Display driver issue - cant find the cause

-JD-

Reputable
Jun 17, 2014
32
0
4,530
OS - Win 7 x64 Home Premium
PSU - Corsair 600W
Mobo - MSI z77A-GD55
CPU - i7 2600k
GFX - Nvidia GTX460 EVGA SC edition

It is a self build and has been stable with the exception of 6 months ago when the PSU died and I replaced it. It was then fine again until Saturday.

At first i was getting a black flickering on my screen and i got the following message when it stopped.
"Display driver stopped responding and has recovered - Display driver NVIDIA Windows Kernel Mode Driver, Version 346.25 stopped responding and has successfully recovered"

After this happened a few times i checked and was due a driver update so I did that, it did not help.

I uninstalled the drivers and re installed an older driver this did not fix the problem.

Im not sure the exact sequence of events but during all these driver changes the problem became more persistent and I blue screened a couple of times.

The code i got 0x000000161 when googled told me it was a graphics card issue.

I then ran the system with a much older card which at first was fine, then after a few hours I got a very brief flicker once or twice. It was enough that when I looked it up in the event logs it had been logged twice since i put in the back up card, the log showing an almost identical error as when using my normal card.

The next day (Sunday) I tried to force the issue again with the older card by opening up lots of streams and videos but it was stable all morning. I then tried uninstalling drivers & putting back in my normal card. The PC booted up, I installed the drivers then it prompted a re-start. I re-started and then it went into a cycle of trying to boot and blue screening without getting to desktop. I stopped it and put back in the old card and it was stable again for a while.

Using my old card for the rest of the day was mostly fine then I had some flickering then a freeze for 1 - 2 seconds followed by a very similar error message "display driver has stopped working and recovered etc etc".

I tried removing the drivers using DDU and putting back in my GTX460, it boots up fine with no drivers but when i install drivers and re start it goes back in to the BSOD cycle mentioned above.

While the problem is much worse with my normal card (GTX460) i am still getting it with my very old backup card. Im very confused!

My thoughts, either the problem is my GTX460 and the reason i got it with the older card is driver conflict and now i have used DDU it will be fine with the old card proving its the GTX460 or the problem is something else, perhaps PSU as the 460 has 2xpcie power and the old card has none. Having replaced the 3 year old PSU 6 months ago makes me think it cant be the PSU and if it is, the fact that ive had 2 fail in 6 months seems strange, and wonder if something else in the system is causing the PSUs to fail.

Thanks for any help you can offer, it is greatly appreciated. I've trawled through many threads and am getting desperate!
 
Solution
The GPU that doesn't require supplemental power is a lot less likely to cause issues than one that does, which may be why it only has slight problems. The GTX460 is still a rather old card so it could be related to the cards age but it's more likely to be related to the PSU since both cards experience the issue to some degree.

Corsair has few good PSU models tiered high enough to be recommended for use with gaming cards requiring supplemental power. CX, CS, VS and RM series units, among others, are all noted for having a higher than normal rate of issues with gaming systems and actually producing sustained rated voltage.

Add to that the fact that some of the GTX460 models require a minimum of a 620w unit, and I think your...
The GPU that doesn't require supplemental power is a lot less likely to cause issues than one that does, which may be why it only has slight problems. The GTX460 is still a rather old card so it could be related to the cards age but it's more likely to be related to the PSU since both cards experience the issue to some degree.

Corsair has few good PSU models tiered high enough to be recommended for use with gaming cards requiring supplemental power. CX, CS, VS and RM series units, among others, are all noted for having a higher than normal rate of issues with gaming systems and actually producing sustained rated voltage.

Add to that the fact that some of the GTX460 models require a minimum of a 620w unit, and I think your superclocked edition is one of them, and you could be very much underpowered in addition to the actual build quality of your unit. Corsair might be a well known name, but it's also well known that they use a variety of OEM builders for their power supplies, and are more oriented to the budget market than to the enthusiast market except on their AX, HXi, HX, GS, RM (Below 650w) and TX series units.

You might be best served replacing your unit, depending on the actual model, with a unit ranked Tier 2B or higher and then considering a newer GPU if that problem is not resolved with a quality PSU which it's probably more likely than not it will.

http://www.tomshardware.com/forum/id-1804779/power-supply-unit-tier-list.html
 
Solution
You should also probably, as a first measure, record the system voltages as reported in the bios, then download HWinfo (Not HWmonitor) and run sensors only and take a screenshot of the system voltages and post a screenshot here along with what was reported in the bios. Doing it while under load and posting a shot of that would also be helpful if you can.
 

-JD-

Reputable
Jun 17, 2014
32
0
4,530


Hi, thanks for your help with this issue.

Bios readings were as follows:

CPU Core Voltage - 1.120/.128
CPU I/O Voltage - 1.040/1.048
GPU Voltage - 0.000
DRAM Voltage - 1.488
3.3V - 3.312/3.328/3.296
5V - 5.045/5.003
12V - 12.056/12.144

The values with / and multiple numbers indicate where there was flickering between these variations.

