Hot Issue

#1

For the past couple weeks or so I guess, my PC has been crashing...

Thinking that it was an XP simply needing to be formatted issue, I dealt with the crashes (which XP reported as memory problems) until I saw the time in my schedule to format...

So, this week, I started to format, back down to reliable and trusty Windows 2000...And just run XP and others off the nifty Virtual Machine program I just got if need be...But for the past couple of nights, I've been waking up to find my computer on, but not functioning...And today, it got worse, I rebooted, and it gave me a profile error, I rebooted again, it gave me a failed system file error, I rebooted once more, and it never made it past post...

The only thing that I could even think it would be is a cluttering of dust in my machine, so I unplugged it, gave it a decent cleaning, and fired it back up....Worked for a few minutes, rebooted itself and never posted...

The way the machine was behaving, is as if I was experimenting with overclocks...But I have a beasty ass heatsink, with not one but 2 90mm fans on the side of it...No way the chip should be overheating...Or so I thought...

So, I rebooted the system to fail safe defaults, which puts the system on a 100 MHZ FSB, instead of the normal 200...I went into the bios and saw that my chip was @ 66 degrees celcius, and this is at idle...So, there is no doubt in my mind that the problem was with it overheating, but what I'm trying to figure out now, is what the hell..? Could it be my chip has gone bad..? Do heatsink cores...Give out..? Broken capacitor or transistor on my mobo..?

My instincts tell me that its the chip, but I could be wrong, and I've never seen a chip fail like that before...But heat would explain the crashes and such...I've also checked my chip it doesn't appear to be cracked or damaged physically...Any suggestions..?

#2

Um, if I were you, I'd rip out your CPU, dis-attach the heatsink, clean it out with some heat-sink removal liquid (get em from places like Tandy and stuff - don't get the $2 shit though), and re-apply the heatsink with some decent thermal paste (ArticSilver 3 is pretty good).

Also, this'd be a good time to clean out the fan if its a bit noisy.. /wink.gif" style="vertical-align:middle" emoid=";)" border="0" alt="wink.gif" />

If you need a tutorial, plus some good products I'd recommend (the actual ones, not just things off the top of my head) I can write you a tut/list em, lol

I doubt it would be anything to do with the chip itself, otherwise your computer wouldn't be workin at all... /wink.gif" style="vertical-align:middle" emoid=";)" border="0" alt="wink.gif" />
I suspect the thermal paste/wax (wax is probably what it is if you haven't changed heatsink ever) has gone all black and hard - it isn't transfering any heat.

-tech,
who.am.i

#3

if you have access to a multimeter, you may also want to check the voltages coming from your PSU. IF the PSU has been running hot for a long time the caps in it could be failing giving you erratic voltages and making your system unstable. Have you ruled out your memory as being the culprit? Booting with minimum amount of peripherals attached (1HD, optical drive, on board video if avail) will also help diagnose PSU issues. Don't most processors have some kind of built in thermal-shutdown-before-meltdown failsafe? Can you boot with a knoppix (or similar) live linux cd and run reliably?

May be time to take the plunge and go for that Achilles upgrade you have been talking about for the past many months!

#4

Oh, I am definitely building Achilles (It may be my birthday present next month) but I didn't want to completely scrap this machine, I was hoping to relegate it to server/router duties...

Well, what I've done is unattached the heatsink and disassembled it completely, cleaned it pretty good, and reapplied some Artic Silver 3...

Right now, my temps say about 38 degrees, with the door off and the FSB still @ 100 MHz...That still seems a little high, but its certainly better than the 60+ I had prior to cleaning the thing...I'm guessing when I put the door back on, and move the FSB back to 200, it will go up about 3 or 4 degrees and be OK...But for now, I think I'm gonna leave it as is...

Thanks for the tips guys...

#5

did you happen to photo-document the process?

#6

Ahhh, no, LoL...But I think its fairly stable @ this point...

#7

So you didn't need my tutorial, good.. lol
I knew you were capable enough, hehe

So, i was RIGHT..! YAY..!
You know what this means, don't you..

#8

Well...

I moved the system to its default 200 MHZ setting...And, after a few minutes, it crashed...Quick flash, and then nothing...

Moved it down to a 166, and it worked for a little while...And then when the computing got a little heavy, it crashed...Moved it to a 133 FSB, and it operated pretty well, but right now I'm back on the 100 FSB to try and figure it it will crash when I multi task it this time...

I swapped out both of the RAM sticks, it crashed with both...But, I do have a memory diagnostic, and the RAM passed w/flying colors...So, I don't think its the RAM, or the CPU getting hot at this point...When I opend my case after a crash, everything was literally cool to the touch except my video card...If it does it this time, I'm all but certain that's what the issue is...

#9

what PSU are you running and how long has it been in service?

#10

Damn e, its almost like you are here troubleshooting w/me...I think it may be my powersupply...I have an Antec True Power, 550 Watt...It's been in my system for about 3-5 years I think....I have an add in card for my motherboard that gives the machine 6 stage power to work with, which aids in overclocking....Well, I woke up and decided to take the card out...The machine is still crashing without warning, but its rebooting, and not just hanging...My guess is that the daughter card was trying to regulate teh power, and when it crashed, it hung the system...

I'm not sure what I could do to give the PSU a few more legs...