Hi Everybody,
Although I've never posted on this forum, I've been reading it for quite a while and I know there's some smart cookies in the community.
Anyways, I decided to install an Arctic Cooling S1 Rev2 GPU heatsink and Enzotech BGA ramsinks for the RAM / VRM's and other IC's. The installation on my 6950 was straightforward ----- the problem is, when I installed the card back into my PC, I wouldn't get a video signal, but the PC booted. Switching to onboard video gave me video back, although Windows didn't see my 6950 in the dev. manager. I also tried another video card, which worked, eliminating the possibility of a bad PCIx slot.
Some background info:
1) All heatsinks were epoxied onto the card with Arctic Silver thermal adhesive.
2) I cleaned the bases all of the heatsinks with alcohol and Q-tip prior to installation. Same goes for the corresponding mounting surface of the IC (RAM / VRM, etc).
3) I bought a spare set of Enzotech BGA sinks and cut them to fit the VRM's with a hacksaw. I used 2 x 5 (pins) pieces for the VRM's - roughly 14mm x 3mm (l x w).
4) I installed the remaining 3 x 5 (pin) pieces onto the 'R23' chips on the board, as I read in a waterblock installation guide that they do generate heat (and were directly in contact with the waterblock I was looking at). I also made sure that none of the heatsink were touching one another / other components.
Here are my thoughts, and I would very much appreciate some advice (or at least optimism
) as to what could be causing the card to malfunction. Although I fear the worst, I've seen some wierd stuff happen (read: miracles) in my PC days 
1) The 2x5 (pin) Enzotech pieces that I used on the VRMs were long enough so that I could use one heatsink for every two VRM's. In retrospect, is it possible that the heatsink(s) are shorting out the VRM's by connecting them electrically? Or am I totally wrong here?
2) Is it possible that the card won't function properly because the onboard fan header no longer has a fan attached to it? I'm using two 120mm fans to cool this setup. I'm doubtful of this, as I haven't seen anything supporting this claim on various forums...
3) It's entirely possible that a copper filing from the cut ramsinks may have fallen onto the card, causing a short somewhere.....time for a can of compressed air.
4) It is possible that some of the epoxy seeped out from between the heatsink and ICs, which leaves the question of whether or not this could cause a fault in the card....is Arctic Silver conductive?
5) I think that it's inevitable that I reinstall the stock cooler as a last-resort troubleshooting step prior to having a funeral service for my card. This leaves the question of how to remove and recover the epoxied heatsinks....any suggestions?
6) Anyone think that an XFX warranty will cover this? They dont seem to have a problem with removing the stock heatsink, but Id imagine that using epoxy on the ICs would classify as physical damage, and rightly so. Has anyone experienced otherwise?
Thoughts?
Although I've never posted on this forum, I've been reading it for quite a while and I know there's some smart cookies in the community.
Anyways, I decided to install an Arctic Cooling S1 Rev2 GPU heatsink and Enzotech BGA ramsinks for the RAM / VRM's and other IC's. The installation on my 6950 was straightforward ----- the problem is, when I installed the card back into my PC, I wouldn't get a video signal, but the PC booted. Switching to onboard video gave me video back, although Windows didn't see my 6950 in the dev. manager. I also tried another video card, which worked, eliminating the possibility of a bad PCIx slot.
Some background info:
1) All heatsinks were epoxied onto the card with Arctic Silver thermal adhesive.
2) I cleaned the bases all of the heatsinks with alcohol and Q-tip prior to installation. Same goes for the corresponding mounting surface of the IC (RAM / VRM, etc).
3) I bought a spare set of Enzotech BGA sinks and cut them to fit the VRM's with a hacksaw. I used 2 x 5 (pins) pieces for the VRM's - roughly 14mm x 3mm (l x w).
4) I installed the remaining 3 x 5 (pin) pieces onto the 'R23' chips on the board, as I read in a waterblock installation guide that they do generate heat (and were directly in contact with the waterblock I was looking at). I also made sure that none of the heatsink were touching one another / other components.
Here are my thoughts, and I would very much appreciate some advice (or at least optimism
1) The 2x5 (pin) Enzotech pieces that I used on the VRMs were long enough so that I could use one heatsink for every two VRM's. In retrospect, is it possible that the heatsink(s) are shorting out the VRM's by connecting them electrically? Or am I totally wrong here?
2) Is it possible that the card won't function properly because the onboard fan header no longer has a fan attached to it? I'm using two 120mm fans to cool this setup. I'm doubtful of this, as I haven't seen anything supporting this claim on various forums...
3) It's entirely possible that a copper filing from the cut ramsinks may have fallen onto the card, causing a short somewhere.....time for a can of compressed air.
4) It is possible that some of the epoxy seeped out from between the heatsink and ICs, which leaves the question of whether or not this could cause a fault in the card....is Arctic Silver conductive?
5) I think that it's inevitable that I reinstall the stock cooler as a last-resort troubleshooting step prior to having a funeral service for my card. This leaves the question of how to remove and recover the epoxied heatsinks....any suggestions?
6) Anyone think that an XFX warranty will cover this? They dont seem to have a problem with removing the stock heatsink, but Id imagine that using epoxy on the ICs would classify as physical damage, and rightly so. Has anyone experienced otherwise?
Thoughts?