T42 freezes occasionally, displays lined screen

[oldsguy]

I can't really seem to determine a triggering factor except maybe for temperature/processor load. The system will do it randomly, with stretches of stability between 10 minutes and 2 weeks of normal functioning.

I replaced the motherboard on this T42 with a new T42p MB recently, with the FireGL 2 gpu. I failed to replace the fan, though, and had some overheating problems occasionally. I think this might be the root cause of my problem.

Whenever it freezes, the screen displays solid white and black lines, and music may play for another 30 seconds. Very occasionally, "VPU Recover" restores the system, but very infrequently.

Is this just a plain case of having to replace the MB again? Or would reflowing the GPU help with this? Since it was a refurbished MB when I bought it, I wouldn't expect this kind of problem.

[zdriver]

What fan are you currently running in the T42, long or short?
The long fan heatsinks the gpu as well.
What temps are the system components running?
If you purchased the board from ebay, refurbished probably means it was a system pull.

[oldsguy]

When I first installed the MB, I reused the short fan that was original to the computer.

After I noticed that the computer was running hot, I installed NHC and a new long fan, which now keeps temps down.

I just set NHC to keep my CPU temps below 50 by running the fan harder. I think this has mitigated the freezing problem, at least temporarily.

I'm thinking that because I mistakenly used the short fan originally, I might have "burned" my GPU.

Would reflowing help me, or is that for mechanical fatigue only?

[madkat]

did you use FRESH an GOOD thermal paste on the CPU when replacing the fan?

[oldsguy]

Yes, I slathered on the paste.

Right now, my CPU temp is 37 and GPU temp is 42. is this an acceptable variation?

I hit 80 on the GPU a few times before installing the new fans. do you think it's permanently damaged?

[oldsguy]

I just realized something interesting --

I got tired of waiting so long to reboot after crash, so I removed a bunch of programs from startup. One of these was NHC. Ever since I removed NHC, in the ~20 hours since, I have not experienced a single crash!

Could NHC have been causing my crash problems?

[visionviper]

It's possible a setting in NHC could have been causing problems. But since you removed other programs as well its hard to say for sure. The only way to find out would be to re-install NHC and see if you have the same problems again.

[madkat]

maybe some of the programs you've uninstalled caused some overheating that triggers the gpu problem...
before the surgery (while i had the original T41 mobo) crashes occured only on AC, when on battery, the speed was lowered, ati powerplay was active, so, less heat - no crash

[oldsguy]

We may be beating a dead horse here, since I haven't been having any problems lately -- but it might be a good discussion for academic purposes.

I took four programs out of the startup sequence via MSCONFIG. NHC, MSN Messenger, AVG virus check and "mystery" program with no name. Could have been a virus -- who knows.

I don't think the subsequent crashing problem is the same as the prior overheating problem. I watched NHC like a hawk and never saw the GPU break 50 -- most of the time, it sat at 42 immediately prior to crash. CPU temps never exceeded 70 (these are all temps after installing the long fan.)

Definitely no crashes occurred before installing NHC with powerplay. I am thinking that this might have been the triggering factor, since it directly accesses the GPU's settings and may either be buggy or accessing a part of the GPU that got fried during the overheating problems.