D930 using Thermalright SI-120 heatsink - issues

Trinten

n00b
Joined
Aug 28, 2006
Messages
6
Okay, so I have the D930 processor, 3.0, and used the Thermalright SI-120 heatsink (and the 5 dollar adapter to be able to mount it to an intel processor, and of course a 120mm fan).

Now this wonderful adapter uses a bracket on the bottom of the motherboard that screws into a set of rails that the heatsink then snaps onto.

When I have the thing snapped on, and stress out my system (or run games), it hard locks up. So I swapped out my system, part for part, stressing it each time. Finally it came to my processor. I put in my friends processor, dressed it, and just set the heatsink on top (giving it a little push down), but didn't snap it on. It ran fine. So I figured my processor was shot.

RMA'd the processor, got the new one. Put it in, snapped everything together... same problem. So remembering the only difference with my friends processor, I unsnapped the heatsink, and started the stress test.

IT DIDN"T LOCK UP! GAH! -- however, after an hour, my CPU usage dropped off as my system throttled down the processor from heating up.

So the only thing I can think of, is that the heatsink is somehow putting bizarre uneven pressure on the chip, causing a barely-there connection that breaks down under stress.

Is this possible? I can't think of any other reason for this unusal trait. The only other thing I can can think of is to loosen up the bracket, snap the heatsink on, and then gradually tighten it back down, trying to make sure it's even tension on all screws.

Any other thoughts about this (besides slapping on the retail heatsink/fan?)

BTW, I used artic silver 5 as my thermal transfer paste, and no, I didn't get it on any connectors, or anyplace else.

Thanks guys (and gals!)
 
I've never had a CPU not work because the HSF was too tight, and believe me, those POS stock Intel heatsinks are tight. It sounds to me like the bracket is shorting something out. Is the bracket plastic or metal where it touched the back of the motherboard?
 
The bracket is metalic, with a plastic coating on it. It's a square, with an 'arm' coming off at each corner and a slight 'downward' angle (away from the motherboard, so the ends of the arms do not touch the motherboard, they are at least a few mm away. That's what generates the tension for the rails that the heatsink snaps into. The screws go through the bottom of the bracket, up into the rails, and tighten it down, the flex of the arms providing the pressure (and the heads on those screws suck, damn near stripped one of them out).

The plastic coating did not cover all of the metal, just the square where it would touch the motherboard and the part of the arms that might touch the mobo after being tightened down.

However, I was paranoid about this, so I coated the thing with liquid electrical tape. Just to be safe.

Also, it's not so much the pressure, but *UNEVEN* pressure that I think might be the culprit.

If nothing else, this weekend I will put on the stock heatsink/fan and stress it, and see what happens. If it still acts up then... maybe there's an issue with my mobo that arrises from the stress of the heatsink pulling up on it.
 
Well, an update to the saga.

I pulled off the old heatsink and fan and the accompanying bracket, dressed the stock heatsink and fan, snapped it on, stressed it, and it locked up. Disconnected the mounts from the mainboard, and it ran fine (until the system throttled down the processor for being too hot about 40 minutes later, still far longer than it runs when the heatsink is anchored).

So it looks like my issue is actually with my motherboard, and the flex that the boar gets put under by the anchoring of the heatsink is causing the issue.

The board is a Gigabyte GA-8N-SLI rev. 1.1, BIOS Rev. F5. Just in case anyone has any similiar problems in the future.
 
Back
Top