GTX 280 and DD Tieton cooling block woes

[newconroer]

I got the block on the GPU, with the copper backplate for the backside RAM.

I used very little Ceramique, and idle temperatures were 42sh, and I knew something was wrong just from that. I let ATi tool give it a burn in and temperatures sky rocketed really fast.

I took it apart, cleaned it, applied a thicker layer, and tried again. This time idle was 38 at full clocks, and with ATi tool going, I peaked at about 65. I though 'ok, well I was hoping high 40s at the most, but let's move on to Crysis.' After less than five minutes, temperatures broke 80c!

Back I go for a third attempt, this time putting a layer of Ceramique more than the original and second attempt combined. Idle showed 35 at full clocks, ATi tool running resulted in 55c peak. Onto Crysis: 65c for a few minutes, then BAM they shot up to 85c+.

Fourth attempt : I used a little less this time, getting a nice even spread as always. I could barely make out the word "Nvidia" underneath, as I always attemp to do. Idle again was at 35/36c, ATi tool run 55c, Crysis 60sh, then started breaking 80c.


The cooling system is a modified EXOS II, which runs only to the GPU itself. I've used it previously on a 640 GTS, 8800 GTX, 9800 GTX and always mainted 42c or lower through any test, any program(s). I've used the same methods in all of the build, paste, pressure, etc.

The only difference is the block I'm using and the card (280).

At this point I can't figure it out, even the copper backplate, which others who have used the DD Tieton say it's 'cool' to the touch, mine is a lot cooler than on air, but I can definatley still feel the heat. The block should be fine, others seem to get away with less than 50c load using lesser sized radiators than mine.

I'm starting to think it's the card. With the stock cooler it idled at 43 and pushed over 95 at times in Crysis even at 80% fan.

I was hoping even if it was the card, the liquid cooling would be able to support the heat output, but I suppose not.


Anyone else think that's the most logical explination?; that it's a faulty card in terms of heat output?

m

[intel igent]

maybe a bad mount? check to make shure the card is seated properly

[newconroer]

Ya, well after four attempts, I'm fairly confident I've mounted it properly. You can only take it off and put it back on so many times. It's pretty difficult I think with modern GPUs to get the mounting wrong. Flush is flush if you know what I mean.

[intel igent]

well it is quite possible that your GAINWARD card uses a non-reference board layout which would explain a poor mount, know what i mean?

[newconroer]

Layout is identical, everything matches up.

I'll take it off and try again. This time I'll go hog wild with the Ceramique.

[intel igent]

too much TIM can act as an insulator.....

[newconroer]

I know, remember my first attempt I went skimpy skimpy, but I admit it was too little. The third and fourth attempt were bang on the amount I'd usually go with, but at this point I don't know what else to try.

I've ordered another GTX and hopefully it was just a bad batch.

I was a bit adamant about doing so, as this card clocked really well, but then again, my fps went up by two so hehe.