PC randomly restarting help

Hi,

Can anyone give me some advice as what could be causing my PC to randomly restart?

PC Specs:

Gigabyte GA EX58-UD4 Rev 1.0
Intel i7 920
Corsair PC12800 Ram - 6GB Triple Channel
Nvidia GTX 680
Samsung 840 Evo Pro 128GB
Seagate Momentus 750GB
PSU - Can't remember, I know its a 800w modular PSU. It's fairly recent (18-24 months old) as the last one blew. It's a big brand though.
Case - Zalman Z11

The problem I'm having is, the PC restarts at random intervals. Whether it's just browsing the web, watching videos online or playing games. It isn't regular and doesn't even happen every time I use the PC, but it does happen often enough to be a problem, especially recently as my Fiancée has been studying for an Accountancy Exam and has lost some work.

There are no error messages, no blue screen and no message when booting to say the usual "the PC wasn't shut down correctly...start windows normally", no message once Windows has loaded to say it's recovered from an error.

It's only started happening over the last couple of months, at least 6 or 7 months after I installed my last upgrade (SSD). I can't figure out what the issue is.

I've never come across this issue before, usually when a computer is misbehaving there's usually an error message or blue screen or something to acknowledge that there is a problem.

The components are pretty new with exception of the Mobo, CPU and Ram which I bought 5 or 6 years ago now.

Has anyone come across this problem and found an issue?

Could it be that my Mobo, CPU and or Ram is on it's last legs?

Thanks in advance.

plavonica,

Thanks for the reply.

I have a program called "Open Hardware Monitor" installed and I've been keeping an eye on it after the before mentioned PSU blow out as overheating was the issue (poor fan positioning on my part).

The CPU temp doesn't go above 55 degrees C even under allot of stress. GPU temps are well within tolerances. I really can't see it being an overheating issue and when I installed the SSD I cleaned out the case, fans etc while I had it open.

Also it reboots when watching a video online or just general browsing, then I'll play a game (Sniper Elite 2 or Bioshock Infinite for example) and it will be completely fine with no reboots which would suggest that the cooling is fine.

Overheating is a good suggestion, just don't think it's the issue here.

Edit: I also have a different cpu cooler than the default one it comes with. Can't remember what it's called but it was recommended by Custom PC Magazine (UK) at the time I bought the parts. It's a tall heat sink with a 12 or 15 cm fan that barely fits inside the case. CPU idles at 30-32 degrees Celsius.

Hmm, I wonder if you've got some kind of malware?

One thing to try would be to narrow down whether it's hardware or software. If you've got a reasonable-sized USB key handy, you could try running a LiveCD version of Ubuntu for a few hours, just to see what happens.

It'll be slowish, running from USB, but it beats the heck out of actually having it on a CD.

Good shout. I am planning on doing a clean install of Windows 8 this week to narrow it down. I've been unable to do it as the missus has been hogging it and has her files all over the place.

It's a bit strange malware just to randomly reboot though but who knows with these things.

I've just basically reset the bios settings as I had the cpu boost set to Extreme rather than Standard or Turbo, so set it to Standard to see if that makes any difference. But it's been set to Extreme for at least 2 years now.

The multiplier was set to x21 which produces 2.8ghz so I've put it to x20 which shows the stock 2.66ghz. Not sure why it was on x21 but again not looked at the bios for years.

Changing the multiplier didn't seem to make any difference as when it boots and shows the bios cycle it still shows the multiplier as x21. Not sure what's happening there as when I went back into the bios after seeing that the setting itself shows x20!

Crazy mobo...

My mother had a similar problem and it ended up being one of the sockets on her power strip. It was flaky enough to cause restarts. Plugging it in elsewhere resolved it. It might also be worth tracing the cable going from the power/reset buttons to the headers on the motherboard. Perhaps they've worked themselves loose or there's some related problem with them or the headers.

The PSU may have also gotten cantankerous over time.

What do you mean by power strip and plugging it in else where? Power Strip to me means the 24pin port that you plug the psu into, but obviously you can't plug it in somewhere else.

I'll take a look at the power button cable and see what it's like. The pc doesn't behave like the power or reset button has been pressed, as if the power button is pressed it would shut down correctly and if I pressed the reset button I get an error when the pc restarts that says "Windows wasn't shut down correctly"

Haha. Fair enough. Worth changing it to a different plug I guess. Got a couple spare.

plavonica wrote:

This is what LZ means by a power strip.

IMAGE(http://i.imgur.com/BdLmVuK.jpg)

Ooh! That method must save a lot on their electricity bill!

