Can't Diagnose Cause of Display Driver Crashes and PC Restarts

[Thalyn]

Hello everyone,

Been having a hard time diagnosing the root of the issue I've been having with my custom-built PC. Two weeks ago, I started getting display driver crashes every time I was in a game. It would crash, then recover. This continued for a few days. I shrugged it off because I am aware that NVIDIA drivers have been having issues with Windows 10 lately. They are prone to crashing. However, a few days later and continued sporadic display driver crashes in-game I had my PC completely restart and then nothing happened. The monitor was blank. The LED's on my fans were on. I tried to power down the PC via shut down. Nothing happened. I had to shut it down by flicking the PSU power switch off. I then flicked the power back on, pressed the power button on my PC...same thing. LED's on fans were on, but the PC wouldn't boot. I opened up my side panel and saw that the CPU fan was not spinning. I powered everything down. I proceeded to take out the CPU 4-pin connector and use the other one (it was originally an 8-pin). The PC then powered on and has been working perfectly until today.

While in-game I had around my display driver crash and recover 5 consecutive times within seconds of the other. Eventually I turned the game off. My PC was normal until maybe an hour or two later I tried to launch another game. This caused my PC to instantly restart after getting past the main menu and loading a save game.

I proceeded to shut down. Used the original 4-pin when this happened the first time. It booted up, so it obviously wasn't a bad connector. Same thing. Tried a different PCI-E connector. Same thing. Then I proceeded to reseat my GPU, RAM, CPU, GPU and mobo power connectors. Well, I am now booted back up, launched Witcher 3 and no crash. Was in it over 10 minutes until I turned it off.

I was moving my PC around the week prior to the first time all this happened. I am wondering if my GPU simply got loose? My mobo doesn't have those things that snap into place, as you would on RAM slots. I also upgraded my RAM at the beginning of September.

Maybe I need a new PSU. Maybe I need a new mobo. I really don't know and it's beyond my abilities at this point, so I am asking for help here. Any help would be greatly appreciated.

PC specs are as follows:

CPU: Intel i5-4590 3.3GHz;

CPU Cooler: Stock Intel;

Mobo: ASRock H81M-HDS Micro ATX LGA1150

GPU: EVGA GTX 960 2GB SuperSC ACX 2.0+

PSU: Corsair CX750;

RAM: Corsair Vengeance 16GB 1600MHz;

HDD: Toshiba 2TB 7200RPM;

Keyboard and Mouse: Corsair K70 Cherry MX Red Keyboard + Corsair M65 RGB Mouse

Temps are all normal. I know the CX line of Corsair PSU's isn't the best. I've had this since around July (around the time I got my GPU) and I don't have coil whine. I don't plan on keeping this PSU forever, I do plan to get a better one but at the moment this is what I have. I DID have coil whine or some sort of sounds when my display drivers were crashing soon as I launched a game. But again, since I reseated my GPU, RAM, etc, it appears to be stable and launching a game no longer restarted my system.

My friend thinks it may be the mobo?

Apologies for the length. I tried to give as much info as I could.

 

[Slacker]

I see that you are having a bad PSU day. If you said you bought it around July, was it this year? If so, do you have a power supply tester lying around? or a friend or someone you know? Your troubleshooting leads me to believe it's your PSU failing really quick. Check with a tester and see if you can RMA your PSU. I can see that all your components are well within the spec of your power supplies capabilities. As for your friends thinking its the motherboard, I think you should check your PSU first. I did that once on my AMD unit and after the 3rd board it was the same issue until I found out that there was a loose connector on my PSU's modular side.

 

[Assimilator]

Apologies for the length. I tried to give as much info as I could.

More info is ALWAYS better when it comes to troubleshooting.

As for your issue, from what I understand you think it may be sorted now? Personally I don't think 10 minutes of Witcher 3 is enough to tell whether it's stable or not, I'd recommend running the game for at least and hour. If your system is stable after that, it probably was slightly loose connectors.

