It has been over two weeks now and I have googled high and low, and searched many a forum, but I'm no closer to finding any resolution. I'm sorry for being so wordy, but maybe a question well asked is half-answered as they say.
I'm currently not having random reboot problems, but instead I have troubles getting the machine to boot up properly. Here is why specs, the symtoms, and what I've done so far to troubleshoot the issue:
AMD Athlon64 3200+ with retail heatsink fan (I could have sworn it was one with a larger L2 cache)
Chaintech ZNF3-150 motherboard
1gig RAm - twin Corsair DDR PC3200
ATI Radeon 9800Pro 128M
80GB WesternDigital HD
Enermax PSU EG475P (470W)
WinXP SP2
I bought all these components a year ago (2/17/05 to be exact. Glad I got an extended warranty). It has been working beautifully up to this point.
One day while it was running it decided to reboot for no reason. Since then I closely monitored the temp but it was always 48C or less under any circumstances. The machine would run fine for an hour or so then reboot again. The reboots kept occuring more rapidly up to the point where it would reboot before it finished posting.
I could find no indication it had anything to do with Windows. In fact it would sometimes reboot while I was in BIOS. I tried changing to automatic shutoff temperatures to 65C or 70C but that didn't seem to change anything. At the time I couldn't check the memory because it would reboot parway though the test at different levels of progress.
I took apart the machine, unplugged all the connectors and re-inserted all the cards. Random reboots still occured. I unhooked everything but the floppy and put in an ancient geforce256 video card and only 1 stick of ram. Random reboots. I tried using the other stick of ram by itself. Random reboots.
One night I decided I'd be brave and take off the heatsink to the CPU and look for something there, maybe the temps weren't being reported accurately. I had some extra thermal grease that came with my mobo so after I inspected the CPU and put it back I cleaned off the gray goo that came with the retail unit and applied the new thermal grease. That worked...for about a week. During which time a ran the mem test fully through 5 cycles and no errors. Temps where always below 45C.
Sometimes the random reboots would cause the machine to freeze during bootup and it wouldn't respond to the power or reset switch. Again, I removed the heatsink and reapplied thermal grease. Again everything worked fine...for a little while anyway.
The random reboots have not come back thus far, but now I will go to turn my machine on and I will hear the CD-ROM drive whir and click as normal, as does the hard drive like it always does just as it boots, but then they just keep doing that. The monitor's light keeps flashing since the machine never gets to a point where it sends a signal.
Last night when I begrudgingly took the heatsink fan off again I noticed a sheen on the CPU board edge. I also noticed it around the edge of the socket in the same place. Almost like a liquid had gotten in there. May have been thermal grease that spooged over the edge (though I was careful to use a tiny drop in the center of the heat spreader). Anyway, I carefully mopped up whatever it was with a qtip, cleaned and re-applied thermal grease and was back in business.
Tonight I go to turn it on and again the CD-ROM and HD did their endlass dance and the machine never really booted. I cycled the PSU's power switch and tried again and everything started up, although the CD-ROM did cycle more times than I think was normal before going silent.
I just ordered some Artic Silver 5 just so that they next time I have to touch the stupid thing I'll at least know I'm putting down some quality stuff instead of whatever generic thermal grease came with the mobo.
But if this was really a CPU overheating problem, why would it not take any time at all to overheat from a cold start and not even get to the normal boot sequence? The random reboots would have made sense for an overheating problem. Perhaps CPU damage from the previous overheating problem? Then why is it working now?
I don't have a spare PSU to try and eliminate that from the equation, but the voltages shown in bios seem to look okay. I also tried plugging the machine into a different outlet (one that was on a different breaker in the house) to no avail. The voltages do fluctuate a little, but I think this is normal, no?
I also don't have a spare motherboard or CPU to switch out to see if those might be the culprit. Maybe the movement from taking the heatsink fan on and off is just enough to temporarily fix a fickle connection between the CPU and socket. No idea at this point.
Any ideas would be greatly appreciated. Again, sorry for being so wordy. I should never post a message after 3 cups of coffee.
