/diy/ - Do It Yourself

>>1148858
The original "obvious setup" describes doing this and rectifying and filtering the output. No filtering should be needed between the triac and the transformer. The alternative as I see it is to rectify the transformer output first, then use DC PWM and then filtering instead. It's probably easier to get the correct rated parts. Twice the caps, but probably still easier.

What you describe in your "other idea" is to use PWM before the current enters the primary. Pic related shows a waveform before and after PWM, and long story short, a sinusoidal wave is the most efficient waveform to put into a transformer. It's got something to do with a differentiated sinusoidal waveform being a sinusoidal waveform, change in magnetic flux, etcetera.

But if it's going to be DC anyways, you might as well PWM it after rectification, it's probably less lossy since the transistors aren't passing through zero all the time, and you could probably do it without filtration caps before the PWM anyway. No need for bipolar caps either.