If the issue persists, my money would be on the PSU - see if you can borrow one from a friend to test.

My mobo doesn't have those things that snap into place, as you would on RAM slots.

It does! It's a little "hinge" that clips over the protruding "foot" of the graphics card when in the upright position. I highlighted it here in cyan: http://imgur.com/tGqL3jN

When it's flipped "down" (toward the bottom of the image) then the graphics card is unsecured, when it's flipped "up" (towards the top) the card shouldn't be able to be removed.

 

[Thalyn]

Thank you for the replies. The PSU was purchased last September and only installed this July. Around the time I installed my 960.

I don't have a spare PSU around. Well, I have a really crappy Sparkle Power Inc Switching Power Supply in storage, but it is only 350W. No other spares.

Wow, could I actually be that dense and it does indeed have one? It is indeed pictured there. I was perplexed that it didn't. I am going to open it up again to check. I am going to laugh if I am this stupid and literally did not notice it. I should add, the mobo was a spare from a pre-built (Cyberpowerpc) Will report back.

I also agree that 10 minutes in Witcher 3 is not indicative enough but prior I would have restarts as soon as I launched a game. The fact that it got that far was a good sign to me.

EDIT: Apologies.

 

[Thalyn]

Okay, so I opened up my side panel and took some pictures. I apologize for the dust, I need a new can of compressed air. Will be doing that shortly.

EDIT: So, it is there but it doesn't have that part that "snaps down", is what I meant. As I see some mobos having. Also, I apologize. Your did link exactly my mobo and I went and linked you mine like a fool. I'm a little flustered right now.
I don't even know if I am describing it properly. It's those clips on the side that make a click. I don't appear to have that. Is this just something ASrock do with their mobos? I have watched videos before and I always hear that click from the clips when people install their GPU.

EDIT 2: Checked with the only other friend I have that is into PC's. Basically, I won't be able to get my hands on another PSU to test things out.

 

[newtekie1]

The clip on that board is weird. Most boards have the connector snap in when you put in the card, that one you have to manually slide over. If it wasn't slid over the card may have been loose, and may have been loosing connection to the slot while the machine was on and causing the crash.

 

[Thalyn]

Manually slide it over? I did not know that. I had a GTX 750 (non Ti) in this PC for a whole year prior to the 960 and never had these sorts of issues. However, it was a very compact card and self-powered, so it did not require a PSU connection.

When I first saw it, it really perplexed me because I didn't notice the conventional clips I am used to. I am not a fan of this mobo and do plan to replace it one day. Currently however, I am between jobs after graduating university so would only be able to budget either a PSU or mobo upgrade if something is wrong with one of them and it is the cause of these sporadic crashes/restarts. If I can recall correctly, I would hear nothing from my GPU when the display driver crashed. Soon as it recovered, I would hear the fans again. And there was a sort of whine to them. Maybe it was indeed coming loose? As of right now, I hear no odd noises from my computer. When I was in Witcher 3, no odd noises.

I just seat the card in the slot (don't hear any clicks), make sure it feels secure and screw it in. The first I put the GPU in, I did look for those clips and simply did not find any (that I knew of anyway) so I figured, well, okay....

EDIT: I am going to test things out with launching a game. I launched a Total War game (Attila) and soon as I did, this...noise came from the GPU. It's not loud. I can't tell if it's a slight coil whine, or just the fans. The game hasn't restarted my PC upon launch, nor have I had a display driver crash yet. We will see what happens. Not a fan of the noise, but maybe it is normal and I just never noticed it before to recall as I had no reason to until now. I do have open vents at the top of my case and "vents" where my side fan is mounted. So it would be more audible.

EDIT 2: Okay, so. When I end my turn and the AI is simulating it's moves, that "noise" goes away and all I hear are the fans. When it gives me control and I can move about the map again, etc, that noise returns. It is not overly loud and may be normal. I am just reporting what I can since this is beyond me right now.