HWinfo - no load - with load

The PSU is only 6 months old and I think it has a 3 year warranty. If you think thats my best first course of action then ill change it. I would rather not fork out for a new PSU with this one being only 6 months old. I will try to push corsair for a refund or if not a paid upgrade but I doubt they will go for it. If that is the case I'm going to end up with a like for like replacement. What would you think of carrying on using that assuming it works or is there a risk of it damaging my other parts? It seems strange that the first one lasted a bit over 3 years i think and this one died after 6 months.

I'll wait to hear back from you on the above before I take action. Thanks again.

 
Your voltages look ok both in the BIOS and in HWinfo. That doesn't mean there can't still be an issue, but on the surface it seems to be ok.

Try complete removal of the current drivers using the Display driver uninstaller, and ONLY the display driver uninstaller, no other driver utility. First, uninstall any Nvidia applications in the Programs and features section of control panel. Use the method outlined in the drivers section of the following link followed by the installation of the most current drivers available for your card at the Nvidia drivers site which is linked to in the article.

http://www.tomshardware.com/faq/id-2492424/laymans-simple-guide-solving-graphics-card-issues.html

If that fails to correct the issue, there may be other problems and we can attempt to address them at that time but for now let's be sure there is a clean driver installation with no chance of registry or previous driver conflicts.
 

-JD-

Reputable
Jun 17, 2014
32
0
4,530
I had already done that but I did it again anyway to be extra sure and had the same result.

Once the drivers are removed as instructions and the machine is off I put back in my normal card, the machine boots as normal except for about 1 second during the booting process where there are white vertical dashes across the whole screen which is not normal. Other than that it boots fine, then I install the drivers which prompts a restart. The computer then gets stuck in an attempt to boot followed by BSOD cycle.
 

-JD-

Reputable
Jun 17, 2014
32
0
4,530
Ok i tried downloading both 1.10 and 1.11 from there. When I try to run them it says "the version of this file is not compatible with the version of windows you're running. Check your computer's system information to see whether you need an x86 (32-bit) or x64 (64-bit) version of the program, and then contact the software publisher. There are no options there for different versions.

I used the MSI "live update" program then afterwards checked in the BIOS on booting and then system information, they both say im running the 1.1 bios, the MSI live update program says i have the 1.10 version. It says the latest version is 1.B0 which is what it was supposed to install the first time i ran it. I ran it again and nothing changed.

Thanks for your continued help.
 

-JD-

Reputable
Jun 17, 2014
32
0
4,530
I've read through it all and I must say I'm a little hesitant because of my inexperience and the potential risk involved. You mentioned in your initial BIOS update suggestion it could help when installing newer hardware on an older board. Since no new hardware has been installed for 6 months and it has been working fine until now is it still worth me updating the BIOS?
 
Flashing firmware from USB has been standard practice on most boards for quite some long time, and I'm not sure why that option isn't present on that board as a default method, so I don't think THAT part is a problem. However, if it has been working fine previously and you've installed no new hardware recently, then it probably isn't the problem.

Make sure you have a clean graphics driver installation by following the steps outlined in the drivers section here:

http://www.tomshardware.com/faq/id-2492424/laymans-simple-guide-solving-graphics-card-issues.html

as windows may have updated the GPU driver and screwed things up or just having installed more than one driver version over a period of time could be the issue. Both Nvidia and AMD recommend full removal of drivers prior to updating to any newer versions.

If that doesn't resolve the issue then it's likely just a GPU that's beginning to fail.
 

-JD-

Reputable
Jun 17, 2014
32
0
4,530
Yea, ive done that as per the guide 2 or 3 times. And im definitely getting the same/similar problem with my backup card. It happened again today. It hasn't yet BSOD'd with the backup but the same screen freeze followed by display driver recovery message. Seems like it might be getting worse.
 
Yes, it does. So long as you selected the Nvidia option during the DDU process. It's more likely that the PSU is at fault since you get the same behavior from two different cards and regardless of the driver used and considering the voltage really should be more stable, not "flickering" around. I'd replace the PSU with a higher quality unit as outlined on the tier list.
 

-JD-

Reputable
Jun 17, 2014
32
0
4,530
Well, there's never any guarantees, and sometimes the process is more difficult than you would think it ought to be. I think there's a high probability of the PSU being at fault since multiple cards present similar issues, but again, there remains the possibility of an unrelated issue. There's also the fact that with gaming cards and budget Corsair units we see higher failure rates than with a lot of other units, so I'll keep my fingers crossed for you.
 

-JD-

Reputable
Jun 17, 2014
32
0
4,530
*sigh* no luck with the new PSU. I have it installed and the PC runs the same as it did with the old one. I did all the previous steps, DDU etc but as soon as I put back in the GTX460 i got the BSOD cycle. I think I will keep the new PSU anyway as i can return my old one for a full refund and this one was only £10 more and according to your PSU guide is a lot better quality.

