GA-880GM-USB3 gives randomly BSOD

[Chem.Assistance]

psu: ANTEC VP 450W 12CM ATX12V2.3
ram: Crucial 4GB (2GBx2) 240-pin DIMM, DDR3 PC3-10600, CL=9, Unbuffered, NON-ECC, DDR3-1333, 1.5V, 256Meg x 64
cpu: AMD Athlon II X2 Dual Core 270 (3.4Ghz/2MB/65W) Socket AM3
hard drive: SEA BARRACUDA 7200.12 1TB SATA2 32MB
gpu: nvidia geforce 210 1GB
gpu: amd radeon hd6870 1GB

with geforce 210 I had 5 BSODs so far:
SYSTEM_SERVICE_EXCEPTION caused by driver ntoskrnl.exe by address ntoskrnl.exe+7f1c0
KMODE_EXCEPTION_NOT_HANDLED caused by driver ntoskrnl.exe by address ntoskrnl.exe+7f1c0
IRQL_NOT_LESS_OR_EQUAL caused by driver MpFilter.sys by address MpFilter.sys+7128
PAGE_FAULT_IN_NONPAGED_AREA caused by driver ntoskrnl.exe by address ntoskrnl.exe+7f1c0
SYSTEM_SERVICE_EXCEPTION caused by driver ntoskrnl.exe by address ntoskrnl.exe+7f1c0

then I switched to amd radeon hd6870 and got 1 BSOD so far:
SYSTEM_SERVICE_EXCEPTION caused by driver atikmdag.sys by address atikmdag.sys+9fb5

I have tested ram with memtest, no errors.
Also I am using Windows 7 for only playing Team Fortress 2 and it crashes 2...3 times a day.

I would like to know how to fix this instability

[absic]

Hi there,

you don't say which revision your motherboard is but are you running the latest BIOS for it? If not, try updating to the latest version to see if that helps.

Your PSU may also be under powered as Gigabyte recommend a PSU of at least 500 Watts for this motherboard. If possible try a larger unit and see if that helps with stability.

[Chem.Assistance]

BIOS is latest (FE; 2011/10/25), motherboard revision number is 3.1

newegg psu calculator gives 396W with amd radeon hd6870 and 240W with nvidia geforce 210

Don't bsod errors hint at anything?

[Vezina]

Looks like RAM issue to me ,but the power supply is not good in my opinion.Use a 550W real watts PSU ,your ATI video card requires this.

Use quality power supplyes ,if Antec use Signature series not the low ends.
For low prices  affordable ones use Corsair builder series or OCZ or Chieftec or Sirtec.

[Chem.Assistance]

Looks like RAM issue to me

But why then no errors in memtest?

[absic]

Your BSOD errors would seem to point to an issue with drivers but these can be caused by other aspects of your PC's configuration. I have seen similar problems on various machines which have been solved by using a PSU that supplies more even power. The figures you quote from the Newegg PSU calculator, are the minimum required for the system and, as I mentioned earlier, Gigabyte recommend at least a 500 Watt unit for your motherboard.

Memory could also be an issue but, once again, this could be due to power rather than faulty modules. MemTest works through DOS and uses basic drivers for graphics and other hardware on your system which requires less power to be drawn than when running in Windows and playing games. This is another reason why I suspect the power supply as the cause of your problem as you have stated MemTest passes your modules when you tested them.

To resolve the BSOD's you have several options:
Try a PSU that has more power - If you can borrow one from a friend to test with this is the first thing I would try.
Do a clean install of the OS to eliminate the possibility of problems with drivers etc. - Time consuming, especially if you have lots of software installed.
Test your system using different RAM modules -  This option will highlight if your RAM is the cause but, once again, you may not be able to do this easily unless you have spare modules or a friends you can borrow for testing.

[Chem.Assistance]

I replaced ram and now crashes in TF2 are gone. So far no BSOD either.

previously I had crucial CT2KIT25664BA1339: http://www.crucial.com/store/partspecs.aspx?imodule=ct2kit25664ba1339

now I have corsair CMV4GX3M1A1333C9: http://www.corsair.com/cmv4gx3m1a1333c9.html/


Still I find it odd that memtest didn't find anything. Was that some kind of compatibility issue with motherboard?

[absic]

Pleased you seem to have solved the problem even if it is down to the RAM.

MemTest is not perfect and there have been several instances posted on this forum, where modules have passed testing with this software only to be found later to be the real cause of issues.

Without knowing your testing methodology it's hard to say exactly why this has happened to you but often running Memtest for a short time (1 - 4 passes) per stick and not testing a complete set of RAM, after testing individual modules is a possible reason.

Running the tests is time consuming and many users don't want to waste a day/night running the tests fully and switching RAM modules around. Letting Memtest run through a minimum of 10 passes per stick of RAM and then again with a full set is often the best way for errors to be shown but, once again, this is no guarantee that the modules are fine, even if no errors are shown, as you have now discovered.

[Chem.Assistance]

I run memtest overnight. 12 passes 0 errors. Are you sure it's not some weird voltage or timing issue instead?

[absic]

There is the possibility the problem is due to timing or voltage issues but without having your PC and your components in front of me, it is hard to say exactly why the errors have occurred. All you can do is experiment (with the original RAM) and see if changing the timings and or voltages have any effect.

Sometimes though, you never find the real answer and have to just accept that you are dealing with a PC and strange things can happen.

[trublubiker]

Mate, I feel your pain and frustration.

I have a similar set up, see specs below, and for six months I was having up to five screen freezes per day. It was driving me nuts and M$ tech was no help, always blaming me and every component in the computer. You are not alone, just google 'Windows 7 64bit screen freeze' and meet about 17,000,000 others with the same problem.

I decided to try and see what was happening when the  problem occurred so I downloaded and installed a little program called 'Core Temp' whose monitor now lives on my screen. It tells me CPU usage, RAM usage and CPU temps.

I very quickly found out that the W7 64 bit OS chews up almost 3GB of RAM just to keep it running, even after pruning all the crap from start-up etc. Try doing a focus stack and it struggles, big time. I was getting so many screen freezes it was driving me nuts. I imagine playing games would have a similar effect.

With the original 4GB of RAM installed I noticed that every time I had a screen freeze, my RAM usage was off the scale. Swapped the RAM for 2x4GB and haven't had a screen freeze since.

CPUID, another free download, alerted me to the fact that my CPU fan was spinning at up to 10.000RPM. Not good for CPU longevity. Hence the upgrade to the Noctura CPU cooler set-up.

My system:
Mobo: Asus  GA-880GMH-USB3                                                                                   
CPU: AMD Phenom II X4 955 @ 3.2 MHz (Black Edition ) - Unclocked
RAM: 8GB G.SKILL Ripjaws @1333MHz - Unganged
CPU Cooler: Noctura Twin Fan
PSU: Cooler Master eXtreme Power Plus 550W
GPU: Radeon HD 5750 with 1GB GDDR5
Case: Antec 300 plus 3 extra inlet fans and 1 exhaust (YATE LOONS)
HD: 2 x Western Digital 500GB + 1 x 2TB
OS: Windows 7 Pro 64 bit
Monitor: Dell 22" TN (My next upgrade is to an IPS)

Obviously I can't guarantee that doubling your RAM will solve your problem, but it sure as hell fixed mine.

Good luck

Kevin

[Chem.Assistance]

In case I wasn't clear enough before: my problem has been fixed after I changed my ram to Corsair.
No crashes in TF2 and no BSODs in Windows