PSU issues, advice needed.

[athos56]

OK, I'm going nuts on this one. I modled a PSU on PSUDII and I thought I had everything worked out but somehow its just not working right. So I tore it down and started again with test leads measuring as I went. The problem is that I'm getting really high voltages no matter what I do. Details:

Transformer = Hammond 278cx (400 - 0 - 400 and 465ma)

Full wave rect made from 1n5408 diodes snubed with some 10omh resistors and some .01uf ceramic caps.

With just that, I measure aprox 400v at the junction of the 2 diodes. When I add a 2.2uf cap between the junction and ground I now measure aprox 615v where the diodes and cap meet. Ok, so I think, lets drop that a bit. I change out the cap for a 1uf cap, I measure somewhere around 610v or so, I change it out for a .22 cap and I read the same high voltage aprox 600v. So no real change for all the different cap values. Now even unloaded there should have been a change right?? So I make a choke input filter with a 18h choke and 50uf cap, what do I measure at the junction of the cap and choke?? You guessed it, aprox 608v.... I think I'm going crazy.

Any Ideas??

[AndrewT]

A capacitor input filter generates sqrt(2) * Vac across the output capacitor when unloaded. Vac is the unloaded transformer secondary voltage at the mains voltage when tested.

A choke input filter generates ~ 0.9 * Vac across the capacitor, only if the load current is in the range that the choke can supply.
If one reduces the current drawn through the choke, the voltage at the output rises, ultimately to sqrt(2) * Vac.

[Eli Duttman]

Originally Posted by athos56

OK, I'm going nuts on this one. I modled a PSU on PSUDII and I thought I had everything worked out but somehow its just not working right. So I tore it down and started again with test leads measuring as I went. The problem is that I'm getting really high voltages no matter what I do. Details:

Transformer = Hammond 278cx (400 - 0 - 400 and 465ma)

Full wave rect made from 1n5408 diodes snubed with some 10omh resistors and some .01uf ceramic caps.

With just that, I measure aprox 400v at the junction of the 2 diodes. When I add a 2.2uf cap between the junction and ground I now measure aprox 615v where the diodes and cap meet. Ok, so I think, lets drop that a bit. I change out the cap for a 1uf cap, I measure somewhere around 610v or so, I change it out for a .22 cap and I read the same high voltage aprox 600v. So no real change for all the different cap values. Now even unloaded there should have been a change right?? So I make a choke input filter with a 18h choke and 50uf cap, what do I measure at the junction of the cap and choke?? You guessed it, aprox 608v.... I think I'm going crazy.

Any Ideas??

You have to load the PSU down to get its true behavior. Without a load, a cap. I/P filter will charge up to (VRMS) (2^1/2). Also, regulation effects in the power trafo raise VRMS if a legitimate load is not present.

What sort of rail voltage are you aiming for?

[athos56]

Originally Posted by Eli Duttman

You have to load the PSU down to get its true behavior. Without a load, a cap. I/P filter will charge up to (VRMS) (2^1/2). Also, regulation effects in the power trafo raise VRMS if a legitimate load is not present.

What sort of rail voltage are you aiming for?

I had a mono block built up with everything in place and a 8ohm dummy load on the output transformer. I went to test the voltages and noticed they were really high (in the 590's). I was aiming for something between 450 and 500 to keep the regulator controlling my B+ dissipating under 13 watts. I figured a psudo-choke input with a small first cap would be the way to go and PSUD seemed to confirm this. With everthing in place I tried a big 20w 47k bleeder resistor to load everthing down a bit but it didn't seem to matter much. What value of dummy load should I use for further testing?

Edit: should I have ran a 1 or 2v signal through the amp as well?

[aardvarkash10]

ideally you would emulate what the amp loads up. 47k is going to present insufficient load on the PS. It would help if you posted the PSUD model so we can see what you have there...

[Johan Potgieter]

Athos,

As hinted by Aardvarkash, that 47K would only load to about 12 mA. Unfortunately you are in the middle of a 'no-go' area in practice. Also beware of correcting with a small capacitor as filter input. It will work, but will land you on a slope of the regulation curve, making regulation between min. load - max. load poorer than either proper C- or L-input filters.

But let us take a closer look:

The 600+V unloaded means that the transformer is giving 426V per sec. h.t. winding. That is the no-load voltage. Calculating backward, to get that down to 400V under load at a spec. of 465mA would land you with about 16 - 20 ohm (dc) per 400V winding - you would probably find that. (I did not look up the transformer specs, but there are also probably heaters to be fed. Remember that transformer voltage specs are for full load on all windings). With a choke of some 30 ohm dc resistance (a little on the high side for that current - still) you might land at below 500Vdc with load, capacitor input filter.

What will your load be (including heaters)?

[PS: Giving advice you did not ask for: Remember that with a tube amplifier - I presume this is for such - the initial h.t. voltage with Si-diode rectification could rise to 600+V before the tubes start drawing current. Hope you reckon with that when choosing capacitors - also interstage coupling caps!]

[athos56]

Originally Posted by Johan Potgieter

Athos,

As hinted by Aardvarkash, that 47K would only load to about 12 mA. Unfortunately you are in the middle of a 'no-go' area in practice. Also beware of correcting with a small capacitor as filter input. It will work, but will land you on a slope of the regulation curve, making regulation between min. load - max. load poorer than either proper C- or L-input filters.

But let us take a closer look:

The 600+V unloaded means that the transformer is giving 426V per sec. h.t. winding. That is the no-load voltage. Calculating backward, to get that down to 400V under load at a spec. of 465mA would land you with about 16 - 20 ohm (dc) per 400V winding - you would probably find that. (I did not look up the transformer specs, but there are also probably heaters to be fed. Remember that transformer voltage specs are for full load on all windings). With a choke of some 30 ohm dc resistance (a little on the high side for that current - still) you might land at below 500Vdc with load, capacitor input filter.

What will your load be (including heaters)?

[PS: Giving advice you did not ask for: Remember that with a tube amplifier - I presume this is for such - the initial h.t. voltage with Si-diode rectification could rise to 600+V before the tubes start drawing current. Hope you reckon with that when choosing capacitors - also interstage coupling caps!]

All my caps are highvoltage Russian caps. I'll do some more tinkering. Thanks for all the info everybody, I'll digest it and see what I come up with.