Random restarts, can't pinpoint problem

M

Mnx4

[H]ard|Gawd
Joined
Jul 29, 2005
Messages
1,039
My last hard drive died, a maxtor 80gb SATA, so i got a nice WD Caviar 120GB SATA (AAJS). My other two hard drives are the first generation IDE Caviar 120GB. This is a complete fresh install of everything with the most updated drivers, service packs, and updates.

Anyway, after the install, I've had issues with the computer randomly restarting, sometimes when it's idle, sometimes after continuous tasks (virus scan, SETI@home). I've had problems with the ram before (crucial ballistix) and i've RMA'd them 3 times, generally once each year, with the most recent RMA in September. Only once had it been caused by overclocking. I've reset all timings to automatic, and they're running 3-4-4-8 2T @ 2.7v. I've run memtest for an hour with no problems. (i'll run it throughout the night tonight and update). It had previously been overclocked, but when i started having problems, I reverted back to stock clocks on everything. I also installed a fan to sit on the ram to keep it from overheating.

All of my temps are pretty good, with the exception of my northbridge, which hits about 54C on high load tasks like seti. I tried applying artic silver to the northbridge, but i don't know if a thermal pad would be better as the NB fan that came with the motherboard doesn't sit snug on top. I've got very little dust. I've got high flow air through the case.

Any ideas on what I can try to figure out the problem? All along i've figured that the motherboard was faulty and causing the ram to die, but I don't know how to test for that.
 
What kind of psu are you using? I had a problem like that with random reboots and it turned out my psu couldn't handle everything, especially my overclock.
 
Ran Memtest for 6:30 hrs and had no errors, which suprised me.

My power supply is a Cooler Master 550w. I had to replace my previous power supply in may, which was also a 550, because there was a short and my power supply's fuse blew out and couldn't be replaced easily.
 
The blue screens of death i've gotten so far have been:
IRQL NOT LESS THAN OR EQUAL
PAGE FAULT IN NON PAGED AREA
ATTEMPT TO EXECUTE NON EXECUTABLE DATA IN MEMORY (can't remember what it said exactly, went by to quickly)

I can't stay on more for than 5 minutes now.

I haven't installed any new hardware, and I've disable some of the hardware i don't need, such as SATA slots 3 + 4 and 2nd onboard lan. I've loaded default bios settings.

I don't know if the stability problem started after a new driver was installed. What's the best way to determine if it is a driver issue and which driver it is?

I've taken out all the cords and hard drive to make sure all the connections were alright. I noticed that the ramsinks that come with the ram have seperated from ram chips. I guess they've done this naturally, but don't know what could have caused it. Could this be causing the problems?
 
Have you tried doing a burn-in with prime95 or occc to see if the processor is returning errors as it heats up? An overclock may have damaged something that's not readily apparent.
 
Mnx4 said:
I don't know if the stability problem started after a new driver was installed. What's the best way to determine if it is a driver issue and which driver it is?
Click to expand...

Boot into safe mode. If no error or problems, than yah, you have a corrupted driver.
 
I did do a burn in at the beginning of the overclock, but now i'm at default speeds.

I booted into safe mode, and the first time i got an IRQL blue screen, theni rebooted and went into safe mode again and i got a Page fault in a non paged area message.

I'm going to run memtest again for a longer time, but last time i did it for 6 hours and no errors came up.
 
Memtest doesn't really put a whole lot of strain on the processor. I'd think if it were a memory issue that happens at every bootup it would show up pretty quickly in memtest. The processor could've been damaged during your overclock and now is unable to sustain even normal stress levels.
 
Is there a way to test the processor only in dos? I can't boot to windows.
 
well, is there anywhere you can get a dual core 939 processor these days? I've tried newegg, frys and mwave, but i can only find lower speed single cores.:(
 
So i just did a fresh install of windows xp. I removed 2 extra storage hard drives, as well as one stick of ram, and any unnecessary peripherals that were attached. After a lot of problems with the Win XP hanging at Registering Components for 13 minutes, I was able to boot into windows for the first time in a while. I ran a couple of runs with super pi, and received no errors. I think I did the 4M benchmark after running the 512k, 1M and 2M.

I tried restoring the files from the formatted drive when I got my first blue screen from the fresh install. It labeled NTFS.sys as the culplrit, and I had gotten other stop codes which pointed to this file. I've run chkdsk /f on it, with no errors found.

Any ideas?
 
How is the voltage on your power supply? Do you have anything to monitor it as you run these tests? At this point I'm willing to vote motherboard along with a couple others who posted, but if you can test all possibilities first (such as a failing power supply) it's always better to be safe than sorry.
 
Miiiight have figured out the problem. I think it's the north bridge. I had loaded default settings in my motherboard, which changed a vital setting. I had previously set the northbridge's fan to be fully on when it goes above 35C, but default is 55C. I've reverted back to 35C and so far, i'm able to enter in windows and I'm running two prime95s to see if something is overheating, and running ntune with logging so i can get an idea as to what exactly is overheating.V

Anyway, as per a previous request, I can't monitor volts from within windows at the moment, but I did get the numbers right after a cold boot from the BIOS.
PSU
3.3V Rail = 3.27V
5v Rail = 4.91V
12V = 11.96V
5V = 4.94
Battery = 3.09


CPU = 1.32V
LDT = 1.2
Chipset = 1.52
Ram = 2.85

Edit: 11:31pm - I ran prime95 for 20 mins, then figured it was stable enough to being recovering formatted data. During the recovery, I got a BSOD for an IRQL error. I restarted, did some updates for norton, then started again with some SETI@home for both processors, as I've had a problem with heat before from the SETI program. Ran that for 30 mins, then began the Large FFT test for prime95. IT's been running for about 2 hours now with no BSODs or errors. Temps are holding stable at around 38C for both the processor and chipset.
 
You must log in or register to reply here.
Back
Top