Win 7, Freezing / Crashing, Critical Kernel-Power Error

[jocasee]

---
Windows 7 64bit
ANTEC 300
ASUS M5A88-V EVO
AMD 1090t x6 Black Edition
CORSAIR Vengeance 16GB (4 x 4GB)
1x MSI GTX 560 TI Twin Frozr II
Thermaltake TR2 RX 850W
1x 64GB Intel SSD
1x 1TB Western Digital HDD
---

Hello All,

First post. Sorry for the length, just wanted to include details. Thanks in advance for any help.

Before I start, please note that I have updated my BIOS to the latest and reverted to defaults.

I built my computer late 2011, and it was working fantastic up until about October of 2012. It started acting strangely around that time.

It will either freeze or power-off completely. This happens on the desktop, when I'm browsing the internet, gaming. It doesn't matter what I'm doing, it will freeze or crash. I will leave my computer on overnight and it will be frozen on the desktop the next morning. Always, I get a Critical Kernel-Power error in my Windows Event Log.

At first I was in complete denial that it could be my hardware because, well I just didn't want to believe it. I really truly thought that it was just my OS that got corrupted or something. So I backed up, reformatted, and the same thing was happening. Either a freeze or a power-off crash. Always with Critical Kernel-Power error in the Windows Event Log.

So OS is not the problem. I then really truly thought that it must be the hard drive if anything. At the time, I was only using the 1TB Western Digital HDD. So I got a small SSD (64GB Intel SSD). I unplugged both SATA connections to the HDD, installed the OS again. Eventually, the system froze again while I was installing a game.

So the hard drives are not the problem. I then was certain that memory had to be the issue. I took out all memory except (1) DIMM, tested with memtest86+ for about 30 hours. No problems with that piece of memory. I booted my system to the OS and watched a little Youtube, and the system crashed. Not while watching a video, but while browsing for a video to watch.

Additional memory testing I have done since includes booting to a Linux OS and running 'memtester' and utilizing my full memory. I then looked for memory errors in the /var/log/messages and /var/log/mcelog files. None there. I also used a tool which I found online called OCCT and did some LINPACK testing, which also did not report any errors. Both LINPACK and memtester ran flawlessly with no crashes. So I feel that I can put memory aside now.

With the LINPACK testing, I also got a good idea that the CPU is not the issue. The LINPACK test utilized all 6 cores to 100%, and the system was very stable. Let me also note that through all of this testing, I was monitoring my temperature and my CPU did not pushed a max of 47 degrees C. I have an aftermarket Cooler Master heatsink & fan which keeps my CPU nice and frosty cool.

I thought it could be my power supply. I work at a place that builds and distributes super computers, so I took my system in to do some testing there. They had a power supply tester, and my power supply was showing all of the proper numbers in the proper places. I ran a GPU stress tester called Fur Mark and that OCCT program pretty much all throughout the day, and my system did not crash once in the 8 hours. I then thought, "Huh, maybe its the power at my house."

So I got a little voltage meter thing from work, took it home. I plugged my computer into it, and plugged it into the wall. It read 120V just as expected. My refrigerator in the garage turned on, and it read 117V. I then thought, if it is a power issue, I should be able to force it to crash by lowering the amount of power on the circuit. So I turned on a space heater which brought it down to 114V, and my vaccuum cleaner which brought it down to 110V, and the system would not crash. So my power at home I think is okay. Just in case though, I went out and bought a small UPS.

The system still crashes when plugged into the UPS Battery portion, so I know that power is not the issue. At least not the power from my house.

The last thing I wanted to believe it was was my GPU, because it is my most expensive component. But I tested it anyway. I removed my GPU, and used on-board video. The system crashed.

So I have tested the PSU, CPU, GPU, RAM, and my home power. Can I safely say that it is the motherboard? Or does anyone have any suggestions as to how I can 100% be certain of which component is failing?

I'm really struggling here guys. These or freezes crashes do not have a pattern. They don't happen when I'm playing a particular game or running a particular application. It happens whenever it wants.

Be my hero!
--
JocAseE

 

[MeldarthX]

Yep you've done everything I'd done; MBs can and do go bad sometimes; from a bios flash.....*which I did to my main board and killed the bios; had to order a new bios chip and back up and running*

Ok question - did this start after you flashed to the latest bios??

