Sign in with
Sign up | Sign in
Your question

Black screen crash during gameplay

Last response: in Video Games
Share
September 28, 2012 10:37:34 PM

Hello,

i read other posts regarding similar issues, but none of the fixes work. I've contacted nvidia and ea tier 2 tech support for help, but neither knows how to fix this. It usually crashes with a black or red screen flicker and a constant sound loop after which i have to usually force a shut down. On rare occasions it recovers and i get an error.

Issue started soon after i got close quarters with premium...it soon made me unable to play bf3 without it crashing my rig. I figured it may only be with bf3...but now it spread to black mesa and borderlands 2.

Here's the info:

" DirectX function "GetDeviceRemovedReason" failed with DXGI_ERROR_DEVICE_HUNG : Device hung due to badly formed commands. GPU "NVIDIA GeForce GTX 670" Driver 30623"

Things I've tried based on troubleshooting guides by the techs:

- reinstall video drivers
- clean reinstall (safe mode with full removal of old) of drivers
- fixing bf3 from origin
- checking for updates for bf3 through origin
- reverting to any older version of drivers (tried everything from 301 up, including beta)
- playing game on different settings such as: high, medium and even low (normally its all ultra)
- cleaning registry (Eusing Free Registry Cleaner)


some other info that might help you obtained from a dump file which sometimes gets created:

bug check code: 0x00000116
param 1: fffffa80`148da4e0
param 2: fffff880`0ff791a4
param 3: ffffffff`c000009a
param 4: 00000000`00000004
caused by driver: dxgkrnl.sys
caused by address: dxgkrnl.sys+5d000
crash address: ntoskrnl.exe+7f1c0



I have to say this is a...BRAND NEW (about 2 weeks old) video card...Its the oc version at 4 GB of ram.

My system specs are:

Intel i7 980 @ 3.36
16 GB of Triple Channel ram @ 1600
GTX 670 oc @ 4gb
win 7 ult x64 sp1
asus rampage formula 3 mobo
sound blaster recon 3D
current driver installed 306.23
latest bios updates on mobo
directx 11

if you need me to run any tests or post any additional info please ask...

Please help

September 29, 2012 1:05:36 AM

Bad luck mate...

and your PSU is strong enough?
OR
you might have a defective card
(yes, some bad VRM's / resistors / FET's /diodes sometimes fail.)

But i will think the first scenario is likely to be the case...
September 29, 2012 1:18:37 AM

doubtfull....my previous videocard was the gtx 580 @ 3 Gb and it had the same issues...

this 670 is BRAND NEW...and issues started right away....

also my PSU is the Antec TruePower Quattro 850W

thanks for the reply...but not even close...
Related resources
September 29, 2012 1:20:16 AM

just as a sidenote: is there a way to test whether or not the psu is faulty?

since you did spring a notion of doubt about the PSU in my head...id try that...if you can point me in the right direction...?
September 29, 2012 1:21:08 AM

beowulf6666 said:
doubtfull....my previous videocard was the gtx 580 @ 3 Gb and it had the same issues...

this 670 is BRAND NEW...and issues started right away....

also my PSU is the Antec TruePower Quattro 850W

thanks for the reply...but not even close...


Ah now you spill the beans ^^

However, was the 580 in the same mobo as the 670 ?
September 29, 2012 1:29:47 AM

lmfao "spill the beans"? ROFL

so yeah i actually went out to upgrade my video card because i thought IT was the issue...when it appears it not

anywho...yeah the 580 was on the same mobo

i should mention i tried different pci-e slots...just in case that was the issue

hope this helps
September 29, 2012 1:33:37 AM

beowulf6666 said:
just as a sidenote: is there a way to test whether or not the psu is faulty?

since you did spring a notion of doubt about the PSU in my head...id try that...if you can point me in the right direction...?



Well, a PSU is not easily tested without a Multi meter.

You will have to take the PSU out the case, plug it in and keep the green and any black lead (from the 24pin plug)shorted, to force it ON.Use a paperclip.
Then use the multi meter and test the voltages.

Here are the standard colors(you prob know, but just in case and some companies use white for +12V)**
YELLOW +12V
BLACK Neutral
RED +5V
ORANGE +3.3V
GREEN power-ON
WHITE** -5V
PINK/PURPLE +5V SB
BLUE -12V

But this is all really moot, as the readings will probably be good. This is cause there is no load on the PSU.

Easiest is to try it in another PC and look for new/similar problems in the test pc..DO NOT USE THE MULTI METER when the mobo is plugged into the PSU.
September 29, 2012 1:38:06 AM

beowulf6666 said:
lmfao "spill the beans"? ROFL

so yeah i actually went out to upgrade my video card because i thought IT was the issue...when it appears it not

anywho...yeah the 580 was on the same mobo

i should mention i tried different pci-e slots...just in case that was the issue

hope this helps


Hehe, ye then it sounds like the PSU. But before we nail the coffin...
are you using the PSU's 6pin vga power cable or an molex adapter?

I had a 4890 with a bad 6pin molex adapter that gave me issues a while ago.

