Random solid colour screen freezes.

[-JT-]

Hi guys!

As I said, I'm getting these totally random solid colour screen freezes.
I have bought a new desktop a few months ago and since the beginning I've been having this.
I have send the desktop back to the company I bought it from. They told me it was a BIOS setting. And when I got it back it was fine for two weeks but soon the problem returned. So it has never been good. And I also like to mention I've never been in the Bios editing the settings before.

I've also had two BSOD the last week but the only info I could read from those where different 2 times.
-Bad pool header the 1st time
-Reference by pointer the 2nd time

But these weren't as annoying as the colour freezes.

I've been watching my temps and just now I got a colour freeze after booting. With temps low at 25°. And I'm kidding you not, sometimes after I give the desktop a little tap the screens jump back to normal. But this doesn't always work.

After checking Toms hardware I saw that this might be a drivers issue, as the crashreports tell aswell. But I've already did a fresh install of the drivers with DDU and the problem still persists.

Here are my systemspecs.
Intel(R) Core(TM) i7-4790 CPU @ 3.60GHz
GIGA-BYTE GA-Z97-D3H
GIGA-BYTE GeForce GTX 780 3GB
CRUCIAL Ballistix Tactical 2x8GB
SEAGATE Barracuda 2T HDD
SAMSUNG 840 series Evo 250GB SSD
Windows 7 Home premium 64bit.

This is one of many crashreports


Source
Windows

Overzicht
Videohardwarefout

Date
‎21/‎02/‎2015 18:16

Status
Niet gerapporteerd

Description
Een probleem met de videohardware heeft ervoor gezorgd dat Windows niet meer goed werkt.
(A problem with the videohardware has stopped windows from working properly)

Probleem met handtekening
Naam van probleemgebeurtenis: LiveKernelEvent
Versie van besturingssysteem: 6.1.7601.2.1.0.768.3
Landinstelling-id: 2067

Bestanden die helpen bij het beschrijven van het probleem
WD-20150221-1816.dmp
sysdata.xml
WERInternalMetadata.xml

Extra informatie over het probleem
BCCode: 117
BCP1: FFFFFA80141B4010
BCP2: FFFFF88005C3A978
BCP3: 0000000000000000
BCP4: 0000000000000000
OS Version: 6_1_7601
Service Pack: 1_0
Product: 768_1


So I'm a bit in the dark here guys. Hope you guys can shine some light on my problem.
Cause I paid quite a few euro's for this desktop and it has givin me alot of shite already.

Let me know if I can give any extra information.

Greets JT!

 

[johnbl]

bugcheck 0x117 means the video driver too over 30 seconds to respond and windows figured it had failed and called a bugcheck.

normally, you would
- Stop any overclocking (software)
- do not overclock in BIOS
- update your graphics drivers
- disable unused high definition audio support by the graphics card (use windows control panel)
- confirm your power supply is rated correctly for your system (600w)
- confirm your GPU is not overheating

- update your BIOS particularly if it is older than july 2014 and you are using a newer low power processor like the one you have.

-if you have to, try and underclock the GPU by 50 or 100 MHz

you can also, change your memory dump type to kernel or full, then run verifier.exe and set flags for deadlock detection. Reboot and wait for the next memory dump and put it on a server and post a link.
how to run verifier.exe:
http://www.sevenforums.com/crash-lockup-debug-how/65331-using-driver-verifier-identify-issues-drivers.html

use verifier.exe /reset to turn off verifier after you have produced the bugcheck.

Here are two more suggestions that sound silly but I have debugged cases to these causes:
- update BIOS, CPU chipset drivers, and USB 3.0 drivers. I have seen directx drivers chained as the last in a interrupt chain behind a old buggy USB driver. The old USB driver got confused and prevented directx from working.
- update your ethernet driver, Nvidia drivers now enable streaming and expose bugs in default ethernet drivers. You want a version that is dated in 2014 or better or you could disable the nvidia streaming software if you don't use it.
(I disabled mine)
- if you updated your graphic card, be sure you go into BIOS and reset it to defaults. This will force a rebuild of the database that the BIOS sends to windows. (all the hardware and interrupt, DMA channel settings,...)
Failure to do this can cause your card not to work because of conflicts.
- since you mentioned tapping your case can fix your screen problem, sometimes fans can get clogged with dust and get stuck or run slowly and your card overheats. Blow the dust out of the GPU, CPU and powersupply. Also, check for poor cable connections, cables can connect and disconnect with thermal expansion and slight vibrations. These can happen several times a second and really cause very screwy problem. (also check the video cable)

 

[-JT-]

Wow, thanks John for your quick and accurate response 😀

I can outrule overclocking, cause I don't do that.
But the thing you said that caught my eye was the last in your list.

Ever since bought this desktop, which has a NZXT Phantom 410 White Case.
The top fan had been making this ticking noise.
I tried to remove it once and noticed a cable was against the fan itself.
But the ticking noise soon returned when I made the fans spin at max, and since you said that might also cause it, I've unplugged the topfan totally.

Might it be broke? I'm gonna try this for a few days, and if the colour freezes don't return I guess I'm buying a new topfan. Or contacting the company that assembled my desktop.

I'll let you know if any more freezes happen in the forthcomming week.

Great thanks to you already John for the big help!

 

[-JT-]

bugcheck 0x117 means the video driver too over 30 seconds to respond and windows figured it had failed and called a bugcheck.

normally, you would
- Stop any overclocking (software)
- do not overclock in BIOS
- update your graphics drivers
- disable unused high definition audio support by the graphics card (use windows control panel)
- confirm your power supply is rated correctly for your system (600w)
- confirm your GPU is not overheating

- update your BIOS particularly if it is older than july 2014 and you are using a newer low power processor like the one you have.

-if you have to, try and underclock the GPU by 50 or 100 MHz

you can also, change your memory dump type to kernel or full, then run verifier.exe and set flags for deadlock detection. Reboot and wait for the next memory dump and put it on a server and post a link.
how to run verifier.exe:
http://www.sevenforums.com/crash-lockup-debug-how/65331-using-driver-verifier-identify-issues-drivers.html

use verifier.exe /reset to turn off verifier after you have produced the bugcheck.

Here are two more suggestions that sound silly but I have debugged cases to these causes:
- update BIOS, CPU chipset drivers, and USB 3.0 drivers. I have seen directx drivers chained as the last in a interrupt chain behind a old buggy USB driver. The old USB driver got confused and prevented directx from working.
- update your ethernet driver, Nvidia drivers now enable streaming and expose bugs in default ethernet drivers. You want a version that is dated in 2014 or better or you could disable the nvidia streaming software if you don't use it.
(I disabled mine)
- if you updated your graphic card, be sure you go into BIOS and reset it to defaults. This will force a rebuild of the database that the BIOS sends to windows. (all the hardware and interrupt, DMA channel settings,...)
Failure to do this can cause your card not to work because of conflicts.
- since you mentioned tapping your case can fix your screen problem, sometimes fans can get clogged with dust and get stuck or run slowly and your card overheats. Blow the dust out of the GPU, CPU and powersupply. Also, check for poor cable connections, cables can connect and disconnect with thermal expansion and slight vibrations. These can happen several times a second and really cause very screwy problem. (also check the video cable)

I also noticed that yesterday my GPU temp stayed lower then it did before.40-50° and reaching 65°max after a few hours of gaming. I sooo hope it was the topfan.
It had been acting up ever since I got it.
Sometimes it didn't spin after a boot until I switched the fan controller switch.

Grts and thanks JT