Gigabyte GA-6WMMC7 (IBM 19K3582)

[linuxguru]

This is an IBM branded, Gigabyte OEM GA-6WMMC7 with an IBM BIOS that boots very fast - that's one of the reasons I'd like to keep it going, apart from the fact that I have a bunch of spare Coppermines lying around.

There are 17 G-Luxon 330/25v caps on the board, as well as 8 Sanyo 1200/6.3v clustered around the VRMs/CPU.

The G-Luxons are slightly bulged at locations TC2 & TC3 near the CPU socket, and at two unmarked locations at the edge of the board next to Q32. The board still boots and runs fine with most Celerons, but has slight instability with a P3 800/100.

I can't see any reason why the G-Luxons are rated at 25v, since there is no voltage higher than +12v on any modern motherboard. I'm reasonably certain that none of these caps actually handles anything higher than 6.3v. I plan to check and replace with 6.3v or 10v rated caps.

Can anybody who has this board check and confirm the values of the large electrolytics? TIA.

[Rainbow]

Re: Gigabyte GA-6WMMC7 (IBM 19K3582)

I have two these boards - both with shorted VRM controllers (probably result of bad PSU).
One board has:
17 CHOYO 330uF/25V
8 CHOYO 1200uF/6.3V
The second one has:
17 GSC 330uF/25V
8 CHOYO 1200uF/6.3V
Although these are all major craps, not even single one is bulged - maybe because the boards were damaged by bad PSU early. They're waiting here now until I get a RC5057M chip from an unrepairable board.

[linuxguru]

Re: Gigabyte GA-6WMMC7 (IBM 19K3582)

>One board has:
>17 CHOYO 330uF/25V
>8 CHOYO 1200uF/6.3V
>The second one has:
>17 GSC 330uF/25V
>8 CHOYO 1200uF/6.3V

Thanks - so they've retained the same values, but changed brands from Choyo/GSC to G-Luxon (marginally better) and Sanyo (much better) on my board. There's no sign of rework, it looks factory-populated at all locations, with a 2000-01-02 date code on the IBM sticker.

Perhaps the VRM controllers failed due to bad Choyos? These have been known to fail open-circuit quietly without a mess. Perhaps there was a reliability issue, and IBM insisted on Sanyo, not Choyo.

[kc8adu]

Re: Gigabyte GA-6WMMC7 (IBM 19K3582)

iirc 1500@6.3 will replace all the 25v caps on gigabyte boards.thats what i use.
the 25v parts were likely the cheapest caps they could buy that would give acceptable decoupling.since i have lots of 1500@6.3 smt nichicons i even used them by bending the leads straight and pulling the plastic part off.i had a gigabyte duallie that was new in the box with the choyo caps leaking!the stink gave it away when i opened the bag.it was in a closeout bin in a mom and pop store near indy that i just happened to spot while on a service call.

[Rainbow]

Re: Gigabyte GA-6WMMC7 (IBM 19K3582)

The VRM is powered directly from 12V, the mosfets are good just the chip is fried (shorted internally). Chips go bad like this after a voltage spike. When a PSU dies, it often kills VRM or super I/O or some 74xxx logic chips like this. So this might be from bad caps but in the PSU, not on board.
GSC is IMHO one of the worst craps (Jackcon too), much worse than G-Luxon.
I'd replace all the G-Luxons with 1000uF/16V (I use only 1000uF, 2200uF and 4700uF for all mainboard recapping, all 16V so they can be used for 12V too). And if some of them are connected in parallel, just leave some of the positions empty.