Raidmax RX-600 ringing noise and 0ms PG

[Dannyx]

Good day folks. A former workmate of mine scrounged up this PSU from a client of us. He had a look at it and found a seized up fan which is presumably the main cause of failure. A couple of caps had bulged on the secondary, so it seemed like an easy fix. He replaced them and today I had the honor of doing the test run with him. We attached an ATX tester to its output and plugged it in: it's not quite there yet. While the supply did technically fire up with no pops or fireworks, it's got a couple of issues. The most prominent one is this "ringing" or rather crackling sound it's making (sound, just like scent, is one of the toughest things to explain in words sadly ). It's pretty loud and we were afraid something might actually pop. We decided to keep it plugged in anyway, at least to check up on our tester. Ironically, all the rails are up and in spec (no way to check ripple on them though), but then we ran into the second issue: 0ms on PG....the bane of my existence, since to this day I cannot work out a definite answer as to why some ATX supplies do this !

There's only two chips on this thing: the combined PFC/PWM IC (a CM6805)and the supervisor IC on the secondary, a ST9S313 for which I'm having trouble finding a datasheet....

First, the squealing is the more concerning issue, so I thought I should focus on it first...what could be causing this ? There could still be a bad cap somewhere, as my colleague only replaced the visibly damaged ones...might be a good idea to recap the whole thing (not sure if it's worth saving though - looks decent and fixable enough). What else could get damaged to cause this noise ? The noise is only present when the supply is "on", so it's not caused by the 5VSB circuit. Haven't tried loading the supply yet to see how or IF the noise is affected in any way. Most likely caused by the main transformer.

The 0ms PG issue: bad supervisor IC ? Perhaps other similar cases you might've come across ? Cheers guys.

[momaka]

Re: Raidmax RX-600 ringing noise and 0ms PG

If the fan had failed and unit overheated, replace ALL the caps, regardless of size. This being an APFC PSU, probably also a good idea to replace the main cap too. I see it's a cheapo brand anyways, so it's unlikely to last. In fact, that could well be the cause of your noise. Bad cap on the APFC can also destroy the PSU in seconds with a bigger load. That said, if you do replace the main cap, go with a 450V one. And maybe something bigger in capacity too. Just look at all that space that the manufacturer left there on the PCB for it, but didn't end up using probably because they went ultra-cheap and grabbed an undersized cap for the job.

As for the 0ms Power Good issue... I've seen that only on cheapo PSUs and I don't know what causes it or how to fix it. But if it's not an issue with whatever motherboard you're trying to run, then let it be. Alternatively, you could try adding an RC low pass network between IC PG output and PG lead to sort of create an artificial delay. But some motherboards may not like the fairly slow "deadzone" between high and low on PG when it's implemented with a simple method like that. Either way, I wouldn't worry much about it. And also I don't trust those PSU "testers". They really are rather pointless. I'd rather connect the PSU to an old useless Pentium 4 motherboard or something similar and test it that way.

[Dannyx]

Re: Raidmax RX-600 ringing noise and 0ms PG

probably also a good idea to replace the main cap too

Good idea. Will try that. Doesn't show any signs of bulging, but then again time has proven that caps can fail even without showing any signs. Would be funny if that also fixes the 0ms PG issue

Will need to do a bit more digging around that IC. See what that PG pin actually reads. Given that the tester shows 0ms, I'm expecting the PG pin to be "stuck on" and come on immediately as the rest of the rails, because if it didn't come on at all, the tester should show "HH" (aka "too high" a delay). The IC is probably sinking the pin (keeping it LOW) until it decides its time to "release" it, at which point it goes back up via a pull-up resistor off the 5v rail...at least that's how I imagine it based on other schematics I've seen. THIS ONE uses that same chip. Notice how R69 is tied to the 5v rail. Most MOBOs refuse to turn on with a 0ms signal, and I've messed around with quite a few at my former shop, that's why I'm trying to fix that as well.

[momaka]

Re: Raidmax RX-600 ringing noise and 0ms PG

The IC is probably sinking the pin (keeping it LOW) until it decides its time to "release" it, at which point it goes back up via a pull-up resistor off the 5v rail...at least that's how I imagine it based on other schematics I've seen.

Yup, that's how it works indeed.
Bad comes to worse, it shouldn't be too hard to build your own delay circuit. Something like an RC circuit and an amplifier or shunt circuit to trigger past a certain level should do.

[R_J]

Re: Raidmax RX-600 ringing noise and 0ms PG

The ST9S313-DAG seems similar to the HY510N

[Dannyx]

Re: Raidmax RX-600 ringing noise and 0ms PG

That IC might be hard to source unfortunately and might end up costing more than the supply is worth, though I haven't looked it up on good ol' AliExpress so far....

