Solved Constant Random BSODs on a 5 days old system

[zipsi]

It's been a while.. I suppose the PC is fine now?.. or just lost hope?

 

[iTitus]

Knock on wood.
Hadn't had an error since the last starcraft fail. And I did play it
Seems to be a workarround/solved

 

[Dave76]

Glad to hear the news.

If you think it's 'Solved' click the red triangle in the upper right corner of your post and put 'Solved' in the message box.

Let us know if you have more issues.

 

[iTitus]

Just got a BSOD again
Don't know whether this is going to continue

 

[Jonathan_King]

I haven't re-read the thread (yet), but have you tried a new GFX card?

...Summary of the Dumps:

[font=lucida console]
Built by: 7600.16617.amd64fre.win7_gdr.100618-1621
Debug session time: Sat Oct 16 11:43:24.063 2010 (UTC - 4:00)
System Uptime: 0 days 6:38:54.060
BugCheck 1000007E, {ffffffffc0000005, fffff880041d553a, fffff8800a0ce598, fffff8800a0cde00}
Probably caused by : dxgmms1.sys ( dxgmms1!VIDMM_SEGMENT::TrimAllProcesses+10e )
PROCESS_NAME:  csrss.exe
BUGCHECK_STR:  0x7E
¨¨¨¨¨¨¨¨¨¨¨¨¨¨¨¨¨¨¨¨¨¨¨¨¨¨¨¨¨¨¨¨¨¨¨¨¨¨¨¨¨¨¨¨¨¨¨¨¨¨¨¨¨¨¨¨¨¨¨¨¨¨¨¨¨¨¨¨¨¨¨¨¨¨¨¨¨¨¨¨¨¨¨¨¨¨[/font]

 

[iTitus]

I havn't tried an other graphics card yet. But after getting this error I played starcraft for a few hours without problems. Strange...
I'm actually running ati 10.9a, the hotfix fixing the overscan issue:
GPU-78: ATI Catalyst

 

[zipsi]

I THINK the issue is PSU related! I googled a bit, such bsods appear on many people.. and all of them are fixed after a PSU upgrade.. I think you bought a too weak one (520W).. a 6 core CPU, a case which has LED's in them and god knows how many hard drives and extras. You are slightly underpowering your GPU, not a complete catastrophe, but enough to make it unstable and cause BSOD's.

 

[iTitus]

No leds and just one hdd
But i'll keep it in mind

Titus

 

[Jonathan_King]

It seems almost certain this is some kind of hardware issue. PSU is possible, although it seems like you have plenty of power from a reliable brand. Of course, any PSU can fail, especially older ones, but I would start checking other components first.

Test RAM again, with all sticks in. In case you threw away your CD or forgot how: http://www.sevenforums.com/tutorials/105647-ram-test-memtest86.html

Test GFX card: http://www.sevenforums.com/tutorials/100356-video-card-stress-test-furmark.html

 

[zipsi]

It could still be a PSU issue, its not the maximum watts I'm worried about, its the amps on 12v rails.

 

[iTitus]

Had another crash...

They seem to occur rarely, although a little more often last time.
I've got the feeling that they are different each time.
What about the 12V rail?

Thanks

 

[Jonathan_King]

I have read through this thread, and I notice a common cause throughout: video.

That can also be caused by bad RAM, which has been investigated earlier.

Run Memtest again, just to make sure.

If you have another graphics card to test, please do so. If you can't borrow one from a friend, purchase one from the local computer store. Be sure it allows you to return the card, should it turn out it wasn't the problem.

 

[iTitus]

On pages 7 and 8 I described the tests with my old GPU, a Nvidia 6700XL
Hope that helps.

Cheers

 

[Jonathan_King]

I guess all I can think of is a bad motherboard. Have you tried a new one?

 

[bwpotter]

Hi iTitus:

I know I'm sounding like a broken record here (reply 1) but I really think you should try a new board. The connection with video issues could be a result of the Northbridge under-power issues noted in replies 2 and 3.

Good luck... Brian

 

[Dave76]

As mentioned previously by bwpotter, you need to bump your NB (North Bridge) voltage to 1.2v. This should help if the cause is the RAM.

Also, verify the RAM is in single channel and un-ganged, leave the DRAM voltage where they passed before.
Is the RAM voltage still at 1.6v?

If this doesn't help, go with bwpotter's suggestion and get a new motherboard.

 

[iTitus]

Still running 1,6V DRAM.
Never did something to the NB

Also I'm running Dual channel un-ganged

Titus

 

[bwpotter]

Yes, for the time being, leave the DRAM at 1.6v and raise the NB to 1.2v; they are independent of each other.

 

[iTitus]

Got another crash, it seems that this problem is returning.
I'll send a few emails to the store I bought the MoBo to see what they think about it and an RMA.

EDIT:
I emailed the shop and he said that I could try to lower the speed form 1333Mhz to 1066Mhz. He also said that if that didn't work I could return it. So I asked how the return procedure would work and what would happen if a possible new set of memory had the same problem(ie new mobo).

 

[Dave76]

Agree with bwpotter, leave the DRAM at 1.6v and raise the NB to 1.2v...

That's going to be your best bet. RMA the MOBO and RAM. Double check the MOBO recommended RAM and get it from their list.

If you want to give it one more shot, set your RAM to run in single channel mode. Your MOBO manual will tell you how, you will need to change the slots the RAM is in.