Computer randomly shutting off

[vigvoodoo]

Hate my computer right now. My computer is randomly shutting off. I was getting a kernel power critical error in event logger and random BSOD on boot. But after a format and OS re-install (win7 pro 64) I'm getting event id 6008, eventlog error. Also amdkmdag is spamming event id 6145 errors. What gives? Thanks in advanced!

 

[Onus]

Is this a Hackintosh, store-bought machine, one you built, ????
Full specs will be needed in order to get a meaningful answer.
First guess, and it is only a guess, is that your PSU is crappy and/or insufficient. What brand/model (not just wattage) is it?

 

[vigvoodoo]

Is this a Hackintosh, store-bought machine, one you built, ????
Full specs will be needed in order to get a meaningful answer.
First guess, and it is only a guess, is that your PSU is crappy and/or insufficient. What brand/model (not just wattage) is it?

System Specs:

Operating System - Microsoft Windows 7 Professional 64-Bit
Processor - AMD Phenom II X3 720 Black Edition (was overclocked to 3.2GHz, not at stock 2.8GHz)
Memory - 4 x 2GB G.Skill Ripjaws DDR3 1333
Motherboard - ECS A790GXM-AD3 AM3
Graphics - Dual XFX Radeon HD 4870 750M 1GB (CrossFireX enabled)
Storage - Intel 80GB X25-M G2 SSD (Main drive, OS, AHCI), Western Digital Caviar Black 640GB SATA HD (AHCI), Samsung Spinpoint F4 HD204UI 2TB SATA HD (AHCI)
Power - Thermaltake Toughpower W0106RU 700 Watt
Case - Cooler Master CM 690 II Advanced Mid Tower

 

[beanoslim]

Does it do it when set to defaults?

 

[Onus]

Take out one GPU and see if it happens. Then try with the other one by itself. If it doesn't happen with one, but does with two, your PSU has fallen down on the job. Do you have good airflow through your case, including through the PSU? If it is overheating, it will shut off.

 

[vigvoodoo]

Does it do it when set to defaults?

I went ahead and cleared the CMOS and left the settings alone, except the time. Running Prime95 and keeping an eye on temps with Core Temp. So far it's been about 40 minutes and still going. Max temp of 55 degrees Celsius.

Take out one GPU and see if it happens. Then try with the other one by itself. If it doesn't happen with one, but does with two, your PSU has fallen down on the job. Do you have good airflow through your case, including through the PSU? If it is overheating, it will shut off.

Airflow seems great. Minimal dust, all wires tucked out of the way. I will try running just one GPU if the computer fails again.

 

[dkant1n]

Be careful when re-configuring the BIOS with the memory settings (clock, timings, voltage).
It seems to me that there was the problem

 

[vigvoodoo]

Be careful when re-configuring the BIOS with the memory settings (clock, timings, voltage).
It seems to me that there was the problem

Well the computer powered off again I'm going to try just one video card.

 

[beanoslim]

If GPU doesn't solve it.

Unplug everything that isn't needed to boot.

Run memtest for 10 passes.

Manually set cpu voltage to default.

 

[vigvoodoo]

Take out one GPU and see if it happens. Then try with the other one by itself. If it doesn't happen with one, but does with two, your PSU has fallen down on the job. Do you have good airflow through your case, including through the PSU? If it is overheating, it will shut off.

Took out one of the GPUs, been running Prime95 for 4 hours now. No issues. So I guess it is the PSU. Right?

 

[hell_storm2004]

"amdkmdag" is something which is related to a failing GPU.

I think it would be best if you could provide the BSOD error codes. That would help us pin-pointing the issue.

 

[Onus]

Try the test with the OTHER GPU too. If that test is successful, then it is the PSU, which can handle just one, but not both of your GPUs. If that test fails, then that GPU is the likely issue.
As hell_storm2004 says, the error message is GPU-related, but whether it is due to a failure in a GPU or one or both GPUs being unhappy with insufficient power is what we're trying to find out with these tests.

 

[avatar_raq]

Interesting issue. Subscribed just to see how it'll be resolved!

 

[vigvoodoo]

Installed a new Corsair 650TX. Ran fine all day playing BF:BC2. Then just a few minutes ago after being on the internet a while I launched BF:BC2, 2 minutes later it powers off. First screen freezes, then the USB stuff powers off, then the PC. So I rebooted and it powered off while I was trying to type this on here. Get a 6008 critical error in event viewer, but right before that I have an information event 89 (Kernel-Power). Also getting a mess of AMDKMDAG event 62464 errors, over 1,000 in the last hour alone.

 

[beanoslim]

Have you manually set your cpu and memory voltages and run memtest for 10 passes then occt for 1hr?

Just one error under memtest can cause all sorts of problems, occt stresses mem and cpu.

 

[vigvoodoo]

Have you manually set your cpu and memory voltages and run memtest for 10 passes then occt for 1hr?

Just one error under memtest can cause all sorts of problems, occt stresses mem and cpu.

No. I will give it a try.

 

[vigvoodoo]

Now it's powering off more and more often. I can't even load anything intensive. Getting critical error 41 (kernel-power).

The computer does not BSOD, it just powers off. The same as unplugging the power.

 

[beanoslim]

Check your cpu temps under load and your shut down temp in the bios.

 

[vigvoodoo]

BIOS is not set to shut-off at a certain temp. The CPU idle is 26°C and 50°C load (Prime95). I'm running a Cooler Master 212 Plus, push/pull fan setup.

 

[beanoslim]

So you have a new psu, its not overheating and its not overclocked.

I'd set it up with a single card, run memtest, run occt, run 3dmark, if it passes them it shouldn't be shutting down.

If it passes try with the other card, then try both single cards in the other pci slot.

I think you've almost got to start again and work your way along testing as you go to see if you can pinpoint the problem, you may have a board error, just seems random...hence the title.

 

[vigvoodoo]

Okay guys I went ahead and removed both HD 4870's (using onboard HD 3200), all memory sticks except 1. The thing is still powering off randomly. Less often but still powering off.

Someone mentioned it might be the video cards so I installed one of my HD 4870s into another PC I just built. It powers off randomly now also.

The only thing I get in both systems is the Kernel Power 41 error in event viewer, but that only tells me that the system powered down unexpectedly.

So is there something wrong with both video cards, the motherboard, or both systems? I'm losing sleep over this please help.

 

[Onus]

Hmmm, I'm wondering if that Thermaltake PSU spiked on +12V and damaged your GPU(s) and/or mobo...
What does the other HD4870 do in the second PC?

 

[vigvoodoo]

How do I use it? It's empty at the moment.

 

[hell_storm2004]

There's this software:
http://www.resplendence.com/whocrashed

Use this. Its simple and use the instructions on how to run it.

 

[vigvoodoo]

I can't get this program to work.