FSP400-60GLN worth fixing?

[Elitist]

Re: FSP400-60GLN worth fixing?

" the diode " - which one? position, value/type, etc?
One of the big problems with all these mass produced structures is that flux removal both after wave soldering and, esp. after-machine-assembly, hand soldered components and wiring may be insufficient. It's usually accomplished by a solvent wash, but this is often too cursory for heavy deposits. Excess flux degrades after many heat cycles, may hydrolyse inter alia and may become conductive. Anyone spending (far too much) cash on their prized electronics would be well-advised to wipe over boards with e.g. a paint brush dipped in a lower alkanol - IPA is preferable but I suppose cheap vodka might do...
If anyone has good PSU transformers surplus, mindful of the above conversations, they would be most welcome here.
Otherwise, not making much progress with repairs. The conundrum is that my dud units worked perfectly for ~2-3 yrs, caps, diodes, overheated Rs all replaced. All these expedients seem to fix PG, +5V, but none of the other lines???

[Stefan Payne]

Re: FSP400-60GLN worth fixing?

Naa, the CX400 was Seasonic.

[Behemot]

Re: FSP400-60GLN worth fixing?

Eek, that one is most likely CWT. I mean for this FSP GLN/BSII:

https://www.badcaps.net/forum/showpo...&postcount=132

[Per Hansson]

Re: FSP400-60GLN worth fixing?

Yes please
If it's this transformer I posted about here for a Corsair CX400W:
Top: VEE19FB10
Side:
SBI4.2
E150436 C

[Behemot]

Re: FSP400-60GLN worth fixing?

Seems that the latest BSII I acquired had the diode so hot it nearly burned a hole in the PCB, again. It atually has green salt on it as it started corroding. I've seen such carnage before so there is no reason to repair this - chances are it will get conductive enough for the high voltage that the whole circuit will be screwed. If it goes down to few kiloohms it's already useless. And cutting the PCB part out and so some other major operations…nah, no way, got plenty other things to work on which actually do have some financial sense.

So it will be available for spare parts, if anybody needs. Per Hansson, yo still want that transformer?

[var123]

Re: FSP400-60GLN worth fixing?

But, beware: have you looked underneath them all?! Prolonged overheating can char the adhesive spot and give partial (or full, even) shorting.

There is only one that is brownish. R42 (2R0) under the hot 5VSB diodes.
I have checked R42 several times and tapped it while measuring.
Guess it could still be charred on the inside and malfunction under load.

The big OST cap is ok. 361uF, ESR: 0.29.
No cold solder joints. PFC coil looks fine.
Could be some intermittent problem with the transformer or a leaky diode, but which one?

R81,C80 and R45A belong to the 5VSB. One would think that these are somewhat critical.
The rest of the missing components seems to be on the 3.3V rail.

[Stefan Payne]

Re: FSP400-60GLN worth fixing?

Has annyone tried replacing the 8mm Caps on minor rails with Polymers?
Can/will that work?

The rest can be refitted with 10mm ones. Well, depending on the revision...

Here my revision of this unit:


And I didn't have good 8mm caps at hand and used what I've got out of other units...

[Elitist]

Re: FSP400-60GLN worth fixing?

"At the next power cycle it suddenly turned on and started working. "
Had almost identical experiences as var123, changes of caps, etc., but as reported above, +3, & +/-12 lines couldn't support sensible loading - only +5V fully functional. So far unable to locate what seems to be a single fault location.
"SMD resistors and diodes mentioned in this thread seems to be fine"
But, beware: have you looked underneath them all?! Prolonged overheating can char the adhesive spot and give partial (or full, even) shorting.

[var123]

Re: FSP400-60GLN worth fixing?

The following SMD components are missing:
D35
R22, R23
R32, R45A
R53, R54, R81
C31, C38, C80

Where to find the correct values?

[var123]

Re: FSP400-60GLN worth fixing?

