Sign in with
Sign up | Sign in
Your question

Faulty mobo or bad RAM?

Last response: in Motherboards
Share
November 22, 2012 8:23:49 AM

Hi everyone,

I've been struggling to troubleshoot random BSODs and system instability for quite some time now.

I had bought a new motherboard (and a few other parts) but kept the same processor, wanted to upgrade next year to Haswell.

Anyway, the rig:

Intel stuff:
Core 2 Quad Q8400
DP43BF mobo
120GB 320 series SSD

Others:
MSI GTX 560 (reference clocks)
2x4GB Corsair Vengeance DDR3-1600 (9-9-9-24) @ 1333 (9-9-9-24)
WD Caviar Blue 320GB
Cooler Master GX550w

Windows 7 SP1 64-bit, fully patched
Had Norton 360 for a long time but in the course of troubleshooting (and an expiring subscription) i've currently got MSE running.

There's a DVD writer, a cooler and a case as well, but i don't think they're relevant here (temps are fine).

After i put this together, the system passed prime95's Blend test for a good 8 to 10 hours, and multiple passes of memtest86+, along with windows memory diagnostic, Sandra, you name it. iirc even with 8-8-8-22 timings for the RAM.

After about a month or more of running this rig with the RAM latencies tightened to 8-8-8-22 (although i could go as low as 8-8-8-14), i had the first BSOD on this rig. So i reverted to the earlier timings (9-9-9-24), and faced no problem for a number of days. I remember getting a "hardware error detected" with the OC RAM at this point using Windows memory diagnostic. I believed i switched both RAM slots to A1, B1 from A0, B0. Memtest never showed any errors.

Then the BSOD and instability army took over. Eventually getting tired of this, i re-installed windows, since WhoCrashed would point to the kernel as the source of the crash, and no drivers were named. I used to get PAGE_FAULT_IN_NONPAGED_AREA (0x50) and once or twice i got a MEMORY_MANAGEMENT (0x1A) error...don't remember anything else. Win mem diag still never showed a hardware issue. Finally when the problems continued after re-installing Windows (and severe ones, like unsuccessful updates, very frequent BSODs, etc), i ran prime95 again, it failed after some time. So i ran memtest, and after 2 passes it found one or two errors.

So i tested the sticks one by one, in all slots, they cleared multiple passes, showed no error.

I put them back in together, and they passed.

I ran prime95, and it fould fail Blend after 1hr 22mins (three of four workers would fail), and after testing i used to get over a thousand errors in memtest.

But, if i switched off standby power too, turned the system back on and ran memtest, there were no problems.

Here i began to think that errors were being induced by either the software or the mobo/CPU.

So i tested with small FFTs, and it would fail instantly. Same thousand+ of errors in RAM.

Then i switched RAM slots, and well. No errors anywhere. Not in memtest, not in prime95 after 2 hours of Blend.

But now it's happening again, more BSODs and this time i got an additional IRQL_NOT_LESS_OR_EQUAL (0xA) code.

Memtest showed a few hundred errors. So i took out one of the sticks, and...no error again.

I suspect my motherboard can't handle dual channel or something. It supports DDR3-1066 and DDR3-1333 (with an OC), but i can't down clock the RAM manually, the system's setting it automatically to run in OC mode, i.e 1333 MHz.

FSB = 1333 MT/s. My mobo also supports a max of 8GB RAM, but i could actually use all 8GB (not 7.x) so i guess it must automatically support memory remapping.

My SSD is fine, performed short and extended tests on it, looked at smart data for both drives. PSU is fine, it's over a year old, provides stable voltages. The GPU hasn't given me trouble either, occasionally used to get driver crashes but then that was usually with BF3.

But never had a BSOD while playing a game or running a benchmark. It's alway while browsing or like yesterday in PowerPoint. Today i asked MSE to do a "quick scan" and it gave me BSOD. the COM service crashed, steam's been either crashing or resetting my Planetside 2 download.

It completely sucks at this point.

One stick's been stable so far. So what you all think? Mobo issue? Have the latest BIOS and chipset drivers. using Nvidia's 301.42 for now, kind of felt that 306.97 was unstable.

Thanks a lot!

More about : faulty mobo bad ram

November 22, 2012 8:27:59 AM

Forgot to add, RAM wan't a kit. Bought 2 separate 4GB sticks.

but i bought them within 2 weeks of each other (had to buy them together but long story, i digress) so i don't think that's causing issues.
m
0
l
November 22, 2012 5:26:39 PM

Update: The stick in A0 also resulted in a BSOD.

Tried the other stick in B0, passed memtest and no problems so far.
m
0
l
November 29, 2012 11:59:45 AM

[solved] was bad ram, 1 stick.
m
0
l
!