3

My computer died around 3 weeks ago due to the Graphics Card failing (Asus Direct CU II 7950, an old 1TB Samsung HDD also failed to power up after this) and just 2 weeks ago we replaced it with a new R9 390. It was working fine for about a week, but recently we've received 4 BSODs in the past 3 days. The first 2 were 0x0000000A and 0x0000001E and after the BSOD, one of the RAM modules stopped being detected (resocketing fixed it).

I was going to run a memtest this evening, however during operation I got 2 more BSODs (relatively quick succession), one being 0x0000003B and then 0x1000007E (BlueScreenView is pointing to hal.dll and Wdf01000.sys for this error). I am not really sure what is wrong with my computer right now, and I need help solving this.

If this is RAM, I intend to run Memtest later to verify, however if it is a driver issue I don't really know how to pinpoint what driver is causing the problem. Hope someone can help shed some light on this. For my system specs:

  • Mobo: Asus Sabertooth 990FXA
  • Processor: AMD FX-8320E
  • RAM: 2x G.Skill 8GB DDR3 1333Mhz
  • HDD: 2TB Seagate, 1TB Seagate (with windows), 160GB WD
  • GPU: AMD R9 390
  • PSU: Coolermaster Silent Pro 700W

The only recent driver change I've done is uninstall the display driver completely (AMD clean uninstaller tool) and reinstall Crimson 16.3.2. I have had no other system changes during this time.

The previous 3 BSODs were kernel dumps so I cannot really upload them here but I will upload the latest (0x1000007E) BSOD dump here once someone tells me how (as I don't see an attachment option here). I really hope someone can help me with this really big problem.

Thank you very much!

Update: Used DDU to reinstall display drivers, still ended up getting 2 more BSODs:

  • KMODE_EXCEPTION_NOT_HANDLED (0x0000001E) pointing to CompositeBus.sys and ntoskrnl.exe
  • NTFS_FILE_SYSTEM (0x00000024) pointing to Ntfs.sys, ntoskrnl.exe and usbccgp.sys

Running a memtest now, hopefully this is really just a RAM issue.

Update 2: Ran memtest and it didn't detect any errors, ran a chkdsk as well and sfc and no errors from them either. Does anyone else have any suggestions perhaps?

Update 3: It is worth noting that the BSODs only happen when I start a game (Dota 2, Fallout 4, Batman Arkham Knight, etc). During normal use (e.g. videos, music, work or browsing) I experience no bsods, but the moment I start up a game it will bsod within a few minutes

  • 1
    In my experience, varying BSoD codes almost always point to bad RAM, followed by motherboard. I'd suspect the RAM module that vanished unexpectedly first. So, as you plan, test your RAM, beyond that, as-is this question is too broad IMO. We're not a "please examine my dump files and tell me what's wrong" service, at least not when you haven't done basic troubleshooting yet to narrow the scope. – Ƭᴇcʜιᴇ007 Apr 5 '16 at 13:53
  • 1
    @Ƭᴇcʜιᴇ007 I understand that, and I am sorry if I made it sound that way. I am doing my best to narrow it down. I am running CHKDSK and memory tests, scanning event viewer and my drivers for any compatibility issues and whatever else I can find online pertaining to my problem. I am not yet running memtest since it will take a long time and I want to do some of the faster checks first. I'm sorry again, but I'll do my best to narrow it down as best I can – BlackVulcan Apr 5 '16 at 15:26
  • 1
    The fact that it happens when when you start a game make me thing it is either thermal, power or a problem with the card itself. Coretemp will let you monitor CPU temperatures, GPU-Z will let you check graphics card temperatures. How old is the PSU? Check that you didn't knock the memory or anything else loose while fitting the card. – Mokubai Apr 6 '16 at 6:12
  • 1
    700watts should be fine, but 5+ years is quite old and the components can degrade over time and have lines brown out slightly causing voltage instability. The new graphics card could be just enough to accelerate an emerging problem with the PSU... Bit of an expensive component to replace just to test though. – Mokubai Apr 6 '16 at 6:23
  • 1
    I'd agree with your plan of action: rule out the PSU first, if possible, then Windows. Hopefully it's not the mobo. Your previous corruption could be a sign of dying regulators on the graphics card, especially considering the sudden death, and if the PSU 12v line is unstable or low it could have been the cause as it would cause the regulators to work harder or hotter. Power to the GPU being bad could cause any number of seemingly random failures. – Mokubai Apr 6 '16 at 6:43
1

In my experience, BSOD is most commonly related to drivers, and specifically, video drivers. However, the corrupt HDD does catch my attention because that would not be a typical result of a failed gfx card. It could potentially damage its PCIe slot though.

