How does this regulator work?

[james_s]

I recently pulled out this TDS460 power supply to take another crack at repairing it and I'm reminded of what a strange design it is and trying to wrap my head around just how this works. It consists of a primary side driving a single transformer at 32kHz which is open loop aside from over voltage shutdown. The secondary side has four very similar sections each with their own winding which are each regulated and adjusted individually without any feedback to the primary side.

At first glance I thought they were simply shunt regulators which would be bizarre but I suppose it could work however I then realized it's just a small signal transistor and these outputs each deliver up to several amps. Then I thought maybe they are creating a DC bias to control saturation of the core, but then how could each output be adjusted individually without affecting the others? I have a couple of these power supplies and on one of them one of the outputs works correctly and adjusts over a range of about +/- 15% without affecting the rest.

One clue is that inductor closest to the transformer is ahead of the rectifier so I wonder if there is some kind of clever resonant thing going on here. Or maybe they are controlling saturation of the inductor?

[Vovk_Z]

It's a kind of magnetic amplifier.
Look here for an answer.
Something like:
"Sending current through them in the other direction via L201 (L202) and diode CR209 reduces this net DC component through both coils, which increases their inductance and therefore their impedance, reducing the output voltage."

ATX power suply 3.3V regulator is done that way usually too

[james_s]

Ah ha! Things are starting to clear up a bit! Maybe now I can actually get these going, I had one with all outputs high, regulation not working at all and someone gave me another one which I finally made some progress and have it running with one output regulating. Trying to understand how the regulation worked has been really bugging me.