Partnered Output Coils - Free Energy

[Vortex1]

FWIW:

An actual transformer model would include what you see in the attachment, and in your case, you could include an additional secondary with it's own associated leakage inductance, distributed capacitance, resistance etc.

As you can see from the model, the distributed capacitance will form tank circuits with the winding inductance and the leakage inductance, damped somewhat by the winding resistance and applied load, if any.

When testing, it is a good idea to do wide band frequency response vs voltage plots of each winding , loaded and unloaded.

As the various tank circuits come alive at their specific frequencies, an increase of voltage may be seen for that particular winding.

Parasitics can be flattened if properly loaded, but testing at only one frequency gives a tiny slice of the total frequency response of a given transformer design, and a small picture of the overall performance.

Depending on the coupling coefficient between the  transformer windings, resonances may couple to other windings to a greater or lesser degree.

As an example a high quality audio grade output transformer my have a power bandwidth from 20 to 20,000 Hz, yet still exhibit some peaky response on it's secondaries, usually in the upper range due to distributed capacitance effects.

Regards, Vortex1

[gotoluc]

Thinking about it,my UPS should have a 1:1 isolation transformer in it. I doubt it was the transformer it self that went,so i will check that out before building one out of MOT's.

Going to do some pulsed DC test on it now,and see how we go with that. Once i have my isolation transformer setup,i will carry out the test MH requested.

Yes, a ups is as good but it would be best to not have it plugged in the grid to have true isolation.

Pulsed DC sounds like a good experiment and don't forget to try your diode.

Interesting stuff mate!

Thanks for sharing

Luc

[MileHigh]

Brad:

You are back to the same issue of using a non-linear load and you are trying to make power measurements.  Your measurements are no good even though you were making true RMS measurements.  Can your DSO multiply two waveforms together?  If yes, the only way to measure power in the setup you show in your diagram is to put the DSO on the input voltage and current and do the multiplication, and then put the DSO on the output voltage and current and do the multiplication.

If you get rid of the FWBR and the LED then all of these problems go away and you can make the measurements with your multimeters.

Here is my suggestion:  Even if there is very low impedance on the input it's not going to hurt your function generator.  You can make R1 a standard 1 ohm or 0.5 ohm current sensing resistor.  Then you define your input power as the power that is pumped into the promary coil only.  i.e.; you ignore the power dissipated in the R1 resistor.   Then try different load resistors and frequencies, whatever you want, on the inner secondary.

If you want to be a keener, you will know the RMS currents in the primary and the inner secondary.  Therefore you can easily calculate the power dissipated in the primary coil and inner secondary coil.

What you are looking for is this:  <power pumped into the primary coil> = <resistive losses in primary coil> + <resistive losses in inner secondary> + <power delivered to load>.  You can define the efficiency as <power delivered to load>/<power pumped into the primary coil>.

You can assume that the power pumped into the primary coil is going to be slightly greater than what goes into the load and the resistive losses in the two coils.  You can speculate about that "missing power" and where it went.  The first place that comes to mind for me is in the hysteresis losses of the core material.   Naturally, as Vortex1 stated everything is frequency dependent.  One can assume the higher the frequency the more unaccountable losses there will be.  Therefore, one suggestion to keep it sane would be to make measurements at 100 Hz and 6 KHz.   Or, if you were hard core (pun intended) you would sweep the frequencies as per what Vortex1 stated.

MileHigh

[tinman]

Brad:


MileHigh

You are back to the same issue of using a non-linear load and you are trying to make power measurements.

Ruddy hell MH-just hold ya horses-->1 test at a time.

Your measurements are no good even though you were making true RMS measurements.  Can your DSO multiply two waveforms together?  If yes, the only way to measure power in the setup you show in your diagram is to put the DSO on the input voltage and current and do the multiplication, and then put the DSO on the output voltage and current and do the multiplication.

The measurements are good-as you will see soon enough.
My scope dose have a math function,and will display the math trace,but i cannot figure out how to get it to display the value of that trace. There have been a couple of others here that were going to see if they could work it out,but it never happened.

If you get rid of the FWBR and the LED then all of these problems go away and you can make the measurements with your multimeters.

Already done,and the video is uploading now. But i still get the feeling that even this will not be good enough,so i will be sending the link to ION and Poynt first to see if they can see if and where i made any mistakes.

Here is my suggestion:  Even if there is very low impedance on the input it's not going to hurt your function generator.  You can make R1 a standard 1 ohm or 0.5 ohm current sensing resistor.  Then you define your input power as the power that is pumped into the promary coil only.  i.e.; you ignore the power dissipated in the R1 resistor.   Then try different load resistors and frequencies, whatever you want, on the inner secondary.

I used 10 ohm resistors for the CVR. It dose not matter if power is being dissipated by the resistors,as it's still dissipated power,and i was careful when i carried out the calculation's-i think?.

If you want to be a keener, you will know the RMS currents in the primary and the inner secondary.  Therefore you can easily calculate the power dissipated in the primary coil and inner secondary coil.

What you are looking for is this:  <power pumped into the primary coil> = <resistive losses in primary coil> + <resistive losses in inner secondary> + <power delivered to load>.  You can define the efficiency as <power delivered to load>/<power pumped into the primary coil>.

I calculated only the P/in to the primary,and P/out across the 3 resistors(in my next video). I have not accounted for dissipated power in either coil as yet,that will just be the cream on the top.

You can assume that the power pumped into the primary coil is going to be slightly greater than what goes into the load and the resistive losses in the two coils.

I assume nothing,and we will see what you assume if and when you see my next video.
Once Poynt and Ion(vortex) have viewed it,and if they find i made a cockup some where.i will post that here. I will still post the video,so as you to can see my mistake-if i have made one.

You can speculate about that "missing power" and where it went.  The first place that comes to mind for me is in the hysteresis losses of the core material.   Naturally, as Vortex1 stated everything is frequency dependent.

I have tried frequencies between 1 and 30 KHz,and see no change in the outcome with various types of loads.

One can assume the higher the frequency the more unaccountable losses there will be.  Therefore, one suggestion to keep it sane would be to make measurements at 100 Hz and 6 KHz.   Or, if you were hard core (pun intended) you would sweep the frequencies as per what Vortex1 stated.

Below 800Hz,this transformers efficiency drop's right off. I have been running test between 3KHz to 30KHz at 250Hz step's,and the performance remains much the same throughout that band range.

[MileHigh]

The measurements are good-as you will see soon enough.

Really?

You have 1.441 V RMS across the primary x 6.4 mA RMS current through the primary = 9.2 mW.

The problem is that the current is not linear with the voltage.  The current increases faster than the voltage rises rendering your calculation wrong and the measurement bad.  The power is actually higher than you are calculating.  The same thing applies to your power measurement across the FWBR + LED.  I mentioned this issue just the other day and Vortex1 echoed what I said.

Note in your diagram you also have a problem on your power out calculation.  You have the power in as being the power that is going into the primary coil which is fine.  But then you are including the power dissipated across R1 as part of the power out which is wrong.  If anything, the power dissipated across R1 can either be ignored by choice to take it out of the picture, or you can include it as part of the input power which is supplied by the function generator.