Articles

Installed a new graphics card, PC frequently crashes. Help!

April 17, 2012 at 05:29:14
Specs: Windows 7 64-bit (no service pack installed), Intel Core i7 920 (Quad), 2.66Ghz, 6GB DDR3 RAM

I'm new to this site, so hey!

To summarise, a few days ago I had a new graphics card installed to replace an old, overheating one (its fan had broken). The new graphics card is an ASUS GTX 560 Ti. However, after the first few days of using this graphics card, my PC began to crash frequently, specifically when any 3D game is being played. There are no signs of overheating, and using temperature monitoring programs, my CPU and GPU temperatures are fine.

When my PC crashes, it usually doesn't produce any artifacts or screen-tearing. The image either freezes then the monitor displays 'no signal', or my PC restarts itself. I have also had a crash result in a simple brown screen. These crashes are sometimes accompanied by my sound hanging and skipping, and usually the only way to recover from them is to power the PC off via removing any power going to it (ie unplugging it from the mains electricity).

My power supply is adequate, I believe - it is 727.8W max.

If anyone can help me diagnose this problem I would appreciate it. :)


See More: Installed a new graphics card, PC frequently crashes. Help!

Report •


#1
April 17, 2012 at 05:36:07

"My power supply is adequate, I believe - it is 727.8W max"

Wattage is not a good way to determine if a PSU is adequate or not. What is the make/model? Does it have a single +12v rail with high amperage or multiple +12v rails? Is it 80 Plus certified?


Report •

#2
April 17, 2012 at 07:02:25

My PSU is just a stock PSU that came with the PC, so the make and model probably won't mean anything to anyone, but here goes...

The make is Delta Electronics, and the model number is GPS-750AB A.

As for the rail question - it has multiple +12v rails, each at 18A. It appears it is 80 Plus certified - it has the appropriate symbol stuck onto its exterior (exterior of the PSU).

Thanks for the help, by the way.


Report •

#3
April 17, 2012 at 07:16:32

make sure your new video card doesnt have a extra place to plug a 12 volt power into some do and wont work without this being plugged to power also.that or your power supply isnt enough for the new card

Davidw


Report •

Related Solutions

#4
April 17, 2012 at 07:37:52

Although the specs look OK, Delta is generally not considered to be a quality brand.

http://www.delta.com.tw/product/ps/...

http://www.ascendtech.us/delta-gps-...

Looks like the card requires two 6-pin PCIe plugs.

http://www.asus.com/Graphics_Cards/...


Report •

#5
April 17, 2012 at 08:49:41

Yeah, I figured that Delta wasn't a particularly good brand. An update, though:

Because of the fact that my old GPU was overheating, I had my computer on power saving mode (via the Windows control panel).

Today I remembered that, and switched the PC to high performance mode, and no crashes so far, so maybe that was the cause?

Anyway, hopefully the crashes don't occur again. :)


Report •

#6
April 17, 2012 at 13:22:43

After using the computer for a longer amount of time, I can confirm that these changes have not fixed the problem. The crashes still occur. It is worth noting that the crashes are occuring even outside of games (just watching a YouTube video, for example), and, instead of just a black screen, the last crash I had resulted in bright purple colours and artifacts all over the (frozen) image.

So I'm not out of the fire yet, I suppose.

I think the best direction to go in at this point would be to try diagnose the problematic component(s), so can you please give me any help as to how to differentiate between a crash caused by a faulty GPU, the PSU, or any other component?

Thanks a lot,

~Lasur


Report •

#7
April 17, 2012 at 14:52:05

Did you uninstall the drivers for the previous graphics card and get the latest drivers for the new one? This might not be a hardware problem at all. Also clean the card edge connectors with a pencil eraser then pop the card in and out a few times to ensure there is no oxide on the socket itself. Even new cards can have oxide on the edge connectors.

Another simple thing to try is to shut down, unplug from the mains then hold the power off button in for about 20 seconds. If it's a laptop take out the main battery too before doing this. Sounds crazy I know, but it helped me out once (suggested by another helper - long story).

Always pop back and let us know the outcome - thanks


Report •

#8
April 17, 2012 at 15:29:59

Thanks for the advice, Derek.

As for drivers, yes, I have ensured that I have the latest drivers for my graphics card installed, and the old ones uninstalled.

After updating my motherboard drivers I haven't had a crash, however, as I have learnt while trying to fix these crashes, does not necessarily mean that the problem is fixed.

It's late at the moment (I'm in the UK), but if any crashes occur again I will give your other suggestions a try. Thanks a lot for the help, I really appreciate it.

Wish me luck!


Report •

#9
April 17, 2012 at 15:34:12

Hope it lasts then. I'm also from the UK (Kent).