 

[Bubbaleone]

Try this; open Device Manager, select Show hidden devices in the View menu, and scroll down to Sound, video and game controllers. You should see AMD High Definition Audio Device, and Realtek High Definition Audio listed. Right-click on each of them and select Properties from the context menu. On the General tab of the Device Properties dialog is listed the Device type, Manufacturer, and Location. If both devices show the same location ie, "Location 0 (Internal High Definition Audio Bus)", disable whichever audio device you're not using, then test to see if the kernel power event stops occuring.

 

[MeldarthX]

Woo; nice call Bub - completely forgot about that fun irq conflict

 

[jocasee]

Yep you've done everything I'd done; MBs can and do go bad sometimes; from a bios flash.....*which I did to my main board and killed the bios; had to order a new bios chip and back up and running*

Ok question - did this start after you flashed to the latest bios??

The issue began happening before I flashed the BIOS. I flashed it because I saw several update logs on the BIOS which said 'Improves system stability'. Man I sure do wish that worked.

 

[MeldarthX]

Check what bub said; it very well could be irq conflict. I haven't seen those in a few years but possible....

 

[jocasee]

Try this; open Device Manager, select Show hidden devices in the View menu, and scroll down to Sound, video and game controllers. You should see AMD High Definition Audio Device, and Realtek High Definition Audio listed. Right-click on each of them and select Properties from the context menu. On the General tab of the Device Properties dialog is listed the Device type, Manufacturer, and Location. If both devices show the same location ie, "Location 0 (Internal High Definition Audio Bus)", disable whichever audio device you're not using, then test to see if the kernel power event stops occuring.

I didn't check specifically for the Location. But I did see that there are 2 sets of audio drivers in there. Realtek HD Audio and NVIDIA HD Audio. I disabled all the NVIDIA HD Audio devices and the issue still occured. I then disabled the Realtek HD Audio and enabled all of the NVIDIA HD Audio drivers. System crashed.

I found this article which has possible causes for the Critical Kernel-Power error:

http://www.sevenforums.com/bsod-help-support/104790-possible-causes-kernel-power-critical-errors.html

Tried them all with no success.

 

[jocasee]

I will check more specifically the location that the devices are trying to access when I get home. At work at the moment.

 

[Bubbaleone]

The bottom line is that kernel power events can have myriad reasons for occuring, and tracking down the exact cause can become very tedious. My bad for listing AMD instead of Nvidia, but the principle is still the same. I'm not following where you say; "...and enabled all of the NVIDIA HD Audio drivers". In the diagnostic procedure I outlined there will be a red X next to the disabled device following a reboot. Is that what you did?


.

 

[jocasee]

Sorry about that, I meant enabled all of the NVIDIA HD Audio devices. There are 4 of them in my Device Manager. I did see a conflict with (2) of the Location 0's. So I disabled the one that I'm not using. As far as the red X, that isn't there. Maybe Win7 is different, I just see a down arrow. Event after a reboot. It clearly says disabled though.

 

[jocasee]

I see that my PSU is a Thermaltake TR2 RX 850W. I thought this was a pretty bad ass PSU, but upon googling, it seems like this PSU has a history of problems. I'm still not sure that this is my problem though. Why would it work for days and days on end, and then crash? And then upon the next reboot it crashes within minutes.

Anyways, I will let you know the results of the Location 0 test.

Thanks for your help thus far.

 

[Bubbaleone]

Sorry, guess I've still got old XP stuck in my old brain...I'm running 7 64bit for gawd sake ! Yes, down arrow is disabled.

 

[paul878]

did you reset you bios to default?

 

[jocasee]

did you reset you bios to default?

Before I start, please note that I have updated my BIOS to the latest and reverted to defaults.

 

[jocasee]

Sooooo I disabled the conflicting sound device and the system hasn't crashed in 2 days so far. I don't want to get my hopes up too much yet.

I found 2 sound devices trying to access Location 0, and I disabled the one that was unused. If anyone is having the same problem, try this and see. I'll post back in a couple more days to verify that this fixed my problem. I'm going to give it a week of being on constantly.

Wish me luck!

 

[jocasee]

So it froze again today. It was running fine for 4 days with no crashes or freezes, but then I turned it off last night. This morning I turned it on, went to make coffee and came back to see my system frozen on the desktop. I force restarted, and verified that the conflicting sound device was still disabled. .

Any opinions on that power supply I mentioned? Thermaltake TR2 RX 850W.

I'm open for options!

Thanks again guys.

