No, I haven't used these ones. Yes, using caps rated for higher voltage will work fine. Also, rated hours will be largely meaningless for this application as they'll be running well under spec.

Anyway, UCC/NCC is generally excellent so long as you stay away from their KZG and KZJ series. Also, these are poly caps, so double plus good.

gdjacobs wrote:

No, I haven't used these ones. Yes, using caps rated for higher voltage will work fine. Also, rated hours will be largely meaningless for this application as they'll be running well under spec.

Anyway, UCC/NCC is generally excellent so long as you stay away from their KZG and KZJ series. Also, these are poly caps, so double plus good.

Awesome, thank you.

I've got the capacitors on order, shouldn't have any issues, this will probably make the board a bit more stable.

Here's the data sheet of caps I ordered.

http://www.chemi-con.co.jp/cgi-bin/CAT_DB/SEA … j=p&pdfname=psg

I've ordered some polymer caps to do a rebuild of a socket A board as well. But just the vulnerable caps around the CPU and chupset.

canthearu wrote:

I've ordered some polymer caps to do a rebuild of a socket A board as well. But just the vulnerable caps around the CPU and chupset.

I am doing something similar, however, I am only replacing the bulging caps for now. Here's a picture of the board I have, the ones I have circled in red are the ones I am replacing.

I suggest also replacing the 3 around the LPT port, as they would be the same kinda crappy caps.

It isn't always good to replace regular low ESR electrolytic caps with solid electrolytic caps aka organic polymer etc. Their capacitance is lower and price higher, though ripple current and ESR are great. They are meant for high frequency multiphase PWM circuits, but those old mainboards have only one to three phases clocked at 100kHz to 200kHz usually. I'd replace those failed caps with known good low ESR Japanese electrolytic caps like Rubycon MCZ/MBZ/ZL/ZLH, Sanyo (Suncon) WG/WA, NCC (UCC) KZH/KZE, Panasonic FJ/FM/FR. There were issues with NCC KZG and Nichicon HM in the early 2000's due to failed experimental electrolyte. I think they are fine now, too.

ph4nt0m, the problems are:

a) Motherboard low ESR electrolytic caps are not produced any more. Rubycon MCZ/MBZ would be great, but since motherboards have moved to solid capacitors, these type of capacitors are not produced anymore.
b) Rubycon ZL/ZLH, Panasonic FR are designed for power supplies. They are definitely more resilient, but their performance is going to worse than capacitors you pull off the socket A/478/775 motherboards.

If the proper capacitors were available for these repairs, I would certainly recommend using them rather than using solid polymer capacitors. But they aren't and here we are.

That's true, Rubycon discontinued all MxZ series in 2011 due to low demand from OEMs. Some NOS could still be found, though those offered by Chinese sellers are most likely fake.

There's one thing to note about ultra low ESR caps. Their electrolyte is very temperature sensitive, so those caps are rated for 2,000 hours usually. Caps with a bit higher ESR, though still considered low, are 5,000 to 6,000 hours rated usually, though some models like Sanyo WA are 8,000 to 10,000 hours rated. High VRM temperatures and hot air exhausted by CPU coolers can make ultra low ESR caps die in just a few years of regular usage.

ph4nt0m wrote:

That's true, Rubycon discontinued all MxZ series in 2011 due to low demand from OEMs. Some NOS could still be found, though those offered by Chinese sellers are most likely fake.

There's one thing to note about ultra low ESR caps. Their electrolyte is very temperature sensitive, so those caps are rated for 2,000 hours usually. Caps with a bit higher ESR, though still considered low, are 5,000 to 6,000 hours rated usually, though some models like Sanyo WA are 8,000 to 10,000 hours rated. High VRM temperatures and hot air exhausted by CPU coolers can make ultra low ESR caps die in just a few years of regular usage.

The capacitors linked below were the ones that were recommended to me, also I think with the temperature that most capacitors are rated at 105*C is quite a bit higher than what case temp is going to be at, my CPU isn't getting much above 60*C. And my plan is to change out the fans in the case for some Arctic Cooling fans to maintain reliable and adequate air flow. So I do not think temperature is going to be an issue.

gdjacobs wrote:

You can go with these 16V rated UCC polymers.
https://www.digikey.com/product-detail/en/uni … 4315-ND/5824538

They have a little bit more capacitance and are cheaper than the Nichicons.

Anyways we'll see how well these work, once I get them in, just need to get some appropriately sized solder wire, I'll probably play on the safe side and go small.

Yup, either 60/40 or 63/37. The stuff I have is 23 gauge 63/37 and works well for this type of job.

gryffinwings wrote:

The capacitors linked below were the ones that were recommended to me, also I think with the temperature that most capacitors are rated at 105*C is quite a bit higher than what case temp is going to be at, my CPU isn't getting much above 60*C. And my plan is to change out the fans in the case for some Arctic Cooling fans to maintain reliable and adequate air flow. So I do not think temperature is going to be an issue.

105*C is when the electrolyte starts to boil and you don't want it to happen. There is such thing as thermal derating. Caps exposed to high temps experience high current leakage and their capacitance decreases faster over time. BTW 2,000 hours for an ultra low ESR cap ain't much. It's just 83 days non-stop. There is a rule of thumb for electrolytic capacitors that their service life doubles for every 10*C below the rated maximum. So, 2,000 hours at 105*C = 32,000 hours at 65*C.

ph4nt0m wrote:

gryffinwings wrote:

The capacitors linked below were the ones that were recommended to me, also I think with the temperature that most capacitors are rated at 105*C is quite a bit higher than what case temp is going to be at, my CPU isn't getting much above 60*C. And my plan is to change out the fans in the case for some Arctic Cooling fans to maintain reliable and adequate air flow. So I do not think temperature is going to be an issue.

105*C is when the electrolyte starts to boil and you don't want it to happen. There is such thing as thermal derating. Caps exposed to high temps experience high current leakage and their capacitance decreases faster over time. BTW 2,000 hours for an ultra low ESR cap ain't much. It's just 83 days non-stop. There is a rule of thumb for electrolytic capacitors that their service life doubles for every 10*C below the rated maximum. So, 2,000 hours at 105*C = 32,000 hours at 65*C.

Well I guess it's a good thing that the caps I purchased are rated for 20,000 hours at 105*C, so I think I am fine there, so at 65*C that rule would give me like 350,000 hours, I think I'm good.

Well, no more power through the board, for now. The old caps have reared their ugly heads and have just slightly started leaking some yellow substance.

HM caps brewing up some trouble. Good times.

gdjacobs wrote:

HM caps brewing up some trouble. Good times.

Indeed, disassembly will start today, so I can identify what specs the VRM caps near the parallel port are, I'll order some more caps for those once I find out.

gryffinwings wrote:

gdjacobs wrote:

HM caps brewing up some trouble. Good times.

Indeed, disassembly will start today, so I can identify what specs the VRM caps near the parallel port are, I'll order some more caps for those once I find out.

During disassembly I decided to remove the northbridge cooler for some more cleaning and found this nastiness, definitely too much thermal paste.

That was fine, but definitely reapply when you reinstall the heatsink.

I've gone ahead and done the polymod on my Athlon XP motherboard. Works fine, even though I replaced 3300uF electrolytic capacitors with 1500uF polymers.

Brilliant! Congrats!

gdjacobs wrote:

Brilliant! Congrats!

I am still waiting for some things in the mail, I might be recapping tonight.