Designing my own delay circuit: since the delay has to be rather short (a couple of ms at most, according to ATX specs) I think it can be done with an R/C combo, with that TAU equation I always find funny for some reason Trouble is I think I'd need some hysteresis there as well, since in an R/C circuit the voltage SLOWLY increases, so the MOBO might not like a SLOW rising signal like that - it might expect something that "snaps" *bang* like that to 3.3v. If it doesn't, then even better: botch a resistor+cap there, otherwise the quickest and dirtiest thing that comes to my mind at this late hour, as I'm lying in bed wondering WTH I'm doing with my life is an op-amp with a fixed reference voltage of around 3v its Inverting pin and the R/C setup on its N/I pin. As the voltage on the R/C circuit SLOWLY increases, the output of the op-amp stays low because I is higher than N/I at this point. As soon as it hits that sweet spot, the output swings positive and that's our PG signal....again, just a scenario.

[Dan81]

Re: Raidmax RX-600 ringing noise and 0ms PG

Just checked AliExpress for you, and they have 8964 lots, so you're in luck.

As for the main cap, replace it as soon as possible. I have a Andyson here (Hiper 525W) that was killed by the main caps, and I just gave up with it because the 3842 was killed by them, along with the primary switchers, the WHOLE APFC board, and a boatload of the caps on the secondary. I don't have any replacement chips, and the wiring was as basic as it could get - 24 pin, 1x 8pin EPS, 2x molex, 1x SATA, and one floppy connector.

[R_J]

Re: Raidmax RX-600 ringing noise and 0ms PG

What about a Texas Instruments TPS3510, might be easier to get. Or maybe a FAN7680

[Dannyx]

Re: Raidmax RX-600 ringing noise and 0ms PG

Just checked AliExpress for you, and they have 8964 lots, so you're in luck.

8964 ? I searched for the other two types of chip (original and compatible) and found both for dirt-cheap, so I might push on with this repair. I asked my buddy to get me a 180uF 450 volter as well, so we should be good

[Dannyx]

Re: Raidmax RX-600 ringing noise and 0ms PG

What about a Texas Instruments TPS3510, might be easier to get. Or maybe a FAN7680

I see you've edited the post. Yeah, one of those two might be easier to source locally. Need to do some measurements first to see what actually IS there. Like I said, I'm expecting PGO to be stuck on for some reason.

[Dannyx]

Re: Raidmax RX-600 ringing noise and 0ms PG

I managed to scrounge up a 220uF 450v cap instead of the original 180uF....hope it's not TOO much It also reads 213uF on a cap meter, which I think is within tolerance, especially at such a high voltage.

[Agent24]

Re: Raidmax RX-600 ringing noise and 0ms PG

Will be interesting to see if the capacitor is the problem.

I posted about a PSU on here once, where I heard a strange faint arcing/sizzling/crackling sound that I tracked down only to the APFC section but could not further discover its source. I had suspected the PFC coil, but could not see any corona in darkness.

I was advised it was likely the primary filter capacitor going shorted internally.

[Dannyx]

Re: Raidmax RX-600 ringing noise and 0ms PG

Yup, one day I'll get around to replacing that cap and I'll be sure to update the thread. The supply is still belly-up at my old shop waiting for me, I just have to find some time to go there for a chin-wag with my former mates. This weekend would be nice hopefully.

The cap I found will require some work because it's got two flat metal blades for leads which are unlikely to match the holes for the existing cap, so I'll have to convince them a little

[Agent24]

Re: Raidmax RX-600 ringing noise and 0ms PG

It may be easier to drill new holes in the board?

[momaka]

Re: Raidmax RX-600 ringing noise and 0ms PG

It may be easier to drill new holes in the board?

Or if you have enough space above the cap... 14-AWG solid copper wire for leads anyone?


^ Picture of my modded and recapped Antec Earthwatts EA-500

[Dannyx]

Re: Raidmax RX-600 ringing noise and 0ms PG

The cap already has some leads soldered to its "ox horns" so with a bit of luck, I'll be able to bend those into shape.

But damn - 500uF ? :O

[eccerr0r]

Re: Raidmax RX-600 ringing noise and 0ms PG

The PG problem reminds me of one cheapo PSU that I repaired many years ago. Turned out a ceramic disc cap shorted out, and tricked the PG to always be no good. So much for annoying debug efforts! At least that PSU was easier to debug since it had no shoe horned daughter boards...

[Dannyx]

Re: Raidmax RX-600 ringing noise and 0ms PG

Was that cap on the PG "wire" directly or somewhere around the supervisor chip which sends out the PG signal ? Sucks it's all SMDs in this one

[eccerr0r]

Re: Raidmax RX-600 ringing noise and 0ms PG

It was near the op amp on the PSU... So I guess supervisor chip on yours. Yeah SMD sucks, double sided SMD sucks more, and double sided daughter boards with SMD is worst.

[momaka]

Re: Raidmax RX-600 ringing noise and 0ms PG

But damn - 500uF ? :O

That's all I had on hand. Well, four of them, actually - came out of some industrial A/C board along with a few other goodies. The board was smashed to pieces in several places, but the caps were OK. Knowing how much one of these caps costs, I immediately snatched it when I saw it in the dumpster.

Yeah SMD sucks, double sided SMD sucks more, and double sided daughter boards with SMD is worst.

Eh. At this point, I don't mind SMD that much. But double-sided PCBs for PSUs, I don't like to work on - becomes hard to trace things and especially hard to remove stuff if it's double-sided with SMD, like you said. This is one thing I do like about cheapo and old PSUs - easy to work on.