Computer restarts without warning

[ldarkwarriorl]

GIGABYTE GA-990XA-UD3 AMD 9 Series Motherboard
AMD FD8150FRGUBOX FX-8150 Processor - Eight Core, 8MB L3 Cache, 8MB L2 Cache, 3.60GHz (4.20GHz Max Turbo), Socket AM3+ (Running at stock)
Crucial Ballistix BLS2KIT4G3D1609DS1S0 Desktop Memory Kit - 8GB (2x 4GB), PC3-12800, DDR3-1600MHz
Crucial CT064M4SSD2 m4 2.5" Solid State Drive - 64GB, SATA 6Gb/s
HDD WD 1.5TB SATA3 7200RPM 64MB Cache
EVGA GeForce GTX 460 SE 1GB
ViewSonic VX2250WM-LED 22-Inch
Logitech X-530 5.1 Speaker System
Lite-On IHAS124-04 Internal DVD Writer - DVD+R 24X, DVD-R 24X, DVD+RW 8X, DVD-RW 6X, DVD+R DL 8X, SATA
Midi Tower Green X-Cruiser
Diablotek DA Series 600-Watt ATX Power Supply PSDA600
12-in-One Internal Card Reader
Windows 7 Ultimate (x64)

I recently built this computer. But it keeps randomly restarting after Windows boots up. It restarts more and more frequently as I use it, but it never restarts before the windows login screen. I also notice it's likely to restart sooner if i move the mouse around a lot in the login screen.

It's not a heat issue because my processor never runs over 55C. There are no warning beeps either, and the fans and lights never stop or flicker. What's odd is that it works flawlessly in Safe Mode.

It's not the RAM either because Windows detects the full 8gigs. And i restarted with each stick independently and problem remains.

I tried swapping HDDS but the issue remains regardless of the harddrive inside. I also swapped my video card with a geforce 8400 GS thinking it might be a PSU problem, but that isn't the case either.

Windows xp and xp 64bit work without problems. But windows 7, server2008, and 8 consumer preview all reset randomly on me.

I've read a few topics using google of people with similar problems. They fix the situation by lowering or raising voltage in their RAM or CPU. Can anybody guide me through this? I'm afraid of burning something or shorting something out accidently since i've never overclocked before.

This guy has the same processor but he overclocked it, mine is using stock settings, http://www.overclockers.com/forums/showthread.php?t=702158

I never overclock but I think RAM or cpu might be over or underclocked by default.

Any help will be appreciated. Thanks.

 

[EarthDog]

55C is as warm as you want to get it... Is that just gaming and messing around or is that temperatures from a stress test like Prime95?

Just because it detects all the ram, doesnt mean the ram is not an issue. That said, test the memory with Memtest 86 and see what happens.

Diablotek PSU? Never heard of it. I hope its not a POS as that could be the problem.

 

[ldarkwarriorl]

55C means im just messing around, computer reboots, then i race to my bios to see what cpu temp is.

How can I run Memtest 86 if the PC doesnt stay powered on? Should I run it in safe mode?

And the PSU I got last summer when I upgraded my graphics card. Hasnt given me trouble on the old hardware but I cant say for sure that it's not the issue.

 

[EarthDog]

Whoa then, thats warm and may be part of the issue (as temps drop when rebooting).

Memtest86 runs from a boot disk so you are not in windows. It doesnt stress the CPU a lot so if its heat, it should run.

That said, it doesnt matter what it did before it could be the issue now. But first I would get those temperatures down. Use Coretemp to check your temps in windows and report back.

 

[Sentential]

Hard reboots like that are typically power related and not heat. Taking a quick glance at your power supply makes me incredibly suspicious that its not within the 5% of each voltage rail

 

[ldarkwarriorl]

Whoa then, thats warm and may be part of the issue (as temps drop when rebooting).

Memtest86 runs from a boot disk so you are not in windows. It doesnt stress the CPU a lot so if its heat, it should run.

