random reboots, no idea why

[squeegee]

So far my new computer has rebooted twice in the middle of a program for no reason (that i can see). This is my first computer i've put together, all the hardware is new:

Windows XP home edition
gigabyte 7vtxe (kt266a) motherboard
athlon xp 1600+ running at default (1.4Ghz)
2x256 crucial pc2100 running at cas2, 133Mhz
Thermaltake Volcano 6CU (4800rpm model) + arctic silver2
IBM 60gxp hard drive
Plextor 16/10/40 cdrw
Leadtek geforce3 ti200 using WinXP 21.83 nVidia driver
PCI devices: (starting from the bottom slot and moving to agp port)
Netgear fa311 nic
SB audigy
USR 56k faxmodem (this leaves 2 free slots for breathing room for the downward facing fan on the geforce3)
Enlight 7237 midtower case with 300W (enlight psu, amd approved)
80mm intake on front bottom, 80mm output on back/top

Currently it idles around 44/33C (cpu/system). The PSU reports 82C usually. This found from using Motherboard Monitor 5, and the utility that came with the motherboard confirms the cpu/system temps.

Opening the case brings the idle down to 41/31 (cpu/system again) so i figured 44/33 is pretty good for being closed. Now for the stress tests...

Case cover is on, normal operating conditions:

Toast at low priority for 23min - 55/37, no crashes
toast is then switched to high priority, ran it for 7 more min with no temp change or crash

CPU stability test 6 - 25min, no problems, 49/35

CPUburn K7 for a while, temp stabilized at 56/37
CPUburn MMX for a while, temp stabilized at 54/37

Also ran 3DMark2001 loops for a while, and Sandra2001 stress test, no problems. maybe i should run them longer?

Something called super pi, calculated Pi to 32million places, took a little over an hour, no crash, wasnt as hot as cpuburn or toast.

One thing i did notice... under a heavy load (cpuburn) the 5V voltage would drop to 4.86, about 3% below. significant?

Reboot 1 was while playing Empire Earth, system didnt seem slow or anything. Reboot 2 was when exiting Dark age of camelot. (if that matters/helps... more info is better right?) the 2 incidents were separated by about 1 week, with the computer on about 8-15 hours a day.

Other possibilities:
Via 4in1 and AGP drivers... I know they weren't installed for the 2nd Reboot. So maybe that would be the problem? I'm putting them on now, and running windows update, hopefully that'll fix things.

anyone else see a potential problem?

 

[Guest]

you have an active mobo monitor that reports temps and voltage etc? if so, was thinking that maybe an alarm is sounding and forcing reboot based on improperly set default settings for cpu or sys temperatures.

If you are running somethin like that, try raising the parameters to more realistic levels. think defaults are pretty low.

Hope this helps.

Bones

Dreaded Bones
HARDWARE JUNKIE

 

[HonestJhon]

yeah, when i first got my socket 7 epox, the defaults were set pretty low for the alarm and the shutoff....
the alarm would go off almost instanty after i turned the thing on, and i thought that i was going to kill my computer...
but then when i looked at the temps it was running at, and what the alarm was set at, i realised that it was wayy to low....so i set it higher, and never had a problem after that.

-DAvid

-Live, Learn, then build your own computer!-

 

[pr497]

im not sure what the problem is but i might be able to help you find out.
first...right click on the "my computer" icon and click properties. (or go to the control panel and click on system...whichever is fine).
click on the "advanced" tab.
under "startup and recovery", click settings.
under "system failure", uncheck "automactically restart".

now...the next time your system "fails", rather than your system rebooting, it will show the almighty blue (sky blusih actually) screen of death.
now...just write down whatever the error is (that is if it is shown) and go to the support section at microsoft's website.
btw...the error could look something like this (but not exactly this):

DRIVER_IRQL_NOT_LESS_OR_EQUAL

or

KMODE_EXCEPTION_NOT_HANDLED
or whatever.

just type in whatever error message you get on the website and you might be able to find out what the problem is.

this is the best i can do right now.

I don't claim to know anything about everything, I just tell people what I know.
-PSB

 

[lhgpoobaa]

temps look ok. and if u can run toast at high priority for over an hour you should be good.

not too sure about the voltage dip in your 5v rail though... 4.86 should be enough, though it does indicate that more juice would be better.
remember that when u play a game, both the cpu and graphics are used alot, and both are renowned for sucking PSU juice.

or it could be a software issue...

hard to pin down.


Excuse me for a moment. I need to drive my ergonomic wheely chair over a sheet of bubble wrap!

 

[squeegee]

***** EDIT ***** EDIT *****
I guess i should have checked the DAoC website sooner... theres a known issue with winXP and nvidia cards when exiting, which should be solved with a different driver. downloading 23.10, hopefully that'll fix it. As for the other reboot i was getting, maybe it was the same kind of bug, or because i didnt have the via drivers installed. thanks for the help everyone.

*** EDIT ***


A few more reboots at the same spot, exiting Dark age of Camelot. It shouldnt be the motherboard monitor shutting things down, since i have to start that manually. After disabling the auto-restart, and checking out the error code: (cut&paste from MS)

STOP: 0x00000020 (parameter, parameter, parameter, parameter) KERNEL_APC_PENDING_DURING_EXIT

Explanation:
This is a Windows 2000 Executive character-mode STOP message. It indicates that a kernel-mode asynchronous procedure call (APC) was found pending during thread termination. The first parameter is the address of the APC found pending during exit. The second parameter is the thread's APC disable count. And the third parameter is the current IRQL (interrupt request level). If the thread's disable count is non-zero, it is the source of the problem: the current IRQL should be 0. If it is not, a driver's cancellation routine returned at an elevated IRQL.

the actual parameters were 0xff900198, 0xfffffffe,0x0,0x0.
So the way i'm reading this, if the 2nd param is non-zero, which it is, then the third param should be 0?

It sounds like its a driver issue, so I'll see if i can download newer drivers for the audio and video cards. I suppose it could be the modem though... I've had really bad luck with modems on this computer. Tried 3 modems, DSI, Creative, and US Robotics. All brand new, winXP wouldnt install their drivers, gave me a "the data is invalid" error when it finished trying to install them. Finally i got around this by installing win98, (shipped win98 drivers worked here), and running the compatibility test on the winXP cd before installing over win98. It downloaded something during the compatibility test, presumably updated drivers for the modem. I'll check the USR website and see if they have updated XP drivers.

I suppose if i can't get it to work i can dual boot win98 or ME with XP, although i'd hate to do that... Thanks for the help so far.

<P ID="edit"><FONT SIZE=-1><EM>Edited by squeegee on 11/30/01 05:56 PM.</EM></FONT></P>