[LLXX]

I decided to do a stress test on one of my older machines, a Celeron 2.4 that's been running at 3.2GHz for the past few months. I started the Prime95 stress test with the temperature monitor on and approximately 10 minutes into the test, the machine started to sing. At first it was a high-pitched intermittent warbling sound, which then developed into a rather loud squealing - like a "pig being tortured". The temperatures were still normal: CPU was at 42°C and CPU fan was still running at ~3600RPM, Vcore at ~1.575. I immediately pulled off the side cover, but everything looked alright. There was no obvious source of the squealing - the whole case seemed to be emitting this rather terrifying sound. All the fans were operating normally, none of the chipset heatsinks were hot, but the voltage regulators slightly warm.

After a few minutes the squealing became a siren-like whine, increasing and decreasing in pitch. The CPU continued to heat - it was at 50°C now. Then, the whining turned into squealing again and bck to the warbling sound it started with, and finally disappeared when the CPU was at 55°C. For the rest of the 4 hours the CPU remained around 60°C and the sound was not heard from again.

I restarted the test after the CPU cooled back down to idle (30°C), and again 10 minutes in the sound reappeared. It seems to be emanating from the case itself, but the case does not seem to be vibrating. All the other components are also silent. Stopping the test, letting the CPU cool, then starting it again returns the sound. I've also tried other CPU-stress-testing programs and they also exhibit this behavior - sound starts ~40° temperature and disappears at around 55°.

System:
Celeron 2.4 Northwood @ 3.2
256Mb DDR400
Gigabyte GA8S655FX mobo

What exactly is going on here?

This post has been edited by LLXX: 21 February 2006 - 02:10 AM

[RJM]

The only thing that I've heard squeal like that in a computer is a power supply at or near it's limit of power.
A SMPS can squeal quite loud when it goes into current limit.

This post has been edited by RJM: 21 February 2006 - 07:51 AM

[ripken204]

ya, power supply is definetly the first thing to look at, have u tried to pin down the noise yet? if you rly want to know then take out the psu and reconnect it so that the psu is out of the case, then try priming again. hopefully you will be able to hear the sound from the psu. if its not the psu then it could be a capacitor on the mobo..

[ssmokee]

Maybe the bearings are going in one of the fans.

[LLXX]

Tried again today. Now in addition to the squealing it's making a rapid fluttering noise like a trapped butterfly.

ripken204 said:

ya, power supply is definetly the first thing to look at, have u tried to pin down the noise yet? if you rly want to know then take out the psu and reconnect it so that the psu is out of the case, then try priming again. hopefully you will be able to hear the sound from the psu. if its not the psu then it could be a capacitor on the mobo..

The sound isn't coming from the PSU - it's a 580W so I doubt it'd be near its current limit. I disassembled the whole system so that all the components are sitting on the desktop, and tested. The sound seems to be coming from below the heatsink...

Removed, cleaned, and reseated the heatsink with some new thermal compound, then tested again. Now it's not squealing or fluttering anymore, it's buzzing. I've inspected the mobo closely, there appears to be no damage to any components on it.

All the case fans weren't on so it's not those either. The sound isn't coming from the PSU so its fans I suppose are fine. The CPU fan isn't the problem either - swapped it with one from another machine and tested, the sound still occurs.

What could be beneath the heatsink that's causing this noise? The CPU itself?

This post has been edited by LLXX: 21 February 2006 - 10:01 PM

[ripken204]

what heatsink is it? that is very strange, could it be a capacitor next to the socket, do any capacitors look like they have fried, oozed something, or are bloated?

[LLXX]

It was the stock Intel 'sink, but I swapped it out with another one and it still makes the same sound.

The capacitors next to the socket look normal, but one of them feels a bit warmer than the others. Could that be it?

[ripken204]

well im not exactly sure how hot they are suppose to be, but they can let out an anoying sound if they are broken or dying.

[LLXX]

I ran Prime95 for 6 hours - CPU remained at ~60°C and showing no signs of instability, although the sound is now louder and instead of intermittently occurring as before it's now a continuous screeching.

Following the priming session I ran BurnCPU for another 3 hours. Again, no instability.

I even lowered the clock back down to 2.4GHz but it still made noise.

Maybe my mobo is dying... it's screeching continuously in the background as I write this post, and it's now making that sound rather loudly and at idle too. I'll test it continuously for another few days to see what happens.

[LLXX]

Update:

Something finally decided to give up - I heard a loud pop followed by a short frying sound just a few hours ago. Surprisingly, the machine continued running, so I shut it off, pulled out the fan and CPU, and inspected the mobo closely around the area where the sound was coming from. There seems to be no damage to any components, it all looked like it did before. I reinstalled everything and turned it back on, it still works (doing some priming now) but so far I haven't heard the sounds nor noticed anything else weird.

This post has been edited by LLXX: 25 February 2006 - 01:43 AM

[CptMurphy]

**** man, maybe ur pc is haunted. That or maybe one of ur components is fried. You should check the status of every one of 'em, HDD, FDD (if any), DVD-CD, video card, w/e. Anything that might spin.