Win some, lose some...

[tazwegion]

Like the old adage goes... win some lose some, so true! recently I pulled an old Abit KT133 chipset m/board out of the 'to-be-fixed' drawer (I was bored ), unfortunately 12+ cap's was too much, and the resulting mayhem & frustration left the board in two halves! LMAO


Enter the next Abit KT7... this is actually out of my secondary rig (semi-gaming & DC), and was suffering from bulging cap's too (only 3-4 thankfully) also to my disgust the NB cooler had given-up-the-ghost, however luckily I had some spare Gigabyte GPU chipset coolers (and thermal goo)... certainly heavier duty than the original fitment (kinda' beefy)



The original motherboard (now in physical halves) was a great source of replacement capacitors (though different V ratings), as were left over ATX PSU boards (2200mF cap's) though a little on the large/bulky side



Then last, but by no means least... the GeForce2 MX400 I purchsed of eBay (minus chipset cooler) was mated (after a slight 3pin -> 2pin mod') to a lovely blue illuminated Vantec IceberQ4 chipset unit, Whoa! check out that blue!

[Jim]

Very Nice!

[aaronamd]

cool!




poor capaciters....

[tazwegion]

Yeah poor capacitors indeed, a local PC tech I sometimes 'annoy' was telling me that Abit (in the past) had used suspect quality capacitors on their boards, and due to an impending law-suit were repairing said items for gratis, however that's no good to me living this far away from the U.S. though it did teach them a lesson in Quality Assurance Vs. product profit margin

The other part to the tale, is a disgruntled employee @ a Japanese capacitor plant left their job (taking with them the recipe for capacitor composition) and went and set up shop in China (or there abouts), the problem was the recipe was incomplete, hence capacitors coming from that NEW plant didn't have the tolerance for heat that they should, thus making them increasingly prone to failure and a liability to any company who purchased from them to save a buck

[Jim]

The whole sad story about that can be had from Gary Headlee who has made a profession of replacing defective caps. (He operates out of Utah).

[tazwegion]

Still Utah is more than a' Coo-eee' away from here... pretty much leaving me to my own devices, or at the mercy of local electronics tech's


BTW it wouldn't be a sad story for Gary... it was his bread and butter

[tazwegion]

Well... believe it or not, I had 8 more capacitors to replace on my Abit KT7 (s462) m/board, I even had to resort to recycling cap's off old MSI s370 M/boards


This must make it about 12+ (out of the 30 total) capacitors since I first aquired the darn thing

Perhaps it would be possible to set fire to the bean counter's abacus

[Jim]

Capacitors are very cheap. Would strongly recommend buying new ones instead of using used ones off dead boards to replace defective ones. Reason being that the "Cap Problem" is caused by the breakdown of fluid within the caps electrolyte into gases. (Which makes them eventually bulge in some cases as the gas pressure within them goes up) Because this is a gradual process You may ( Probable) be replacing dead caps w/ half dead caps.

Note : if caps are hard to get where you live, you could probably make an arrangement w/ somebody living in a city in Australia to send you the kind you need. Heck, I could do that from here even.

[tazwegion]

The reason I sometimes re-use cap's is because 6.3v 2200mF versions are not available anywhere is Australia... I'm lucky to find a 10v variant, seeing as space is @ a premium (in the pcb layout) a larger replacement will sometimes just not fit

BTW for the record... I've yet to replace the same capacitor twice... it's always a different location on the board, heck I'm getting so paranoid now when I pull a board for inspection, I'll change every cap' that even thinks about bulging (early signs)

I can source most of what I need locally.... and for what I can't there's always -> RS Australia

Interestingly enough... Skt. A has been the biggest offender, with barely any SS7 boards requiring replacement, but we all know the history of the IT industry @ that point in time

[tazwegion]

Backed up all my Jpeg collection to the network server during a format/OS reinstall... only to have the drive cause the computer/server to shutdown on reboot?

That's right... it wasn't the PSU, RAM, V/card or BIOS... it was the HDD (fujitsu), how bizarre I thought... only to discover the drive was corrupted & files missing

Yet another example of hard copy beats magnetic storage, OR back up everything twice!

[Jim]

Surprized you hadn't heard. Fuji produced millions of drives w/ defective controllers. They packup and die after a while. They also lost a class action suit and were offering refunds to owners of those drives, complete w/ up to $500.00 for data recovery. You used to be able to download complete info from their site. Maybe still can.

[tazwegion]

Thanks Jim, it was a second hand 4GB unit... only registers as a 2Gb unit now

So I placed one of my trusty Quantum Fireball's on that system... only to find a (blue screen) failed attempt to write to drive error... what's up with this? I've since added an ISA IDE controller card (and disabled on-board primary) now we shall see what's going on...

[Jim]

Odd. I'd try hooking both drives up to another mobo to see if suddenly they are normal again. If so you may have another cap prob, or even just a bad cable.

[tazwegion]

Well since I've added the IDE controller card I've had no more issues (yet)... the problem now is the card doesn't support DMA and the drivres are @ least ATA33

Dunno' about further faulty cap's Jim, but a suspect IDE (on-board controller) IC seems to fit the bill pretty well, and replacing which is far-and-beyond my humble expertise

[Jim]

Hmmmmm. You arn't giving many details about this machine; but : 1) If only the Primary controller is shot, you can boot from the secondary provided the primary is not used; - though that would limit you to 2 IDE devices. 2) Alternately you could use a newer IDE controller card which supports higher speeds. 3) Junk the mobo and use another. 4) Send the mobo to Gary Headlee who does have the expertise, (and equipment), to fix it.