Sign in with
Sign up | Sign in
Your question

Random freezing due to unknown hardware issue (likely overheating)

Last response: in Systems
Share
July 24, 2013 3:07:12 PM

Hi, y'all!

First time poster, here. Most of the googling for answers to my problems led me to this site. So far, I haven't been able to solve it (duh), so I decided to sign up to ask this community directly. I hope I'm in the right subforum. If not, well, I hope there's a mod that can boot me to where I belong ^^

So, my computer has been experiencing random freezes since about three weeks ago. They've gradually become more frequent (about daily in the beginning, till about half an hour after booting up, now.) I've been trying a bunch of things to figure out what it is, but I just don't know anymore. It's probably overheating, but can't get much further...

Specs:
mobo: GA-Z68X-UD3H-B3
CPU: Intel Core i5 2400 Sandy Bridge
SSD: OCZ Agility 3 120GB
RAM: Corsair XMS3 DDR3 1333 2x4GB
PSU: XFX Core Edition Pro550W
GPU: XFX Radeon HD 6850 XXX

A list of things I have excluded from my suspect list:
- OS or software: freezing problem also occurs in BIOS
- PSU: tried with a friend's PSU, but didn't fix anything
- GPU: keeps freezing up when I remove GPU and plug monitor into mainboard (although freezing is sometimes replaced by a black screen, followed by the screen turning a variety of flashy colors)
- SSD: disconnected it, still froze
- BIOS: after it started to freeze, I tried updating my BIOS from F8 to F12. Could still go to U1L (UEFI BIOS), but BIOS doesn't seem to be the issue.

What remains:
- RAM: tried testing with memtest86, but apparently that can also freeze. On occasion, I manage to get a pass, and no errors are ever found, but can't run memtest for over 8 hours, like is often suggested. Freezing also keeps occuring if I put in only one stick of RAM.
- CPU: no idea how to test it properly, no direct access to a second one to try the old switch-a-roo
- mainboard: also no idea how to test

My main suspect is overheating. The problems started with a burst of very hot weather, and I can't cool my room very well. At night, the problems seem to be less prevailing, and if I put a big fan on it, it takes a bit longer to freeze up. Underclocking also seems to help some, but either I should underclock more, or it just won't cut it.

The problem with this theory is, there is no proof: RealTemp, HWMonitor, SpeedFan and BIOS all indicate normal operating temperatures (when idle it sticks below 50°C, and it'll freeze just the same when idle).

Sensors might be faulty, but they do indicate massive increase in temperature when I stresstest my CPU. It'll hit 100°C without any freezing, but when it's cooling down again, it might just as well decide to freeze when the temperatures measure about 60°C. I've also tried a feel, but neither the RAM nor the CPU consistently feel hot right after a freeze (sometimes neither of them will feel hotter than what I can only describe as "pleasantly warm").

This all leaves me wondering, is something overheating, and if so, what? Can I fix things with a better cooler (have just the standard that came with the CPU), or would it be a waste of money? Have both sticks of RAM gone bad together? Or is it my mainboard or CPU, and if so, is there any way to figure out which one it is without running off to the shop and buy these parts anew?

I hope the people here can help me find answers, because this is getting past frustrating.

Cheerz,
NecRock

More about : random freezing due unknown hardware issue overheating

July 25, 2013 6:46:50 PM

First off, thanks for the advice ^^

I'm afraid the problem still isn't fixed, though. I tried replacing with a new bar of RAM, but no luck there. I guess that pretty much rules out memory issues (three pieces, one of which newly bought, being bad seems highly unlikely).

Also bought the suggested cooler, also didn't help much. After that, I ensured the CPU fan was always running at full speed in BIOS, to rule out sensor issues, but even in a relatively cool room without heavy load on the CPU and fans at full speed, I kept getting freezes. Maybe overheating isn't all of the problem? I'm beginning to think trying to replace CPU is inevitable.
m
0
l
July 26, 2013 5:23:12 AM

Installed only the new RAM module, left out the ones I had. When I took out the CPU before, none of the pins looked bent. I am now trying to run on just one out of 4 cores, see if that solves anything (it's a long shot, but it's a shot).

I'll ask around if any of my friends happens to have a CPU I could use ^^
m
0
l
Related resources
Can't find your answer ? Ask !
July 27, 2013 8:23:29 AM

I guess it's not the CPU. Found someone who had a CPU I could try, memtest got stuck after 40 minutes running.

I guess this means it's the mainboard, as everything else has pretty much been cleared. The problem is testing, don't know anyone who's got one spare that'll fit, and asking someone to take apart his system so I can try it, is a bit much. Is there anything I may yet have overlooked?
m
0
l
August 5, 2013 8:14:59 AM

Sorry for the long wait. It seems the motherboard was indeed faulty (it took a while for the shop to verify this, but evidentially they also got freezes with parts of their own on my mobo). I now have a new motherboard, so here's hoping that'll be enough, rather than finding out there were several issues and not only the motherboard. If my system survives the next 24 hours without problems, I'll consider this solved ^^

As for the CPU, it was a weak one from someone who's into servers. It ran fine on a smaller board, but that one wouldn't be able to handle my CPU or something like that, so I couldn't test that way.

I had updated pretty much all drivers in Windows, but beyond that, I only updated BIOS and SSD firmware. I had pretty much eliminated GFX card as being the (only) problem, though, by taking it out for a bit.
m
0
l
!