TechSpot - PC Technology News and Analysis

I have had a problem with my system overheating for as long as Ive owned it.
I blow it out for dust almost daily (its a habit now), all my fans are working, I have replaced the stock thermal paste with arctic silver once myself and once by the local computer store.
My system specs;
Gigabyte k8nf-9 v2 mobo
AMD Athalon 64 3200+
AMD ATHLON64 certified CPU fan & heatsink
Nvidia Geforce 7800GT 256 mb 16x pci-e
NZXT PF-500 500Watt Power Supply
2Gb Corsair Value Select Dual Channel ddr
Aspire X-Navigator Aluminum ATX Mid-Tower Case
I had an extra fan mounted to the northbridge sink because the guy at the local shop noticed how hot it ran but unfortunatly this didnt fix the problem.
I took a couple screens of speedfan at idle and again after a short while of gaming. I had to stop gaming quickly to catch the temp as close to failure as I could.

http://www.techspot.com/vb/attachmen...1&d=1189654518
idle temp.JPG

http://www.techspot.com/vb/attachmen...1&d=1189654518
gaming temp.JPG


As soon as the one listed as core gets to 120 my whole system shuts down.
I also run Core Temp 0.95 and It lists only one temp and it is always the same as the one listed as core.
The only way I can play my game is to sit the system on top of the A/C vent in the floor with the a/c unit running. My problem now is that its getting cooler outside and the a/c isnt running as much so I crash alot sooner and more often.
Please, any help would be apreciated.

Attached Images

	idle temp.JPG (38.8 KB, 16 views)
	gaming temp.JPG (39.3 KB, 15 views)

Your 5v rail is way off. I suggest replacement of the PSU before you do anything else.

Are you saying that your cpu reaches 120 C? That should not happen at all. Try another cpu fan assembly. Your entire system is in danger of other permanent damage until your temps get right.

I don't see how your CPU core temp could be that high. I've had my computer crash with temps as high as just 70C. I think there is some error involved here, but AlbertLionheart is correct. It is new PSU time pronto.

This shows the onhttp://www.techspot.com/vb/attachment.php?attachmentid=22144&stc=1&d=1189805889
temp addresses.JPGe listed as core with a pci address so maybe NB or Vidcard?

Also, while watching the voltages I notice they fluctuate alot, is that the problem you guys are reffering to?

Attached Images

temp addresses.JPG (36.2 KB, 4 views)

I've never seen PSU voltages that wacky on a running machine. Better get a new one. Other than the one goofy reading (120 C) the temps seem about normal. This is a guess but, if anywhere in the CPU actually hit 120 C it probably would have failed by now.

I'm gonna agree with Albert and say that it's most definitely times for a new PSU. Also, 75C on a CPU is not good either. It's not complete danger zone yet, but it's not safe either. I would look into getting a new heatsink and/or fan. 120C doesn't seem likely for anywhere in a machine. The CPU is usually the hottest part and I've never seen one go above about 70-75C without shutting down.

I really dont think the computer would even boot with those voltages. Try running Everest and see what it reports:

http://www.majorgeeks.com/download4181.html

I would concentrate on getting those temps down first. Open the case and make sure all the fans are clean and running properly. If they are, remove the CPU heatsink and fan assembly and clean of the old thermal compound using alcohol. After cleaning both the CPU die and the heatsink, reapply a small amount of new thermal paste and reassemble the heatsink and fan