Sign in with
Sign up | Sign in
Your question

Computer Restarts On Its Own

Last response: in Systems
Share
March 7, 2012 3:49:27 PM

Computer I just built is restarting on it's own. Running Windows 7 Professional, clean install. No apps other than utilities that come on the drivers CD (such as a CPU temp monitor, and the TV Tuner app, and antivirus).
I'll list my specs below and the troubleshooting I've done.


Case: http://www.newegg.com/Product/Product.aspx?Item=N82E168... - Cooler Master HAF 912
MOBO: http://www.newegg.com/Product/Product.aspx?Item=N82E168... - ASRock Z68
CPU: http://www.newegg.com/Product/Product.aspx?Item=N82E168... - Intel Core i5 2500k
SSD (OS is installed here): http://www.newegg.com/Product/Product.aspx?Item=N82E168... - Crucial M4 64GB
Cooler: http://www.newegg.com/Product/Product.aspx?Item=N82E168... - EVGA M020-00-000234 Superclock CPU Cooler
Memory: http://www.newegg.com/Product/Product.aspx?Item=N82E168... - G.Skill 2x4GB sniper series
HDD (for data): http://www.newegg.com/Product/Product.aspx?Item=N82E168... - Seagate Green 2TB
HDD (used, for data): http://www.newegg.com/Product/Product.aspx?Item=N82E168... - Western Digital Caviar Black 2TB
TV Tuner: http://www.newegg.com/Product/Product.aspx?Item=N82E168... - Hauppage WinTV HVR1850


What I've tried:
Checked event log, only thing it shows is "The previous system shutdown was unexpected."... nothing more than that, unfortunately.

Voltage: Everything looks normal. i'm not at the computer now (i actually wrote up this entire post last night and 2 seconds before I hit submit).... +12V shows about 12.098V, +5V shows about 5.097V, etc.

No Overlocking has been done to RAM or CPU

Heat on CPU is steady between 36-38C, MB is around 32-33C

I've run the computer with only Windows services running and it still reboots on it's own. It doesn't always do it at a set time, never 'just 5 minutes after login', it's either 5 minutes or could be 45. Could be when i'm using it heavily to when I'm not even touching it.


I'm running short on ideas, so I'm turning to the experts. If anyone can give me some help, it'd be awesome.

More about : computer restarts

a b B Homebuilt system
March 7, 2012 3:55:49 PM

soounds like either a power supply issue or a major driver issue.
March 7, 2012 3:58:49 PM

Does it do it in Safe Mode? (you elluded to that, but not certain that that is what you meant).

Test your ram.
Related resources
March 7, 2012 4:00:33 PM

Check your BIOS; sometimes the default RAM timing (the x-x-x-x and so on number) and voltage are not ideal for your RAM. I would say you have to much overclock/not enough volts, but no OC... so start with RAM timing/voltage.

My POS motherboard will let me manually set the RAM timing, but always reverts to a faster speed than my original RAM was rated for, so unless I entered the BIOS EVERY boot, then I would have a restart issue like you are having. The mobo manufacturer was no help, so I just got faster RAM. Crappy solution, but it is faster, so...
March 7, 2012 4:24:23 PM

Rockdpm - I have used the power supply before with another PC, so I know the PSU is good, but I'm not sure if maybe it working with this particular MOBO is good (however, the voltages look fine).

How could I check drivers for this? I've gone thru to each manufacturer site and updated drivers for each piece of hardware.


@danny2000 I'll try safe mode. All I tried was 'msconfig' and checking "Diagnostic Startup" to only load windows services. But I'll try safe mode specifically.

@jfby - I'll check that and post back up tonight.

If anyone else has suggestions, let me know. Thanks!
March 8, 2012 11:54:47 PM

danny2000 said:
Does it do it in Safe Mode? (you elluded to that, but not certain that that is what you meant).

Test your ram.


Computer ran in BIOS/UEFI config just fine.
Ran in Safe Mode all day, no issues.
Started up in diagnostic startup (under 'msconfig' and selected 'Diagnostic Startup' loads only core services and programs)... 6 minutes and it reboots.

Since it doesn't work in diagnostic startup (not even network services running) but it does in Safe mode... I'm kinda stumped?