Always pop back and let us know the outcome - thanks


Report •

#10
April 17, 2012 at 20:26:22

I believe that the problem is in the fact that the power supply has four 18Amp 12Volt rails. and this does not translate to even current distribution throughout your system. Even though you have a total of 72Amps on your 12Volt rails if they could be combined, since you cannot, you may be starving your graphics card for current while there is unusable current 'available' elsewhere.
You have 4 PCIe connectors and your graphics card apparently needs two of them to run. try swapping out your PCIe connectors one at a time and experiment to see if some combination of them will work for you. With 4 rails and 4 PCIe connectors, ONE of those connectors will be sharing current with your CPU and possibly other power hungry components so you may hit on a combination that will work for you. In the long run you would be better off with a power supply that has a SINGLE 12Volt rail so that ALL of your Amperage will be available to whatever parts of your system needs it.

You have to be a little bit crazy to keep you from going insane.


Report •

#11
April 18, 2012 at 14:17:34

After doing some testing, the situation seems to be getting better but still not fixed. I haven't had any of the crashes I had been experiencing at all now - I used the PC (and gamed on it) for approx. 15 hours today as a test, and had no crashes while gaming.

Instead, the only crash I have had today was while watching a YouTube video, just after setting it to fullscreen. In addition, last night I experienced a crash after switching a YouTube video to fullscreen too. This makes me (with my layman's perspective) believe that whatever was causing the crashes while gaming has been fixed, while something else is causing these YouTube crashes. All of these YouTube crashes have caused a lot of artifacting and strange, bright colours to appear on the monitor, instead of the usual frozen image/black screen.

Any help from you (exceedingly helpful and generous) people would be greatly appreciated, after this sudden change in symptoms my PC is exhibiting.

Thanks a lot,

~Lasur


Report •

#12
April 18, 2012 at 16:31:15

Have you tried DirectX diagnostics? Type dxdiag in the search box and make sure no errors are reported in the Display tab.

You could also see if there are any errors showing against Display Adapters in Device Manger (usually red or yellow marks against the icon).

Depending on the card there might be a hardware acceleration setting too.

Can't think of much else other than the 12V rail situation suggested at #10, although I would have expected this to affect gaming.

Always pop back and let us know the outcome - thanks


Report •

#13
April 24, 2012 at 09:36:54

After a few days of use of my PC, the crashes during gaming seem to have disappeared. The only crashes I have experienced have been rare, and have all involved watching YouTube videos in fullscreen. Instead of the usual symptoms the crashes displayed (ie a black screen, or just my PC restarting), these result in heavy artifacting and colour distortion on the frozen image, requiring a hard reboot.

I've tried your suggestions regarding DirectX diagnostics and checking for errors in device manager, both of which have yielded no results.


Report •

#14
April 24, 2012 at 11:32:31

Well, I'm down to long shots now. Sometimes you can get some strange symptoms if there is oxide on the RAM edge connectors. Nothing lost by cleaning them with a pencil eraser, then popping them in and out a few times to clear any oxide off the sockets themselves. This usually happens after you have been working inside the computer - slightest movement of the RAM puts it onto an oxidised area (even though it looks reasonable). Same goes for the graphics card and any other edge connectors, such as SATA HD signal and power.

About the only other thought (another long shot) is to run a registry cleaner. I don't normally advocate them because they can get things wrong and usually invalid entries cause no problems. If you want to take that risk the the one in CCleaner is one of the safer ones IMO.

Always pop back and let us know the outcome - thanks


Report •

#15
April 24, 2012 at 19:39:41

I still think it is your power supply. Borrow a better one or purchase one with the condition that if it does not fix your problem you can return it for a full refund (some computer retailers are willing to do this). Make sure that it have the single 12Volt rail, 80% efficient, active PFC, is a name brand and has at least two PCIe 6pin connectors as required for your card.

You have to be a little bit crazy to keep you from going insane.


Report •

#16
April 25, 2012 at 07:51:03

Yep, it would be as well to eliminate the PSU from the equation as it is quite a possibility.

Always pop back and let us know the outcome - thanks


Report •

#17
April 25, 2012 at 13:39:17

I'll do that - when I get the chance. Thanks for the advice. However, buying a new graphics card leaves a dent in the finances for a while, so when I get the chance I'll grab a new PSU and see if it fixes the problems. Thanks. :)

Report •

#18
April 25, 2012 at 14:12:39

OK - thanks for the feedback.

Let us know the outcome on here if you get a chance. The threads stay working for months.


Report •

#19
April 25, 2012 at 15:14:21

I plan to, Derek, thanks. :)

A quick update - after using my PC and gaming for about three days and having only one crash, I have had two in the space of the past two hours-ish. Both of these crashes have been the kind that involve the frozen image the PC crashes on being heavily artifacted and colour-distorted. And again, I have experienced no crashes while gaming.

This seems incredibly odd to me. Can any of you guys shed some light on how, exactly, a PSU could cause crashes... when your PC is doing some of the most low-power operations it does, and not while doing taxing 3D rendering? Not doubting your advice, it just seems odd to me. :P

Thanks a lot. I really appreciate the help and advice.


Report •

#20
April 25, 2012 at 15:29:52

I quite understand your view - it does seem more than a little odd that you can do intensive gaming without crashing yet it freezes on YouTube.

Is it only on YouTube that you get this freeze up or does it happen at other times?

Did you try my long shot about cleaning RAM edge connectors?

