Restart, no BSOD, no overheating

Custom / CUSTOM
October 10, 2009 at 22:26:51
Specs: Microsoft Windows Vista Ultimate, i7 920 / 6gig DDR3
Hey guys...

Having a wee bit of a problem with my build.

Build info:

i7 920 cpu (stock)
Asus P6T6 WS motherboard
Asus Matrix GTX 285 x2
2G Corsair DDR3 1600 RAM x3
Cooler Master real power pro 1000w power supply

All drives are western digital
500gb 32mb (for OS)
750g 16mb
1t 16mb

The problem:
When I play games (ie: Crysis, L4D) the computer randomly restarts. There is no BSOD, no warning and completely random. Can happen 2-5 minutes into the game play, can happen 30-40 minutes into the game play.

What I have tinkered with and tried
The RAM i got is a 1600 (and is on QVL for mobo) and, as the mobo says, it can support it in the OC. Dug through some forums and was suggested to just change the ram to 1600 in BIOS, was fine. Then changed the DRAM voltage to 1.64 ( i did 1.64 as I did not like the warning message at 1.65...although supposedly its fine, but still) and QPI to 1.4
Everything was alright.

What I did to try and fix it.

I am 99.999999% sure it is not overheating...
I have been monitoring the system as that was my first thought and while gaming and doing 3d mark test, CPU hovered around 51-60C while GPUs were never above 71C (on Crysis, everything on ultra high). The case is a HAF good airflow as well...

Computer worked fine, then while gaming randomly restarted, a messaged popped up saying "overclocking failed" I went into BIOS and everything was back on auto. Did some gaming, random restart. Went back into BIOS, set the ram at 1600, clicked f10 to save and exit, random restart as soon as i pressed i thought (aha...bad ram). Went back into bios again, ram was sitting at 1600...i set it to auto, f10....random restarted again as soon as i pressed it. I thought that was very starange so I decided to test the system, gamed.....restart. I thought maybe something up with bios so I updated it and flashed it and set all to default factory. Gamed some more, restarted....I started to get annoyed. I completely reinstalled my GPU drivers, problem continued. At this point I was about 95% there is something wrong with the ram because maybe my tinkering did something to it and also as soon as I touch it in BIOS and attempt to save it, computer restarts. I made sure to change other minor bios settings and then save and exit, and the restart does not happen. Then I thought that maybe when the RAM goes to high frequency it encounters I set it to 1333 which is also supported by the board and without any OCing....f10....again, restarted. I went back into bios and it was sitting at 1333...

I did a windows memory test which did not detect any errors.
Also performed a directx test on the problem there either.

I also checked the event viewer in windows and at first, it was giving me the following after the restart:
\SystemRoot\SysWow64\Drivers\ULCDRHlp.sys has been blocked from loading due to incompatibility with this system. Please contact your software vendor for a compatible version of the driver.

I got rid of the program which was dealing with the file (it was a Ulead Burn.Now 4.5 SE) thinking it would help..but no, same problem. Now when I view the even logger all it says is that the system was shutdown unexpectedly. (drivers and bios update was done after I uninstalled the software) but...same thing.

Did a mem test, 2 passes...0 errors...

See More: Restart, no BSOD, no overheating

Report •

October 10, 2009 at 23:02:06
Disable the "automatically restart" feature:

Report •

October 10, 2009 at 23:05:26
I did before.
There was no BSOD at all...just a total shut-down.

Report •

October 11, 2009 at 07:59:55
"There was no BSOD at all...just a total shut-down"

Then it's probably a hardware problem rather than a software problem, *possibly* power related or heat related.

Units with multiple +12v rails are not the best choice anymore & haven't been for some time. On a multi +12v unit, the amperage is "trapped" on the individual rails...what that means is you need to "hand balance" the load by distributing the hardware across multiple rails. You don't want one rail overloaded while others are under utilized or not used at all. Units with a single high amperage +12v rail are the best way to go because ALL the amperage is available & no balancing is required.

Cooler Master isn't one of the major players in PSUs & you certainly don't need anything close to a 1000W. You could have done much better for a lot less money. For example, here's an 850W Corsair with 70A on a single +12v rail for just $120:

Double check the way you have things connected.

Report •

Related Solutions

October 11, 2009 at 09:16:49
Yea will do....although I don't think there are many ways I could have screwed up the connections...its only what...mobo, 2 cards, 3 HDDs and a fan (other 2 r on mobo). I got 1000w because I was planning on upgrading later...but....gah I will have to talk to New egg about refunding me or replacing this unit with another one....

Report •

Ask Question