Strange graphics card problem (Im stumped)

[Aurek]

So a while ago a strange error started apprearing at boot, the computer would emit a series of long audio tones and a message telling me that my graphics card was not recieving power would appear. The PC would not boot. It went away after a while so I forgot about it. Fast foward a bit and now when my computer is being stressed (i.e playing a modern game) the catalyst control centre informs me my graphics card is overheating, though the system returns to safe temperature when idle or during simple tasks. The strange thing about this is that is began to happen only after all the computers in my computer room had been moved and the room cleaned, care was taken with the computers however and the only thing that changed after that was how the computers are plugged in. This overheating warning would appear at seemingly random times though with no real consistiency as to the circumstances.

Now It will help if I describe my situation and my computer:

Dell dimension 8400
P4 3.0ghz
1GB pc 3200 ram
PCI-E X800 XT
80GB SATA

The only changes to this system after purchase have been the addition of a 120GB SATA hard drive and a DVDR/RW.

My computer is situated in a room with three others, two of which are high end gaming computers, the fourth an older Dell. Unfortunatly all the computers (and anything else) are plugged into one wall socket. We have no choice in that regard, its an old house. There are long four plug adapters plugged into each other, that each pc is then plugged into. As I said there has been no real consistiency as to when this overheating occurs, others PC's on or off, room hot or cold. It would happen every now and then before disappearing for a while, though now its constant and I cant play anything.

So far I have updated my graphics drivers, cleaned the fans and heatsinks in my pc with a low power vacume cleaner, varified all the fans are spinning and that all the cable connections are sound. Most recently I have taken the CD drive that shipped with the pc, out, in order to improve ventilation. All to no avail. I have never overclocked the system. Im at a loss and any help will be very greatly appreciated (as I want to get back to playing oblivion!)

 

[bluntside]

-K- this might B UR problem, U can only plug in so many at atime, after that The wall socket becomes a bottlenech for certain powerflow, U might want 2 becareful cause it might damage all of ur comuters that R conneted 2it. Might try 2 check on the power connection on your gpu to make sure all 4 pins are connected together.

Good Luck

 

[RichPLS]

If you have several high end gaming PC's off one outlet, surely you ponied up for a decent 1500VA UPS to insure clean power going to your rigs?!?

 

[Aurek]

I had wondered if it could be problems with power coming from the wall socket. I would have thought that if we tried to draw too much current from the socket the fuse would simply blow? but if the power is fluctuating....

I checked the ATI website, and in the customer support area they talk about the x850 cards throwing up overheating warnings with catalyst 5.6+ drivers. They also say that the problem may arise with other ATI products. I changed my drivers back to 5.5 and I no longer recieve the warnings, after playing a graphically intensive game for a while I exited and the catalyst control centure listed my cards temperature at just over 100 degrees c! That cant be reporting right, as a temperature that high above safe operating temperature would surely cause graphical artifacts and bad performance very quickly?

 

[Aurek]

I just downloaded Ati tool 0.25b14. It reports my GPU temperature as climbing into the same ranges (100C+) when under load, however when scanning for artifacts it brings up nothing. Reading the internet, people have talked about getting artifacts at 85C. Im not getting choppy performance, no random lock ups or crashes. Could the GPU thermal sensor have become faulty somehow?

 

[SuperFly03]

Man you really do have an issue, ok im stating the obvious I know. Have you checked the seating of of the GPu's heatsink? Maybe it got knocked loose and isn't in full contact with the GPU core?

You many want to think about buying an aftermarket cooler for it, if you're that concerned. 100c would concern me, I usually get artifacts at a much lower temp than that.

I honestly don't know much about power distribution problems in a house, but 4 pc's on 1 outlet sounds intense. You Also may want to check for dead fans in the case, not just the GPU cooler. I never have owned your specific Dell, but they aren't exactly known for optimum air flow in their cases.

I'll let ya know if I think of anything else

 

[Aurek]

I usually get artifacts at a much lower temp than that.

Thats why I think it must be reporting the temperature incorrectly, If people are normally getting artifacts at 80C+ there is no way I should be able to run my system at a "reported" 100C for an hour with no unusual effects at all. At that temp the card should be fried surely?

 

[SuperFly03]

I don't know about fried. The GPU Core shutdown temp on my 7800GTX is 120c so 100c may be sustainable, though I probably wouldn't recommend it. Ifyou had a temperature probe you could toss it on there and see what it really is. :tongue: