Solved Possible PSU Failure?

[JackALope52x]

Hi All,

I have been experiencing some issues with my computer recently and have a feeling it is my Power Supply. The main problem is that I have been getting random and sporadic crashes / shutdowns without any reason or warning. Weather I am browsing the internet for 5 minutes or playing a game for an hour, my computer will just crash to a black screen and restart or just shutoff completely. There are other times when I boot the computer and a black screen appears with white lettering informing me that there is no power going to my GPU. Also it seems that my boot time has gotten much longer. Typically my computer boots and is ready to go in about 25-30 seconds, now since I have been having these issues, there are times it can take up to a few minutes.

Again, all this is random with no rhyme or reason. There have been a few times when my computer will boot and run flawlessly until I turn it off, but most of them time not. I was wondering if anyone has ever experienced these problems before and if so, was there a way to fix it outside of buying a new PSU? Now I am not 100% positive that it is my Power Supply so if you have any other ideas please feel free to share. Any help or information to shed some light on my situation would be much appreciated.

Thanks in advance for all the help,
JackALope

 

[Indianatone]

The kind of issues you are experiencing sound like a power supply issue but that is one heck of a supply you have there. Do you have a voltmeter to check the 12 and 5volt supplies from the PSU? You could start by making sure the connections on the modular supply are fully home. I would get the SF diagnostic tool
http://www.sevenforums.com/bsod-help-support/96879-blue-screen-death-bsod-posting-instructions.html
and get together a report for the BSOD team in case it is some hardware issue other than the PSU such as CPU or main board. A time like this is when you need a spare PSU to replace the suspect one with.

 

[linnemeyerhere]

Is the PSU mounted fan side down drawing air in? If so when was the last time it was blown out from both sides?

This could be a good investment:

http://www.newegg.com/Product/Product.aspx?Item=N82E16899261023

 

[JackALope52x]

The kind of issues you are experiencing sound like a power supply issue but that is one heck of a supply you have there. Do you have a voltmeter to check the 12 and 5volt supplies from the PSU? You could start by making sure the connections on the modular supply are fully home. I would get the SF diagnostic tool
http://www.sevenforums.com/bsod-help-support/96879-blue-screen-death-bsod-posting-instructions.html
and get together a report for the BSOD team in case it is some hardware issue other than the PSU such as CPU or main board. A time like this is when you need a spare PSU to replace the suspect one with.

Thanks for the quick reply. Yeah I actually bought this PSU new off of a friend for cheaper than it's 850W counter part, so I couldn't say no. Anyways, no I don't have a voltmeter. I was planning on getting one but I just wanted to see if I could diagnose the problem before I go out and buy one.

Also, can I get a BSOD error report even if the the computer is crashing to a black screen or even just shutting off? It hasn't blue screen yet which is why it never even crossed my mind.

Is the PSU mounted fan side down drawing air in? If so when was the last time it was blown out from both sides?

This could be a good investment:

Rosewill RTK-PST Digital LCD Power Supply Tester - Newegg.com

Hey thanks for the reply linnemeyerhere. Yeah the PSU is mounted fan side down. I do have animals but I don't think dust or hair would be an issue because I clean out my computer every month or so just to avoid this issue.

Also thanks for the link. I am in the market for voltmeter and I had no idea where to start. That tester is just what I need.

 

[Indianatone]

Also, can I get a BSOD error report even if the the computer is crashing to a black screen or even just shutting off? It hasn't blue screen yet which is why it never even crossed my mind.

Run the tool and it will gather together all the files (explained in the link) Windows will report something happened even it there is no blue screen. Post it here so as not to start a new thread and we can ask one of the BSOD guys to take a peek. Like I said "sounds like a PSU " but could be a MB or CPU issue where the data times out and the machine goes awol.

 

[JackALope52x]

Awesome, thanks for the information Indianatone. I will run the test tonight when I get home and post my findings.

Oh I also went ahead and ordered a Power Supply Tester. Whether it is my PSU or not, I figured it would be a good tool to have around on hand.

 

[JackALope52x]

So I was able to run the diagnostics tool and gathered all the dump files. I have attached the zip file for you to look at. How should we go about getting one of the BSOD guys over here to take a look at it? Let me know if any more information is needed.

Thanks,
JackALope

 

[Britton30]

You can also read voltages in BIOS/Monitor. This sounds like a PSU or board issue as suggested by Tony. Have you unplugged and re-plugged the power cables to the GPU?

