Bifilar pancake coil overunity experiment

[F6FLT]

...
Would you clarify little more precisely how you connected the bifilar (copper strip) coil? Are the other two ends
of the coil left floating independently or they are connected together and left floating?
From your measurements, if I understood it correctly, could it be deduced that your bifilar coil as you connected it
behaves like a series LC circuit (at 1.5 MHz) resonance?
I mean as if a series (lumped) LC is driven at one end from the "hot" output of a generator and its other end is
connected to a 47 Ohm resistor across which your scope probe is connected and the ground of the generator is
connected to the cold end of the 47 Ohm (and to the ground of the scope of course).
...

Hi Gyula,

It's not a bifilar coil but monofilar. My first tests were made with a bifilar coil that gave bizarre results, so I simplified for a monofilar coil (that gives the same bizarre results).
My method is to go towards the most elementary phenomena with the simplest possible configuration. Now I just want to qualify the properties of a coil having a very high distributed capacity (> 200 nF for mine, like this one but wider : https://www.amazon.co.uk/6mmx30-Copper-Horse-Irmband-Adhesive/dp/B016Y58PPG/ref=sr_1_5 ).

All your other considerations are perfectly correct. I use the schema given above by ayeaye.

My main question is: why is there a resonance at 1.5 MHz while L and C are obviously too big values to reach this frequency?

(It's may be interesting to know that with my bifilar coil of same length, which was 0.6 cm wide instead of 1.5 cm for the monofilar and not so tight turns, the resonance frequency was 2.3 MHz when both wires are not connected together and no resonance when in series).

[itsu]

Hi F6FLT,

My main question is: why is there a resonance at 1.5 MHz while L and C are obviously too big values to reach this frequency?

Why do you think L is to big here?  Did you measure it?  What was the value?

According to http://www.1728.org/resfreq.htm , the inductance at 1.5Mhz with 200nF is 0.056uH (56nH).

According to https://www.emisoftware.com/calculator/stripline/ , the biplanar inductance of a stripline
with Width=15mm, seperation=1mil and length=30m is 6.384e-8H = 0.06348uH = 63.48nH

I know that the biplanar configuration is not like your coiled up strip, but it confirms that striplines are
known for their low inductance as you will know.

Itsu

[ayeaye]

Channel 1 measures vs, and channel 2 measures vr. Typical calculation of power from voltage, nothing complicated. Divide by 1000 as voltages are in mV, to get mW.

p = v * v / R

The input part.

vl = vs - vr
pl = vl * vl / R / 1000
e += pl

And the output part.

pr = vr * vr / R / 1000
e += pr

The following things are important to consider.

* We calculate input and output power only for the coil.

* We calculate instantaneous power for the input part, as the power consumed by the coil.

* We calculate instantaneous power for the output part, as the power generated by the coil, that was consumed by the load resistor.

Because the voltage on the coil has to be calculated from the voltage on the signal generator and the voltage on the resistor, both the voltage on the resistor and the signal generator voltage, have to be measured.

Unless it is known that during all the input part the signal generator voltage didn't change, and this voltage is known. Then only one channel, the voltage on the resistor, has to be measured.

In the case above, the signal generator voltage did change, thus two channels had to be measured, and included in the calculations.

To fork Trinket Python code make a small unimportant change, like add a space to the beginning of a comment, which after making a fork can be changed back. When making a fork after that, the link provided will be different from the page where the original code is. This is inconvenient, with otherwise so nice tool.

[F6FLT]

Hi F6FLT,


Why do you think L is to big here?  Did you measure it?  What was the value?

According to http://www.1728.org/resfreq.htm , the inductance at 1.5Mhz with 200nF is 0.056uH (56nH).

According to https://www.emisoftware.com/calculator/stripline/ , the biplanar inductance of a stripline
with Width=15mm, seperation=1mil and length=30m is 6.384e-8H = 0.06348uH = 63.48nH

I know that the biplanar configuration is not like your coiled up strip, but it confirms that striplines are
known for their low inductance as you will know.

Itsu

Hi Itsu

The thickness of the winding is 20 mm and there are about 95 turns. We have 20/95=0.2mm for the thickness of the strip, including conductor + insulation. Reporting width=15mm, separation=0.2mm, length=30m in your site, we get an inductance of 125 nH.

This value being not so far from a resonance at 1.5 Mhz, I decided to calculate more precisely the capacity for which I had given the minimum value of 200nF.
I had measured 55 nF for the bifilar coil (tape width 6 mm, same length). So my monofilar coil  (width 15 mm) would have a capacity of 55*15/6 = 137 nF. But the monofilar coil has turns much more tight, the thickness of the winding is 2 cm while it is 5 cm for the bifilar coil. The capacity of a plate capacitor being insersly proportional to the thickness of the dielectric, this leads us to a capacity of 137*5/2=342 nF.

This give us a frequency of 770 KHz, about the half of what is observed. Nevertheless you are right. Taking into account the margins of uncertainty, we are in the correct order of magnitude, unlike what I thought. I must have been mistaken by my habit of using capacitors of only a few hundred pF to obtain resonances at about 1 MHz which is the AM MW radio band, and with much smaller coils (in volume) but probably not in value due to the ferrite core.... 

So there is no mystery here. This flat capacitive coil works strictly like a normal coil tuned with an external capacitor.
Probably we can explain the absence of propagation of a pulse in the line because we are really in the approximation of quasi-stationary states: the capacity concerns the whole winding as a block, and each coil turn shares the same magnetic field, so that at a given time, currents and voltages do not depend on the position that is measured in the circuit.

So now I have to go back to the bifilar coil.

Thanks

[itsu]

F6FLT

concerning the pulse, i am not sure if one could use such a Time Domain Reflection setup:
https://en.wikipedia.org/wiki/Time-domain_reflectometer on a single wire.

Its used on transmission lines (coax etc.) which have 2 wires.
The reflected pulse, when shorted, comes back as a negative pulse and with an open at the end as a positive pulse
(hmmm, could that be considered as a single line?   I don't think so as we miss the distributed capacitance and
inductance between the 2 wires).

Itsu