• Competitor rules

    Please remember that any mention of competitors, hinting at competitors or offering to provide details of competitors will result in an account suspension. The full rules can be found under the 'Terms and Rules' link in the bottom right corner of your screen. Just don't mention competitors in any way, shape or form and you'll be OK.

xfx GFX 260 black problem

Associate
Joined
26 Dec 2008
Posts
1,211
Location
Scotland
xfx GTX 260 black problem

Was playing GuildWars when suddenly the screen goes purple/pink and a few seconds later the pc restarts. GW doesnt exactly have cutting edge graphics, so I was slightly worried.

The card was bundled with 3DMark Vantage, so I ran that to see if it could complete without crashing. It got as far as test 2 when the screen went orange.

Retested with Vantage after I underclocked the default frequencies to standard 260 settings and it was able to complete a run.

@With Stock clock 666/1150/1404

At idle:
GPU: 63c
VDDC: 54/56/56c

point of crash:
GPU: 82c
VDDC: 89/91/89c

@underclocked 576/999/1242

At idle:
GPU: 62c
VDDC: 52/55/54c

peak temps:
GPU: 80c
VDDC: 83/86/84c

Theres not really much difference between the two sets, how do those temps look?

What should I do now?

I dont like the idea I have to underclock the card to make it work, nor am confident it wont decide to fail completely in the near future.
 
Last edited:
Lots of problems with these XFX 260's :(

Experiment to find what the highest clock you can get before it crashes is, also up the fan manually or set fan profiles in rivatuner and see if that helps.

If after testing you're happy with it at the speeds you can get then fine, or you might be able to get it back to 666 by keeping the temp lower.

Overall though, it isn't capable of working as it's intended and you could go down the RMA route. But if it's from the same place as others who've had problems then the returns procedure might cause a headache :(
 
Have you disabled automatic restart? Might give you an error message if you do.

Under Vista go to Computer>System Properties>Advanced System Settings>Startup and Recovery using the "Settings" button and untick "Automatically Restart"

I always do this on a new installation of Vista, means at least if something does go wrong then you can usually see a error or BSOD message to help diagnose the problem.

Might also help if we knew the rest of your spec, power supply is quite key. Also try using Rivatuner to monitor the temperatures and set the fan to 100% to see if it's an overheating issue.
 
Yeah I probably wouldnt have bought the card if i'd seen the problems other people were having.

I do have automatic restart disabled. When the pc reset, there was no BSOD, just a corrupt screen.

Power supply is a bequiet 750w.

Im thinking the automatic fan speed of 40% is too low for these cards. Obviously I could manually set the fan speed to another value but that would remove the benefit of increased cooling with more heat.
 
I really don't think these cards are capable of running stable at the factory overclock speeds once they hit 80C+, mine crashed pretty quickly once it hit those temps.

Try running at a constant high fan speed as already suggested just to see if it confirms this. Also, try setting just the core or memory at different speeds one at a time to isolate the problem. With mine it was core clock that was the problem, the memory was fine at 2300.
 
I'd get Rivatuner to monitor temps while you are gaming, that way you can tell exactly how the card is doing.
 
I've had exactly the same problem with mine last week -

http://forums.overclockers.co.uk/showthread.php?t=17969619

Were not alone though, there are plenty of reports of this problem on this forum and around the web. Imho These xfx cards are terrible, mine would only run stable if I underclock it to stock 260 speeds or run the fan at 100%. I've returned mine for rma but as it's 7 weeks old I'm stuck with the xfx rma procedure and there is no telling how long that will take.

One thought, you have removed the protective film from the card? I don't see how not doing so would cause a problem but it's best to check!

If you do go down the rma route (I'm guessing who you bought it from here) make sure you take pictures of the card and the packaging you put it in including all the bits in the box as the company are terrible for finding any excuse not to help.
 
That pink screenshot you made in your thread sepulchre, its what I saw when mine crashed.

The protective film was removed, 2 seperate pieces of it.

If the card is older than 7 days am I entitled to a refund?

Also, for the former xfx 260 owners, what card did you finally get to replace it and does it work without problems?
 
