- Joined
- Oct 23, 2002
- Location
- Muscle Shoals, AL
I'm not really sure where to post considering the three components I've narrowed the problem down too, but since all the errors involve the display driver I'll post here.
First off system specs:
Power Supply: 400 Antec SmartPower
MB: A7N8X-Dlx Rev 1.04; 5.10 Nforce Drivers
CPU: Mobile Barton; 1.6v; (load temps 44-47C depending on room temps)
Video: XFX 6800GT; various drivers (idle 57C, load upper 60s to low 70s, shows 65-67 immediately after a 30 min 3dmark loop); stock clock settings
Ram: 2x512; Geil Value Series 3200 GE512PC3200B; default timings
2xWD 80g IDE drives
1 SATA Hitachi drive
NEC DVD Burner
Lite On CD Burner
Previously I had the above system minus the SATA drive and 6800GT (9700np). Everything was stable at 200FSB and 11x multi. Then I swapped cases, added the SATA drive and 6800GT.
That's when the problems began. It started out with random reboots playing Doom3. Then started getting reboots while playing videos with Media Player. Usually the reboots would be accompied with a "system has recovered from a serious error" or driver stopped responding type message and in each case would point to the graphics driver. So I set out to try and find a way to make the errors recreateable (fresh install so there are no left over ATI drivers). The only way I've found is to start looping 3dMark05 and wait. No particular tests cause the errors so it's basically start it and wait.
Originally I assumed it was the graphics driver itsself causing the errors (66.93). So I swapped to the previous offical release (61.77). That caused texture corruptions in 3dMark01, and when left looping overnight caused a system reboot. Then I went to driver version 67.03 since I read a lot of posts of people getting good results with those. Those would run 3dMark01 with no corruptions. 3dMark05 passed a 6 hour loop without issue. But then I was trying to watch a movie with Media Player and the system rebooted about 1 hour into it. I was then greeted with the same driver error message I had been getting. Leaving 3dMark05 looping overnight (100 times) caused a reboot.
Finally I started looking at other sources for the problems. First thing I did was run Memtest86. That froze at test 6. Which definately wasn't a good sign.
I was able to pass memtest by lowering the FSB to 166. So then to test the ram, I ran with FSB async so the ram was at 200mhz and FSB was 166. Memtest was able to pass ok.
At this point I decided to see what 3dMark05 would do with 166 bus. Left it looping overnight and everything passed ok (100 times).
Oddly enough the system is 24hr prime stable at 200 and 166 FSB. Also I've had this same motherboard for over 2 years at 200FSB and it used to be able to pass Memtest86 at 200FSB. So now I'm not sure where to go with this.
Just from the results of my testing I'm pretty sure it's going to turn out to be the motherboard. Is it possible I bumped something when moving components between cases? Or is it more of it starting to show its age and the swapping of cases was just a coincidence?
Also does anyone think the video card could still be at fault? The only reason I'm still considering it is that the errors are so random that it wouldn't surprise me if it was able to pass 12 hr loops and still be faulty.
Finally the power supply is a distant third on my list of possible causes. Voltages show lower than what they should be in MBM, but they've been that way for two years. Considering how often software voltage monitoring is wrong I just check for stablility of the voltage rails more so than the actual readings. The voltages are pretty stable according to MBM log when stressing the sytem. However given the addition of a power hungry video card and an additional hard drive I feel I should at least consider it as a possible cause.
Well that's the full rundown of the problem and what I've done so far to narrow down the causes. If I was forced to pick the problem right now I would probably say the motherboard. I would like to hear some other opinions as to possilbe causes and maybe some other tests I should try before I start looking at new motherboards. Also if someone could suggest a reason why the motherboard would start failing at 200FSB when it had previously been stable for 2 years+ at that clock speed. I know most of those boards had trouble reaching 200 so I should probably consider myself lucky it went that long. Also do you think the motherboard is on its way out or is probably still good, just not at 200 anymore? Reason I ask is I would want to sell it when I change boards, but I don't want to sell a defective board.