Oh and if your gfx uses 2x 6pin power plugs(1x psu and 1x molex), try to find a molex plug on a different 12V rail.
September 29, 2012 1:45:26 AM

thanks for the reply

i'll try what you suggest, but yeah...i figure the psu would have to be really REALLY damaged to get any low/off readings...

i will try taking the psu to a different comp but i think it'll have to be one of similar specs otherwise the psu might not get solicited enough to cause the same issue (if it is the psu)

i might have to go out and get another psu...just in case...and rma this one...

i assume there is no software that stress tests the psu...as such ill try these 2 options and get back to you in a few days...

thanks again for the reply...

btw...if i may ask...since the error log says the dx driver is at fault...how did you come to the psu conclusion?

any others that might have more insight...please reply...id be most grateful
September 29, 2012 1:46:28 AM

im using the 6 pin GPU designated connectors (2 of them)
September 29, 2012 2:02:45 AM

beowulf6666 said:
thanks for the reply

i'll try what you suggest, but yeah...i figure the psu would have to be really REALLY damaged to get any low/off readings...

i will try taking the psu to a different comp but i think it'll have to be one of similar specs otherwise the psu might not get solicited enough to cause the same issue (if it is the psu)

i might have to go out and get another psu...just in case...and rma this one...

i assume there is no software that stress tests the psu...as such ill try these 2 options and get back to you in a few days...

thanks again for the reply...

btw...if i may ask...since the error log says the dx driver is at fault...how did you come to the psu conclusion?

any others that might have more insight...please reply...id be most grateful



Well, dx drivers are special software libraries that literally give applications the processing commands the GPU is capable of, thus they do not have to be concerned "how" its done, just, that that it can be done.
In other words, these drivers and even your nVidia drivers communicate to the GPU an a hardware level.
It is like built in instructions(hence DX-9-10-11 etc)

and since you had an gaming issue, dx libraries was loaded and executed, and then you got that error log with the dx errors after a while as the dx drivers was unable to comunicate further with the GPU.

aaaand... since the GPU is new it had to mean it was probably not at fault.And the "crash" was the card stopping itself from drawing to much/little current bringing the PSU into the picture... ok?
September 29, 2012 2:06:06 AM

thx man...ill try to replace the psu and test the old one...

ill let you know what happens.
October 8, 2012 1:06:49 AM

UPDATE

ok...so:

- i got a UPS (for the power outages)
- i bought a corsair ax850 gold psu

and still same error

i tried increasing my page file size (even though its a little hard to believe it could be the problem since i have 16 gigs of ram)

and still same problem...play for a random amount of time and crash

tried the bcdedit /set increaseuserva 2500 command under admin and still same issue

any ideas how i could fix this?

please help
October 8, 2012 3:44:31 AM

UPDATE 2

ok, so after another 5+ h of crawling through forums i couldn't find anything that works

what i decided to do was try to underclock my gpu back to the specs of a regular gtx 670

i used evga precision x, and had no positive result.

FINALLY...i decided to try overclocking an already superlocked card...just bec. after the process of elimination...the only thing left was the ridiculous

so...it worked.

i know...im surprised...but hey...i raised the voltage up just a tiny bit...and yeah...tested it for 2 hours...no crash...no overheating of the VC no abnormal behavior...nothing.

i will try to test it out tmr as much as i can...if it handles a bunch of good hours of testing we can close this thread...but i want to make sure first.

thanks for your help.

BTW...if you think what i did was just a cheap workaround...and u know how i can fix the issue...please reply :) 
October 19, 2012 4:47:25 PM

Soz I was away for a while.

But good to see it was voltages!
HF.
October 21, 2012 2:35:28 PM

hey man, it appears i was wrong...i was really really busy these past couple weeks, whole bunch of new projects popped up.

problem is, the friggin game still crashes...well don't know if its the game that crashes due to the dx driver stopping or vice versa...

what i can tell you for sure, is that i am not the only one with the issue (the battlefield forums are littered with reports of the same issue)

i know this is tiresome...i personally ran out of ideas...

i ran my software from the geek squad (m.r.i 5.7) and couldn't find any issues with any of my hardware...including the videocard which i, even though its brand new, decided to fur mark test test it

any clues?
October 26, 2012 6:14:43 PM

Then I will suggest waiting for a software patch from either BF or GFX vendors.

I found in the past that disabling/lowering some advanced gfx settings and running games "as Administrator" in some games stopped some of the crashes...fiddle around until you find a "playable" scenario. And don't forget to close unused apps while gaming.
October 30, 2012 7:38:54 PM

hey...maybe you can tell me what the stock voltages are for the gtx 670 so i can underclock it myself.

maybe lowering it to whatever specifications will let bf see it as not overclocked.
November 10, 2012 7:58:11 PM

ok so i found my actual stock voltages...i just dont know how to underclock the card to them...

http://www.geforce.com/whats-new/articles/introducing-t...

maybe some1 can help me with that? i tired using precision X but its not an exact numbers thing with that...just slide some dials and percentages...i need to make it work and see if its still causing issues...

please help
January 5, 2013 11:07:21 PM

Dont know if it will work, but try MSI's Afterburner....
!