Don't think they are shorted.
The reason is I measured the 5VSB again today and it was only 4.68V.
The last few times I checked it was 5.05, but the PSU didn't start.

At the next power cycle it suddenly turned on and started working.
+12V: 12.27
+5V: 5.03
+5VSB: 5.07
+3.3V: 3.50 (high, but I only had bulbs and resistors on 5/12V)
Hmm....

[Behemot]

Re: FSP400-60GLN worth fixing?

Check if some of the output rectifiers has not gone shorted, that sometimes happens especially with all caps bad. I have actually even seen the -12V diode bad couple times (mostly in different units, but still).

[var123]

Re: FSP400-60GLN worth fixing?

Hi,
I have a FSP500-60GLN that was used 24/7 for many years before starting to act up.
Eventually the PC wouldn't boot and it even lost the network connection at standby.

Pretty much every electrolytic cap was bulged and bad. I have changed everything except the big OST 390uf/450V.
Now, the +5VSB is stable but the PSU doesn't turn on.
D41/D42 have been replaced with UF4007. It was clear that this area had been running hot.

All cables are directly soldered to the board including mains and fan, which makes it a real pain to work on.
I don't have an isolation transformer and poking around in this with power on does not feel safe.
The SMD resistors and diodes mentioned in this thread seems to be fine, and the optocouplers have also been changed to rule them out.

Any suggestions on what to check for from the experts?

[Per Hansson]

Re: FSP400-60GLN worth fixing?

The site does have a SSL certificate so you can access it via HTTPS.
That will get rid of the warning but redirection is not active.
So it will switch back to HTTP when you click certain links.
Here is more info: https://www.badcaps.net/forum/showth...491#post748491

But please don't discuss this in this thread about the FSP PSU!

[eccerr0r]

Re: FSP400-60GLN worth fixing?

Yeah, this seems to be new behavior in firefox 52... It is usually correct (when trying to POST login/password over a non-https link), but very annoying...

--- EDIT ---

Odd, looks like the problem went away for one situation where the login/password field was actually encrypted, but nevertheless gave the warning anyway. Fixed in 52.2 it seems. In any case you should still be able to use it, despite it being very obtrusive.

[Elitist]

Re: FSP400-60GLN worth fixing?

Ff !

[stj]

Re: FSP400-60GLN worth fixing?

what browser are you using??

[Elitist]

Re: FSP400-60GLN worth fixing?

This website. Something like "This website is not secure, etc, etc..."
It overwrites the log-ins which seem to be required every time now.
The pop-up obscures the log-in panel but just keep typing and press <enter>.

[Behemot]

Re: FSP400-60GLN worth fixing?

What website, what message?

There is another zener just by the DM311, but I don't think there's something for the CM6800 and the supervisor so it should be there, and OK

[Elitist]

Re: FSP400-60GLN worth fixing?

Your website is now showing insecure - message overwriting the U/N, P/W area?!

Well, I did the caps & D's + few R's - SB OK, all lines showing correct V s on o.c.. Is Z41 so mission-critical? Yes, already reported above some sme s had overheated and charred the adhesive spot underneath, probably making them 0R s. Why is my +5V restored to full capacity, but not other lines, esp. with clean transformers? Why is this possible if DM311 controller misbehaving. Still a can o' worms - where to turn next? How do you build in obsolescence for ~3yrs? What collateral damaged have the leaky caps perpetrated? And all the other unfathomable questions...

[Behemot]

Re: FSP400-60GLN worth fixing?

I just realised, there is some change though. Besides normal link to the input filter and so on, there is also some strange wiring from the AC switch directly close to the SB controller. I am not sure what it does or how it works yet, gonna have to investigate.

AS for recapping procedure - well I started doing that with these terrible things, but as you, even when you think all is fine, it is not in the end…sp I have to drag my load tester to half done, opened-while-being-worked-on unit each time and test it under load? I'd rather toss it to fire, really