First thing is to try reseating all connectors (HDD, RAM, PCIe cards). If youve already tried this, or it fails, I would start narrowing down the possibilities, beginning with the gfx card. Remove it and use the on board graffics. If this works, concentrate on drivers. If fail, test any other PCIe cards in similar fashion. If fail, remove all but two sticks of ram. If fail, swap the RAM with different modules. If fail, try a fresh OS install on a separate disk. If this works, it was a software issue on the prior installation. If fail, you may have a bad motherboard.

I once had DIMM slot 1 die on a mobo, and as a result the rest of the RAM would not initialize properly. I blamed it on a capacitor that looked extra bulgey and moved on to new gear. Good luck.

In response to Update#3 This screams video driver issue. Possibly a 3D/DirectX/OpenGL config somewhere. Make sure DirectX is up to date and check out the dxdiag tool. Try different/older versions of the driver, newer is not always better.

  • I also believe that BSODs are mostly driver related however the weird thing is the different types of errors I have been receiving, so far they've been: KMODE_EXCEPTION_NOT_HANDLED, NTFS_FILE_SYSTEM (this one is alarming, but a CHKDSK didn't find any errors), IRQL_NOT_LESS_OR_EQUAL, and even SYSTEM_SERVICE_EXCEPTION – BlackVulcan Apr 6 '16 at 5:59
  • I agree, that is weird, but I'd still bet on it being the graphics driver. I did see one mention of hal.dll which often comes up in gfx issues. The OS could be falsely reporting the cause of the crash, or maybe the driver failure is happening while these other operations are trying to occur. That would be an explaination best left to someone with more Windowsy experience than me. – elCrash Apr 6 '16 at 6:27
  • Okay, will try reinstalling the display drivers again (did it once already with DDU but maybe I could try an older version of the driver). Thanks! – BlackVulcan Apr 6 '16 at 6:29
1

BSOD in ~90% of the cases is caused by drivers. I have also met some extreme cases when the hardware failed. I have even ran into the problem when the computer case itself caused a short circuit (this took me 2 days to find out), you might need some patients to solve this problem.

So lets start with the drivers since that is most likely to cause the problem. There are some official methods that you should try first before troubleshooting the hardware.

Make sure all your drivers are officially signed by Microsoft. You can do this by starting sigverif.exe by searching it from the start menu or typing it into Run . This should be all right, however we should rule this out since this won't take long.

Next, there is an official driver troubleshooting software included in Windows named verifier, you can start it the same way as you did in first step but you have to start this as an administrator.
Here is a nice step by step guide of how to use it. If you get BSOD while the test is running, boot Windows in safe mode and try again.
If you find a problem related to any driver, you should try to disable it from the device manager first, than check if the problem has been solved.

To check the RAM, you can also try another MS program included in Windows, called mdsched.exe (also ran as an administrator the same way as we did before)

One of these methods should help you find the root of the problem in most cases. If not, I would start replacing the hardware in my computer one by one and check if the problem still persists. You should start it with your video card. If your motherboard has an integrated video card, use that instead for a while and check if the problem is still present.

If you have found out anything with these methods or have any questions, give a feedback and I will update my answer.

  • I tried verifier, but didn't get much out of it. I've updated all my existing drivers (though I didn't touch some of the hidden ones, at least the ones I don't know what they're donig). I've tried checking the RAM multiple times with windows memory diagnostic, and even tried memtest and there were no errors present. I did however, try taking parts out and found out it still was the memory causing it, I just replaced the RAM today and it's working fine again. Thanks for all the help! – BlackVulcan Apr 10 '16 at 12:31
  • @BlackVulcan - forgot to mention before, that you should optimize the stability and performance of RAM by loading the XMP profile in bios if it is supported. Had freezes on my own computer in the past and this solved it for me. – Divin3 Apr 10 '16 at 12:41
  • Yeap, loaded the XMP profile at first, though decided to manually set the DRAM timings and it's working fine now – BlackVulcan Apr 10 '16 at 14:32
0
  1. Uninstall the drivers again using DDU
  2. Clean up the registry and delete temporary files (ccleaner or glary utilities will do)
  3. Reboot (let it boot without the drivers). Now check for BSOD, if it's not happening then it was 100% driver problem.
  4. Goto your GPU manufacturers website and download latest drivers(do not update Windows. The driver which is installed by Windows update are always out of date). Install. And you're good to go.

Your Answer

By clicking “Post Your Answer”, you agree to our terms of service, privacy policy and cookie policy

Not the answer you're looking for? Browse other questions tagged or ask your own question.