Thats a close call...

[Cygni]

So heres the story:

I woke up this morning, and my computer was off. My computer hasnt had any kind of reboot for days (since i last installed the new Cat's), and before that had been running for weeks without any shutdowns. So to find it off was quite surprising. I figured Housecall, which was running when i left for the night, had caused a shutdown or something. So i rebooted, system powered up and sounded fine, and went straight into XP as per normal. 5 minutes in, a hard freeze. I figured i had simply overclocked too far when i put in the new Omega Cats, so i boot and ease the memory timings and multiplier back a little, as well as taking down voltage and GPU timings... reboot, same thing, hard freeze 5 minutes in.

So i figured something was pretty wrong. Boot into the bios, and checkout the PC Health menu... sure enough, Core Temp: 83 degrees C! I immediatly reach for the switch and before i can even hit it, the computer (set to auto kill at 60C) kicks off and a bright red LED comes on on the board. Im using a ThermalTake TMD fan, which is impossible to hear over the whine of the Antec case fans, so i didnt notice that it had somehow sized completly. How a TMD fan, with no moving parts other than the blades themselves, siezes in the middle of the night is beyond me.

Fearing the worst, i slapped the stock fan from a 1.0 Duron on the Applebred... and sure enough, she survived! I downclocked the core and took the voltage way down, and the little HSF seems to be alright at this level until i can get a new fan. How the little core survived is beyond me... in fact, im typing this using it right now. I cant help but remember that old Toms Hardware video of the 1.4 Athlon frying when the HSF was removed, AMD's comeback video, and the outcry from the community for boards supporting AMD's auto shutdown technology.

Well, i for one am quite glad that Shuttle decided to include that little bit of tech on my AN-35N U, cause it saved me a potentially costly deal. Now I gotta write a mean email to Thermaltake/ YS Tech.

Important bits of the system from the story:
AMD Duron Applebred 1.6ghz 1.5v @ 2.128 (16x133) 1.65v
ThermalTake TR2-M2 TMD HSF
Shuttle AN-35N Ultra (Nforce 2 Ultra)
Antec tower w/ 2 too-loud case fans

Max normal core temps for the system are around 46 degrees after a night of Prime, btw.

 

[tribbles]

Glad to hear your system is okay. That would scare the crap out of me.

 

[VirtualLarry]

Interesting story, I thought at first that you were talking about a P4 system. Good to hear that AMD's thermal-shutdown solution (implemented in the mobo) is working as well. I had the chance to test the P4's thermal-shutdown too, recently, and it behaved nearly identical to how you describe it.

As for the TMD fans, how is the one that you have, attached to the heatsink? I had heard that they should use *plastic* screws, or plastic washers, to prevent crushing/shorting the metal coils inside the fan-housing that allow them to operate. There is/was an issue with completely shorting the mobo too. I've never used a TMD fan myelf; hopefully they've fixed those issues by now.

 

[Cygni]

The TMD fan itself (made by a company called Y.S. Tech, not ThermalTake itself) was held onto the HSF with four black long metal screws. I didnt notice any rubber washers, but its possible the holes through the metal casing on the TMD were lined with rubber or an insulator of some sort. The fan itself never gave me any grief. It was really quiet, cooled great, and i never tampered with it in any sort... which is part of the reason its catastrophic failure was really a surprise to me. I sent ThermalTake an email, and i will post the reply if i get one. The fan really is sized solid like the whole thing is rubbing all the way around. Very strange.

 

[Cygni]

Just found this on Newegg:

"We bought 10 of these to put in rigs for our gaming center. 9 of the 10 are working fine. The last one however, after 4 months of usage seized up. We opened up the casing and discovered that the magnetic strip around the perimeter of the fan was starting to come up. We pressed it back in and hit the power. It ran for a few seconds then the strip popped up again causing the fan to click or completely seize up. Fortunately the systems we have shut down automatically if the CPU fan stops running, otherwise this could have been a very expensive design flaw. None of the others have done this, here's hoping they don't. Otherwise it was a handy little HSF."

Im gonna have to disect mine when i get home... see if its the mag strip comming up on mine too.