Totally bizarre WC Issue

w3bbo · 24 Mar 2009 at 19:26

Are you sure the copper on the block isn't conducting electricty by shorting out on the card somewhere. I had this once with a chipset block, temps were skyrocketing but all watercooling equipment seemed fine. Turned out it was a nut catching on a mosfet :eek:

. Luckily everything still worked fine afterwards. Something you might wanna consider.

WJA96 · 24 Mar 2009 at 22:05

w3bbo said:
Are you sure the copper on the block isn't conducting electricty by shorting out on the card somewhere. I had this once with a chipset block, temps were skyrocketing but all watercooling equipment seemed fine. Turned out it was a nut catching on a mosfet . Luckily everything still worked fine afterwards. Something you might wanna consider.

That's interesting as the power figure in GPUz was very high indeed - something like 77A whereas the other cards in the system were only pulling 25-35A. I have three of these blocks fitted to BFG cards and they're very good, but this is an original EVGA card bought from the 'bay. I'll take the block off and check, as the fact that EVERYTHING was very hot just leads me to believe that everything was working, it's just that the system wasn't able to dump off the heat.

Alexdaredevilz · 24 Mar 2009 at 22:56

there isn't a flow direction on the card block that might be a problem?

Alex :-)

Bubo · 25 Mar 2009 at 08:44

WJA96 said:
That's interesting as the power figure in GPUz was very high indeed - something like 77A whereas the other cards in the system

Only 924 Watts then. :eek:

Short on the card seems very likely.

gurusan · 25 Mar 2009 at 11:26

Bubo said:
Only 924 Watts then. Short on the card seems very likely.

3d clock voltage is 1.18V so that would be about 91W

Bubo · 25 Mar 2009 at 12:38

gurusan said:
3d clock voltage is 1.18V so that would be about 91W

Gcard uses 12V direct from PSU doesn't it? 25A at 12V for a normal graphics card equals about 300W which sounds about right for a properly working graphics card. By your calc a card running at a normal 25A would only be using 30W of power. Something wrong there surely?

LupoLover · 25 Mar 2009 at 12:50

I had my GTX with a EK full cover block. Rad was cool and so was GPU-z Loading 47C. Must be a problem

WJA96 · 25 Mar 2009 at 18:30

LupoLover said:
I had my GTX with a EK full cover block. Rad was cool and so was GPU-z Loading 47C. Must be a problem

Yes. I agree. Major problem. The EK blocks are good, but the EVGA ones are better, just this one looks like it needs RMAing unfortunately.

Bito · 25 Mar 2009 at 20:40

Are the inlet and outlet of the waterblock offset from one another ?

gurusan · 25 Mar 2009 at 20:49

Bubo said:
Gcard uses 12V direct from PSU doesn't it? 25A at 12V for a normal graphics card equals about 300W which sounds about right for a properly working graphics card. By your calc a card running at a normal 25A would only be using 30W of power. Something wrong there surely?

GPU-Z reads the VRMs which feed only the GPU as far as I know, so it doesn't account for the power consumed by the rest of the card.