Seasonic S12II PC PSU Noise

[AKM]

I`m a beginner, so please bear with me.

I have a Seasonic S12II 520W PC power supply, the modular version, which has started making weird noises. I have recorded the noise with an old PC microphone and uploaded it to Youtube:
At first, it happened once in January, then again in February, then few times in June, and finally many times in July. Sometimes the noise starts with the machine, sometimes during high load. Sometimes I can turn off the PC and turn it back on, and the noise is gone. But sometimes it persists even if the PC has been unplugged for hours. And even weirder, sometimes it starts with the machine, but stops after just a few seconds.
The noise modulates with load, higher load increases its frequency; and I can hear popping noise on my speakers.

I have looked on the Internet, and there seems to be a few similar cases with no solution.
I have contacted Seasonic, and they offered me to return the unit. But I need the PC for daily work, so I couldn`t send it.

I don`t have an oscilloscope, which would probably make this much easier, just a multimeter and an ESR meter.

So I took the unit apart. Cleaned the dust and inspected it. There were no obvious faulty components. All output capacitors looked and measured fine. The main filter capacitor had a bit lower capacity, but nowhere near the +-20% manufacturer tolerance. I have resoldered a few solder joints on the larger components that looked even a tiny bit suspicious.

After a few days of rest, the noise returned.
Then I measured the outputs(should have done it first) while it was making noise. Every single one was clean. Stable voltages and almost no ripple under light and heavy load; and no sign of the noise on the multimeter bargraph on any of them. So I concluded that the switching side and the secondary side were OK. Only filter and PFC sections were left.
I planned on bypassing the filter section, to pinpoint it further, but then the noise started every time with the PC for 3 days, and on the 4. It stopped. 

Those 3 days were the hottest this year, so heat came to suspicion. Took the unit apart, started heating the whole board gently and powered the PC. Nothing, so I started heating different parts of the board without any success. I couldn`t get it to start making any noise again. 
Then it came to me that those days were also very humid, and it would explain why I couldn`t get it to make the noise by heating it.

I noticed that some SMD components, on the main board underside and on the main PWM/PFC daughter board, were conformally coated.
I don`t know exactly why I did it, but I scraped the coating with a nylon spudger only on the underside of the main board while it was still warm, cleaned it with ethanol alcohol and dried it with hot air.
I don`t know if it was due to the components, the glue used to hold them or the conformal coating; but the noise has almost stopped. Or even if the problem is on the daughterboard, and I dried it indirectly.
I say almost, because if I switch the overclocking profiles in BIOS, the noise starts and then stops when the OS boots for the first time after. It doesn`t start again even under heavy load while the PC is overclocked.

What I plan to do is to remove the all the conformal coating, clean those areas with Isopropanol, dry it with hot air, and apply a few layers of conformal coating.
The only conformal coating I can get is Plastik 70 (acrylic coating) from Kontakt Chemie, everything else would cost more than 60$ for even the smallest ammount.

Now, to my question. Is this an OK thing to do, will it work if the problem is due to moisture; or am I completely wrong?
If anyone has any similar experience, problems or experience with conformal coatings; any information would be highly appreciated.

I don`t have any pictures, but there are a few reviews of the non-modular version(only difference, besides extra connectors,  is that they don`t have conformal coating on the underside of the main PCB):
http://www.jonnyguru.com/modules.php?name=NDReviews&op=Story4&reid=185
http://www.hardwaresecrets.com/seasonic-s12ii-bronze-520-w-power-supply-review/2/

[wraper]

Certainly sounds like something wrong with fan.

[AKM]

Certainly sounds like something wrong with fan.

Nope. Had stopped the fan slowly with a cotton swab, no change. Unplugged the fan, and the noise persisted.
It is definetly the PFC section. Popping noise could be heard on the speakers plugged on the same power strip, but no sign on the 12V, 5V, 5Vsb or 3.3V PSU outputs.

[wraper]

Then check if PFC is working at all (under load). There should be about 380-400V on the big electrolytic capacitor. If PFC is not working, PSU still could work, especially if not heavily loaded.

[Marco]

Conformal coating shouldn't cause it to draw fast current spikes from mains (ie. audible on your speakers).

PS. damn it, I just bought the 620W non modular version of this.

[mos6502]

Funny enough, I have the exact same PSU (the 430W version, SS-430GB instead of SS-520GB). A few weeks ago my PC wouldn't power up anymore. All I heard was clicking/sparking noises from the PSU. Opened it up, expecting the worst, and guess what? The NTC (big green thingie, inrush current limiter) had desoldered itself. The solder from one pin was completely gone and it was arcing between the pin and the surrounding pad. Desoldered the NTC, cleaned everything up, soldered it back in and it's been working like a charm ever since. Easiest fix ever.

[Marco]

So either they have bad QA/soldering or they have some bugs in the PFC circuitry pulling current surges, neither is making me feel good about my purchase :/

[AKM]

@Wrapper
If I remember correctly it was 380V.

@mos6520
Nice to hear. But, this doesn`t seem to be like that. 

@Marco
The PSU has been running for a few years without any problems. The soldering is pretty good, quality components, nice layout; everything you could wish for in a PSU.
I think mine was exposed to a bit too harsh enviroment than it could take. Probably the PCB got contaminated due to dust, moisture and whatever else was in here.
A proper test would probably be to breathe a few times on the suspected area, and see if the circuit starts making trouble.

I`ll try to buy IPA and the coating tomorrow and apply it.
Thanks for your replies; I`ll keep you updated.

[jitter]

Hmm, for a little over 6 months now I have the S12II Evo 520 W (modular) without any problems, so far, but I'll keep an eye... erm... ear on it.

It seems to be quite a common problem, googling on it turns up a fair number of hits with this kind of clicking noise coming from (Seasonic) PSUs. On one forum the clicking noise was attributed to a faulty cap and the unit was replaced.

I wonder if the conformal coating has anything to do with it. It's my experience that intermittent faults have a tendency to temporarily disappear when a pcb is handled, especially if some stress is applied to it. I hope your reapplication of conformal coating works for you, though.

[AKM]

Looks like it is fixed.

Took it apart, desoldered the daughterboard, put on latex gloves, removed as much coating as possible, cleaned it with IPA, dried it with a hot air tool @120°C with maximum air flow, taped the pins with paper masking tape, applied 3 thin layers of acrylic coating with 10 minutes drying time after each, removed the masking tape and cleaned the pins with IPA, dried it with a hot air tool, soldered it back in place, inspected it, cleaned the rest of the PFC sensing circuitry on the main board with IPA, dried it, taped everyting around it, applied 3 thin layers of acrylic coating with 10 minutes drying time after each, removed the tape, dried everything with a hot air tool, put everything together, turned the PC on and it works.

No noise even when changing overclock profiles or when the PC is overclocked and under heavy load.