In reply to jfby:
Ram timing is running at 9-10-9-28

a c 78 B Homebuilt system
March 8, 2012 11:58:37 PM

Download and run a program called MemTest86+ when you are sleeping tonight and see if it reports any errors.
March 9, 2012 1:15:14 AM

Raiddinn said:
Download and run a program called MemTest86+ when you are sleeping tonight and see if it reports any errors.


Pass 1 went thru fine, 0 errors...

It's still running (I can see the top portion of the screen, numbers are still running and the \|/-- is still spinning, but does this mean it failed (or tell you anything)


March 9, 2012 1:40:44 AM

Based on that screenshot it is almost certain that your memory is defective (even if memtest does not report it as such.)

Unless I am mistaken (which happens sometimes ;)  ) you are using the integrated CPU graphics. The corrupted display is an indication that the system memory that has been reserved as video buffer isn't working as it should.

Since that part of the memory has been assigned to the video buffer, I believe memtest will not test it.

The reason it worked in safe mode is because for graphics a different area of the frame buffer is in use. The reason it doesn't work in normal mode is because more of the frame buffer is used in normal mode than in safe mode. When that "more" part is used the system crashes.

See if you can boot the machine with only one stick of memory. Try each stick at a time. Unless they are both bad (unlikely but possible), the machine should work fine when the good stick is in use.

HTH
March 9, 2012 3:09:21 AM

440bx said:


See if you can boot the machine with only one stick of memory. Try each stick at a time. Unless they are both bad (unlikely but possible), the machine should work fine when the good stick is in use.

HTH


Tried one stick at a time.... failed both times (two sticks)... does that mean bad mobo? or incompatible memory with mobo? (I posted specs in opening post)
March 9, 2012 3:33:00 AM

The GSkill sniper is compatible with the Z68. Therefore that part is ok.

If you have a friend who could test your memory in his/her system, I'd do that. At this point, the candidates are memory or motherboard.

It's hard to believe you get two bad sticks of memory together, though possible, and for that reason I'd have them tested in another system (if that is possible for you.)

If the memory is good then the motherboard is almost certainly at fault.

Just to be thorough and leave no loose ends, I'd like you to repeat the test you just performed (one stick at a time) but this time, remove the TV Tuner from the system before you perform the tests.

a c 78 B Homebuilt system
March 9, 2012 2:43:20 PM

Sounds like motherboard or processor to me, after looking at that Memtest result.

Every time I have ever seen bad RAM and good motherboard and good processor it did not ever present results like that.

I know it is a strech, but if you know anyone with an 1155 processor you could borrow, or a board you can put your 1155 processor in, that would be the logical next step.
March 9, 2012 3:22:06 PM

Raiddinn said:
Sounds like motherboard or processor to me, after looking at that Memtest result.

Every time I have ever seen bad RAM and good motherboard and good processor it did not ever present results like that.

I know it is a strech, but if you know anyone with an 1155 processor you could borrow, or a board you can put your 1155 processor in, that would be the logical next step.


There is a chance that the problem is the processor. On the other hand, defective intel processors are so rare (I've never had one) that I concentrate testing on the "other" components (in this case the memory and the motherboard, which are easier to test as well). As you pointed out, testing a processor is cumbersome.

The results he got (corrupted screen) are not very unusual. I have seen that a number of times in video cards with bad memory. In this case the "video card" is using system memory reserved for the frame buffer. Therefore the memory is highly suspect at the point. The other possibility is a problem in the motherboard's memory slots, that could prevent good memory from working properly. Hopefully, it is one of those two and not the processor.




a c 78 B Homebuilt system
March 9, 2012 7:17:16 PM

I looked up corruption during memtest86+ in google and another poster said they got the problem to go away by downclocking their processor speed by half.
March 9, 2012 9:57:48 PM

Just tried two brand new sticks of memory (PC3-10600, a little slower, same voltage and CAS latency, though). System ran fine for about 20+ minutes, so I decided to run HeavyLoad (a stress test) to see how it held up.

The stress test ran for about 30 minutes, putting my CPU at 100% usage the entire time as it allocated more and more data to RAM. At about 90MB of available memory, I figured it was pretty safe to say that the memory was good. I clicked 'Pause' to stop the test and my machine immediately crashed.

