User error or bad hardware?

[TechnoPro]

After years of not building anything, I just put together a new system, specs as follows:

Motherboard : Abit IP35-E
CPU : Intel C2D E6550 (stock cooler used)
RAM : Crucial 2GB (CT25664AA800)
Video : MSI RX2600XT
HDD : WD 500GB SATA (WD5000AAKS)
PSU : SeaSonic S12 II SS-500GB
Optical : ASUS DRW-2014L1T
Case : Lian-Li PC-60BPLUSII
OS : XP Home SP2

Windows loaded just fine, as did the drivers, and other esential software (Office 2003, Adobe Reader, and AVG). Seems to run fast, with stability as the real issue.

Symptoms:

* System has BSOD twice now, each mini-dump file pointing to a different issue. Once before any drivers or other software was installed, the other time afterwards.

* Video signal will simply die out, NumLock on keyboard turns unresponsive (leading me to believe it's not just a bad video signal), and computer needs to be rebooted to regain functionality.

* I received that NT Authority Error (LSASS.EXE) once that was so prevalent with the Sasser worm. I am behind a router.

* Another boot, I received a WinLogon.exe error.

* ATI Control Center kept crashing. I later uninstalled it, and the other issues are still there.

Troubleshooting I've done:

* Updated the BIOS to the latest version and the problem still persists.

* Reviewed the Event Log for issues. Only the 2 BSOD's were recorded. All other failures go undocumented.

* CPU temp according to Core Temp: Core0 23 degrees C / Core1 20 degrees C (measured at idle when the PC fails)

* 3.5 hours (7 passes) of MemTest86+ w/ no errors

I've always believe that erratic, unpredictable problems like this indicated a hardware issue. Trouble is, which piece? Or did I simply pick out sub-par components for this build?

I'm grateful for any advice that can be given, thank you.

 

[LOUISSSSS]

my first guess is bad ram. u need to download and burn memtest86+ to a cd and boot from it and test each of your ram modules alone in each ram slot on your mb. this will eliminate any openings for problems. isolate which ram is bad/which motherboard ram slot is bad, then retest. configure ur bios to the ram specifications and reinstall windows (your current windows is probably already corrupted by bad data.)

 

[TechnoPro]

Originally posted by: LOUISSSSS
my first guess is bad ram. u need to download and burn memtest86+ to a cd and boot from it and test each of your ram modules alone in each ram slot on your mb. this will eliminate any openings for problems. isolate which ram is bad/which motherboard ram slot is bad, then retest. configure ur bios to the ram specifications and reinstall windows (your current windows is probably already corrupted by bad data.)

I'm with you, great advice. The only part that confuses me is about configuring the BIOS to RAM specifications. I have everything set to auto, or the functional equivalent. Is this wrong? Thank you.

 

[Roguestar]

Are you sure temperatures are fine? What are you using to check temperatures? From that it sounds to me like it may be overheating (unless it's a virus as well, that could explain the file corruption); check both graphics and CPU temperatures. Use CoreTemp for the CPU and ATI Tool for the GPU; consider using speedfan to up the graphics card fan speed if it's the one overheating. Run a virus scan too, with AVG or your favourite flavour of AV program. A router isn't enough to stop getting malware.

Edit: Can't believe I didn't think RAM. Stick with LOUISSSS and run memtest from disc and see what happens.

 

[LOUISSSSS]

Originally posted by: TechnoPro

Originally posted by: LOUISSSSS
my first guess is bad ram. u need to download and burn memtest86+ to a cd and boot from it and test each of your ram modules alone in each ram slot on your mb. this will eliminate any openings for problems. isolate which ram is bad/which motherboard ram slot is bad, then retest. configure ur bios to the ram specifications and reinstall windows (your current windows is probably already corrupted by bad data.)

I'm with you, great advice. The only part that confuses me is about configuring the BIOS to RAM specifications. I have everything set to auto, or the functional equivalent. Is this wrong? Thank you.

its probably not "wrong" but since you're having problems, it might be the cause of it. set your vRam and timings specified by your ram vendor in your BIOS, clear your CMOS, then run stability tests starting with Memtest86+

 

[TechnoPro]

Originally posted by: Roguestar
Are you sure temperatures are fine? What are you using to check temperatures? From that it sounds to me like it may be overheating (unless it's a virus as well, that could explain the file corruption); check both graphics and CPU temperatures. Use CoreTemp for the CPU and ATI Tool for the GPU; consider using speedfan to up the graphics card fan speed if it's the one overheating. Run a virus scan too, with AVG or your favourite flavour of AV program. A router isn't enough to stop getting malware.

Edit: Can't believe I didn't think RAM. Stick with LOUISSSS and run memtest from disc and see what happens.

After 3.5 hours of MemTest86+, no errors.

CPU temp according to Core Temp: Core0 23 degrees C / Core1 20 degrees C (measured at idle when the PC fails).

ATI Tool fails. Specifically, when launched, it loads only in tray and chews up 50% of the CPU and fails to actually open. Double clicking or right-clicking tray icon have no effect.

I also tried ATI Tray Tools. Won't give me GPU temp - option is grayed out. Perhaps because this GPU is passively cooled?

 

[SpeedEng66]

didnt Abit IP35 have a issue with this? that they fix via bios update>?

 

[TechnoPro]

Originally posted by: SpeedEng66
didnt Abit IP35 have a issue with this? that they fix via bios update>?

I experienced issues under both the BIOS revision it came with as well as the latest. Then again, is it possible that Windows would have been corrupted when installed under he original BIOS if indeed it was faulty? I should just reinstall for the sake of ruling this out...

 

[Billb2]

Originally posted by: TechnoPro

After 3.5 hours of MemTest86+, no errors.

CPU temp according to Core Temp: Core0 23 degrees C / Core1 20 degrees C (measured at idle when the PC fails).

Memtest-86 doesn't stress memory, it just tests it. Try Orthos.

Idle temps are useless to us. Do your problems occur at idle?

Check core temps after Orthos/OCCT/Etc. runs for at least 30 min.

And default BIOS Vcore and Vdimm usually default to very conservative settings. Check the manufacturers specs for both and set them manually in the bios.

 

[TechnoPro]

Originally posted by: Billb2

Originally posted by: TechnoPro

After 3.5 hours of MemTest86+, no errors.

CPU temp according to Core Temp: Core0 23 degrees C / Core1 20 degrees C (measured at idle when the PC fails).

Memtest-86 doesn't stress memory, it just tests it. Try Orthos.

Idle temps are useless to us. Do your problems occur at idle?

Check core temps after Orthos/OCCT/Etc. runs for at least 30 min.

And default BIOS Vcore and Vdimm usually default to very conservative settings. Check the manufacturers specs for both and set them manually in the bios.

Thanks for the suggestions. All problems occured at idle, so idle temps were what I measured. I will, however, give Orthos a try.

I may have solved the problem already, though. I never cleared the CMOS at any point in my previous troubleshoot attempts. Since I have done this, it has not failed. I will, however, run it through a battery of stress tests before I can say "problem solved".