 

[Matt1970]

Hard Drive, Ram Tests and even power supply tests are good indicators but sometimes they will miss problems. I would really point to that power supply. And yes, it does have some bad reviews on the net.

 

[meeshu]

Although the PSU was tested at work (thoroughly?), it may be an idea to swap out the PSU with another PSU anyway, and see what happens.

If all is well, then the original PSU may be faulty. If there are still problems, then that suggests a faulty motherboard.

 

[Steltek]

I'd definitely swap out the power supply. Though your testing shows that it puts out the proper voltage at the time you are testing it, the test can't account for an intermittent component failure unless it just happens to flare up while you are testing.

If that doesn't fix the problem, I really feel you have to accept you may have a motherboard fault. If it isn't under warranty and you have to replace it, were it my system I'd swap both the motherboard and the power supply together. Thermaltake makes decent cases, but just isn't a good brand for power supplies (they are like Antec used to be a few years ago, selling very spotty at best OEM units). If you go with a new power supply, get a new Corsair, Seasonic, XFX, or the newer model Antecs and you won't go wrong -- there have been some very good deals especially on Corsair and even Seasonic units off and on lately.

 

[jocasee]

So I'm just trying to convince the wife to allow me to spend some of our tax return on a new PSU.

Lets see how it pans out!

I'm looking now at a CORSAIR HX1050. On sale at FRYs for $159, and seems to be well reviewed. I know HX1050 is a bit overkill, but the CORSAIR HX850 is the same price, and why not have a bit extra power? Maybe I wanna get a GTXTITAN or 2 somewhere down the line. Also, the HX850 I gotta pay shipping from newegg.

What say you bondsmen? Go with the more critically acclaimed HX850, or get the extra power of the HX1050?

I'm totally up for options as well. Let me know.

And thanks again.

 

[jocasee]

I'd definitely swap out the power supply. Though your testing shows that it puts out the proper voltage at the time you are testing it, the test can't account for an intermittent component failure unless it just happens to flare up while you are testing.

If that doesn't fix the problem, I really feel you have to accept you may have a motherboard fault. If it isn't under warranty and you have to replace it, were it my system I'd swap both the motherboard and the power supply together. Thermaltake makes decent cases, but just isn't a good brand for power supplies (they are like Antec used to be a few years ago, selling very spotty at best OEM units). If you go with a new power supply, get a new Corsair, Seasonic, XFX, or the newer model Antecs and you won't go wrong -- there have been some very good deals especially on Corsair and even Seasonic units off and on lately.

Im perfectly willing to accept that it could be my motherboard. I would be soooo happy if that was it. but I dun think so.


Posted from Anandtech.com App for Android

 

[jocasee]

I'm looking now at a CORSAIR HX1050. On sale at FRYs for $159, and seems to be well reviewed. I know HX1050 is a bit overkill, but the CORSAIR HX850 is the same price, and why not have a bit extra power? Maybe I wanna get a GTXTITAN or 2 somewhere down the line. Also, the HX850 I gotta pay shipping from newegg. What say you bondsmen? Go with the more critically acclaimed HX850, or get the extra power of the HX1050? I'm totally up for options as well. Let me know.

Anyone got any input?

 

[Scultimidonus]

Try this; open Device Manager, select Show hidden devices in the View menu, and scroll down to Sound, video and game controllers. You should see AMD High Definition Audio Device, and Realtek High Definition Audio listed. Right-click on each of them and select Properties from the context menu. On the General tab of the Device Properties dialog is listed the Device type, Manufacturer, and Location. If both devices show the same location ie, "Location 0 (Internal High Definition Audio Bus)", disable whichever audio device you're not using, then test to see if the kernel power event stops occuring.

I've had that same crash occur about six times over the past few months, I followed your advice last night after the latest crash and disabled 'AMD High Definition Audio Device'. So, here's hoping this fixes it for me, I'll post again if anything significant does or doesn't happen. Many thanks!

 

[Scultimidonus]

I've had that same crash occur about six times over the past few months, I followed your advice last night after the latest crash and disabled 'AMD High Definition Audio Device'. So, here's hoping this fixes it for me, I'll post again if anything significant does or doesn't happen. Many thanks!

Damn it just happened again. Any help or advice would be much appreciated.

 

[Ketchup]

Damn it just happened again. Any help or advice would be much appreciated.

Suggestion: either continue here or start a new thread (recommended), telling us what issues you are having (instead of 'the same as that guy') and give us some information about your hardware.