I'm going to try to stress test again with my previous/original memory here in a couple of hours, but I have to run out for a while.
March 10, 2012 2:28:19 PM

If you can, I suggest you test all the memory sticks in a different system (a known one to be working ok.) Hopefully you know someone that has a Z68 motherboard based system that can do that for you.

A P67 based motherboard should work as well but, it would be best to find a motherboard based on the same chipset as yours.

a c 78 B Homebuilt system
March 10, 2012 9:13:52 PM

Any motherboard will work to test RAM. Even if it is an AMD motherboard.

It helps if the RAM is on both QVLs, but most of the time if RAM works in one computer it works in a different one as well with the same settings.
March 11, 2012 3:04:41 AM

Raiddinn said:
Any motherboard will work to test RAM. Even if it is an AMD motherboard.

It helps if the RAM is on both QVLs, but most of the time if RAM works in one computer it works in a different one as well with the same settings.


True. As long as the memory is compatible in both.

Whenever possible, I try to test on the same type of hardware but, as you pointed out, it is not required.

March 11, 2012 2:48:47 PM

I haven't had a chance to try the original (G.Skill) memory in another PC. But using the new memory, my PC has rebooted on it's own several times.

I've uninstalled audio and display drivers and it still reboots (with new memory).

I've now installed Windows provided display and audio drivers and have uninstalled the Hauppage WinTV HVR application/service.

If this doesn't work, I may just reformat the drive and reinstall Windows without anything added on, and see if it reboots then... then that will narrow it down to software or hardware.
a c 78 B Homebuilt system
March 11, 2012 5:16:32 PM

If you can't get through MemTest86+ then your problem 90% likely has nothing to do with drivers.

If you tried different RAM and the same thing happens in MemTest86+, then its 90% likely not a RAM problem.

I am inclined to think processor or motherboard, but someone else mentioned the graphics card as a possibility.

If you can't test those things by replacing them with something else, then I would consider RMAing the processor as the next step.
March 11, 2012 5:38:50 PM

floydtheduck said:
I haven't had a chance to try the original (G.Skill) memory in another PC. But using the new memory, my PC has rebooted on it's own several times.

I've uninstalled audio and display drivers and it still reboots (with new memory).

I've now installed Windows provided display and audio drivers and have uninstalled the Hauppage WinTV HVR application/service.

If this doesn't work, I may just reformat the drive and reinstall Windows without anything added on, and see if it reboots then... then that will narrow it down to software or hardware.


I suggest you physically remove the Hauppage card from your system for all the testing you are doing. I suggest this because the Hauppage has the potential to cause video problems and _may_ have been the cause of the garbage you saw on the memtest screen.

The memtest failed, that means Windows is not part of the problem. The one thing that is certain at this point is that the problem is caused by hardware. The motherboard, the memory and the processor are the prime list of culprits.

You've tested with different memory and the problem still occurs. This almost rules out memory as being the cause.

The candidates left are the motherboard and the processor.

Given the very rare occurrence of bad processors, I am much more inclined to suspect the motherboard as being defective.

Do have access to another motherboard you could use for testing (e.g friend's system)? (if you did that would be extremely helpful.)

HTH
March 12, 2012 5:23:07 PM

I have another PC at work I can use for testing my G.Skill RAM in. I'll try that tomorrow.

I went ahead and reinstalled Windows (7 Pro) as I hadn't done much with it yet due to the issues.

Without loading any drivers or adding any hardware, the computer has been running for 2 days now. I'm going to try loading Windows drivers, next. Once I get drivers loaded I'll see how it runs, and take things one step at a time from there.

This still doesn't explain the strange occurence with the Memtest86, though... I thought perhaps I had a bad ISO download/burn or maybe a scratched disc? Not sure, but regardless it is interesting the computer has been running fine after a reinstall without any added drivers.


I'm going to add the drivers tonight and put it to some use. I'll let you all know how it goes.
March 12, 2012 6:07:13 PM

floydtheduck said:
I have another PC at work I can use for testing my G.Skill RAM in. I'll try that tomorrow.

