Weird Crashing

[Guitarman0]

I have a question from my friend and I would appreciate it if you guys could have a look and see what you thing for me please.


Hi,



I am currently experiencing issues when loading/ just after loading games. Often when I have just loaded a game my computer crashes with a pinstripe affect like in the picture, or sometimes just a black screen - on both occasions i must manually shut down. It most often occurs if I try move around in a game just after loading. This to me seems like a graphics card issue/ graphics driver crash. However, on other occasions, if I am very cautious about my actions in the first five minutes then the computer will not crash and will run perfectly for hours without issues.



These are my specs:

CPU: i5 4670k

GPU: ASUS R9 280X 3GB

Motherboard: MSI Z87-G43 Gaming

PSU: Corsair CS650M 650W

RAM: Avexir 8GB (2x4GB)



The computer runs perfectly at all other times and I have done Furmark tests and it does it perfectly with no issues. I find it very weird that many games (not all) cause this crash around the loading phases of the games, and not under any other circumstances. I have reinstalled the games to see if they were the issue, but they all still did crash afterwards. The games themselves never produced any crash files so I assume it’s not to do with them.



I have reinstalled windows, I have reinstalled drivers and nothing seems to have worked. I have taken the graphics card out and used the integrated graphics which works fine with no crashes in the same circumstances (laggy though, of course).



Does anyone have any idea what this issue could be? I had my previous R9 280X replaced due to a much worse issue where there were graphical artefacts etc appearing and was a dead card. Am I unlucky and got another dead card or is there something I am missing? I have never overclocked anything on my computer.



Thanks

Photo of the crash page:

 

[CinderzFTW]

Have you ran a MemTest to check your memory... one of my old cards wouldnt game but it turned out to be one of my sticks of mem

 

[Guitarman0]

Thanks didn't think of that, I'm just running the test now. I hope it is the memory otherwise I have no idea what it is - will post results when I'm done.

 

[CinderzFTW]

well you can rule out a driver issue due to the fact you formatted and reinstalled OS... lets hope its the memory hey

 

[Guitarman0]

well you can rule out a driver issue due to the fact you formatted and reinstalled OS... lets hope its the memory hey

I've tried running the memtest and made 15 passes without any errors :/ I'm really stuck as to what it could be.

 

[AthlonXP1800]

I remembered read about AMD GPUs pinstrip screen of death few years ago.

This mean 280X is faulty as it not worked with games.

Try downclock and upclock both 2D and 3D see what happen.

If 280X model is either R9280X-DC2T-3GD5 or R9280X-DC2-3GD5, here is VBIOS update on ASUS site that could fix the issue.

 

[CinderzFTW]

Go into Catalyst control panel and gpu clocks... push the power limit up to max... helped with my 7870 crashing

 

[Guitarman0]

Thanks for the suggestion, I do have the R9280X-DC2-3GD5 and so tried update the VBIOS via the ASUS website. When i went to install it however it said that the update was already present.

I upped the power limit via catalyst. I thought that had solved it, I opened up a game 25 times with no crashes at all just to make sure i wasn't imagining things. However today when I tried it has gone back to the usual crashing after loading with the exact same settings applied. I've tried a small voltage boost using MSI afterburner, when i put it back to default settings the screen started producing artefacts. This was resolved when I restarted the computer.

I'm going to do one more test using the integrated graphics and use the computer on it as I would normally to make sure it is the card and not some other component that's causing these issues.

If the integrated graphics work fine without any faults, do you guys reckon I should send the card back for an RMA?

Thanks

 

[AthlonXP1800]

Hmmm your card boost clock is 1000MHz, 150MHz overclocked from base clock 850MHz. You did tried upped power limit and voltage but not tried downclock yet.

Try downclock to 950MHz boost and see what happen.

 

[Guitarman0]

Thanks, I've Just tried downclock it down to 950, it crashed within seconds of playing a game. I did it again a few times but it still crashed. I then did the base clock of 850 and the same result :/

What do you think?

Aside from that I've been using the integrated graphics over the past few hours and been running all the games perfectly, albeit low settings.

 

[Guitarman0]

Hi,

Unfortunately I still have no resolution to the issue and I am rather desperate.

Nothing much has changed with my computer except I have a new larger SSD and have installed it and the other hard drives from fresh. However, I still have the same issues and cannot explain what is wrong.

I sent the graphics card back to OcUK and they said there was nothing wrong with my card, so they sent it back to me. But when gaming I still get the pinstripe/blank screen crashes.

