Ms-6570 K7n2g

[bgavin]

The board died.

This is a real pity, considering it is 100% populated with Rubycon MBZ, and Chemicon KZE, KZG caps. There are no junkers on this board anywhere. It has a nice nForce2 MCP chipset and 8x AGP.

I can't spot anything fried, and no smoked land patterns. It won't power up with a known-good PSU and CPU.

[pentium]

Pictures would be really nice.
Are you positive everything is socketed right?

[bgavin]

The board is pristine.

Using a known-good PSU, it won't even bump the fan when I jumper the momentary Berg strip contacts. I checked the manual, and am indeed using the correct pin pair.

CPU inserted, or absent, memory inserted or absent. Fans don't even twitch. The poor thing is 100% DOA. Examination with a magnifying glass does not show me any apparent smoked land patterns or components. There is no smell I can detect.

[danten]

Stupid possibility: bad BIOS? Or a corrupted data in FlashROM.

The IC FlashROM can die or a range of data get corrupted without a explanation, it is rare but possible.
It is little hard to test it without proper hardware ( IC eprom-flash programmer like WILLEM with hub-lpc adapter to 3.3V flash I suppose needed ). Or with another same working motherboard to test if it is ok or not.

[davmax]

You said the board died. How long did it last??? Did it ever operate? I notice you now say DOA. Please clarify.

[bgavin]

The board ran for several years in a client's machine. The PSU is one of those hinky non-standard types: irregular shaped frame, non-standard mounting, and a fan that protrudes out the rear. Avance Technologies is the PSU manufacturer.

I've been through a few PSUs in both this machine and others that belong to the same client. Mostly replacing failed PSU with those cannibalized from similar (retired) machines.

It just dropped dead one day. Swapped in (yet another) PSU, and it stayed dead. Put in on the bench with a known good PSU, etc, etc, and it remains dead. Drained CMOS, replaced battery, still dead.

Normally I avoid board repairs, but I am now learning to recap good boards and power supplies. I hate to toss out a good 462 board, as that leaves me with a pile of good 462 processors. Especially considering this board is 100% Rubycon and Chimican capped.

[bgavin]

Jumpering ATX-14 to COM starts the PSU and fans.

I attempted to trace back from ATX-14 to the Berg strip connector, but it gets lost very quickly.

I don't have any idea of what the startup circuit looks like, and don't have any schematic. I can't spot any smoked land patterns.

[gdement]

Have you tried looking for any shorted mosfets?

[bgavin]

I'm not that sharp (yet).

The southbridge nForce chip started heating up to red hot after I forced the board to power up.

Pity. Nice board, but shot. I found another one on eBay with a 30 day warranty. From the photo, it is laid out the same as this board, and will probably be all good caps like the dead one.

[pentium]

The southbridge nForce chip started heating up to red hot after I forced the board to power up.

If it was alive before that it most certainly is crisped now.

[tazwegion]

Scrap the old board for it's capacitors, jumpers, DDR retainers & anything else you can salvage

[bgavin]

I found an eBay replacement with 30 day warranty. I'm pretty sure all the K7N2G are populated with similar premium caps.

[Rainbow]

Originally Posted by bgavin

The southbridge nForce chip started heating up to red hot after I forced the board to power up.

What PSU was used when it died?

[bgavin]

Avance Technologies ATX12V-250MNH, crappy little micro ATX supply. I opened it up to find the usual Fuhkyyu bulgers and leakers.

My client inherited a number of these systems when they bought out this business. Most of the PSU have died, and in this case one board went with them.

I'm getting a replacement board from eBay, and will use a standard case and recapped Sparkle when I redo this system for the client. I figure the HAL should work (same chipset), and his existing Windows configuration should boot right up.