Yes Tinman,
                  it's all down to closed surface integrals,
                           John.

Quote from: tinman on May 26, 2015, 07:20:23 AM
This is great TK-thanks. An actual device showing a similar effect.
Now,if we take your's and poynts scope shots,and compare them with mine,we can see the difference in the current line. I have all my windings incased in core material,where as you have yours on a spool inside a pot core. So you would have an air gap-correct?.

My scope shot below shows the yellow trace across the outer secondary,and the blue trace across the inner secondary-both with same value loads. If i disconect the inner coil load,i can get that outer coil trace to follow what you and poynt show-a decline in current through the cycle. But i am interested in this(almost) non changing current value throughout the 95% off time-as can be seen in my scope shot below-->blue trace. I can now get this to happen with any resistive load value.

Your waveforms suggest that you have much longer L/R time constants than poynt99 and TK.

Quote from: MarkE on May 26, 2015, 10:51:10 AM
Your waveforms suggest that you have much longer L/R time constants than poynt99 and TK.

Indeed

Tinman,

It would be interesting to see just how long it takes the inner sec. winding to return to zero after your MOSFET turns off.

Consider increasing the time between your MOSFET drive pulses, while maintaining the same positive going pulse width and amplitude, until the you are able to see the secondary output return to zero during the off time.

Consider externally triggered 1 cycle bursts if necessary.

PW

OK, then... here's another pot-core, with a "secondary L2" of 90 turns of #34 wound onto the small bobbin first, then a "primary L1" of 30 turns of #27 wound over that. The load is the same 3 ea. 100R carbon resistors in parallel.  Again, I've shown the trace with the peaks (24 V) clipped off to emphasize the negative portion of the trace. I've also turned up the brightness of the grid!

So now the time constant has increased and the negative portion is more horizontal and still significantly negative by the time the next pulse comes along.


@TinMan: as far as I can tell the two halves of these little potcores mate perfectly with no airgap either at the central core or the outer edges. Of course the bobbin means that there is some airgap inside between the windings and the core itself.