I went ahead and reinstalled Windows (7 Pro) as I hadn't done much with it yet due to the issues.

Without loading any drivers or adding any hardware, the computer has been running for 2 days now. I'm going to try loading Windows drivers, next. Once I get drivers loaded I'll see how it runs, and take things one step at a time from there.

This still doesn't explain the strange occurence with the Memtest86, though... I thought perhaps I had a bad ISO download/burn or maybe a scratched disc? Not sure, but regardless it is interesting the computer has been running fine after a reinstall without any added drivers.


I'm going to add the drivers tonight and put it to some use. I'll let you all know how it goes.


I concur. The fact that reinstalling Windows solved (or seems to have solved) the problem would indicate a driver/software problem of some kind. As you stated, that leaves what happened with memtest a bit of a mystery.

Your plan of action is sound, we'll see what happens as you perform additional testing and add drivers to your new Windows installation.
a c 78 B Homebuilt system
March 12, 2012 7:42:42 PM

Run MemTest86+ again, see if it does the same thing now that it did before the reinstall.

Also, wait a long time on the drivers. You can do windows updates, indeed you should do all of them, but wait on the updated drivers for hardware devices for at least a week.
March 13, 2012 3:54:08 AM

Well... I think I've found the culprit. I used the CD to install the LAN driver (manually thru device manager browsing to the CD) so that I could get Windows Updates (this is like a bad movie where you can already see how it will end)... I finish installing the driver 'Successful'... and sure enough, I click Start, point to Windows Update to finish the updates, Beeewwwwwwwwwww (that's the machine shutting down sound).

So, I downloaded a fresh copy (and different version) of the LAN driver off of the manufacturers website. So far, it's been running for 30+ minutes just fine. I'm not running any more updates just yet, going to let it sit for 24 hours and then take it from there.
March 13, 2012 11:43:04 AM

Well, even with a fresh copy of the LAN driver from the manufacturers site, it still crashes. It crashed as soon as Windows Update started downloading. After the crash, it went back to the downloading screen (it had almost finished) and it started installing. It installed all the updates without crashing...

Is it possible that with the LAN connection that any sort of network traffic could cause it to crash? That just seems weird to me...
a c 78 B Homebuilt system
March 13, 2012 4:13:58 PM

It is certainly possible that LAN traffic can cause your computer to crash, especially if you don't have a firewall active.

In fact, it is pretty common. "Script Kiddie" hackers often just set their computers to do port scans of huge swaths of the internet to see what they can hack into. Many of the programs that scan to see if a computer is vulnerable to a certain security risk work by actually trying to use the exploit and see if it works on the computer or not. Many of those cause a computer to crash, and some of them allow lines of code to run on the remote system if a crash has occurred.

Not to rile you up or anything, because it could be something different, but it is highly suggested that the first thing anyone ever installs is a firewall and an anti-virus program. Windows comes standard with firewalls now, but older OSs didn't.

Even the firewalls available now aren't necessarily 100% up to date on the initial install. It is only as new as the time it was programmed. Some have to have their default rules changed at a later time when it is discovered the default rules don't account for an attack discovered later.

Anyway, if you don't have a firewall and antivirus active yet, I would definitely get them asap just to shield yourself against this sort of eventuality.

Microsoft Security Essentials is available for free from Microsoft's website and it is a reasonable choice and both free and unintrusive. They can't bundle it with the OS, because Symantec and other anti-virus makers have said they will sue them if they do so you have to go to their website to get it, but you should have the firewall installed automatically. That would keep you safe long enough to get MSE.

Anyway, the security risk stuff out of the way, it is entirely possible other stuff crashes you over the wire too. An old file server type of device could do such a thing, or a different computer on the same network sending out bad packets or it could be a problem in a download.

Internet speeds are in the megabits per second range which means millions of 1s and 0s every second, and if one of those is received incorrectly it could cause a crash potentially. Usually networks are pretty resilient against this sort of thing, but not 100%.

It could just be a bad motherboard port too or a bad cable. Or it could just look like it had to do with the network and it really didn't.

There is no way to tell at this point what the cause is, that I can see anyway.
March 13, 2012 4:41:42 PM

Raiddinn said:
... it could just look like it had to do with the network and it really didn't.

