I7 system randomly resets

[pokornym]

Hello! My son & I just built our first system. The system started randomly resetting itself for no apparent reason, mostly when the system is idle. Here are our components:

Gigabyte GA-EX58-UD3R Rev 1.6
i7 920 running at 2.66GHz (no overclocking)
Zalman ZM750-HP
6GB (3x2GB) Kingston DDR3 memory
EVGA GeForce GTX 260 Core 216
Seagate Barracuda 1TB 7200rpm SATA
Sony Optiarc 24xDVD/CD SATA
Koutech IO-RCM620 USB Card Reader
Edimax EW-7128G PCI Wireless NIC
Windows 7

After the initial build and software installation I noticed the problem. During one reboot, the Windows7 system repair scan said an unidentified device might be installed incorrectly, so I've run system restore multiple times and ended up back at the basic installation with just the NIC installed. That appeared to give me a stable system, so I added the peripherals back again, one at a time, running the system repair tool again after each. Once the peripherals were installed with no errors, I started with software.

I only have a few programs installed:

Adobe Flash Player 10 ActiveX, Avast! AntiVirus, Browser Configuration Utility, Cisco EAP-FAST Module, Cisco LEAP Module, Cisco PEAP Module, Edimax RT6x Wireless LAN Card, EVGA Precision 1.8.1, Logitech Webcam Software, Logitech Webcam Software Driver Package, Multimedia Card Reader, NVIDIA Drivers, NVIDIA Physx, NVIDIA Stereoscopic 3D Driver.

At some point in my troubleshooting I started the BIOS and left the computer run for several hours. While there I checked the system temp, which was stable at 33C and the CPU temp which was 44C - 45C. It seemed to remain constant while I monitored it (periodically for a few hours). The system did not reset during this time.

The problem seems to appear when I install Office 2003. I am trying to rule out the hardware. Everything on my system appears to be installed correctly...we triple checked during the build. Is there anyway I can verify the hardware is not the problem? One thread suggested a power supply problem, but I don't know how to check that the PS is working properly.

Thanks for any advice!

 

[o1die]

Check your kingston ram specs; you may have to adjust the voltage in the bios. Normal voltage is 1.5-1.65 volts.

 

[pokornym]

Check your kingston ram specs; you may have to adjust the voltage in the bios. Normal voltage is 1.5-1.65 volts.

Thanks o1die. I checked in the BIOS and the DRAM is set to Auto at 1.5V, which matches the Kingston spec for the modules.

I do have some additional info from the Adminstrative Event Logs:

Unable to start a DCOM Server: {8D9A64F2-357D-40C9-97CD-69FA7E64A518}. The error: "193"
Happened while starting this command: C:\Program Files (x86)\Common Files\LogiShrd\LVMVFM\LVPrS64H.exe -Embedding

The Ralink Registry Writer service failed to start due to the following error: Ralink Registry Writer is not a valid Win32 application.

The Ralink Registry Writer 64 service failed to start due to the following error: Ralink Registry Writer 64 is not a valid Win32 application.

The NVIDIA Stereoscopic 3D Driver Service service failed to start due to the following error: NVIDIA Stereoscopic 3D Driver Service is not a valid Win32 application.

Unable to start a DCOM Server: {D0B7C734-2D1B-461D-93C6-8264DA4F038B}. The error: "193"
Happened while starting this command: C:\Program Files (x86)\Common Files\Logishrd\LQCVFX\COCIManager.exe -Embedding

No messages appear on boot up, and the Action Center does not flag anything as needing attention, but these error messages are consistent whenever the system boots.

 

[pokornym]

I've played around with power settings, but that does not have an effect. I've also booted to Safe Mode (without network support), but the problem showed up right away there. I can't duplicate it at will, but it has shut down at least twice when opening MS Outlook (2003)...that's as close as I've come to duplicating the problem "on demand" so to speak.

I am of the opinion that this is a software issue with the OS. Anyone have any suggestions that might point another direction? Can anyone suggest a Win7 registry scanner?

 

[trep-]

What computer case you're using ?

I had some static discharge problem with my Antec case because of the front bezel. The computer would just reboot when the cats would run in front of the case or i would walk or vacuum near it.