Sorry for the long read, but I try to give all the information I have to save anyone the trouble of having to ask.
First off system specs:
Power Supply: 400 Antec SmartPower
MB: A7N8X-Dlx Rev 1.04; 5.10 Nforce Drivers
CPU: Mobile Barton; 1.6v; (load temps 44-47C depending on room temps)
Video: XFX 6800GT; various drivers (idle 57C, load upper 60s to low 70s, shows 65-67 immediately after a 30 min 3dmark loop); stock clock settings
Ram: 2x512; Geil Value Series 3200 GE512PC3200B; default timings
2xWD 80g IDE drives
1 SATA Hitachi drive
NEC DVD Burner
Lite On CD Burner
Previously I had the above system minus the SATA drive and 6800GT (9700np). Everything was stable at 200FSB and 11x multi. Then I swapped cases, added the SATA drive and 6800GT.
That's when the problems began. It started out with random reboots playing Doom3. Then started getting reboots while playing videos with Media Player. Usually the reboots would be accompied with a "system has recovered from a serious error" or driver stopped responding type message and in each case would point to the graphics driver. So I set out to try and find a way to make the errors recreateable (fresh install so there are no left over ATI drivers). The only way I've found is to start looping 3dMark05 and wait. No particular tests cause the errors so it's basically start it and wait.
Originally I assumed it was the graphics driver itsself causing the errors (66.93). So I swapped to the previous offical release (61.77). That caused texture corruptions in 3dMark01, and when left looping overnight caused a system reboot. Then I went to driver version 67.03 since I read a lot of posts of people getting good results with those. Those would run 3dMark01 with no corruptions. 3dMark05 passed a 6 hour loop without issue. But then I was trying to watch a movie with Media Player and the system rebooted about 1 hour into it. I was then greeted with the same driver error message I had been getting. Leaving 3dMark05 looping overnight (100 times) caused a reboot.
Finally I started looking at other sources for the problems. First thing I did was run Memtest86. That froze at test 6. Which definately wasn't a good sign.
I was able to pass memtest by lowering the FSB to 166. So then to test the ram, I ran with FSB async so the ram was at 200mhz and FSB was 166. Memtest was able to pass ok.
At this point I decided to see what 3dMark05 would do with 166 bus. Left it looping overnight and everything passed ok (100 times).
Oddly enough the system is 24hr prime stable at 200 and 166 FSB. Also I've had this same motherboard for over 2 years at 200FSB and it used to be able to pass Memtest86 at 200FSB. So now I'm not sure where to go with this.
Just from the results of my testing I'm pretty sure it's going to turn out to be the motherboard. Is it possible I bumped something when moving components between cases? Or is it more of it starting to show its age and the swapping of cases was just a coincidence?
Also does anyone think the video card could still be at fault? The only reason I'm still considering it is that the errors are so random that it wouldn't surprise me if it was able to pass 12 hr loops and still be faulty.
Finally the power supply is a distant third on my list of possible causes. Voltages show lower than what they should be in MBM, but they've been that way for two years. Considering how often software voltage monitoring is wrong I just check for stablility of the voltage rails more so than the actual readings. The voltages are pretty stable according to MBM log when stressing the sytem. However given the addition of a power hungry video card and an additional hard drive I feel I should at least consider it as a possible cause.
Well that's the full rundown of the problem and what I've done so far to narrow down the causes. If I was forced to pick the problem right now I would probably say the motherboard. I would like to hear some other opinions as to possilbe causes and maybe some other tests I should try before I start looking at new motherboards. Also if someone could suggest a reason why the motherboard would start failing at 200FSB when it had previously been stable for 2 years+ at that clock speed. I know most of those boards had trouble reaching 200 so I should probably consider myself lucky it went that long. Also do you think the motherboard is on its way out or is probably still good, just not at 200 anymore? Reason I ask is I would want to sell it when I change boards, but I don't want to sell a defective board.
Sorry for the long read, but I try to give all the information I have to save anyone the trouble of having to ask.