Magnetising current in a Full-Bridge SMPS?

[sawreyrw]

Just when I think I understand coupled inductors I run into something like this. I have to admit I had to think about it a little. Of course, it makes sense once you understand it. As an excuse, my initial response might have been somewhat confused because I was thinking about leakage inductance as well as magnetizing inductance. The primary leakage inductance is not coupled to the secondary so that energy must be be dissipated in the primary circuit, and there will be a reverse spike due to it.

Hopefully, this message does not add to the confusion :-)

[Sch3mat1c]

Full wave, discontinuous current forward converter works correctly as long as the transformer's secondary-referred inductance is larger than the rectifier load inductance. If this condition is not met, the circuit will behave somewhat like a full wave flyback, with correspondingly poor performance.

The same is not true of full wave conduction (e.g., full bridge phase shift, or a similar circuit could be arranged in push-pull), where the transformer's voltage is under control of the inverter, or of continuous current (large output filter inductor), where the inductor's current accumulates over several cycles and therefore must always be greater than the secondary magnetizing current, which must reset every half cycle.

Since transformer inductance is very easy to have very high, this doesn't concern anyone...

Far more critical is the leakage associated, causing the primary current to persist after the switches turn off, requiring damping of some sort.

Tim