Please, anyone seen what I written? I wait for a answer.
also, i have a guarantee that after recapping can I hope that it POST again?
Also, replacing the 10V 1200uF with 10V 1500uF does hurt?

I'm not an expert but I've had some problems when using diferent capacitance electrolitics, usually if you have a 10v cap, the voltage is 5v or less, for example, the cpu works at 1.6v so on the cpu vrm 2.5v caps are enought, ddr memory works at 2.5v so 4v caps are enought. I usually measure the voltage in the caps with the board working when doing polymods because is easier to find low voltage polymers, and its easy to get then from newer dead boards.
Againg I'm not a expert.

so mine is not polymod. It is traditional electrolytic caps. Someone recapped the 10v 1200uF with 10V 1500uF and didn't had problems.

also it seems that due to weather conditions and the university exams i need to do, I will recap it after March 2022.

Do you know which model capacitors you have? They probably need to be low ESR types.

i don't know if LOW ESR. I put also on a MB a no-mentioned LOW ESR but worked

If you can post photos of the capacitors showing the markings then people should be able to work out what they are.

Also, to actually answer your questions, generally for power supply smoothing capacitors having a larger capacity won't matter (might even be good). The thing that matters is the voltage and temperature rating. The further away you can stay from either of those then the longer the capacitor will last (2x voltage is good, 3x is better, think there's less improvement after that). But as long as the voltage rating is at least as high as the old capacitors then you should be ok. And I'm afraid there's no guarantee that this will fix the problem, just that other's have had similar problems and the existing capacitors came from a time when many capacitors failed very early, so it seems like a likely cause. If the capacitors could be tested then you'd know if they were out of spec, which would give a bit more confidence that changing them would be useful. CPU should have survived the power cycling, although if the capacitors have failed then the supply to the CPU would have spikes in it, and if they were big enough then it might have been damaged (or the RAM, or the northbridge). In my limited experience though, they've always been ok.

so if I recap that MB, it should POST again? Also i don't see any physical damage on northbridge or CPU. No burn marks.
And also, if the original CPU Voltage regulator capacitors are 10V 1200uF, and the replacement is 10V 1500uF, it can hurt or it works properly?

RetroPC_King wrote on 2022-01-02, 08:05:

so if I recap that MB, it should POST again?

Can't be sure. It's a fairly common cause of problems, so I think there's a reasonable chance, but the problem could always be somewhere else.

Also i don't see any physical damage on northbridge or CPU. No burn marks.

If the capacitors fail to smooth the voltage supplies then there can be spikes or dips in the supply. They won't show up as any physical damage, but can damage input transistors and stop chips working.

And also, if the original CPU Voltage regulator capacitors are 10V 1200uF, and the replacement is 10V 1500uF, it can hurt or it works properly?

That's fine. As long as you replace a 10V part with at least 10V, and a similar capacitance then you should be ok. Some circuits doing timing and frequency stuff care about the actual capacitance value, but these are just for smoothing.

One thing to watch out for is the ESR (effective series resistance). The lower this is, the better the capacitor will be at removing any disturbances in the supply. So to be really sure you need to try and check the ESR of the old part and compare with the new part. If you can't find datasheets for the old part, then look to replace them with capacitors sold as 'low ESR' or 'low impedance', e.g.: https://www.farnell.com/datasheets/2321292.pdf (Rubycon ZLH series).

The markings on the capacitor should tell you the manufacturer and the series type, along with voltage and capacitance. You can then look up for the manufacturer datasheet for that series of capacitor and read the ESR value.

Rubycon, Panasonic and Nichicon and Nippon Chemicon are first choice.
I have never heard of Sancon and Elite.

Replacing 1200µF with 1500µF will be o.k. - it´s inside the tolerance interval.

OK. So due to auto-oscillation of on-off cycle can damage the SOUTHBRIDGE AND NORTHBRIDGE or not? Also CPU and RAM?
Also in March I will try recapping if I can.
Also with those capacitors (original) sometimes it auto powers on and do not turn off (and also no display), or sometimes does this on-off cycle with blinking DIMM_LED frequently.
Also I heard that ELITE is DECENT

RetroPC_King wrote on 2022-01-02, 19:56:

OK. So due to auto-oscillation of on-off cycle can damage the SOUTHBRIDGE AND NORTHBRIDGE or not? Also CPU and RAM?

'fraid I really can't say. Failing capacitors allow the voltage supplies to be unsteady. If the spikes and dips are big enough then that can cause damage to anything connected to those supplies.

From what's known so far (unsteady board, make of the capacitors, other peoples problems with the board) it seems like there's a reasonable chance that the capacitors are a problem. But without being able to actually measure their capacitance and ESR then we don't really know for certain. Replacing them won't hurt. Until the power supplies on the board are steady then you can't find out if anything has been damaged. Any of the problems you're having could be caused by the capacitors, so we need to know that they're ok before going any further.

Best I can say is that I had a board with awful capacitors (rubber seals blown out of the bottom, electrolyte leaked everywhere, burn marks...) and it's still fine.

If you can, put it aside until you get a chance to do the recapping, then see how it goes. If you haven't already, probably also worth lining up a decent spare PSU to test it with.

Essentially, get anything power related to a known good state, then at least you've got a good foundation to work on.

also changed PSUs and SAME RESULT! So RECAPPING IT is good?

Ah, sorry, forgot you'd already said you changed the PSU.
I'd say that recapping isn't bad, and is probably good. It's not guaranteed to fix the problem (that might be caused by something else), but it won't make the problem worse.

I don't know what experience you have with recapping, but unsoldering the -ve/Ground pin of the capacitors can be a pain because heat from the soldering iron spreads out in to the PCB ground plane very easily. Warming the area with some hot air first can help (even just a hair dryer). Be gentle and don't pull too hard on the capacitor or you can damage the via that connects the capacitor to the ground plane.

I gonna do that warming with a hair dryer (I have one at home), but at what speed? Mine have 2 speeds. Should I use speed 1 or speed 2?

RetroPC_King wrote on 2022-01-04, 09:42:

I gonna do that warming with a hair dryer (I have one at home), but at what speed? Mine have 2 speeds. Should I use speed 1 or speed 2?

Don't actually know. I doubt a hair dryer is going to get very hot (at least, not hot enough to damage any components), so I'd guess whichever is hotter.

After finishing exams, and at 1 month after the last post, i plan to recap this.
I got a solder wick, soldering iron and soldering paste. Can I use soldering paste for soldering capacitors or not?

I done today a partial recap. Replaced only the capacitors around the CPU Socket and from the MOSFET zone.
It still does the tick-tock cycle sometimes, but not frequently. Results improved somehow. But STILL NOT POST.

I gonna replace also the AGP and DIMM caps.

any opinions about this please?

Have you replaced the AGP and DIMM capacitors? Also, did you have a go at checking the datasheets to look up the ESR of the capacitors you have (smoothing capacitors normally need to be low ESR)?