Southbridge chip burning itself out...

[harry_liddle]

Hi,

I am facing a really strange problem since about a month and I still havn’t been able to figure out the cause:confused: .

My older system’s motherboard is Asus K8S-MX, with a 965L southbridge, and I have been running two sata HDDs on it in RAID 0 since i can remember, about 2 years now, as well as a DVD writer and an occasional USB device as well. since about a month, the system started crashing randomly, I ran a lot of tests to catch the culprit, be it the RAM or the HDD, but the tests were always fine. One day I just opened the cabinet and kept it on my table for even closer inspection, and it was then that I almost burnt my finger touching the southbridge, and since then my system is lying on my desk cooling its southbridge with a table fan. The chip(Sis 965L) does not have a heat sink and was running fine till now. But now it has a discoloured spot on its face, and touching it there tells me it is really overheating, thoug the rest of the chip is fine. And even now, I can’t have a HDD intensive operation like a torrent client as it does crash the system, though the defragger runs fine, but still scares me.

I am going in for another motherboard in this ones place after the chip is fried, but i really want to know what could be the reason for this and so suddenly:confused: , as I have already made a custom active copper heatsink:cool: for the chip and wont be able to use it, it took some real effort as you can see the AGP slot shadows the chip on one side and the chassis fan connector on the other, you will also be able to see the discolored spot that gets Real Hot.
http://www.picscrazy.com/thumb/burning965l.jpg
Could some one help me out here, also will need a new motherboard for my Athlon 64, so some recommendation will also be needed there as well, the board should be good at overclocking, as the processor has run that way since day one.

Thanks a lot

[rimmer]

I would suggest buying some RAM heatsinks off ebay, preferable copper ones, and doing what I’ve done: Placed RAM Heatsinks on every chip on my motherboard and all the chips on my hard drives, as they get exceptionall hot. Just make sure theres no metal on metal with the heatsinks and sensitive parts of the circuit boards.

But it looks like an expansion slot, (looks like an AGP one) is in the way of any realistic piece of cooling going into place on that chip. You may be able to fit 2 RAM sinks on it, but not 3, meaning you will only be covering 2/3 or half of the chip.

[harry_liddle]

@rimmer

I would suggest buying some RAM heatsinks off ebay, preferable copper ones, and doing what I’ve done: Placed RAM Heatsinks on every chip on my motherboard and all the chips on my hard drives, as they get exceptionall hot. Just make sure theres no metal on metal with the heatsinks and sensitive parts of the circuit boards.

But it looks like an expansion slot, (looks like an AGP one) is in the way of any realistic piece of cooling going into place on that chip. You may be able to fit 2 RAM sinks on it, but not 3, meaning you will only be covering 2/3 or half of the chip.

Yes, Rimmer, I actually fabricated an U shaped heatsink from a copper plate and a spare small fan, and its working fine, still I do check the chip status in about 2 or 3 days, it still gives off a lot of heat and is burning itself out, very slowely now. I also got a PCI SATA Card so i could connect my other HDDs to this motherboard as it only has 2 SATA slots on it and was trying to backup the RAID 0 inthe last few days and so i wasn’t able to reply to you earlier.

All I want to know is why is it doing such a thing and that too now when the 300GB RAID 0 running on it is almost full of critical data.