Backov

First off, I am not a hard core overclocker, but I often lurk here when I need hardware questions answered, you guys know your ****.

That said, I recently did an upgrade. I wanted to go to 8 gigs and RAID0 to speed up my Warhammer Online experience. So, to do that I needed to replace my aging P5B with a P5Q Pro and some new ram.

So, I get the P5Q Pro, move over my E6600, get the RAID working and windows installed, and it's golden. It runs like a champ, for about 4 hours. After 4-5 hours, if I boot into memtest86, all of the ram fails on just about every test. (It also blue screens like crazy in windows, with lots of services crashing first.)

If I put just one stick in to one slot, it fails.
If I put a stick from my previous mobo in, that stick fails.
If I put a stick in from my media center pc, that stick fails.
It doesn't matter what slot I put it in.
I put the new ram in my media center pc and ran memtest86 overnight, no errors.
I put my old mobo back together and ran memtest (only about 4 hours, should have run it longer in hindsight) and no errors.

So, when that happened, I was thinking "bad mobo" - so, I got the dudes at my local shop to look it over. They did a piss poor job of diagnosing the problem, that's another rant, but they did replace the board and send it back. Exact same problem.

So, I read that sometimes the NB/SB heatsinks on ASUS motherboards can be poorly applied and overheat, so I got yet another new board and popped the NB/SB heatsinks, cleaned them, put new AS5 paste on and reseated. I also did the same to the CPU. Temps were fine (Speedfan) during operation, never going above around 40ish core (IIRC). Same failure after 5 hours.

On the latest failure, I unplugged all of the HDDs and tried it, same error.

So, the equipment in question is:

E6600
8 gigs Mushkin (EM2-6400, 5-5-5-12)
Corsair HX520W (1.5 years old)
P5Q Pro motherboard (3 different ones now)
8800 GTS 640mb - unfortunately I don't have another vid card to use instead to take the load off the PSU, and there's no onboard video.

Now, I'm kind of at a loss about where to go from here. The tech where I bought the hardware is "special" and had my machine for 2 weeks with no fix, so it's not going back there. I'm considering getting a multimeter to test the PSU with, but I don't imagine it's at fault.

So my only guess left is the CPU, but that seems odd - the very little I've read says that it's "very rare" - and it seems strange that it would only cause problems after 4-5 hours when the temps don't change.

Any help?

Cheers,
Jason

Moheevi_chess

Did you make sure you had the right voltage set for the RAM?

__________________

Backov

I was using auto, which is 1.8 on this mobo, and then I upped it to 2.0, same error. My thought was that if it were voltage, it wouldn't fail after 5 hours and then wouldn't continue to fail even with only one stick of ram.

The Master Chief

Bad dimm slot.

__________________
Overclocking stress testers: Prime95 64bit32bit AnotherlinX

Overclocking Essentials:CPUz Core Temp Everst

GPU Overclocking Essentials: MSI Afterburner GPUz Furmak stress test
MINNESOTA OVERCLOCKERS | The Climate Phenomenon

Rpg2

Up the RAM voltage to 2.1v as well as bumping up the Northbridge (it might be called MCH on some boards) voltage by +.1v.

Let us know how it goes.

Backov

On three different p5q pros?

I'll give those voltage bumps a try now.

glenbuck1914

Disable usb legacy support in the bios. That causes memtest to fail on some Asus boards. It's a shot in the dark, but may work.

Edit: Is the error always 00000000000??

Backov

When I was still letting it blue screen, it was a bunch of different ones.. memory manager, 1a, etc - the googling told me they were all ram related.

glenbuck1914

No worries, but disable usb legacy support in the bios and then run memtest again and see if it works out.

Backov

Yep, I did that the second I read your post, thanks. Now I'll probably have to wait a few hours to see.