[ale] Motherboard capacitors bursting

[Warren Myers]

I don't know how many of you read Electronic Design, but there was an
article about 2 years ago covering the capacitor problem. The
technical info is available from a coulpe different websites,
including: http://www.pcstats.com/articleview.cfm?articleID=195
http://www.spectrum.ieee.org/WEBONLY/resource/feb03/ncap.html.
Basically, it turns out to be a case of corporate espionage.

A South Korean company stole an electrolyte formula from a Japanese
company, but the formula was mistranslated, which caused lots of
mboard problems.

I had this happen to a buddy of mine whom I built a computer for back
in 2000. He ended up replacing the board twice, the first was an Abit,
the second was a FIC (I think). Hasn't had any problems since.

WMM


On Thu, 07 Oct 2004 07:25:17 -0400, Van Loggins <vloggins at turbocorp.com> wrote:
> >
> > Message: 9 Date: Wed, 6 Oct 2004 19:04:27 -0400 From: Vincent Fox
> > <vf5 at plm.gatech.edu> Subject: [ale] Motherboard capacitors bursting
> > To: ale at ale.org Message-ID:
> > <20041006230427.GA5188 at gypsy.coon.gatech.edu> Content-Type:
> > text/plain; charset=us-ascii The Abit BX-133 board was famous for
> > this. They used cheap electrolytic capacitors and a marginal design. I
> > had probably six of these boards flake out on me. They don't usually
> > fail in a complete and obvious way they just start locking up or
> > rebooting occasionally when the 1st capacitor fails. As the load is
> > taken up by remaining capacitors in the string, they are under more
> > stress and then another one fails, and it becomes a bit less stable.
> > Continuing until 3 or 4 have blown at which point it becomes obvious
> > you have a hardware problem since it will no longer boot. Have a PC
> > that seems to be flaking out? Open it up and look for capacitors with
> > bulgy tops. No single manufacturer seems to be guilty of this. I have
> > had one electronics type tell me that electrolytic capacitors all have
> > a finite lifetime anyhow but they are far cheaper so everyone uses
> > them. So don't expect any board to last 100 years in operation just
> > some will last longer than others.
> >
> >> > Anyway, I seemed to remember someone on this list bemoaning poor quality
> >>> > components in computers, and now I can testify to it.
> >>
> >>
> 
> I've had two boards to fail on me like this.
> 
> The first one was a MSI K7T Master-S Socket A workstation/low end server board with built-in adaptec U160 SCSI.
> 
> Luckily It was still under warranty but I had to fight with MSI to get it replaced.
> 
> The second one was a Epox 8RDA+ Nforce 2 Socket A motherboard. It never gave me any indications that it was failing up until when It died. I had left my computer running and when I sat down in front of it the next day it had a blank screen like the system had went into power saving mode. Despite all of my attempts to wake the system up I couldn't so in desperation I hit the reset button. The system wouldn't post and the Post code indicator kept stopping on the check memory part of POST.
> 
> I figured it had a bad memory stick or a bad video card or Power Supply. all of those checked out ok when I tested them in other systems, however when I removed the video card I discovered a busted and oozing capacitor, and upon closer inspection the capacitors around the memory slots had started to bulge.
> 
> Unfortunately that board had been out of warranty for about 4 months, so I purchased a inexpensive Amptron Nforce2 motherboard from Computer Geeks http://www.compgeeks.com
> 
> If you're good with a soldering Iron and have access to junk motherboards you can replace the capacitors and make the board work again with a little work. I'm very bad with a Soldering Iron though so I haven't tried to fix my old Epox board yet.
> 
> --
> Van Loggins        vloggins at turbocorp.com
> Assistant System Administrator - ESC Dept
>       _
>      -o)
>      /\\
>     _\_v
> Linux User #316727
> 678-989-3052
> Turbo Logistics
> http://www.turbocorp.com
> 
> 
> 
> _______________________________________________
> Ale mailing list
> Ale at ale.org
> http://www.ale.org/mailman/listinfo/ale
>