These may help to pin it down. Get Core Temp or Real Temp to monitor your temps. Just carefully read the installer windows so you don't install any bundled, unwanted software. One uses the InstallIQ download manager which can be uninstaller later.
http://www.alcpu.com/CoreTemp/
Real Temp - CPU temperature monitoring

Graphics card test:
Video Card Stress Test with Furmark ​

Hard drive(s)test(s):

Pick one for your HDD brand or use Seatools.
7 Free Hard Drive Testing Software Tools​

RAM (memory) test:

This will require memtest on a bootable CD or USB stick.
RAM - Test with Memtest86+​

 

[Indianatone]

Have asked the BSOD guys to run the files thru the grinder so to speak. Someone will post the results and if it is inconclusive I would suspect the PSU.

 

[JackALope52x]

You can also read voltages in BIOS/Monitor. This sounds like a PSU or board issue as suggested by Tony. Have you unplugged and re-plugged the power cables to the GPU?

These may help to pin it down. Get Core Temp or Real Temp to monitor your temps. Just carefully read the installer windows so you don't install any bundled, unwanted software. One uses the InstallIQ download manager which can be uninstaller later.
Core Temp
Real Temp - CPU temperature monitoring

Graphics card test:
Video Card Stress Test with Furmark ​

Hard drive(s)test(s):

Pick one for your HDD brand or use Seatools.
7 Free Hard Drive Testing Software Tools​

RAM (memory) test:

This will require memtest on a bootable CD or USB stick.

RAM - Test with Memtest86+​

Thanks for the reply Gary. Yeah I unplugged an re-plugged in all the PSU cables last weekend, along with dusting out the whole tower. My temps have been normal and I don't have any major fluctuations in them which is what I expected.

I will try to run some of the stress tests that you posted but I am almost positive it will cause the PC to crash. It seems that anything that creates a power fluctuation will cause the system to crash. It also seems that it is getting progressively worse. I hope the Power Supply tester comes in soon because I don't want to damage any of my components with the excessive crashing.

 

[JackALope52x]

Have asked the BSOD guys to run the files thru the grinder so to speak. Someone will post the results and if it is inconclusive I would suspect the PSU.

Thanks Indianatone! I really appreciate all your help with this.

 

[x BlueRobot]

[COLOR="Red"]BugCheck 116[/COLOR], {fffffa80098a34e0, fffff88009910f50, [COLOR="SeaGreen"]ffffffffc000009a[/COLOR], 4}

Unable to load image \SystemRoot\system32\DRIVERS\nvlddmkm.sys, Win32 error 0n2
*** WARNING: Unable to verify timestamp for nvlddmkm.sys
*** ERROR: Module load completed but symbols could not be loaded for nvlddmkm.sys
Probably caused by : [COLOR="Red"]nvlddmkm.sys[/COLOR] ( nvlddmkm+8f0f50 )

6: kd> [COLOR="SeaGreen"]!error ffffffffc000009a[/COLOR]
Error code: (NTSTATUS) 0xc000009a (3221225626) - Insufficient system resources exist to complete the API.

The NTSTATUS error code is usually related to pool leaks caused by graphics card drivers, however, reading through the thread and looking at the stack, I think it may be a PSU issue too.

6: kd> [COLOR="SeaGreen"]!stack[/COLOR]
Call Stack : [COLOR="Red"]7 frames[/COLOR]
## Stack-Pointer    Return-Address   Call-Site       
00 fffff8800296fa48 fffff88007c71054 nt!KeBugCheckEx+0 
01 fffff8800296fa50 fffff88007c44867 dxgkrnl!TdrBugcheckOnTimeout+ec 
02 fffff8800296fa90 fffff88007c70e63 dxgkrnl!DXGADAPTER::Reset+2a3 
03 fffff8800296fb40 fffff88007d4003d dxgkrnl!TdrResetFromTimeout+23 
04 fffff8800296fbc0 fffff8000391abae dxgmms1!VidSchiWorkerThread+101 
05 fffff8800296fc00 fffff8000366d8c6 nt!PspSystemThreadStartup+5a 
06 fffff8800296fc40 0000000000000000 nt!KxStartSystemThread+16

The stack indicates that the graphics card driver has attempted to reset the graphics card, but failed to do so within the allocated time interval, and therefore caused a timeout with the following Stop Code 0x116. More Information - Timeout Detection and Recovery (TDR) (Windows Drivers)

 

[Britton30]

Did you have a look at voltages in your BIOS/UEFI?