 

[pokornym]

Interesting, I had not even thought along that line. Now I'm wondering if I have some metal touching the MB somewhere. The case is an AeroCool AeroEngine Plus (http://www.aerocool.com.tw/case/ae-plus/ae-plus.htm).

The restarts happen even in the middle of the night when nobody is up (supposedly...I do have 5 little ones ) Doesn't quite fit your description, but I used twist ties for cable mgt. It's possible the coating was broken at some point. I've intended to get some zip ties...I'll try that and see what happens, and let you know. Thanks!

 

[jsc]

It would be a lot of work, but you can take the system apart and breadboard it:
http://www.tomshardware.com/forum/262730-31-breadboarding

That would certainly eliminate the case as a problem.

And our standard checklist and troubleshooting thread:
http://www.tomshardware.com/forum/261145-31-read-posting-boot-problems

 

[pokornym]

Thanks, JSC. That does sound like a lot of work, but I'll keep it in mind as a last resort. I'd rather try the zip ties first.

To be honest, I'm not convinced this is a HW problem yet. My wife was up late a few nights ago reading and the computer turned itself on! She said it was talking...to itself of course. She thinks it is possessed! I found this morning that I have like 36 event triggers and I had set the OS to allow events to start the machine. I'm wondering if these triggers could "bump into" each other so to speak, and cause the machine to restart. I've disabled the trigger restart option, but have to research to see what the heck they all are.

I read through the troubleshooting checklist. We did (or didn't do, as appropriate) all the things I found there. Plus, since the system boots normally, it seems like the checklist is solving a different problem.

BTW: What does "Sticky" mean? Yes, I'm a newb too!

 

[trep-]

Stickied means the thread is "stuck" to the top of the forums. Even if no replies are added in a long time (or the thread might be even lock), the strickied threads will be the first one you'll see when browsing the forums.

Good luck trying to find the source of the reset ! A simple way to remove any software issue would be to install maybe winxp on a drive and let the machine run for a while with it installed (or do a complete reinstall of win7). Also, you might be able to do a system restore (maybe some software is causing the issue) and go back in time hehe

 

[pokornym]

trep-

I like the XP hard drive idea. I've used System Restore to go all the way back to the beginning when I installed Win7. At first I thought it might have been Office 2003 causing the problem, but the problem repeated even with just the OS and wireless network driver installed. I've started to think about a complete re-install too. I'm wondering if I do the same thing as I did with the initial installation if I'll get the same results though.

The other thing that I've been thinking lately is to have windows repair itself...I think that's a pretty easy step.

 

[naim32]

What PSU are you using?
I had similar problems although tehy appeared about 3 months after the build.
I bought a cheap supply ~50ukp once I replaced it with a quality brand ~150ukp all my problems disappeard.

The link to my thread on the issue --
Another Random Shutdown

 

[pokornym]

Zalman ZM750-HP. Last weekend I think I've ruled out hardware. I started a profile using SyncBack SE from TwoBrightSparks to transfer data from my old computer to the new one. During the course of the set up, I realized that very rarely has the computer shut down when I was actually using it. It's always done it when I wasn't paying attention to it. Now aside from the possibility that my computer is female (sorry ladies, I couldn't resist) I wondered if opening an application would keep the computer from shutting down.

So, I opened MS Word and the Windows7 Resource Monitor. The system ran for 30 hours transferring the files happily without a single shutdown.

While you cannot prove a negative, the system had been shutting itself down every 30 - 45 minutes when idle. So, I'm still wondering about the cause, and I appear to have a work-around. Any suggestions?

 

[naim32]

Does the system restart as a Windows restart or does it just cut the power.
If it's cutting the power then I would say It's likely to be a hardware issue.

Dodgy drivers, software and memory tend to result in BSOD in my experience.

Another thing you could look at are the NB and MOSFET heatsinks around the CPU.
I originally used a tower style cooler on my I7 and noticed that there was very little airflow around these heatsinks. These were getting very hot, so much so that it was difficult to touch them. This was helped when I put the stock down blowing cooler on.

I used to get these shutdowns when I tried overclocking, the CPU seemed OK ~50 deg with the Noctua cooler but the NB was too hot to the touch ( My BIOS doesn't report NB temps ).

You could try running Prime95 to stress ( heat up ) the CPU and the NB and see what happens.

