BSOD with 2 Ram sticks | Faulty Mobo suspected

[Snortan]

I'm trying to revive this old pc and I'm stuck with these BSOD's. It used to work fine and haven't changed any hardware, only got new ram because 1 stick died.
PC Specs:

• MSI P43 Neo-F
• Q9550
• 2x2GB Generic Ram DDR2 800 5-5-5-13-1.8V
• 640GB WD HDD Blue
• MSI GTX 460 1GB
• Vantec Ion2 450W
• Windows 7 x64 SP1(fresh install

The PC runs just fine with just one stick of ram. When I add the second I can get into Windows with no problems and if I just leave it idle it most likely won't BSOD but sometimes it will do. What I found to crash it almost instantly is to install something (in this case it was nVidia driver, tried something else too) and it would crash or sometimes lock without a minidump even before it starts installing the files, like it would lock when it's unzipping the installer or checking requirements.
Also it will crash if I copy a large file from one location to another.
Same exact thing happens in safe mode.

Weirdly enough it will not crash if I put the system under stress. I've tried stress testing using Aida64, Prime 95, ROG RealBench for about 6 hours each and played Far Cry 2 for about an hour. This with both ram sticks installed. Temps remain in normal bounds, under 80.

What I've tried so far:

• checked the ram sticks separately using memtest for 6 hours each and then together without errors
• Tried leaving the ram settings on auto, also tried putting their manual timers 5-5-5-13 1.8V
• no overcloking and all bios settings set to AUTO, also tried increasing the volts a bit over default
• sfc /scannow says everything is fine
• tried 2x1GB DDR2 800 ram sticks and same story, 1 stick is fine, 2 is trouble
• updated all drivers/chipset/bios and windows updates
• there are no obvious blown caps on the motherboard
• Scanned the HDD with chdks and it says 0 errors/bad sectors

What else am I missing?
Could the CPU be making this issue? I think I might still have an e8400 somewhere but need to get some thermal paste too.

They only thing that I can think of that still remains "untested" is the PSU and the motherboard(don't have a spare for either). But I imagine the PSU is fine since under stress test everything works just fine.
So that leaves the motherboard being faulty with the ram slots?

Edit: Forgot to mention the bug string from BSOD's are PAGE_FAULT_IN_NONPAGED_AREA, BAD_POOL_HEADER, DRIVER_CORRUPTED_EXPOOL, IRQL_NOT_LESS_OR_EQUAL and DRIVER_VERIFIER_IOMANAGER_VIOLATION blaming HIDCLASS.sys and ntoskrnl.exe when I actiavted Driver verifier. Got this info from BlueScreenView, don't really know how to properly read minidumps.

 

[Wraith]

Check your Sata cables dude.

 

[Snortan]

Check your Sata cables dude.

Just got a new Sata cable and replaced it. No luck, same thing happens.
I've also tried changing form IDE to AHCI just to see maybe that would give something, but again no luck.
Any other ideas?

 

[agent_x007]

U tried upping Memory Voltage and/or NB Voltage ?

Other things worth considering are :
1) Checking SPD profiles, to see if timings are correct and trying to loosen them up a bit (to 5.5.5.18 or 5.5.5.15, as last resort try 6.6.6.18).
2) Doing Single Channel (just use two different colour memory slots for each module)
3) ReFlashing BIOS (if last update damaged it somewhat - unlikely but it can happend) or doing "Reset to default" and saving changes.

If all above will fail - it means that this RAM is incompatible with Your MB and/or setup.
Best way to solve this, is to change Your RAM to more "specific" and less "generic" one (for example like Kingston, GoodRAM, OCZ, A-Data, Corsair or anyother better known brand).

 

[NeverBackDown]

Make sure you are using the right cables. If you don't then you will BSOD or the psu will turn off due to failsafe features as it thinks it's not secure. That's happened to me before with my Corsair PSU.

 

[Snortan]

U tried upping Memory Voltage and/or NB Voltage ?

Other things worth considering are :
1) Checking SPD profiles, to see if timings are correct and trying to loosen them up a bit (to 5.5.5.18 or 5.5.5.15, as last resort try 6.6.6.18). However it seems to be less easy to reproduce
2) Doing Single Channel (just use two different colour memory slots for each module)
3) ReFlashing BIOS (if last update damaged it somewhat - unlikely but it can happend) or doing "Reset to default" and saving changes.

If all above will fail - it means that this RAM is incompatible with Your MB and/or setup.
Best way to solve this, is to change Your RAM to more "specific" and less "generic" one (for example like Kingston, GoodRAM, OCZ, A-Data, Corsair or anyother better known brand).

1)Yep. SPD says 5-5-5-13 1.8V, I set it to 6-6-6-18/20 1.9V (tried even looser 6-7-7-23) and upped NB volts. Same thing.
2)Again same thing, run 1 stick no matter the timings/volts and it's fine. Add the second and it locks and crashes.
3)Done that too, even tried and older BIOS version.
I have another set 2x1GB 800 Mhz Corsair and again 1 stick is fine, 2 it goes nuts.

Make sure you are using the right cables. If you don't then you will BSOD or the psu will turn off due to failsafe features as it thinks it's not secure. That's happened to me before with my Corsair PSU.

Yea double checked this, all cables are the correct ones. Even if I wanted to, couldn't use the wrong ones.

I'm going to try to run a copy of windows 7 from a usb and see what happens then, maybe its the hdd acting up but under chdks said its fine and s.m.a.r.t says it's fine too.

 

[Wraith]

Chkdsk isn't that good to be fair and S.M.A.R.T is all well and good so long as the data hasn't been reset or flashed, I've had feebay HDDs that report as 100% healthy but sound and perform like a bag of spanners mainly Seagate drives but I've had the occasional WD do it too.

 

[Snortan]

Any favourite tool to check for healthy or bad sectors then?

 

[Wraith]

Crystal Disks, Active SMART, Argus Monitor All good monitors and info checkers for drives.