 

[Indianatone]

I have one more request for you before you replace the PSU. I have had a machine that has been doing weird things since the last NVidia driver came out. A new one dated 21st October just been released. Try that and see if the problem goes away. Unlikely but humour me. I thought my wife's machine had a bad motherboard or cpu so swapped them both out. That machine started to go wrong after running for nearly 3 years a month ago after I upgraded the video driver. FWIW http://www.sevenforums.com/bsod-help-support/308719-tammy-s-machine-bluescreens.html
Free is also cheaper. The suspect driver is this one
GeForce 327.23 Driver

https://forums.geforce.com/default/...ver-has-stopped-working-fix/?offset=9#3949966

 

[JackALope52x]

[COLOR=red]BugCheck 116[/COLOR], {fffffa80098a34e0, fffff88009910f50, [COLOR=seagreen]ffffffffc000009a[/COLOR], 4}
 
Unable to load image \SystemRoot\system32\DRIVERS\nvlddmkm.sys, Win32 error 0n2
*** WARNING: Unable to verify timestamp for nvlddmkm.sys
*** ERROR: Module load completed but symbols could not be loaded for nvlddmkm.sys
Probably caused by : [COLOR=red]nvlddmkm.sys[/COLOR] ( nvlddmkm+8f0f50 )

6: kd> [COLOR=seagreen]!error ffffffffc000009a[/COLOR]
Error code: (NTSTATUS) 0xc000009a (3221225626) - Insufficient system resources exist to complete the API.

The NTSTATUS error code is usually related to pool leaks caused by graphics card drivers, however, reading through the thread and looking at the stack, I think it may be a PSU issue too.

6: kd> [COLOR=seagreen]!stack[/COLOR]
Call Stack : [COLOR=red]7 frames[/COLOR]
## Stack-Pointer Return-Address Call-Site 
00 fffff8800296fa48 fffff88007c71054 nt!KeBugCheckEx+0 
01 fffff8800296fa50 fffff88007c44867 dxgkrnl!TdrBugcheckOnTimeout+ec 
02 fffff8800296fa90 fffff88007c70e63 dxgkrnl!DXGADAPTER::Reset+2a3 
03 fffff8800296fb40 fffff88007d4003d dxgkrnl!TdrResetFromTimeout+23 
04 fffff8800296fbc0 fffff8000391abae dxgmms1!VidSchiWorkerThread+101 
05 fffff8800296fc00 fffff8000366d8c6 nt!PspSystemThreadStartup+5a 
06 fffff8800296fc40 0000000000000000 nt!KxStartSystemThread+16

The stack indicates that the graphics card driver has attempted to reset the graphics card, but failed to do so within the allocated time interval, and therefore caused a timeout with the following Stop Code 0x116. More Information - Timeout Detection and Recovery (TDR) (Windows Drivers)

So it is my graphics card? Is the PSU causing it to malfunction which triggers the crash? My GPU is only 5 months old so I can't imagine it being the one causing all this.

Did you have a look at voltages in your BIOS/UEFI?

I just got home and was able to get these when I booted up the computer. Not sure if it is good or bad, but this is what I found.

CPU Voltage - 1.12V / 1.20V (It would fluctuate back and forth)
3.3V Voltage - 3.296V
5V Voltage - 5.080V
12V Voltage - 12.288V

I don't know if this was the information you were looking for exactly but just let me know if there were any other readings you needed.

 

[JackALope52x]

I have one more request for you before you replace the PSU. I have had a machine that has been doing weird things since the last NVidia driver came out. A new one dated 21st October just been released. Try that and see if the problem goes away. Unlikely but humour me. I thought my wife's machine had a bad motherboard or cpu so swapped them both out. That machine started to go wrong after running for nearly 3 years a month ago after I upgraded the video driver. FWIW http://www.sevenforums.com/bsod-help-support/308719-tammy-s-machine-bluescreens.html
Free is also cheaper. The suspect driver is this one
GeForce 327.23 Driver

https://forums.geforce.com/default/...ver-has-stopped-working-fix/?offset=9#3949966

That is funny because my current driver is the 327.23. I will update to the new driver (331.58) and test it to see if there are any changes. Thanks for catching this!

 

[Indianatone]

Those voltages are perfect. It is normal for the CPU voltage to go up and down as the speed changes or cores go on or off (power saving when idle).

 

[x BlueRobot]

The PSU most likely seems to be causing the GPU to crash.

 

[Indianatone]

The PSU most likely seems to be causing the GPU to crash.

I think so too, but that driver gave me the runaround on one of my machines so its worth a try since it free.

 

[x BlueRobot]

The PSU most likely seems to be causing the GPU to crash.

I think so too, but that driver gave me the runaround on one of my machines so its worth a try since it free.