GA-8IG1000MK with bloated Rubycon MBZ (pic heavy)

[pfrcom]

Acquired this recently, and found hard to believe ...
... five bloated Rubycon MBZ around PCI and AGP slots


Perhaps not too surprising when one sees heat discolouration underneath the FETs between AGP & PCI1


Matched by discolouration beneath VRM


Latter weird, considering VRM has Rubycon input and Sanyo output


Was advised board had been liquid cooled, and stopped working reliably if multiple DIMMs were used, after a leak
So I also removed caps between memory slots, to find



Bloated MBZ replaced by Sanyo WG - also replaced two non-bloated MBZ


DIMM area, after clean-up


Board overall, after cleaning off coolant stains with wet microfibre cloth



Could be worse though


Unfortunately multiple DIMMs are still not reliable - board's BIOS swaps between OK & Fail for 2.5V health status

However two DIMMs were OK for four passes of Memtest86+

And did a clean XP install without an issue

Then Prime95 didn't even get through a single pass

And stress test of Aida64 also didn't go too far

Aida64's sensor display showed some weird numbers for 2.5V - swapping from ~0.8V to ~1.7V

However Aida64 also showed 5V as 4.1 - but DMM measured 5.03 on spare Molex connector

PSU is a recapped Antec SL300S

Caps between AGP & PCI1 heat up to temperature most unpleasant to put finger on

Where to go from here ?

[Wester547]

Re: GA-8IG1000MK with bloated Rubycon MBZ (pic heavy)

The attachments don't work. The link in red shows bad 2004-2005? Nichicon HM and KZG. No surprise there. The software sensors are never reliable, best not to place any trust in them (but the low +5V reading could be a pin or two on the ATX 20-pin connector, if it really is low). Is there any discoloration underneath the DIMM area? Any chance of trace corrosion (as the resulting effects of the leak from the liquid coolant) or oxidization that may be causing only one DIMM to work, even after cleaning the board? The two linear regulators by the AGP slot, I would imagine, do run precariously hot. No surprise there either.

[pfrcom]

Re: GA-8IG1000MK with bloated Rubycon MBZ (pic heavy)

The attachments don't work

They're hosted on Badcaps, not sure what could be the problem

Can anyone else confirm it's an issue for them please

[Dan81]

Re: GA-8IG1000MK with bloated Rubycon MBZ (pic heavy)

They're hosted on Badcaps, not sure what could be the problem

Can anyone else confirm it's an issue for them please

He's right,I can't see them either.

[pfrcom]

Re: GA-8IG1000MK with bloated Rubycon MBZ (pic heavy)

.............

[pfrcom]

Re: GA-8IG1000MK with bloated Rubycon MBZ (pic heavy)

I should have known better ...

... the penny finally dropped that I was using a 3.0 Northwood with a missing pin

So I swapped in a 3.0 Northwood with all pins intact, and you've guessed it ...

... self-tests 1024K and 8K of Prime95 passed, and continuing

[pfrcom]

Re: GA-8IG1000MK with bloated Rubycon MBZ (pic heavy)

Clicking on the IMG tags in question brings up a page that says the attachments are invalid. Uploading the attachments without the aforementioned IMG tags should work.

I phrased too briefly - originally I followed FAQ and clicked mountains? icon for each inline image, then pasted each upload's URL - which yields IMG tags

FWIW, I cleared Waterfox cache, then I couldn't see images either

[Topcat]

Re: GA-8IG1000MK with bloated Rubycon MBZ (pic heavy)

I fixed the images.

[pfrcom]

Re: GA-8IG1000MK with bloated Rubycon MBZ (pic heavy)

I fixed the images.

Thank you, thread now much cleaner

[ChaosLegionnaire]

Re: GA-8IG1000MK with bloated Rubycon MBZ (pic heavy)

well i guess thats the issue with water cooling. due to that, the fallacy is to neglect cooling or ventilating the board properly. so what i think happened was that there was no airflow over the board due to water cooling being used. so the mosfets all overheated and discoloured the pcb. the bloated caps over a large area of the board is a telltale sign of a poorly ventilated pc case.

for normal air cooling, the cpu fan will give a slight breeze over the board. i've seen a number of cases in overclocking forums whereby when someone upgraded to water cooling, the overclocking got worse or more unstable. the culprit was found to be lack of airflow over the board and overheating mosfets during overclocking. after some mosfet heatsinks were added and a side panel fan installed blowing air over the board, it was business as usual overclocking.