Help finding failing component

[hunter315]

Tried to boot up my desktop about half an hour ago, it woudnt send any signal to the monitors, opened it up, redid all of the wiring, now it makes it to the windows loading screen and BSODs. Reset bios settings back to default, managed to log in, 5 seconds later it BSODs again with looping sound so i decide to check the memory.

I was running memtest, it made it for about 2 minutes, then it had "Unexpected interupt -- Halting". How bad is it when memtest crashes?

I built this system last summer so everything is still under warranty if i need to send something back. It was working fine yesterday, i played Mass Effect 2 for 4 hours last night, no problems at all. Im not sure what would have such sudden onset and be so crippling. Im open to trying just about anything.


Full system specs:
AMD 955 C2 @ 3.2GHz
ASUS M4A79T Deluxe
Antec EA500
Seagate 7200.10 250GB
WD Caviar Black 1TB
XFX HD 4850
GSkill 2x2GB DDR3 1333MHz CL9

 

[mosox]

Try with one stick of RAM, and then with the other one.

 

[Niklas_13]

Sometimes when i OC, my system wont show video when it comes out of sleep mode. I have to clear CMOS then go back into the bios set opt defaults. Did you reset your ram timings volts and speed in the bios after you got into your bios before running memetest ? Memtest Should be able to run for weeks without errors, so its not a good sign its showing errors after 2 hours. make sure you reset your ram.

 

[tulx]

Yeah, overclocking can do all kinds of funny stuff to your machine. Mine, for example, has some BIOS problems ever since I OC'ed my Athlon 9850 BE form 2,5 to 3,1 GHz. Now every second boot fails (meaning - I restart the PC, the boot fails (no BIOS settings are saved etc) but windows loads. I restart again and it works. Restart again, the first step repeats.) Clearing CMOS doesn't help either, but I can live with it. Will buy a new MoBo someday anyway.
But hunters problem does sound like a memory fault. You just never know when a component will decide to die. Can happen just like that. Hope the warranty thing works out for you.

 

[hunter315]

My problem isnt that memtest is throwing errors, that easy to diagnose as a memory issue. My concern is that memtest is crashing. Just got home so i figured i would give it a second shot, it went for about 3 minutes, then my screen shut off and said there was no video input anymore. Im going to try one stick at a time after it crashes out this time and see if i can isolate it to a single stick.

I also know that its not a timings issue, up side to the Gskill kit, it only needs 1.5 volts and runs at standard timings, i didnt modify my ram timings for a few weeks after i first built it so i know its not an issue with the defaults in BIOS>

 

[nforce4max]

If you aren't overclocking them the the ram shouldn't be overclocked. As with the other suggestions one stick at a time if that fails underclock both ram and cpu. Also check the power stages on the the board. How good is the air flow around your ram?
I had one kit fail on me when I had managed to run a meager 533 kit at 800 clocks 5/5/5/15 so after replacing it with a tried and tested I did a fan mod hehe no more failed ram yet.

 

[Niklas_13]

Really could be a number of things, as you know. I guess if it was me, i would

Check temps on the CPU and GPU

Try one stick of ram at a time

Try a different PSU

Make sure the mobo, PSU are grounded

Re seat the GPU

RMA mobo

RMA GPU

Maybe a bad DVI cable Shrugs

 

[hunter315]

Well its been working for the past 9 months so a few things on that list can get ruled out.

CPU temp according to bios sits around 41C, when im in windows its normally about 35 idle and hasnt broken 62 under load. GPU idles around 45 in windows, and i have never gotten it to break 70C even with furmark running for a couple hours.

It has some how made it for a full 32 minutes on this memtest pass so im going to wait on the one stick at a time unless it fails this time around.

Different PSU isnt really an option as im at school and dont have many parts to pick from, i do have a multimeter here though that i can use to check voltage levels if need be.

Motherboard and PSU havent been removed from the case since the initial build so i find it unlikely that they managed to unground themselves.

Reseated the GPU earlier because it wasnt giving me video so i suspected that was my culprit.

Im outside of the RMA window so i will need to do full warranty replacements likely.

Im running dual monitors(they go into clone mode before windows loads) and they have both been behaving identically so i know its not the monitor or the cable connected to the monitor.

 

[Niklas_13]

Well its been working for the past 9 months so a few things on that list can get ruled out.

^ Yes that is true, just shoot *** off

If you saved the BOSD you can look at this link http://www.maximumpc.com/article/features/blue_screen_survival_guide?page=0,0

 

[hunter315]

It has been kind enough to BSOD a bunch again so that i could keep track of the error codes this time around, i have gotten 3B, 7F, 24, and 1E in the last hour

I ran memtest last night and it made it though 8 passes error free, im not sure whats going on. It will sometimes crash pre-windows so its got to be hardware, the question is, what piece?

Anyone know of a definitive way to check and see if the CPU is busted? Keep in mind i dont have many spare parts on hand to work with.

 

[Niklas_13]

A quick look on the web points to RAM, display drivers, and file system errors with the BSOD Numbers.

If memetest came back clean, maybe try to check your HDD for errors and wipe your display drivers and reinstall them.