https://www.electronics-notes.com/articles/el … es-markings.php

It doesn't have 4J in the list, though. 🙁

If you can't figure out the meaning for markings, you can try measure the voltage, as a last resort.

As a rule of thumb, I would use a value that's a little bit higher.
Because, voltage spikes regularly happens with all digital components (on/off toggling creates power surges and spikes,
even if we're in a micro second or nano second scale).

So if there's 5v applied, a 6.3v cap is near its limit already. I'd use a higher voltage model.
Same with 12v and a 16v cap. I'd use a 25v cap here.

Of course, that's just me. Most people would argue that's not necessary.
And that ESR is so much more important, et cetera pp..

pics

I'm guessing it's 680µF 4V 5% tolerance.
Likely on the 3v3 rail.

Excuse my ignorance, I can safely solder some 1000µF 6.3V to test if the board isn't dead?

weedeewee wrote on 2023-01-15, 08:57:

I'm guessing it's 680µF 4V 5% tolerance.
Likely on the 3v3 rail.

One is shorting between +5VUSB and -12V (I think it's because of a NPN or PNP not closing some MOSFET / gate / xxx when V is applied - I found this situation before, until I check "live" I'm assuming it's okay).
The other one I can't retrace it for the same reason I presume, until a gate is opened.

I guess +12V, +5V have some good circuitry (caps + mosfets), maybe +3V just has a cap and one mosfet.

At first the mobo would power on, then after some retries and in the end 1 every 20 times and then just inst shutdown.
Probably the caps (from my basic experience). I just don't wanna mess it up as I like this little board 😀

Nexxen wrote on 2023-01-15, 14:35:

Excuse my ignorance, I can safely solder some 1000µF 6.3V to test if the board isn't dead?

yes. if it fits you could even put a 400V in there., just don't go too far overboard with the capacitance. best to stick with, as close to, the original value present.

Nexxen wrote on 2023-01-15, 15:59:

weedeewee wrote on 2023-01-15, 08:57:

I'm guessing it's 680µF 4V 5% tolerance.
Likely on the 3v3 rail.

One is shorting between +5VUSB and -12V (I think it's because of a NPN or PNP not closing some MOSFET / gate / xxx when V is applied - I found this situation before, until I check "live" I'm assuming it's okay).
The other one I can't retrace it for the same reason I presume, until a gate is opened.

The 'one is shorting between +5VUSB and -12V' seems off to me. That would be a voltage difference of 17V, and I also do not see any reason to put a capacitor between those two.
Let's hope it's just some measurement oddness.

I would expect all the grounds of the capacitors be continuous to ground, and the positive to a voltage rail. could be 3v3, could be some lower voltage used for ram or cpu. likely DDR2 1.8V .

I found a voltage regulator with a shorted mosfet, shorts with all leads (G/D/S).
I may have put the wrong +12V in the 4-pin header. It's the only thing I can think of.

weedeewee wrote on 2023-01-15, 16:46:

The 'one is shorting between +5VUSB and -12V' seems off to me. That would be a voltage difference of 17V, and I also do not see […]
Show full quote

Nexxen wrote on 2023-01-15, 15:59:

weedeewee wrote on 2023-01-15, 08:57:

I'm guessing it's 680µF 4V 5% tolerance.
Likely on the 3v3 rail.

One is shorting between +5VUSB and -12V (I think it's because of a NPN or PNP not closing some MOSFET / gate / xxx when V is applied - I found this situation before, until I check "live" I'm assuming it's okay).
The other one I can't retrace it for the same reason I presume, until a gate is opened.

The 'one is shorting between +5VUSB and -12V' seems off to me. That would be a voltage difference of 17V, and I also do not see any reason to put a capacitor between those two.
Let's hope it's just some measurement oddness.

I would expect all the grounds of the capacitors be continuous to ground, and the positive to a voltage rail. could be 3v3, could be some lower voltage used for ram or cpu. likely DDR2 1.8V .

Most definitely something died somewhere, board is shutting down immediately.
I'm desoldering the dead mosfet and see what happens.

What a wonderful sunday afternoon 😀

Ok, it was a dead mosfet on the VRM.
Once removed it works well.

The caps I removed were all above 20% loss and half the rated value, I guess it was just good to replace them.
Unless anyone wants to say anything this is solved.

Thanks for playing. 😀
----

🤣 second time I fired it up I got some magic smoke from another mosfet, I have only one vrm line left...

The first that died put some strain on the other two and I guess the last one is waiting to die as well.
Nice 😀

--- THE END ---