Random Resets and occasional BSoD

[Plaster]

I've got an AMD K62 450 running at 4.5x100 at 2.4v. I have an Epox EP-58MVP3C-M main board with 80mb of PC66 SDRAM. Also a Voodoo3 2000 PCI. Nothing is overclocked, not the RAM, the CPU, or the Voodoo3. On occasion my computer will just reset on it's own, or give a fatal exception. It didn't do this with my last CPU and mobo. The CPU never gets above 115ºF so I don't think it's a heat problem. This is also an AT board, so it's not the soft power doing it. Anybody have any idea why it would do this? Could it be RAM problems? Would bad RAM cause a reboot? Or could it be a failing power supply? Any help will be appreaciated. Thanks.

[jokostel]

id check the memory out,amd reccommend that you use at least 100mhz memory with the 400mhz and above......i don't think thats really an issue though.. it does sound like you have a bad dimm....... try some other memory and see what happens (pc100) if accessable....
i had a very simmilar problem with some pc100mhz memory that i had to give to my fathers machine to make my machine run properly again...(traded the memory back that i gave him,long story....hehe)
good luck.... jokostel@usa.net

[Plaster]

Well, I tested each DIMM seperately with Nuts&Bolts and it said each DIMM had errors. I have 2 32mb DIMMs and and one 16mb DIMM. the 2 32mb Dimms gave this result::
RAM Error 1B77 Physical RAM-Butterfly
RAM error 1B76 Physical RAM-Butterfly
RAM Error 1B7A Physical RAM-Butterfly
RAM Error 1B7B Physical RAM-Butterfly
RAM Error 1B78 Physical RAM-Butterfly
RAM Error 1B79 Physical RAM-Butterfly

The single 16mb DIMM had theses errors::
RAM Error 4E3 Physical RAM-Butterfly
RAM Error 4E4 Physical RAM-Butterfly
RAM Error 4E6 Physical RAM-Butterfly
RAM Error 4E7 Physical RAM-Butterfly
RAM Error 4E2 Physical RAM-Butterfly

Since both the 32mb Dimms had the same errors, It makes me doubt the validity and the accuracy of these tests. What is a good utility for testing for physical RAM errors?

[jokostel]

i wouldnt reccomend testing them yourself....they have to be tested in a machine specifically designed too... you need to take them in to a place that will do this for you.... (best buy....computer renassiance,etc.etc.....)

[Plaster]

I've taken parts into Computer Renisance before to have them checked, they ream you good. I think the last thing I took in was an HD and it was $40 just to find out it was completely fine. I can get 64mb 100mhz RAM for that price. I need to upgrade it to. When I overclock my RAM, my system performance increases by roughly 35%. I get a good 15fps leap in Quake 2, the only problem is, it doesn't run very long when overclocked.

[jokostel]

have you tried voltage settings for the cpu,ram speed settings,etc....

[Plaster]

Yes, I've tried them all. My mobo allows the RAM to run at 66, or at CPU bus speed. I've run my CPU voltage at 2.2, 2.3, 2.4 and 2.5. It's marked for 2.4, but the most stable seems to be 2.3. I think that has to do with the fact that the voltage is rounded down from a decimal. For a while I thought it may be my winmodem, but it started crashing or resetting after I exited games of Quake2, and the Fatal Exceptions usually occur when it's idle. It's annoying to, because it doesn't name the VXD. It's alwasys the same error too.
An Exception 0E has occurred at 0028:C0014DC8 in VxD ---
This was called from 0028:C0015B15 in VxD ---.

Not very helpful when it doesn't name the VxD. I've searched Nathan's site and this one isn't mentioned at all.

[ENVY]

What kind of sound card do you have?
You might try removing the sound card from the device mgr while in safemode.Then let windows redetect and reinstall.
I had A similar prob but it wouldn't show it anywhere in windows.When I removed the sound card I immediately got the exact same vxd error as The one that was popping up.
The safemode removal cured it.Might try it.
OBTW,this was a soundblaster live that was kickin up the fuss.

[Plaster]

My sound card is a Diamond Sonic Impact s70 PCI. I'm not to certain that it would cause this problem, but I'll check it out.

BTW, I tested each RAM stick by itself and all had that weird error. I had then in as 32, 32, 16. When I put them back I put them in this order, 32,16,32. The middle 32 is now in bank 1 and the first 32 is now in bank 3. I haven't had a single crash reboot or lockup since then(around 3am). I also o/c the RAM to 100 and before it would crash every running program in a matter of minutes. Not one crash in 14 hours. I slept some in there, but the crsh would have been there when I woke up. How weird is that? Does the order of the modules have anything to do with stability? They are all the same generic RAM, same maker, just different sizes.

[This message has been edited by Plaster (edited 07-05-99).]