To refresh our memories on the current state of affairs:

At first with my main card installed I was getting "display driver has stopped responding messages" then eventually BSOD. But now with main GFX card installed with NO drivers the PC boots as normal EXCEPT after the MOBO logo flashes up on the screen there is a brief 1 - 2 seconds where the entire screen is black but covered with what looks like this symbol |||||||||||||||||||| but in white. Its too quick to take a picture but it looks something like this. It then boots to desktop. I install the drivers and re-boot. It boots the same as I just said but once it gets to the windows loading bar on the screen it immediately BSOD's and then gets stuck looping restarting and BSOD.

With the backup GFX card installed it boots normally, with no |||||||| symbols that I can see BUT I do get occasional freezing with "display driver has stopped responding" messages which is how this started with the main card.

So it isnt the PSU as its the same with 2 different units and I dont see how it can be the GTX460 as although I get the problem much less frequently with my backup card it does still happen.

So im guessing that leaves one of - MOBO, CPU, HDD, RAM ?

What can I do to eliminate some of these possibilities? My guess at this stage is MOBO as even with my limited knowledge I would think that the fact that different GFX cards at least allows me to boot up means it probably isnt HDD/CPU/RAM? If we think its MOBO is it worth me trying a fresh windows install as if I have to get a new MOBO ill have to fresh install anyway. Also if we do think MOBO is it worth me going back and again trying to update the BIOS? Although there is no new hardware would it still be worth trying? Ideally I want to do anything possible to test stuff to make sure I know whats the issue before ordering any new hardware. I appreciate this isnt always possible.

Seriously big thanks for persevering with me this far!
 
Is ANYTHING overclocked at all? CPU, RAM, GPU or base clock? Have we tried resetting the BIOS to default settings? Check to see if you have the most recent BIOS version installed. Also try all of the following.

Run Seatools for Windows on all connected drives one at a time.


Run Memtest86 on each memory module with ONLY that module installed. Make sure individual module operation is correct by installing the module in the DIMM 1 slot. Run Memtest for 7 passes. The number of passes is configurable in the options prior to running the application. The application will need to run in a DOS environment so you'll have to burn the program to disk or create a bootable USB flash drive to run the program. It won't run in windows. All the necessary instructions are available along with the download at the Memtest86 website.

Download it here: http://www.memtest86.com/download.htm

Instructions are here: http://www.memtest86.com/technical.htm


Download and run CPU-Z and post screenshots here of the memory and SPD tabs. Do this prior to running Memtest. On the SPD tab there is a drop down box to select each module. Take a screenshot with each module's data shown, for comparison.

The latest BIOS firmware for your motherboard is version 1.11, so if you have a version older than that, I'd update it. You can find it here: http://us.msi.com/support/mb/Z77AGD55.html#down-bios
 

-JD-

Reputable
Jun 17, 2014
32
0
4,530
Ok thanks, ive got a busy day today, hopefully ill be able to get some of it done later if not then tomorrow.

As far as I know nothing is overclocked except the GPU which is factory overclocked.
 

-JD-

Reputable
Jun 17, 2014
32
0
4,530
Ok, I've just run SeaTools on my HDD's, I ran all the following under "Basic Tests" - SMART check, Short Drive Self Test, Short Generic & Long Generic. All came back passed. Ill move on to the next suggestions as soon as I can.
 
There seems to be a problem with your RAM settings. Your modules are reporting the 9-9-9-24 timings and 1600mhz speed but those specs are for use with 1.65v. Your RAM voltage as you reported from the BIOS earlier in this thread is 1.488v. There are two things to try here.

One, go into the BIOS and enable XMP profile for the RAM, if it's not already. If it IS, see if there is a second XMP profile available and try that one.

Two, if that doesn't change anything, manually set your voltage to 1.65v, since that's what it's supposed to be running at with those timings and speed.

This might very well be the problem. And it might not, but it's certainly not correct as is.
 
You can find the specifications for your memory modules in PDF format here:

http://www.google.com/url?sa=t&rct=j&q=&esrc=s&source=web&cd=1&ved=0CCAQFjAA&url=http%3A%2F%2Fwww.kingston.com%2Fdatasheets%2FKHX1600C9D3_4G.pdf&ei=ZBThVOKfENH2oASd84C4Aw&usg=AFQjCNFnj6ltWAv23QDHxmLYwO38IlXeag&sig2=ifC4LmsUS0McQh1y5tZbtw


It might also be helpful to post screenshots of your BIOS settings which you can do by inserting a flash/USB drive into any available USB port and pressing F12 when you are in the BIOS at the appropriate screens.

How to post images on Tom's hardware:

http://www.tomshardware.com/answers/id-2173703/post-images-tomshardware-guide.html
 

TRENDING THREADS