I'm currently not having random reboot problems, but instead I have troubles getting the machine to boot up properly. Here is why specs, the symtoms, and what I've done so far to troubleshoot the issue:
AMD Athlon64 3200+ with retail heatsink fan (I could have sworn it was one with a larger L2 cache)
Chaintech ZNF3-150 motherboard
1gig RAm - twin Corsair DDR PC3200
ATI Radeon 9800Pro 128M
80GB WesternDigital HD
Enermax PSU EG475P (470W)
WinXP SP2
I bought all these components a year ago (2/17/05 to be exact. Glad I got an extended warranty). It has been working beautifully up to this point.
One day while it was running it decided to reboot for no reason. Since then I closely monitored the temp but it was always 48C or less under any circumstances. The machine would run fine for an hour or so then reboot again. The reboots kept occuring more rapidly up to the point where it would reboot before it finished posting.
I could find no indication it had anything to do with Windows. In fact it would sometimes reboot while I was in BIOS. I tried changing to automatic shutoff temperatures to 65C or 70C but that didn't seem to change anything. At the time I couldn't check the memory because it would reboot parway though the test at different levels of progress.
I took apart the machine, unplugged all the connectors and re-inserted all the cards. Random reboots still occured. I unhooked everything but the floppy and put in an ancient geforce256 video card and only 1 stick of ram. Random reboots. I tried using the other stick of ram by itself. Random reboots.
One night I decided I'd be brave and take off the heatsink to the CPU and look for something there, maybe the temps weren't being reported accurately. I had some extra thermal grease that came with my mobo so after I inspected the CPU and put it back I cleaned off the gray goo that came with the retail unit and applied the new thermal grease. That worked...for about a week. During which time a ran the mem test fully through 5 cycles and no errors. Temps where always below 45C.
Sometimes the random reboots would cause the machine to freeze during bootup and it wouldn't respond to the power or reset switch. Again, I removed the heatsink and reapplied thermal grease. Again everything worked fine...for a little while anyway.
The random reboots have not come back thus far, but now I will go to turn my machine on and I will hear the CD-ROM drive whir and click as normal, as does the hard drive like it always does just as it boots, but then they just keep doing that. The monitor's light keeps flashing since the machine never gets to a point where it sends a signal.
Last night when I begrudgingly took the heatsink fan off again I noticed a sheen on the CPU board edge. I also noticed it around the edge of the socket in the same place. Almost like a liquid had gotten in there. May have been thermal grease that spooged over the edge (though I was careful to use a tiny drop in the center of the heat spreader). Anyway, I carefully mopped up whatever it was with a qtip, cleaned and re-applied thermal grease and was back in business.
Tonight I go to turn it on and again the CD-ROM and HD did their endlass dance and the machine never really booted. I cycled the PSU's power switch and tried again and everything started up, although the CD-ROM did cycle more times than I think was normal before going silent.
I just ordered some Artic Silver 5 just so that they next time I have to touch the stupid thing I'll at least know I'm putting down some quality stuff instead of whatever generic thermal grease came with the mobo.
But if this was really a CPU overheating problem, why would it not take any time at all to overheat from a cold start and not even get to the normal boot sequence? The random reboots would have made sense for an overheating problem. Perhaps CPU damage from the previous overheating problem? Then why is it working now?
I don't have a spare PSU to try and eliminate that from the equation, but the voltages shown in bios seem to look okay. I also tried plugging the machine into a different outlet (one that was on a different breaker in the house) to no avail. The voltages do fluctuate a little, but I think this is normal, no?
I also don't have a spare motherboard or CPU to switch out to see if those might be the culprit. Maybe the movement from taking the heatsink fan on and off is just enough to temporarily fix a fickle connection between the CPU and socket. No idea at this point.
Any ideas would be greatly appreciated. Again, sorry for being so wordy. I should never post a message after 3 cups of coffee.
but don't be shy about kicking the memory voltage up to 2.7 volts if the issue comes back and you have reason to suspect the memory is unstable. I have the same modules myself.
Facebook
Twitter