That said, it doesnt matter what it did before it could be the issue now. But first I would get those temperatures down. Use Coretemp to check your temps in windows and report back.

Trying bto run coretemp in safe mode i get a message saying "Driver has failed to load. This program will not continue".

Going to run memtest now.

 

[Sentential]

Trying bto run coretemp in safe mode i get a message saying "Driver has failed to load. This program will not continue".

Going to run memtest now.

Don't bother with memtest, not yet; earthdog would agree with me that the PSU is most likely your culprit, he probably overlooked it initially. Go look at your voltage rails in BIOS; they must be within 5% of rated.

 

[ldarkwarriorl]

So is it too few watts? Or bad brand? Or was once good and now is old and weak? How can we be sure its theb psu? I really hope your right, but i dont want to buy a new psu and find out something else is the problem.

 

[Sentential]

So is it too few watts? Or bad brand? Or was once good and now is old and weak? How can we be sure its theb psu? I really hope your right, but i dont want to buy a new psu and find out something else is the problem.

Both, go look at your PSU rails in BIOS, specifically the 12v+; it needs to be within 5% of 12, ie over 11.5v on load. If it exceeds that it will litererly pop the PSU breaker and cut the PC off.

 

[RollingThunder]

So is it too few watts? Or bad brand? Or was once good and now is old and weak? How can we be sure its theb psu? I really hope your right, but i dont want to buy a new psu and find out something else is the problem.

Check it this way:

http://www.overclockers.com/forums/showthread.php?t=403837

Sounds similar to a power supply with bad or leaking capacitors but using a digital multi-meter is more accurate, it's quick and an easy way to eliminate or prove adequate / inadequate voltage.

 

[EarthDog]

Both, go look at your PSU rails in BIOS, specifically the 12v+; it needs to be within 5% of 12, ie over 11.5v on load. If it exceeds that it will litererly pop the PSU breaker and cut the PC off.

The problem with this is that software voltage readings are notoriously off. So even if it is reading out of spec, I wouldnt trust it unless you use a MM of some sort (See RT's post above me).

Also, plenty of wattage for your system (assuming it puts out what its rated for).. it shows 40A (Im guessing MAX) on the 12v rail. That said, its terribly inefficient coming in at 78%. Ouch. That woeful efficiency and the fact I dont know who the OEM is makes me think that PSU is a POS. It may/may not be the problem. I would look in to replacing it ASAP.

Coretemp cannot run in safe mode, you need to boot to windows..

 

[ldarkwarriorl]

Am i supposed to use Prime95 forf this test? If so, can someone please link me? Having trouble finding it.

 

[EarthDog]

P95 for what test? To look at voltages? No. Its in your bios (again, take it with a grain of salt).

 

[ldarkwarriorl]

Ok then according to my bios, 3.3v= 3.360v. 12v= 12.048v

 

[EarthDog]

Assuming its accurate (heh) thats a good reading. Try booting to windows and running coretemp and see what idle and load (THIS is where you need to use P95, Torture test, Small FFT). After that, try creating a bootable iso of memtest86 either usb/disk and run it to rule out the memory.

I wouldnt raise voltages to combat this problem yet, seeing where your temperatures are just messing around.

 

[ldarkwarriorl]

installed tempcore in safe mode. got a quick snapshot...

 

[EarthDog]

Ouch.. wow.

See if you can run memtest86, and report back. Make sure your memory speed, timings, and voltage are set accurately in the bios.

It may be the PSU, but its going to be impossible to tell without another one to test it with. This is a solid choice (Corsair CX500 V2)

 

[ldarkwarriorl]

I have a spare psu, but its a 420watt and incompatible with my harddrives. No sata power.

 

[EarthDog]

If memtest works, you need to somehow lower the temps on that thing, if you do both those things, I would look at the PSU being the culprit.

Temps are OK at idle... but idle doesnt matter. See if you can run P95 in safe mode.

 

[trents]

Can you borrow another PSU temporarily to check out that possibility. That PSU you got is a bottom feeder as far as quality goes.