RM1000 PSU Not Quite Dead, Not Quite Alive

[brassicGamer]

Hi folks. Long-time lurker, finally a tricky enough problem that I thought I'd ask some advice.

This PSU seems well known. I picked it up 2nd hand about a year ago to replace another that had died. It was working fine for a while then stopped. Short story is that it was being run off a 240v generator which developed a problem leading to possible overvoltage. However, it passed a physical inspection - nothing burned, popped or unusual there (photos to follow).

I've taken measurements on the primary side up to the transformer and it appears to be working fine there but it won't power up. I'm measuring +5v on the standby pin of the ATX connector (pin 9) but, oddly, I'm also getting +2.5v on pin 21, which is supposed to be +5v when powered on.

When I try to power it on (jumping pins 15 and 16) nothing happens. Some kind of protection must be kicking in though, because +5vsb reads 0, although pin 21 still reads +2.5v. Actually I've just noticed that pins 9 and 21 are adjacent to each other, so maybe I should start with following traces and looking for a short? Nothing is visibly wrong so far.

Thanks in advance.

[brassicGamer]

Re: RM1000 PSU Not Quite Dead, Not Quite Alive

[brassicGamer]

Re: RM1000 PSU Not Quite Dead, Not Quite Alive

I measured all the Zener diodes on the underside. All check out except D8 (furthermost top-right in photo), which was measuring all but short on the board. Removed it and it checked out fine. Is this a clue?

[brassicGamer]

Re: RM1000 PSU Not Quite Dead, Not Quite Alive

Still no visibly identifiable symptoms on this PSU - anyone got any ideas on where I can start?

[RJARRRPCGP]

Re: RM1000 PSU Not Quite Dead, Not Quite Alive

What are those caps with the "X" vent stampings?

[momaka]

Re: RM1000 PSU Not Quite Dead, Not Quite Alive

I've taken measurements on the primary side up to the transformer and it appears to be working fine there but it won't power up. I'm measuring +5v on the standby pin of the ATX connector (pin 9) but, oddly, I'm also getting +2.5v on pin 21, which is supposed to be +5v when powered on.

That indicates the 5VSB circuit is working properly - at least without a load.
All other outputs, except for PS_ON pin, should be at 0V, or close. Check your measuring to make sure you didn't make an error. 2.5V on the 5V rail seems odd.

Also try attaching the PSU to a motherboard (that doesn't self-power On, as some motherboards do like to do that) and measure the 5VSB then. Is it still 5V? If yes, 5VSB circuit is working, and there may be something going on with the main PS (12V, 5V, and 3.3V rails.)

When I try to power it on (jumping pins 15 and 16) nothing happens. Some kind of protection must be kicking in though, because +5vsb reads 0, although pin 21 still reads +2.5v.

You probably won't be able to operate this power supply (jump PS_ON to ground) without a proper load. Need to have PSU attached either to a motherboard or a large load on the 12V rail to make it work without a motherboard. 12V incandescent car bulbs usually work OK for this (1-3 bulbs on the 12V rail and one or two bulbs on the 5V and 3.3V rails should be enough.)

What are those caps with the "X" vent stampings?

The green ones look like TAICON. (PW series?)

The black ones attached to the modular PCB with "KF" writing are probably CapXon KF.

Not the best selection of caps, certainly, but I don't think the output caps are the issue here. The start-up cap(s) may be bad, though... or at least worth checking out / changing by the O/P as a first or second step in troubleshooting.

[brassicGamer]

Re: RM1000 PSU Not Quite Dead, Not Quite Alive

Thanks for the guidance - I'll plug it in to a mobo. I did have a 12V fan attached for load, but maybe that wasn't pulling enough.