 

[newtekie1]

That sounds like coil whine to me.

 

[Thalyn]

I reckon it is that, too. Pretty sure I have heard it before and cannot 100% recall. I am just paying more attention now than I usually would to any odd noises/occurrences and mentioning anything that could be of note. I spent about 30 minutes in Attila and did not have a display driver crash or system reboot. Prior to reseating my GPU, RAM and GPU, CPU and Mobo power connectors it would reboot my PC soon as I launched a game. Witcher 3 right on launch and Attila when loading a save game. Prior to it escalating to that today, I would get display driver crashes, seemingly out of nowhere, while in Attila. As I said, temps are normal as I did monitor with GPU-z and RealTemp.

Launched Witcher 3 for merely a minute after 30 mins in Attila. I simply wanted to see if it would reboot my PC or crash the display driver. It was perfectly fine and what I suspect is the coil whine noise, began once I loaded into the world.

Possibly a stupid question but I want to make sure: how would one go about sliding over that clip? Do I actually even have one? It just looks odd to me. Is there somewhere I can find an example? Because I am pretty sure my GPU is not clipped in. As I said, taking out a GPU; I've simply unscrewed it from the slot and pulled it out carefully. Installing a GPU into the slot; I just carefully push it in and then screw it until secure into the back slot. Nothing clips. Nothing is slid over.

At this point, could it very well be the GPU disconnecting from the slot and causing these crashes/reboots? I've had the GPU since mid July. Display driver crashes started mid-August but those were nothing like this. It would crash and recover. Very sporadically, whereas now when it happens it is consecutive. Restarted maybe once and sounded very much like driver issues attributed to Windows 10 issues with NVIDIA drivers. I reinstalled Windows 10, used DDU to get rid of the Windows Update NVIDIA drivers, and installed the latest ones. Didn't happen again until two weeks ago, where I had the situation I described in my original post. Now it only seems to happen in-game. Back in August it would occur mainly in-game but also during normal usage. Again, I settled on that legitimate driver issues with W10 and NVIDIA. But maybe it wasn't?

Thanks so much for all the replies, by the way. I really do appreciate you guys helping me troubleshoot.

 

[Thalyn]

It has been stable since last night but I will continue to wait and see. One question: what could have been the cause of the crash two weeks ago where my PC would not boot up. It would power on, but go no further. The LED fan's in my case would light up, but the CPU fan would not spin and when I tried to power off by the power button, nothing would happen. Peripherals like my keyboard and mouse were also lit down. Monitor was dark. I could only shut down at this point by flicking the PSU power off at the back. Would a loose GPU in the PCI-E slot cause something like that?
That incident is what led my friend to believe it may be my motherboard failing? It was very odd and I don't know what to make of it, as I've never had that happen before. I "fixed" it by taking out the CPU 4-pin and inserting my other 4-pin. It booted up the first time from there and the crashes/reboots went away until yesterday.

 

[BarbaricSoul]

dieing motherboard or PSU, more likely the PSU

 

[Thalyn]

Is there a way to rule one out? I know for the PSU people will likely suggest that I get my hands on another one and see how my system fares with it. I am not able to get my hands on a spare PSU or borrow anyone else's.

I'd like to shrug this all off to bad GPU seating, but what happened two weeks ago as described above is what worries me and makes me think it's something more serious; mobo or PSU dying. I am only able to replace one at this moment in time and I really need my PC in general, not just for gaming. If I am going to buy another PSU, I will be aiming towards something better than the CX line of Corsair. Any suggestions for good decently priced PSUs? I am in Canada, so I'd likely be buying off NCIX or newegg.ca. A PSU would be easy and quick enough to replace. I really hope it isn't the mobo.

Thanks again for the help guys. I appreciate it.

 

[BarbaricSoul]

Is there a way to rule one out? I know for the PSU people will likely suggest that I get my hands on another one and see how my system fares with it. I am not able to get my hands on a spare PSU or borrow anyone else's.

