Sign in with
Sign up | Sign in
Your question

system unstable at 133fsb

Tags:
  • CPUs
Last response: in CPUs
Share
December 23, 2007 6:15:30 PM

I have a PcChips m811LU v3.1. It has the ECS K7VTA3 v.6.0c loaded, fixing the Ultra ATA issue. CPU= 2100+ (AX2100DMT3C), Mem= Kingston pc2100 (kvr266x64c25/256), HDD=Western Digital 30GB (WD300), Graphics Card= GeForce2 MX400 (AGP 4x 64Mb), PSU= AGI 300W (HEC-300AR-T) The board posts fine. To get XP Home installed I had to set the fsb ,via jumper, to 100mhz. I also had previously set the BIOS to "Failsafe Default" in an early attempt to resolve the installation problem. With the exception of enabling the onboard LAN, the Bios is still in "Failsafe mode". If at 133 the install process produced many BSoD's with errors like, "page_fault_in_nonpaged_area" or "irq_not_less_or_equal" or "driver_irq_not_less_or_equal". I now have XP Home installed with SP-2 and all other critical updates, all hardware updates as well as some optional updates installed. All from Windows Update. I have NOT installed the software/drivers from the PcChips CD, I read that the embedded drivers on XP ran as well or better. OK here's my problem: When set fsb to 133 the system becomes unstable. Windows will boot normally, but after a few min. the system restarts as if I hit the reset sw. I let the system reboot and the "restarts" occur increasingly more frequent.

More about : system unstable 133fsb

a b à CPUs
December 23, 2007 7:29:52 PM

Sounds like a heat issue? What is your CPU temp in bios? Let it cool down to ambient, turn it on and watch the temp in your Bios.
December 23, 2007 8:12:28 PM

I would look for bulging capacitors. I had a similar problem with an ECS board.
Related resources
December 23, 2007 8:15:54 PM

I had the same thought about the temp. CPU temp maxed out at 52c. I have checked after it starts "restarting" AND just let it set in bios and monitored the cpu temp. I also phisically checked the graphics card and...well everything I could touch. Nothing is getting warmer when running windows than when idling at bios screen. Oh ya I checked at both 100 and 133 fsb.
December 23, 2007 8:28:17 PM

Rook - I've already checked for bulging (and leaky or loose caps). I had a M810 that I successfully recaped. I did notice that this board had the original caps. Thanks for the ideas and help. Please keep them coming.
December 23, 2007 8:59:04 PM

Next I would try setting the FSB to 133 and running Memtest. Do you have a PSU tester?
December 23, 2007 9:02:47 PM

Another thing I tought of. Under System> Advanced> start up and recovery> System failure> Make sure automatically restart is unchecked.
December 23, 2007 10:55:57 PM

I have tried a different mem stick and psu.....no change.

I left the system on at a bios screen for an hour or so. Then tried to reboot. It made it to the "F8 Safe Mode" screen since windows hadn't shut down properly. No matter which I selected it would very shortly (2-4 sec.) revert to the graphics bios and try to reboot again, if left alone it would continue this cycle, since the "F8" screen has a timer. All of this seems to support the hardware overheat theory to me.

So while the system is still "HOT" I very quickly changed the fsb jumper back to 100mhz. It just booted windows normaly. I was now able to check and the Automatic Restart was in fact selected. I have now disabled that option. Windows is running and seemingly stable with the cpu underclocked @100mhz.

December 23, 2007 11:11:54 PM

Bios gives acccess to fsb as follows. When jumped @ 100mhz, 100 thru 132. and when jumped 133, 133 thru 165. Both single digit increments. I have not found access to the multiplier.

I may try setting jumper to 100mhz and slowly moving up in bios to see if there's a failure point. Right now 132 sounds good. If I do this, does that effect other functions by limiting the fsb to 100mhz while the bios raises the cpu bus speed or does the jumper just tell bios which options to display, and actually let me adjust the fsb.

It's been nearly 45min and the system is still stable at 100mhz.
December 24, 2007 2:14:29 AM

I did run memtest @ 133MHz. It ran for almost 1 1/2 Hrs and reported 0 errors.
!