Windows stopped giving the BSOD (Blue Screen of Death) because of all the negativity surrounding it. Most people didn't read it or know how to interpret it anyhow.

This does not mean that Windows does not record the crash and the reason why...

Start Menu > Control Panel > System and Security > Administrative Tools > View Event Logs

Click on Windows Logs on the Left > System

On the right, "Filter Current Log", Event Level Critical

Look at the date and time, find where there was a reboot around the issue and google the bugcheck. You should get a Microsoft Community page with details on what it is. If the codes are all 0x0... Then try running the system in safe mode with a benchmarking (prime95) or something. Check it in an hour and see if it's an issue. I would call it likely a driver conflict.

Good rule of thumb is to update drivers from the Gigabyte and Nvidia websites directly. Flashing BIOS is extreme and not recommended unless it's your last option. Even then, make sure you back up your old BIOS file first. Got to also use FAT32 formatted thumb drives for extra precaution. With the Gigabyte boards I've installed, it has been video related.

There are also what's called "dump files (.dmp)" that are created when a system recovers from a crash.
http://support.microsoft.com/kb/315263

Goldendog,

That's brilliant info. No idea where the Windows 8 logs were kept (why is everything so hidden in Win 8?)

Wish I knew that before. I just reinstalled Windows. Gone back to Windows 7 though as I found my disc and CD Key by chance looking for the Windows 8 stuff.

After using 8 I think 7 is a better experience with a mouse and keyboard. I love it on my laptop though.

I'll see if the random restarts are still happening, but so far after around 6 hours use on Windows 7 it hasn't restarted itself once (although I am installing updates etc and restarting manually)

Thanks everyone for the help. I appreciate it.

***I typed this all out before I saw what GoldenDog had posted but he is right so I will just reiterate.***
If it does start again make sure to take a look through your "Event Viewer" and see if you can see what exactly is happening when the computer shuts down. Also, take a look at "Task Scheduler" make sure a piece of malware hasn't thrown a rouge task in thats mucking the system up. (the later should have been taken care of with a re-install though :))

Other things that can affect a reboot are a failing power supply, although I have only seen it once. Normally they go out completely. Power from the wall receptacle (not just the surge protector) can cause this too but you would notice everything loosing power and from what you said thats probably not the case.

How goes things so far? Its been a bit.

Yeah sorry for the lack of update.

As my misses didn't like Windows 8 that much and I don't think it's that good for a Desktop PC, I switched back to Windows 7. It's been a couple of weeks, the pc has been used at least once a day and there's been no random restarts which leads me to believe that it was a software issue. Could have been malware as has been suggested.

I'd have liked to have known what the issue was but am just happy that it's sorted and the hardware is ok.

Thanks for all the help everyone.

I'm having a problem with my computer resetting when playing Borderlands 2. Sometimes I can get 10 minutes in sometimes only 2 minutes. Here's the kicker: I have 85 hours rakked up (booo) in the game and now it's just starting to do it. 85 hours of gameplay with the graphics settings set to max for that matter.

Playing along then black screen. Sometimes the fans remain running and it'll reset itself, sometimes it's a full shutdown (fans stop) and I'll have to hit the power button to reset.

Upon restarting, Windows prompts with the "windows encountered an error, do you wish to search for a solution" blah blah. Just for kicks, I set the graphic settings to low (performance) via Nvidia's GeForce Experience but it still does it.

It started back in July so I thought MAYBE it was the temps. My desk is in the basement, which is generally cooler. Here's what I'm running for temps (Celsius):

CPU temps (as per Core Temp)
Idle – low 30s
Max during Borderlands 2 – high 60s
During FurMark Burn-in test (1920×1080 fullscreen, 2X MSAA, @ 10000 ms) - mid 60s

GPU temps (as per MSI's Afterburner)
Idle – mid 30s
During Borderlands 2 – low 60s
During FurMark Burn-in test (1920×1080 fullscreen, 2X MSAA, @ 10000 ms) - 80 firm w/ 48% fan speed

--
Intel Core i7-3770 3.4GHz Quad-Core
ASRock Z77 Extreme3 ATX LGA1155
Patriot Viper 3 16GB (2 x 8GB) DDR3-1600
Seagate Barracuda 1TB 3.5" 7200RPM
MSI GeForce GTX 770 2GB TWIN FROZR
Rosewill 630W ATX12V / EPS12V
Microsoft Windows 7 Home Premium Full
Asus VE248H 24.0"

I've overclocked nothing.

I build the computer in the spring of 2013 and have always kept my video/sound, etc. drivers up-to-date. After reading the reviews for the PSU recently, I found that it crapped out on users about a year after using it. They also recommend not to use it for intensive gaming machines. I'm leery about it. Is the computer resetting a symptom of a dying PSU?

