coupled output inductor in dual smps

[schiller]

I am modifying a pc supply for dual outputs (+/- 12v). I must rewind the -12v winding with thicker wire. It seems to be wound in the same direction as the +12v, but the output from the rectifiers is fed to the opposite end of the winding. Is this normal, or i make somewhere a mistake?

For additional filtering at the outputs, i want to use a pi filter for each one. Is there a potential disadvantage with this that i may be not aware of? The inductors of the filters are placed very close to each other, should i place them so that mutual coupling is minimized, maximized, or it doesn't matter?

Any advise would be very helpfull, i have done a lot of searching in the web with no results, maybe because these are elementary questions and people think that everybody knows the answers :-)

[Sch3mat1c]

Exactly what power supply is this, and what are you doing with it?

Tim

[schiller]

Originally Posted by Sch3mat1c

Exactly what power supply is this, and what are you doing with it?

Tim

An old pc psu. Just want to have an auxiliary dual supply for the bench to perform "quick and dirty" checking to power amps for example.

[Sch3mat1c]

Ok. The way they make -12 in those is a small winding on the filter choke. So that needs to be bigger. The diodes supplying it also need to be bigger (ideally you could maybe replace the 5V schottky with a negative-common 12V model, changing the circuit to suit). This will effectively make a choke-input FWB rectifier on the 12V winding, which is how it's supposed to be.

As for the second round of filtering, it's not a big deal where they go because they are small signal parts. Ideally, if close, they should be phased mutually, so they are coupled as they would be if on the same core (mind that the magnetic field of a solenoid is upside-down at a distance from the core, relative to the field inside the core itself). Otherwise, they should be as far as possible.

The effect is small, since the coupling factor of adjecent rod type inductors is going to be on the order of 10% or less. So you'll have more variability from capacitor ESR and wiring stuff.

Tim

[schiller]

Thanks a lot for the advise. I did rewinded the -12v of the output inductor with thicker wire, added a common anode diode pair with the same rating as the one for +12v, filter caps for the negative rail and the output filters.

I have made some space for the common anode diode on the heatsink replacing the diodes for the 3.3v with the small ones originaly used for the -12v. ( i am not going to use it, no need for current).

Haven't changed anything else in the circuit. Unfornutately, it doesn't work properly. It starts and after a while shuts down by itself. In the mean time, it seems to function :-) voltages are correct, only regulation seems poor (5% with 3A per rail). I must have made some mistake, i'll check it again.

Anyway, thanks again for the help.

Greeitngs,
konstantinos

P.S One of the funniest and realy interesting ideas i have read so far is your PWM tube amp. Respect