I'd like to shrug this all off to bad GPU seating, but what happened two weeks ago as described above is what worries me and makes me think it's something more serious; mobo or PSU dying. I am only able to replace one at this moment in time and I really need my PC in general, not just for gaming. If I am going to buy another PSU, I will be aiming towards something better than the CX line of Corsair. Any suggestions for good decently priced PSUs? I am in Canada, so I'd likely be buying off NCIX or newegg.ca. A PSU would be easy and quick enough to replace. I really hope it isn't the mobo.

Thanks again for the help guys. I appreciate it.

My suggestion would be to try another PSU, and if the problem continues, it's not a PSU issue.

As for a PSU suggestion- http://www.newegg.ca/Product/Produc...0 &IsNodeId=1&bop=And&Order=PRICE&PageSize=30

 

[Crap Daddy]

Did you buy the mobo new because every one has to have that retention mechanism for the PCI express slot? Cleaning the dust is a good idea. A bad seated GPU could've been the culprit. That PSU should be good enough and they just don't die so soon. That card requires an eight pin PSU connection, do you use that?

 

[geon2k2]

Don't want to put more salt on the wound but I don't really like your mobo.
In the long run I think you should anyway buy a better one, it's not ok to pair an i5 with a low end H81 motherboard.
Also I see 3 power phases, and un-sinked. (It should have 4 but probably 1 is for the iGpu anyway).

Also Haswell had issues with some PSUs, as it required I don't know what very low current capability from the PSU. A WA was to disable from bios some advanced power saving features of the CPU (like C6). Google it a bit and you will find. BTW it is possible that this itself will damage the PSU.

So, first I would try to disable the power saving features and see if there are still issues.

Next I would make sure the platform is stable. I would maybe start with the integrated GPU (just to remove the GPU from the equation) and do some cpu tests (OCCT, Prime, IntelBurnTest). I really like IntelBurnTest as it tends to fail very fast in case of issues (minimum 20 iterations). Although with your VRM ... even this might be risky and might permanently damage the MB.

If these tests passes, then there are chances that the MB and PSU are actually good and there are issues with your GPU, although I don't really think you will reach to the GPU testing.

 

[Thalyn]

The mobo was a spare that came from a pre-built PC (CyberpowerPC). I have been planning to replace it at some point because I know it's a pretty bad mobo. I never got around to it because of having to prioritize other things over PC hardware. However, if it is indeed the issue and is what is preventing me from using my PC, I'll just have to go ahead and replace it now.

I don't think it's the GPU itself. I've had it since mid-July. I am doubting it is the PSU. I purchased it last September but only installed it mid-July.

I am really leaning towards the mobo. I am looking at mobos on newegg.ca and thinking I should maybe just go ahead and buy one. Not an expense I needed right now, but I need my PC working. I am looking at this one:
http://www.newegg.ca/Product/Product.aspx?Item=N82E16813130770

What are some reliable mobos?

 

[geon2k2]

The one you mentioned looks really nice.
Asus has also some nice gamer boards similar to the one you choose, like z97 pro gamer about the same price and the h97 pro gamer a bit cheaper, since anyway you don't have a K cpu.

http://www.newegg.ca/Product/Product.aspx?Item=N82E16813132415

 

[Thalyn]

I think it may be my motherboard after all. I was updating my resume and all of a sudden my PC made a noise different from what it usually does - it wasn't loud - and then everything froze. My HDD light was on, not blinking, just constantly on. I pulled my external HDD out of the USB 3.0 at the front of my case and everything instantly unfroze. I know it's the USB header, but it plugs into the mobo, so.

Thanks for the mobo recommendations. I will take a look. I just want a nice and reliable board from a good brand. I don't want to cheap out, but I do have a budget.

 

[Thalyn]

Well, I went for it and bought a new motherboard. This was a long-time coming anyway:
http://www.ncix.com/detail/msi-z97-g45-gaming-atx-lga1150-48-97044-1111.htm