Has anyone done this before?

[keantoken]

Hi.

I suddenly had a burst of inspiration and concocted this amazing circuit...

I'm not sure how practical the inductor values are, but here it goes.

This is a DC regulator is unlike VERY, VERY, MANY others.

No transistors, just an extra 4 diodes and transformer. Extra stages can be added for better regulation, with 1/2 the ripple for two stages 1/3 for three, etc....

The bad thing I suppose though is that it increases the amount of higher order harmonics, so it may take extra filtration to be practical for a HIFI amplifier PSU...

However, with only about 2.1V ripple, a solid-state transistor regulator could be added to the end of this which would give very good output indeed...

Theoretically, the minimum output ripple is about 1.2V, since this is how much voltage it takes to go through two silicon diodes.

However, with increased modification and insight, I believe lower values can be obtained...

- keantoken

[Eva]

Do you really expect a transformer to handle the full DC output voltage across one of its windings?

[keantoken]

The output of this circuit is just below 60V. However, the current through the secondary is only about 600 mA because it is already pushing its own voltage.

However, if there is a capacitor connected across the rails, this will put 60V across the secondary when the unit is turned off - assuming that this voltage is not also conducted into the primary via C1.

If the primary inductance is high enough, then in the end there should only be at max 1.2V across the secondary.

However, if this fails, then we can still add a diode in series with the secondary, to prevent explosions.

In the case of the primary, however, I'm not sure. I'm assuming your meaning the primary, since it is in the most danger. I'm not aware of the availability of transformers that handle currents like these. You'd be the expert on that, not me.

This version works better, and has fewer parts. Doesn't look as cool though... The graph shows current in the primary.

- keantoken

[megajocke]

That transformer (L1 L2) is going to need a biiiig core! (infinite size actually if you want the circuit to work for a nonlimited time)

[Eva]

You don't understand transformers. Every winding becomes a dead short (actually the wire resistance) at DC. You can't apply DC to any winding.

[keantoken]

Alright, I've seen through my own pitiful delusion.

Sorry to bother Y'all.

I thought it was a dandy idea but then I tried it with just the 4700u capacitor and the regulation was better.

- keantoken

[Eva]

Find a way to AC-couple the transformer (if there is one)