 

[pokornym]

naim32,

The system just cuts power, then restarts after a few seconds delay, kind of like you pressed the reset button. I have not seen the BSOD. My system does not monitor the NB temp either. I don't know what the MOSFET heatsinks are exactly...I'll have to look at the MB to see if I can tell where they are. I do have three cooling fans in addition to the CPU fan: 120mm, 80 mm, & 210mm. Could I really still have an airflow issue with that many fans?

I will look at running the Prime95 stress test this weekend.

Seems odd to me that the system would be stable with Word open, but not otherwise. If this were a heat issue, I think it would have to appear regardless of the software running. Also, the system has shut down right after first power on when it was cool. It did it tonight...I turned the system on after being off all day and started Outlook. Left for a couple of minutes and came back to find it had restarted. That doesn't sound like heat to me, though I understand that everything was heating up as a result of being started.

Thanks,

Poke

 

[naim32]

naim32,

The system just cuts power, then restarts after a few seconds delay, kind of like you pressed the reset button. I have not seen the BSOD. My system does not monitor the NB temp either.

This is what started happening to me, the system would do a hardware reset for no reason. This then progressed in having to turn off the supply before if would restart. A few days after it needed to be switched on/off several times to get going.

I don't know what the MOSFET heatsinks are exactly...I'll have to look at the MB to see if I can tell where they are. I do have three cooling fans in addition to the CPU fan: 120mm, 80 mm, & 210mm. Could I really still have an airflow issue with that many fans?

The MOSFET heatsink will be the long thin heatsink(s) around the CPU. The North Bride is the square one. If you notice they all have their fins perpendicular to the CPU. The downdraft from the CPU is meant to aid their cooling. A tower cooler won't do this effectively.
With regard to fans I have 140mm Front Fan, 140mm top fan, 120mm rear fan, 2x 120mm side fans and things still get hot ( this is with 2xGTX 260s).
In my system one of the GPU is in the slot next the the NB restricting some of the airflow.
Another potential for heat problems could be very badly seated heatsinks so even if they don't feel hot the die temp could. I would be a matter of seconds to reache overtmp.

Seems odd to me that the system would be stable with Word open, but not otherwise. If this were a heat issue, I think it would have to appear regardless of the software running. Also, the system has shut down right after first power on when it was cool. It did it tonight...I turned the system on after being off all day and started Outlook. Left for a couple of minutes and came back to find it had restarted. That doesn't sound like heat to me, though I understand that everything was heating up as a result of being started.

Thanks,

Poke

For the 'word open' fix, is this repeatable or just a one off.
I would agree that this doesn't sound like a thermal issue and sound more like an emergency shutdown of the PSU.
The PSU will usual shutdown due to overtemp or short circuit and overload protection.

Have you used all the PCB supports for the MB? Could there be any possible shorts from the MB to the case?
It's already been suggested to breadboard the system so you might want to try this. It should eliminate over heating and shorting problems.

Otherwise you could just have a duff supply. I assume it's brand new so it could just be a manufacturing fault.
It's hard to test the PSU, mine could be fully loaded with 2XSLI 3XHDD GPU and CPU at 100% etc and still run. It would then die with just the windows desktop.

Have you witnessed the resets?
It's a long shot but, could some of these be down to Windows Updates.

 

[pokornym]

naim32,

Wow, there's a lot to think about. The Word fix is repeatable and consistent, and it's a good part of why I'm skeptical as to hardware. I have witnessed them. I don't think they're related to Windows Update...it has happened within the last couple of days and the last update was 1/28.

I'm planning to open the system up this weekend and replace the wire ties with zip ties. I will see how hot the MOSFET heatsink is.

Thanks for the tips!

 

[pokornym]

Yesterday I replaced the wire ties with zip ties. It did not stop the resets.

Naim32, I also checked the MOSFET heat sink and it was just warm to the touch. I had let the system run for several hours (using the Word workaround!) to be sure it was good and heat soaked.

I tried running chkdisk at startup, but the system would restart before getting through the chkdisk. So, I guess I've found a way to "force" a restart.

Next steps: Prime95 to stress the system, re-install windows, then breadboarding the system. If it restarts while out of the case, I'll replace the PSU.