PSU causing GPU to stop responding?

theinfra

Honorable
May 16, 2012
5
0
10,510
The system:
Motherboard: MSI 890FXA-GD70 - http://www.msi.com/product/mb/890FXA-GD70.html
CPU: AMD Phenom II X4 965 3.4GHz
GPU: Sapphire Vapor-X 5770 (ATI Radeon HD 5770) 1Gb - http://www.sapphiretech.com/presentation/product/?psn=0001&pid=305&lid=1#
PSU: Thermaltake TR2 600W - http://thermaltakeusa.com/Product.aspx?C=1247&ID=1898#Tab0
Windows 7 Prof 32 bit (clean install, nothing else installed but drivers and games for testing)

The problem is that while I'm on Windows on non intensive tasks such as web browsing, periodically I will receive "Display driver stopped responding..." errors after the PC freezes for about 4-5 seconds then recovers.

When playing a game, this will happen more frequently, and it will send me to a Blue Screen crash, often citing atikmdag.sys or dxgmms1.sys for the cause of the error, although I've seen many more windows files as the cause (like tcpip.sys, ntfs.sys, etc).

Recently I started switching components between other machines to weed out the problem, but the only thing I can replace is the GPU card with another pc's nVidia Quadro FX 3500. With this card the errors still keep happening, except they behave differently: I never see the "Driver stopped responding error..." on Windows, only when playing games it'll freeze for a couple of seconds, audio gets garbled and then it recovers with no error messages. For a second after the freeze I notice that the Frames Per Second go all the way down to 1 (the game until then is steady at 50-60 fps).

I've yet to try swapping out the CPU and PSU for other working ones (I don't have ones that are compatible with the mobo, and I don't want to buy new ones just for them not to be the problem), but now reading the PSU guides and the importance of 12V rails and everything else I got the doubt that the PSU may be the problem (and also the PSU is the only thing I didn't research when buying the system), I just need anyway to confirm this, and if so, which PSU would be the best to replace it?

I plan to add to the system another GPU or upgrade to a better one when I need it (now almost every game I play runs smoothly on max settings) and maybe add a fan controller (I got 4 case fans in there).

Thanks for reading and hope you can help me with my problem.

Edit: using the [ url ] tag screwed up my links
 


have you loaded something like CPUID hwmonitor to see where your temps and voltages are?

the free version should be fine. if your temps are too high or your 12volts rails are bouncing in terms of voltages it could cause stability issues in all of your components. you want your voltages 11.8 or higher and your idle temps to be close to 40c or lower. this does not hold true on all hardware but it is a good mean.
 
ive seen them before... i work as a computer repairist and ive seem some weird things. but yes it is more then likely a driver issue.

i used to use a 7600gt that the fan stopped working on it and it got artifacts and BSOD that gave driver errors and wierd driver and if this is the first time messages frequently until i realized the fan was i jury rigged it and cooled off the card then the messages went away.

im not stating that this is the same problen he is having just saying it could be a possibility.
 
and he also states that he swapped video cards from an AMD card to a Nvidia card and got similar messages which leads me to believe it could be more than just a driver issue.
 
Overheated card, yep, that'll do it



In both cases the wrong DX version could be the culprit ( dxgmms1.sys )
Should be asking if the drivers are from Windows update or from AMD/Nvidia and if the DX drivers are the latest ( something often overlooked on a clean install )

Far too often the psu is blamed when the evidence given doesn't point to it

 


good point!
 
Thanks for everyone's input!

I made some play-throughs with hwmonitor, and these are the stats I recorded. Every time the video driver stopped responding, I would go back to hwmonitor and see if anything would go down or up but as you can see from the min/maxes, nothing really is out of the ordinary (for what I know) except the FPS dropping to 0 and the Display driver stopped responding error

* The 12V voltage I one saw a minimum of 11.82V, couldn't get a screenshot.
* The CPU cores temp is way lower normally (hangs in the 50-55's), since I haven't put thermal paste to the heatsink (since maybe today maybe I'll get another CPU and change them to see if anything changes)

http://imgur.com/AC7GK

If these volts are normal, then I'm stuck with the motherboard or CPU being defective or Direct-X issue.

Normally the Direct-X drivers are installed with the first game that uses them, but this time I installed them with the Direct-X end user runtime http://www.microsoft.com/en-us/download/details.aspx?id=35

Is there a better way to install this? Any way to reinstall or repair Direct-X on Windows 7?

Also, should I take this problem to the GPU forum? or Windows 7 general? Or can we keep on going here even though is not likelly a PSU issue?
 


with that being said everything sounds in tolerable ranges so your PSU is not to blame. i would say more then likely this is a software issue. i wold start with a system clean IE antivirus, anitmalware, antipsyware. once that is done then i would download all the latest drivers and reinstall for your major components like your chipsets, VGA especially.

have you tried the GPUs in another computer to make sure they work fine there?

Edit: any component that can be tested in another "know good" machine will help narrow down if it is a hardware issue.

Worst case would be trying an OS reinstall.
 

Not to keen on software voltage readings, but for what they are they look ok
I'd try a clean video driver install, using Driver Sweeper in safe mode to get rid of all the junk that must have accumulated by now, and not the very latest drivers
 


This is on a fresh OS install. It happened when I had everything installed, and then I installed W7 from scratch and only installed drivers and the games I play.

I already tried the GPU (the only part I could swap) on another comp and it's been error free for two weeks. The only other things that could be causing a hardware issue are the CPU and the mobo itself.



As I said, fresh OS install, and I already installed drivers clean with Driver Sweeper and CCcleaner, both running on Safe Mode

 
you got a power problem on that psu or your motherboard under your core temp for your phenom processor the power package should read all the same your value and minimal read 16.00 watts and the maximal is 117.60 they should be all at 117.60 ,also appply thermal paste between cpu and cooler
 


just for curiosity have you tried a different power supply in your trouble system? it could read normal voltage but not supplying enough amps to meet your watt requirements.
 




Aha! So we're back to PSU problems

I haven't tried another PSU since I don't have another one that would work (this mobo needs 24pin+8pin main power, the other PSU got a 24pin+20pin). I wanted to check before with these readings if it;s indicative of a faulty/incompatible PSU before buying one. Tomorrow I'm gonna go to a place that they told me I can bring the whole build and try out a different PSU and see if the errors stop occurring.

Hope that fixes it. Thanks for taking the tie to help me!
 

😱
Is that an AMD thing ?
Never saw an Intel processor show the same values

Just checked, my tablet must have a power issue
amdpower.jpg


Or, do you get it now ?
 

Turn on Cool and Quiet or whatever the power saving feature of AMD is, then the cpu won't run at full power all the time and you'll see similar numbers to what you've seen in this thread, it doesn't mean there's a power problem
mycpupwr.jpg