Simple to build isolation transformer that consumes less power than it gives out

[Jack Noskills]

There will be a test report. I have one that has one trafo setup described. In that current limiter bulb uses 1.5 watts and load bulb around 20 watts. But all the meters show it is not OU, so same effect you have. Last test is to test this in using isolation. There are two trafos back to back because suitable 1:1 was not available, so first stepdown and then step up so there are some losses. This setup also did not show OU, when taking power it was seen at input. Only difference to my setup is now the resistance of the coils, which was 165 ohms in mine. Impedance was not measured, I will ask and permission to post the test report.

I refer once more to two trafo test I made with iron trafos in the beginning. First just two trafos in series, I got bulb on primary of first trafo then at output of second trafo. I put load I got light on first trafo and little less light on second trafo. This is normal. Then I connected the second trafo differently, I got no light on first trafo and more light on second trafo compared to test one. So I got two different behaviour, first test is obviously not OU. What is happening in this second test ? This is why I started this thread in the first place, to get explanation to this effect from others that know things better.

[Jack Noskills]

Here is test report as promised. Now that inductance figures are known it is clear why input 'sees' load. I think inductive reactance should be well over 10 kohms, higher is better. Ideally < 1 mA should flow through one coil when no load. I still don't understand why bulbs are unequal. If coil is leaking' towards source, then the first bulb sees sum of two currents but going in opposite direction (low voltage difference) lamp does not get power. Or something like that.

Does anyone have a small iron trafo that is used to light small bulbs (not leds) and measure L of that ? Maybe we could compare.

[Jack Noskills]

Below you find my views how to build high frequency version. Read it and comment about it. Send comments using PM and I can then update document. Let me know if you want/dont want your name to show in case changes are done.

[leo48]

@Noskil

The file report.zip not open
Leo48

[Jack Noskills]

Explanation from EE for the sawtooth waveform of voltage at output when load connected:

There is harmonic content till 1KHz. The strongest harmonic is the third with amplitude about 1/10 of fundamental (50Hz).