuring the same current!!

So something is not right in this setup.

If i remove all probes (grounds) and only measure across R2 together with the current probe i DO get both signals
to be of the same value (45mA) plus of the same phase see screenshot 2.

thoughts?

Like if i besides using the current probe value (44.4mA),  i also use its phase offset (-62° again), i get as input:

Pin = 1.786 * 0.0444 * Cos -62°
Pin = 37.2mW


Which makes a COP of 0.98  hmm...


Itsu