Perhaps I should try a different game that's as graphically intensive to perhaps prove it's not just BL2. TF2 with max settings runs like a dream with no issues, but that's no BL2 graphics-wise.

Any other suggestions/thoughts?
Thanks!

Well, my first thought was 'failing power supply', because the machine shutting off suddenly under load is the single most common symptom of that, and then I saw this:

Rosewill 630W ATX12V / EPS12V

... which pretty much clinched it for me. Rosewill isn't great. It's not terrible, but their supplies are so-so at best. I think your hunch is probably correct.

I don't think it would still be under warranty, but even if it is, I'd suggest buying something else. Corsair's CX series are cheap but reasonably good, and their TX and HX series are built well. Seasonic supplies are expensive, but built to super-premium specs.

You don't need a ton of supply for what you're doing; with no overclocking, a 430-watter would be enough. If you want room for overclocking, 500 to 550 would be better. If you ever plan to go SLI, you'd want to go up to about 750 watts.

Perfect, gentlemen. Thanks for the advice. Gumbie, trigger pulled! I'll keep you posted over what transpires.

brouhaha wrote:

Perfect, gentlemen. Thanks for the advice. Gumbie, trigger pulled! I'll keep you posted over what transpires.

After finding out I had a bustbuy gift card from way-back-when and finding they had what I needed at the local store, I canceled that amazon order. Got a CX 750, threw it in there and played BL2 for a good half hour.

Thanks again gentlemen! All set for the Pre-Sequel!

Excellent! Always nice when the problem is straightforward and easy to fix.

I'm having random errors in windows. I've tried a new processor, video card, memory. I think it's the PSU but I get different readings from the bios and when running windows. The bios gives me 11.5-11.7 on the 12v rail. Speedfan is telling me it's at 9.x. Is 9 volts on the 12 volt rail even going to start up?

Well, once you get Windows running, the video circuitry is pulling more load than it does in the BIOS, and the CPU may be busier, too. I don't know if 9V is possible, but I tend to think the voltage must be below normal, particularly since the BIOS is telling you 11.5. The BIOS reading should be a nice stable 11.95 to 12.05 or so.

Do you have a physical voltmeter? That would be the next thing to try. Barring that, I'd suggest maybe not using the computer anymore until you replace the supply. It could be damaging other components with that low voltage.... and if it suddenly gives up the ghost, could do a hell of a lot of damage.

I swapped out the PSU and it is still the same. Reads about 11.6 in the bios and fanspeed says 9.5 again. I ran a chkdsk and it restarted during chkdsk but then I believe it finished. It's possible the motherboard is bad. I was able to run it last night for about an hour watching youtube. If I try to update steam, though, it dies almost instantly.

I've swapped out the following:
PSU
Video card
Ram (technically, I've not completely swapped the ram but put in a single stick and run a test on the ram)
Hard drive
CPU

I've run it with only the HD (no dvd, second hd, usb).

The only other thing I can think of is an issue with the motherboard.

Can anyone think of anything else I can check before getting a new motherboard? It's an i5 2400 which, at this point, is pretty old. My big issue is going to be getting MS to transfer the license to a new board.

I'm also unsure what would go bad on the motherboard as it has solid state capacitors (which is usually what goes bad on a mobo). Not sure what else it could be.

I think I tried different sata cables. I'll try those again. The bios 12v rail voltage is a bit low but not technically out of spec.

I went through and retested everything tonight. I got out the multimeter and tested the 12 volt and 5 volt rail and they seemed good (hard to tell exactly what they were at since my multimeter is analog and has settings for 10 or 50 volts). It looked like it was about 12. At 10 it pegged the needle and at 50 it looked to be at 12. Unplugged the drives except for the c: drive. Unplugged all the USB stuff I could (except mouse/kb). Took out 3 of the 4 ram sticks. I think some of the ram is bad. It ran after all that and I added back the other stuff and it seems to be good so far. My son only plays a few games and they are older (Gary's Mod, Dungeon Defenders, Terraria). I ran Gary's mod for about an hour and it was good. Haven't put any of the 3 sticks of ram back in, yet. Going to let him mess with it tomorrow and see if it stays up. If so I'll add another stick and see if that causes an issue. I've seen board in the past that work fine with ram in 2 slots but not 4.

Yeah, sounds a lot like either a bad stick of RAM or a bad slot, so I think you're on the right track. Now the trick is fiddling around to figure it out.

I'd memtest each stick in the known good slot to check if it's any of those. If those pass then try out the good sticks in different slots.