As I understand the soga after 7 days you have to give the retailor a reasonable period of time to resolve the issue. Resonable is the key word here as it's open to interpretation. The retailor I bought from first test the card, if it fails they then send it back to xfx who then take around 3 weeks to test it (so I've been told). They tell me if they both find it faulty I will be sent a direct repalcememt, not something I'm happy about to be honest. My plan is to let the retailor test the card and then once try have told me it's failed I will see where I can go from there as I don't see why I need to wait for xfx given the contract of sale is between me and the retailor, not the manufacturer.

If anyone has a better understanding of the soga maybe they can advise us both?
 
That pink screenshot you made in your thread sepulchre, its what I saw when mine crashed.

The protective film was removed, 2 seperate pieces of it.

If the card is older than 7 days am I entitled to a refund?

Also, for the former xfx 260 owners, what card did you finally get to replace it and does it work without problems?

You can get a refund within 28 days (as long as the RMA is registered within this period), subject to them testing it first. They should also pick it up for you too so won't cost to return. As Sepulchre says cover your back with photos, etc. so they can't wriggle out of things.

Definitely elect for a refund, at the very least you'll be able to buy the same card new again but likely at a cheaper price in a week or so.

As I understand the soga after 7 days you have to give the retailor a reasonable period of time to resolve the issue. Resonable is the key word here as it's open to interpretation. The retailor I bought from first test the card, if it fails they then send it back to xfx who then take around 3 weeks to test it (so I've been told). They tell me if they both find it faulty I will be sent a direct repalcememt, not something I'm happy about to be honest. My plan is to let the retailor test the card and then once try have told me it's failed I will see where I can go from there as I don't see why I need to wait for xfx given the contract of sale is between me and the retailor, not the manufacturer.

If anyone has a better understanding of the soga maybe they can advise us both?

Slightly different process was explained to me - if you send it back within the first 28 days and they test it and is found to be faulty they will replace or refund; if it is found to be ok they will offer to send back to XFX for their opinion if you are still not happy.
 
Last edited:
Run FurMark in stability mode. I did and at 1920x1200 the max my XFX GTX260 Black got to was 87c.
 
furmark stability at 1280x1024 hit 81c for me when it crashed to the pink screen, be intersting to see if the ops does the same. Annoyingly I ran furmark not long after getting the card and it was fine, only after 4 weeks of use did I start to see problems :(
 
ok well i just ran FurMark @ 1920x1200 as suggested by Speed
The screen has turned brown when it reached 85c

I should point out that the card was underclocked to 576/999/1242 for that test. I was expecting it to pass tbh, maybe crash with the default clocks.

I'll retry with the defaults and see what temp I get.

Edit: With the stock speed of 666/1150/1404 it crashed at 69c
The increasing temp graph was a lot steeper this time and it died faster.
 
Last edited:
Definately sounds like a faulty card or power supply then.

I just ran FurMark Stability Test again with the "Xtreme Burnign Mode" selected and MSAA @ 16x with 1920x1200 and it went to 90c. No crashes or anything though. Must be COD5 at fault in my case.

Try putting the clocks back to default and running it at your normal desktop resolution.
 
What gfx card and psu do you have, Speed?

Currently runing FurMark@ 1920x1200 with 576/999/1242 and the fan manually set to 61%.

The temp graph was a lot flatter this time, but again it crashed @ 85c.
 
XFX GTX260 Black like you and Corsair TX750W.

I personally wouldn't recommend setting the fan to anything, when I had it running it was on 100% from about 80c upwards. You risk damaging the card if you don't have it set high enough. So reset it to normal temp controlled variable speed and try again.
 
I doubt it's relevant but I have the same psu as the op, maybe just a strange coincidence were both seeing exatly the same issue?
 
Ive only set the fan manually once, just that last test, to see if a higher default fan % would help the stability. Before, it was on auto control.

Interesting you should mention the fan speed, when it was crashing with Vantage, even though the final temp was 82c, the automatic fan was only at 59% 19xx rpm.

I think the autofan isnt provinding enough cooling when the card needs it.
 
What were you using to monitor the fan speed? I had Rivatuners Hardware Monitoring up in the back ground while FurMark was working. That way you can see what happened over a length of time after you've finished the test. It should be @ 100% if it's setup correctly.

You may have to go into the Rivatuner hardware monitoring setup and select "Fan Duty Cycle" to see it.

Or you could manually set it to 100% and see what happens with FurMark then.

It should also be @ 40% approx 1345 on the desktop, check that as well.
 
Last edited:
Back
Top Bottom