I'm trying to diagnose a problem I'm having when games switch to fullscreen 3d. It's a nasty BSOD with
STOP: 0x0000007A KERNEL_DATA_INPAGE_ERROR
with a code of 0xC000000E
Now, ordinarily this error seems to be caused by a failing hard drive and difficulty with the paging file. But I've done a disk check and everything looked good, and run 6 hours of MemTest86. The problem seems to occur within seconds of a fullscreen 3d display (for example, in World of Warcraft, I can see the login screen for less than 1/2 sec before reset; with Serious Sam 2nd Ed I can go through the loading bars but crashes right after the logo).
It's an A7N8X Deluxe, Barton 2500, Antec TruePower 450. But MBM5 echoes the bios readings pretty well:
Check out in particular the +5V line, which has varied from 4.52 to 4.57, and the entry in the PSU faq on this board saying to watch "especially the 5V line in AMD systems" and wanting them to be within 5% of tolerance. I seem to be out of that.
I'm running an ATI 9800 Pro (with a secondary power connector hooked up). Is it possible that the power spike caused by firing up fullscreen 3d combined with a resolution switch is enough to cause system instability?
Tried starting up some 3d games in windowed mode to monitor voltages; 5v line varies from 4.6 (high end) to as low as 4.38 (low end). Now to me, 4.38v seems like about a 12.4% variation from nominal levels--I would think that much below that may very well cause problems for normal operations.
So it SOUNDS like the PSU may be the culprit (and here I was all happy about my Antec TruePower). I'll poke around a bit more and see what I find.
Sounds like your PSU to me - 4.38 is just too low IMO for stable operation. While specs say 10%, I think 5% +/- from 5.0 is what I would consider the safe limit, and your voltages shouldn't fluctuate during operation at all (maybe .05, but definitly not .22)
Well, I talked to Antec and my customer support fellow called back with "well, I talked to my technical department and they said the lowest that 5v rail is ever supposed to be is 4.75v"
...so I'm guessing that running at 4.57v continuous and 4.38 lowest is probably a bad thing.
Dang it! I don't think I have the receipt any more; may get to buy a new PS at this rate... grr... money I don't really want to be spending. But at least I think that's the problem now.
So much for the Antec True430! Funny thing is, my other two computers (NOT my personal workstation) are running off damn-near-indestructable PowerMan beasts from probably pre-1996 and still going strong. Maybe I'll try OCZ or some such.
Yep, sounds like you got a bummed PSU. I'd try to RMA it if you can.
OCZ, Fortron Source, Enermax - all good names. Fortron Source (aka FSP) makes good quality and affordable PSU's - 400w can be had for under $75US, the 500w is I think under $100. That should ease the bite if you can't RMA your Antec.
The Antec Truepowers are supposed to be good PSU's. Could be you just got a bad one - so its not necessarily bad to give them a second chance. (though I'd feel the same way too - guess it depends on how much you want to trust them and give them the benefit of the doubt)
I'm doing so. I checked--I didn't have the receipt. But happily, I ordered it from NewEgg, which has online printable copies of all receipts from all orders I've done with them. The RMA has begun; hurrah!
> Fortron Source (aka FSP) makes good quality and affordable
Funny you should mention. I pulled a PSU from a friend's dead box for the sole purpose of keeping my computer limping for the week of the RMA. It's a 250W FSP unit, bare-bones, no frills. All voltages stay at great levels with no droop for 3d games... I'm still putting the True430 back in there when it gets back, but that was encouraging... I'd thought that particularly with both the Athlon and the 9800 pro, 250W was just too low, but it's handling it with aplomb.
Looks like the problem's fixed (system is really stable again). Thanks for all the help!