Does anyone have any ideas what it could be? Is it possible my power supply is not capable for the r9 280x. Or is it the PSU is instable and become faulty over time, bearing in mind when I had my original r9 280x I never got hard crashes like the ones I am experiencing now.

I have linked a video showing the crash happening in Rome II.

Any help would be greatly appreciated!

Here's the link: https://www.youtube.com/watch?v=iUQfdfO-xIg

 

[Jonnygrunge]

Have you tried the card in another PCIe slot? If it is an ASUS card are both the green power LEDs visible on the VGA card?
Are you OCing the VGA card? Try and set power limit settings in overdrive to 20%.
Good luck!

 

[Guitarman0]

I have just tried moving it to another PCIe now and it still crashed. Both LED lights are on and I am not overclocking the VGA Card. I set the power limit settings that high before and it still did not work :/

Thanks for your help, is there any other suggestions you could think of?

 

[AthlonXP1800]

I read all the way back on this thread, wow since Sept 2014.

Corsair CS650M 650W PSU has 51A +12V rail, got plenty of amperage power to powered Radeon 280X which the graphic card required 30A with minimum 550W PSU.

Have you been updated your MSI Z87-G43 Gaming motherboard to latest A.6 BIOS?

If you updated it already then I would suggest you to borrow PSU from somebody to test see if it stop games crashes.

 

[Guitarman0]

I know its killing me!

I looked and realised my BIOS was not updated, so I was excited. I just updated and went to see if it had worked... unfortunately not. Same crash.

Also, I have no way of getting a power supply. Is there any way to test it myself?

Thanks

 

[AthlonXP1800]

I know its killing me!

I looked and realised my BIOS was not updated, so I was excited. I just updated and went to see if it had worked... unfortunately not. Same crash.

Also, I have no way of getting a power supply. Is there any way to test it myself?

Thanks

I figured out a way yes you can test yourself find out if your PSU faulty by check +12V readings with HWiNFO sensor status.

http://www.hwinfo.com/

Download, install and run it then click on sensors button to open status window with readings, click logging button to start logging to file and launch game until it crash. Let me know the +12V readings from beginning at idle and end at full load to determine your PSU is faulty if +12V performed less than minimum +11.400 or over maximum +11.600.

http://pcsupport.about.com/od/insidethepc/a/power-supply-voltage-tolerance.htm

 

[Jonnygrunge]

What are the temps like on your card?, you could monitor them with Afterburner/Riva Tuner.
There are many latest ASUS VGA BIOS's here do you have the correct one for your card?
Yeah agree above, it could also be a 12V rail issue, potentially be dropping too much with Vdroop (under full load) I would check that voltage level also.

 

[Guitarman0]

I have checked the 12V readings on my PSU now using that program. I left he computer for 5 minutes just sitting there and got a constant reading of 11.968V. Then I went crazy on Watch Dogs at full settings and the Voltage basically never budged apart from a few small decreases to 11.880V. I assume that was to provide more current but it the card didn't seem to need it that much. Ive attached a screenshot of the graph made by HWinfo. The game didn't crash during the time I did testing though.

Then I tried Rome II to try do the same tests. The same voltage readings were obtained. There's a lot of info but the GPU VRM Current Out (IOUT) [A] and GPU VRM Current In (IIN) [A] caught my eye as they were fluctuating wildly with the IOUT fluctuating between 3.75A and 79.75A in the space of 200ms for example. The IIN fluctuated between around 0.3A and 9A. Now I don't know if this is an issue but the PSU can deliver 51A and this was well over, although I don't know - this could be an entirely unrelated thing. It did seem to do this wild fluctuation when I played a game.

Rome crashed but I cannot see why, I've got the logging file that records until the crash.

As for the temperatures I logged that too and the maximum the card got to was 67, and averaged 65 under full load. When idle it was 36. I believe I do have the correct BIOS now so I don't think that should be an issue any longer.

Really appreciate the help!

 

[new boy]

Could try underclocking the vram very silightly

edit - I had a hd7850 that couldn't do stock vram.speeds and my symptoms were very similar to yours.

 

[Jonnygrunge]

Could try underclocking the vram very silightly

edit - I had a hd7850 that couldn't do stock vram.speeds and my symptoms were very similar to yours.

Yeah thinking that too now! Try under-clocking the VRAM slightly, down from 1600 (6.4GHz Effective) to 1550 (6.2GHz Effective) and give that a try. Or try increasing the VRAM voltage very slightly. I would probably go for the RAM under-clock first.