Always pop back and let us know the outcome - thanks


Report •

#21
April 25, 2012 at 16:16:33

Update your graphics driver with the latest for your card from the NVidia website.
Update Flash Player since I believe it is used for YouTube video playing.

You have to be a little bit crazy to keep you from going insane.


Report •

#22
April 26, 2012 at 06:18:31

I'm going to clean the RAM edge connnectors later today, I think, Derek. In response to Fingers, I already have the latest graphics drivers and Adobe Flash Player.

A fairly important update, though: I just had another crash... this time while gaming. This comes as a surprise. I don't know whether this would change what you think of the problem and its possible cause(s)?

One more question I have that I believe could be pretty important: Do you think these crashes could actually damage my PC when they occur? I just want to know if I should minimise when I'm using my PC, even to just read and post in this thread.

Thanks a lot,

~Lasur


Report •

#23
April 26, 2012 at 11:20:35

Freeze ups don't usually cause any real damage although at times they can cause the odd software glitch. If you have your restore points you should be OK even if that happens. Mostly restarting gets things back to normal (or maybe I should say abnormal in this case LOL).

Always pop back and let us know the outcome - thanks


Report •

#24
April 26, 2012 at 13:29:56

Okay, thanks for the reasurrance, and the cheap chuckle. ;)
Going to see about your 'long shot' before I go to bed. Time to find a pencil with an eraser, I guess.

Report •

#25
May 3, 2012 at 08:09:51

I'm taking a look at PSUs, and this one caught my attention. From what I can tell, it fulfills the specs that Fingers recommended, but since I'm somewhat terrible at the hardware side of PCs, I wanted to run it by you.

It's a Corsiar Enthusiast PSU - it is max. 85% efficient, has a wattage of 750W, and is 80 Plus bronze certified.
Edit: Forgot to link it. I'm a moron.

http://www.scan.co.uk/products/750w...

Thanks for all the help, by the way.


Report •

#26
May 3, 2012 at 13:11:33

After a bit of research, I believe these crashes involving artifacts may be a driver issue. While I tried rolling back my drivers while I was experiencing the gaming, non-artifact-inducing crashes, I don't believe I tried that for these particular crashes, and I have found evidence on the 'net that other people are in the same situation with the GTX 560 Ti.

So, wish me luck, I guess.


Report •

#27
May 3, 2012 at 13:15:03

I'm going AWOL for a bit so I hope you get it sorted.

Always pop back and let us know the outcome - thanks


Report •

#28
May 3, 2012 at 14:13:35

I hope your commanding officer doesn't execute you in the latrine.

I'm kidding. You have no obligation to help people with their first-world problems relating to their PCs. Enjoy whatever you're going to do, and thanks for the help.


Report •

#29
May 3, 2012 at 20:03:15

I check back now and again and I am sure others do as well so if you have a breakthrough or additional symptoms, someone will be around.
Let us know how the driver roll back works.
As far as the TX750 PS, it looks like a good one, though I like the websites in the US as far as the details on the product and the reviews (I'm in the NYC area) so here is where I would typically go to get details and compare products (though they only ship to US and Canada):
http://www.newegg.com/Product/Produ...

You have to be a little bit crazy to keep you from going insane.


Report •

#30
May 4, 2012 at 03:05:36

Yeah, I know of Newegg, and I would order from them, but as you said, they only ship to the US and Canada, and I'm in the UK, so that isn't really an option. A shame too, but I think Scan is probably the closest the UK have to Newegg in terms of a somewhat-specialist computer hardware retailer.


Report •

#31
May 4, 2012 at 04:37:07

Even though you cannot purchase from them, they do make a good resource for evaluating hardware and getting detailed specs, which is why I pointed them out to you.

You have to be a little bit crazy to keep you from going insane.


Report •

#32
May 5, 2012 at 10:48:11

Yeah, I can appreciate that, Fingers.

So far so good, no crashes yet. Let's hope the driver rollback sorted the problem out.


Report •

#33
July 3, 2012 at 14:14:26

Yeah, Arkinshade. I'm in the same boat as you, man.

Been rocking my Point of View GTX 560 Ti for about 4 weeks now but in addition to the odd crashes and Purple Screens of Death my GPU temperatures seem to be off the charts in certain cases during game play.

Now your link is the highest listed in Google regarding this issue, so if you come across any solutions, be sure to list it here because people such as myself are still curious as to your findings.


Report •

#34
August 4, 2012 at 00:49:52

I am also having the same problem. max temps of 60-66 and its been quite some times since i had a purple screen. I think removing the power cable has something to do with the crashes you get. there is a thread I read while trying to fix this and if you unplug your computer for a short time and then plug it back in you will be crash free for days or weeks or more but they will come back eventually and more frequently until you unplug again. now this may be totally off and I have not tested this at all. I do belie it is possible that power supply are the cause again I have no solid data to back that up but I'm going to test my psu soon with a multi meter. I am running a gtx 560ti 2gb. they also seem to be a common element related to this issue but no clue why. and I have a thermal intake I have seen that brand pop up a few times when I find a thread that is close to what my problem is.

If you learn anything please share. or if you replace your psu and still have the problem that would probably save us all the trouble.


Report •


Ask Question