Random BSODs and applications crashes on a fresh Win7 install

[R4ndom]

Thanks for this in-depth reply writhziden. The only worry now is if RMA will accept it, since no errors are detected by seatools and it's obligatory to attach the error code to the RMA I will test the drive again with seatools, this time as slave drive.

I must say this was the quietest drive I've ever had, will be hard to match I think, I start thinking of ssd as a system disk, but prices are high and it seems that they're failing at the same rate as the traditional ones.

 

[R4ndom]

Seatools log:

--------------- SeaTools for Windows v1.2.0.6 ---------------
2012-06-19 21:12:41
Model: HD502HJ
Serial Number: xxxxxxxxxx
Firmware Revision: 1AJ10001
SMART - Pass 2012-06-19 21:12:41
Identify - Started 2012-06-19 21:14:11
Short DST - Started 2012-06-19 21:15:12
Short DST - Pass 2012-06-19 21:16:59
Short Generic - Started 2012-06-19 21:18:29
Short Generic - Pass 2012-06-19 21:19:44
Long Generic - Started 2012-06-19 21:19:56
Long Generic - Pass 2012-06-19 22:33:24

 

[writhziden]

Interesting and confusing.

How has the system been with your friend's drive in?

 

[R4ndom]

Interesting and confusing, as you say with my friend's drive in, I haven't had a single appcrash, apphang or whatever. No BSOD either. Event viewer is clean, as opposed to before when it would spit errors every minute. I'm using exactly the same cables as with the failing disk and same sata slots on MB. Below smart readings, which seem ok as well.

Red crosses on the s-shot were green points in fact

 

[R4ndom]

Hi all,

after the initial success with the other drive, PC started failing again. I sent it for repair, returned (they didn't confirm the issues), happened again, sent again for repair, they changed the MB. Plus I'm runnig a different PSU now for testing purposes, so it seems like the combination of both was causing problems as you suspected. The PSU used now is a little weak for the config, so I'm planning to get a Corsair TX 650M, which has pretty good reviews and still doesn't cost a fortune and since I won't be using it at more than 50-60% max, it could maybe survive for the next build. The PC seems more stable now, transfers between disks and flash drives got much better and stable, screen doesn't go weird, etc.

The only problem is when I choose to restart the system, it doesn't always wake up at first try, it doesn't post, I need to power off and power on again. Plus a few app hangs, but nothing like red event viewer and constant intermittent bsods. So far so good for about 5 days.

Thanks to all for your help.

edit: the PC just BSODed again without any apparent reason (I launched skype) then it wouldn't get to windows bsoding before the "loading windows" screen, I disconnected the secondary disk, swapped the cables, BIOS wouldn't find the boot device... ok, I used the same cable as before on the main system disk and it booted, I got to windows no problems

I've been running the PSU (different from the previous) outside the case for some days and didn't get bsod, then I put the PSU in + probably touched MB or other parts and PC BSOD.
So I took out the PSU again and no BSOD.
This makes me think that it's something shortcircuiting somewhere between the MB and the case? PSU? backplate? which could explain the randomness and intermittence of the problems.

The other day I realized that when mounting the CPU I tightened the screws too much, pc wouldn't boot. I took the radiator off and booted without RAMs - beeps. So I mounted it again with RAMs, but stopping half way before the screws were fully tightened, PC booted to Windows normally.

 

[Dave76]

Glad to hear you found the cause.

Hope this issue is resolved.
If you feel that it is, then mark the thread 'Solved'.

 

[R4ndom]

Hi, I haven't resat everything yet, pc just bluescreened with the PSU outside the case. The only component I haven't replaced so far is the case and the CPU, but I don't know if it can give such symptoms, several hours under prime didn't confirm that.
Windows error reporting is blaming ntfs.sys, which doesn't mean anything.

I purchased a phenom 4 695 to test, but if I unpack it and mount it, and it will not be the cause (didn't mount it so far, as the PC had been stable for 5 or 6 days w/out a single BSOD), I won't be able to return it unfortunately, so I think I'll first re-mount the whole thing.

 

[R4ndom]

Ok, so I re-built all from scratch, put in Corsair RAMs which I got back from RMA, PC BSODed after 5s in windows it's a second pair that fails memtest immediately - this time it's one stick though.

So I put back the OCZ, and I'm still getting BSODs + applications crash immediately and randomly when I click on the executables. Firefox stopped working bla bla.

What I noticed it's always the same scenario as in the 10 previous reinstalls: if I succeed to format, install, update, it runs fine for 3-4 days, then all starts again. Interestingly, it's always once svchost_sysmain errors start to appear in the event viewer that I start getting BSODs. What's even more weird is that I think this time I didn't see these errors appear in the log, but now they're there. The system is just getting corrupted over time and the problems snowball progressively.

This time while mounting the board in the case I took special care to see if there are no rivets short-circuiting with the board, I applied 1mm washers (non-metallic of course) on the standoff screws, but in a way to assure grounding to the board. This way I was hoping to further limitate the possibility of contact between the case and the board.


According to you, is it possible that it's the CPU causing such kind of errors? svchost_sysmain being responsible for prefetch data? something not quite right with its cache, memory controller?