Bad caps are unlikely, as they don't heal when heated. Probably a cold solder join.

No, that original poster is right track of diagnosis, capacitors when they start to fail, will perk up when heated. Replace them.

I know this from my electronic repair background, seen this all the time and I do have a specialized tool to measure ESR of capacitors get lower (low ESR is good, high ESR is no good), when capacitor is lightly warmed.

Cheers,

thanks for your posts. truth is this mobo is 20+ years old. it was given to me for free and i dont know about how it was treated or stored. now it is known that a capacitor can fail even if shown visually ok just like in my case. heat can restore some capacitance and esr levels, but sold solder- crack solder isnt something uncommon. i will flux and solder fresh the components in the suspicious area, that wont hurt and if same things happen after that, whitch is most likly i will recap it

It's a QDI, so yes that means it will need to be recapped.

ok guys i refresh the solder on components on the suspicious area. and same thing happening. looks like im gona need recap im thinking of replace all large 1000mf 6.3 volt (some are away from cpu but are same as the ones near)
i have a shop near my house that sales lelon low esr caps. im looking for a repair that will last although im not working this retro pc everyday and i dont leave it open overnight but i am willing to do as good as i can. what your thought about the replacement? the original caps are from licon.

Generally Japanese caps are highly regarded. I try to use Nichicon or Panasonic whenever I can.

truemaster wrote on 2020-01-12, 19:49:

ok guys i refresh the solder on components on the suspicious area. and same thing happening. looks like im gona need recap im thinking of replace all large 1000mf 6.3 volt (some are away from cpu but are same as the ones near)
i have a shop near my house that sales lelon low esr caps. im looking for a repair that will last although im not working this retro pc everyday and i dont leave it open overnight but i am willing to do as good as i can. what your thought about the replacement? the original caps are from licon.

If you can you identify the brand, series, voltage, capacitance, and case size of the caps you've got, I can identify some replacement options. Don't use Lelon -- they're not very good.

the original caps are licon 6.3v 1000mf other things i read on cap are ltm vent. i would like brands like panasonic but the shop i visit dont have them. lelon vishay are the 2 i found. other option is import from mouser

truemaster wrote on 2020-01-12, 21:30:

the original caps are licon 6.3v 1000mf other things i read on cap are ltm vent. i would like brands like panasonic but the shop i visit dont have them. lelon vishay are the 2 i found. other option is import from mouser

Try to stick to Japanese caps like suggested Rubycon, Nichicon, Panasonic. I order from digikey.ca

and btw its definately a case of
capacitors need replacing. Dying caps will work once warmed up. Time to replace

thanks for your suggestions. the only thing bother me is the fitting. the original are 7mm diameter 10 height

ok i found the ones i need with corect diameter lead space height. there are other caps also 18 in total 100mf 16v some near also cpu and elsewere these are normal? or low esr matters too?

truemaster wrote on 2020-01-12, 21:30:

the original caps are licon 6.3v 1000mf other things i read on cap are ltm vent. i would like brands like panasonic but the shop i visit dont have them. lelon vishay are the 2 i found. other option is import from mouser

Vishay are okay, but their selection will be limited. Licon LTM were replaced by Fujicon TM series as per the Fujicon product guide. Please confirm the case size as that might not be accurately reflected in the datasheet I'm using. I don't want to specify caps that won't physically fit.

Datasheet equivalent:
Fujicon TM 1000 uF 6.3V 8x11.5mm ESR= 0.36Ω Ripple= 645mA

Also, please specify the make, series, capacitance, voltage, and case size to the best of your ability for the other replacements you'll require.

You should be OK using Panasonic FR/FS series capacitors on this board.

They are probably the best performance caps you can readily get now days, with polymer caps taking over the reallly high performance jobs.

I'd probably go with Panasonic FR 1000uF 10V capacitors, 8x15mm ESR=0.041 ohm capacitors if I could find them.

Pentium 3's use low frequency switching circuits typically, so capacitor choice isn't as critical as newer boards.

canthearu wrote on 2020-01-14, 01:36:

You should be OK using Panasonic FR/FS series capacitors on this board. […]
Show full quote

You should be OK using Panasonic FR/FS series capacitors on this board.

They are probably the best performance caps you can readily get now days, with polymer caps taking over the reallly high performance jobs.

I'd probably go with Panasonic FR 1000uF 10V capacitors, 8x15mm ESR=0.041 ohm capacitors if I could find them.

Pentium 3's use low frequency switching circuits typically, so capacitor choice isn't as critical as newer boards.

I was thinking FC or maybe Nichi PW instead, although it's still not a great match. Unless you go over every capacitor application on the board to ensure there's no feedback issues, it's best not to stray too far from the original ESR.

The specs on those original caps were really quite lousy.

thanks for take time and help me. i measure 7mm diameter but most likely is 8mm. the other 100 mf capacitors are LRM series again licon
edit the lrm series found as rm from fujicon they are not low esr

gdjacobs wrote on 2020-01-14, 06:15:

I was thinking FC or maybe Nichi PW instead, although it's still not a great match. Unless you go over every capacitor application on the board to ensure there's no feedback issues, it's best not to stray too far from the original ESR.

The specs on those original caps were really quite lousy.

Panasonic FC caps are expensive and not so widely available here in Australia, and are an older, lower performance range. FRs are much more affordable and easier to get for me.

I wouldn't worry too much regarding feedback. The VRM circuits in question on this board were likely originally designed to use capacitors with performance characteristics similar to the FC/FR caps, and they will somehow tolerate capacitors with far worse specs and keep working for ages despite their awful performance. It is certainly possible to get feedback problems, but you would have to be using capacitors that completely do not match the job. (eg Polymer capacitors in a PSU designed for mid-range electrylics)

Edit: Although, looking at the original caps vs Panasonic FR, there is a huge difference in ESR. It would probably still work, because I can't find any other good branded capacitors that behave so poorly. Not even non-low ESR caps have specs that bad.

yeah the original caps are crap 0.36ohm=360mohm panasonic fr series with desired diameter heigh 8x11,5 is 0.0560hm=56mohm the original ones have x6.4 higher the esr value. but it will work they are repairs like mine and they go as far to polymers and are ok

canthearu wrote on 2020-01-14, 08:01:

Panasonic FC caps are expensive and not so widely available here in Australia, and are an older, lower performance range. FRs are much more affordable and easier to get for me.

FRs are not necessarily a compatible capacitor. They can react differently in circuit.

An alternative to the FC is the Nichi PW series. For the 1000uF caps, perhaps the PW 820uF caps in 8x15mm case size. For the 100uF caps, 120uF PWs have a matching case size. Still not a perfect match for the Fujicons, but certainly closer than the FRs.

nichicon pw is low esr for the 1000mf a 820mf is good they will need a very little bent to oposide in cpu area. the 11.5 heigh of the fujicons leave 2mm space before hit the back plate of the slot cpu. are you sure the 100mf caps need replacement with low esr also? they are licon LRM found as RM at fujicon and the datasheet show them as normal