There is no way to tell at this point what the cause is, that I can see anyway.


Well that stinks...
As far as security, I'm actually majoring in Information Systems Security, so I've had all kinds of fun times setting up a secure network, firewall, etc. I don't think it's that.

As for a cable, I'm using the same one as I did previously.

I do have a NAS device and a few other computers on the network, no issues with others though, but I do understand what you mean about the traffic being misinterperated by the CPU and causing it to crash... that just seems odd knowing the reliability of technology nowadays.


Is there anything else I can try? If this is hardware related I'd really like to know soon so I can still return whatever is defective and get 85% of my money back (NewEgg restocking fee) within the 30-day window.
a c 78 B Homebuilt system
March 13, 2012 4:50:24 PM

Do you have a network card laying around? You could shove it in the PC and use that instead and see if the crash happens.

If you did and it didn't, it could be that there is something wrong with the motherboard network port.

OR it could just be a random one time thing and it never happens again regardless of what you do.

Sometimes it is hard to root out an intermittent problem in a short time frame like you want to do. It is just something that has to be accepted sometimes.

All we really know, from where I sit, is that the crash happened during the update process, that doesn't really guarantee it is even related to it.

You could try checking the error logs. They should be viewable through the control panel, probably under admin tools. This is not something fun to do for a lot of people, though. Trying to get some kind of meaning out of error logs is often times pointless and its always painful.

Something free and cheap is downloading MemTest86+ and running it while you are sleeping. That should at least be able to tell you if it is a RAM problem.

You could try Prime 95 after that if the RAM comes out clean.

Both of those don't do network stuff, but they should be able to show you if it is related to a non-network component at least.
March 13, 2012 5:23:07 PM

Ok, I'll try the PCI LAN card and see where I get, I think I have a few laying around at work.

I've checked the Event Viewer, all it shows is miscellaneous information messages, and then 15 minutes of no messages, the next one shows "the previous system shutdown was unexpected."

I've done troubleshooting with computer rebooting before, usually it shows "Application xxx has caused the system to become unstable." or something to that effect... but this time I get nothing.

Before I left for work today I left the computer on and connected to the Internet, we'll see if it's crashed or not when I get home.

Overall, my theory so far is the Mobo, since the RAM was a little funky on my last Memtest86 run, and now I'm having LAN (allegedly) issues.
March 13, 2012 5:35:54 PM

I'm also inclined to believe that the motherboard is in some way (maybe more than one) the cause of the problems.

Trying a different NIC is a good idea, if it works it would point to a bad port on the MB.

Given that there is/was a memory problem and now there seems to be a problem with the MB NIC, I would request a replacement from Newegg. If you bought it less than 30 days ago and, all you ask is for a replacement of the same board, you have very good chances that they won't asses you a restocking fee.

HTH
March 18, 2012 8:36:31 PM

440bx said:
I'm also inclined to believe that the motherboard is in some way (maybe more than one) the cause of the problems.

Trying a different NIC is a good idea, if it works it would point to a bad port on the MB.

Given that there is/was a memory problem and now there seems to be a problem with the MB NIC, I would request a replacement from Newegg. If you bought it less than 30 days ago and, all you ask is for a replacement of the same board, you have very good chances that they won't asses you a restocking fee.

HTH



I've RMA'd the motherboard. Once I get a replacement I'll try it out and post up and let you know.

Thanks for all the help guys. I'll keep you posted.
April 11, 2012 6:31:52 PM

Well, I got the RMA for the MOBO... and the computer still crashes :(  I tried Linux and Windows, crashes both. I've checked the outlets on the wall and those are fine (and like I said, it will run on the BIOS for days, it's only booting into an OS)....

Would it be safe to say the CPU is defective?
April 29, 2012 7:05:58 PM

I replaced the CPU and the problem still resides. It seems the computer will only run while a video card is in one of the PCI-E slots. I don't have to use the video card (i'm currently running off of my mobo video), it just has to be physically in the slot (not even have drivers installed).

Any explanation for that?
May 10, 2012 12:05:06 AM

I keep getting reminded to select a best answer, but I've had the last 3 (now 4) posts in the thread for the last couple of months... not sure what to do.
!