Chasing a wild goose in my computer

[stjabr]

Hi everyone,

I've just spent the last 5 days trying to solve what in the world has gone wrong with my computer. I'll start from the beginning:

I custom built a computer for myself just over a year ago, using high quality components only. This ran pretty much flawlessly for a year, then around 3 weeks ago my Corsair HX750 destroyed itself. Annoying, but not in itself a disaster - Scan replaced it within a couple of weeks, very helpful service.
In the mean time I used my old Hiper 580w PSU - system continued to run flawlessly, although I was reluctant to push it too hard just in case (PSU is about 4-5 years old).
Last friday, brand new Corsair HX750 arrives. I fit in, turn on, all seems fine.

But then blue screens pop up a few times in the next 24 hours, culminating in yet another PSU failure. I'm guessing it was faulty to begin with, irritating, but otherwise just unlucky. In goes the old Hiper PSU once more...

This time though, I have been experiencing what can only be described as a multitude of different blue-screens and crashes. Initially I was led towards believing that the RAM had been damaged by the second PSU blowout, but the more I test and diagnose the less it makes any sense at all.

I've run Memtest86, tested all memory sticks in all ports, windows memory diagnostics - all with no errors. The windows based Memtest, however, showed up a few errors - just before blue screening again. Every time memory usage is high-ish the computer crashes - but only while being used in windows, never whilst starting up or shutting down. If the problem was the memory or motherboard slots then I'm guessing that would have shown errors in the DOS-based memory diagnostics.

So that then suggests driver/OS based errors. SFC /scannow shows no errors. Registry has been cleaned, drivers updated, the lot.

All the way through, the most consistent and quickest way of forcing a crash was by running the Windows Performance Index test. Every time the crash would occur while switching from testing CPU performance to testing RAM performance.

I have even created a clean install of Windows 7 on a different partition - this is where it get's even more confusing:
Straight away, I ran the performance index. Passes first time, no crash. Video card gets a score of 1.0 due to no drivers being installed obviously. So I install the newest drivers to match my main windows install - Performance Index crashes.

Now I should mention, the one thing I have changed on my main install since the PSU blowout the first time is the video drivers, as I was worried that was causing the problem (which turned out to be the second power supply on its way out).

So I install the older video drivers that have worked for the last 6 months without problem, on my NEW windows install. Still crashes until I update the chipset drivers correctly, which seems to fix it. So I copy the move by installing the old video drivers in the old windows install - appears to fix the problem.

To verify I go back to the new install, run the performance index again. It crashes. Back to old install. It crashes. Blue screen report suggest that this time it was a wi-fi driver error - which I haven't touched, first time that error has caused a crash at all. I run registry cleaner and verify windows install once more. Re-run performance index, it works. Re-run again, it crashes.


By now my mind is well and truly boggled. I cannot even pinpoint whether the problem is hardware or software based. As a result, I can't even be specific about which piece of hardware, or which software/driver is the issue, and therefore cannot begin to go through an RMA/fix the problem.

Does anyone, by any chance, have any clue as to what could be causing my problems? The second PSU blowout I believe was the trigger for the start of the errors, but as I say, I cannot pinpoint the problem, or any potential hardware damage so far. Even less helpful are the blue screens which have covered the following so far:

Old install (In order they occurred):
unknown blue screen - bug check code 0x124
MEMORY_MANAGEMENT 0x1a
SYSTEM_SERVICE_EXCEPTION 0x3b
MEMORY_MANAGEMENT 0x1a
UNEXPECTED_KERNEL_MODE_TRAP 0x7f
SYSTEM_SERVICE_EXCEPTION 0x3b
DRIVER_VERIFIER_DETECTED_VIOLATION 0xc4
SYSTEM_SERVICE_EXCEPTION 0x3b
SYSTEM_THREAD_EXCEPTION_NOT_HANDLED 0x7e
KMODE_EXCEPTION_NOT_HANDLED 0x1e

All except the SYSTEM_THREAD_EXCEPTION_NOT_HANDLED implicate ntoskrnl.exe as the culprit. But on the new install, which should be free of that software problem, the same ntoskrnl.exe is implicated.


I'm very sorry for the ridiculously long post, but I'm at my wits end trying to figure this out. The more I look the further away from solving it I seem to get! I would appreciate ANY help anyone could offer, I'd be happy to provide dump files and other information if wanted/needed.

Just to clarify, all hardware is at stock speeds, no overclocking has been attempted at any point yet, so that is not an issue.

Stu

 

[badkarma11]

I would look at CPU, GPU and chipset temps as overheating can definitely gum up the works. It could be, for example, that when swapping out PSUs, you blocked an air path somewhere. Additional possibility might be that when the first PSU died, it marginalized some motherboard component, thus shortening it's life (probably not, but worth considering). When the PSUs died, did they just quit, or did they explode? I had one blow up once is why I asked. It didn't breach the 'walls' of the PSU, but the internal bits were physically wrecked. Blew out my UPS but didn't seem to damage the mobo.

 

[DeanP]

http://www.sevenforums.com/crashes-debugging/96879-blue-screen-death-bsod-posting-instructions.html

 

[stjabr]

Dean - Sorry about that, only just noticed those instructions after posting! Zip is attached now.

badkarma - I had the same thought about temps, but CPU and GPU are actually running cooler than previously because I cleaned the dust out of the mesh filters in the case when I changed the PSU. Only temps I don't know specifically are chipset, but as the PSU is bottom mounted I can't see any way it can be blocking airflow any more than before. I can't rule out damage obviously, but have no way of confirming that one way or the other.
As for the PSUs, the first one went semi-spectacularly - initially power just cut out, but on trying to start the PC up again it definitely spat and crackled quite nicely! Not quite an explosion though thankfully, nothing even loose inside the PSU, just a strong smell. The second PSU was less spectacular, power initially cut out, and then gradually lasted less time at each attempted restart - after 3 or 4 attempts, it refused to start up at all. No spitting, or exploding, or anything.
If there are any hardware faults, it's almost certainly as a result of the PSU blowing, but I'm also pretty certain it's as a result of the 2nd one not the 1st one, as the computer ran without any signs of problems on my old PSU in between the 2 blowouts.

 

[stjabr]

Is anyone able to gave a look at the crash dumps for me? I'm not quite Windows error-code literate yet!