1. Computer problem? Tech Support Guy is completely free -- paid for by advertisers and donations. Click here to join today! If you're new to Tech Support Guy, we highly recommend that you visit our Guide for New Members.

Solved: So I'm thinking my gpu or psu is dying, but which?

Discussion in 'Hardware' started by seconds, Oct 26, 2014.

Thread Status:
Not open for further replies.
Advertisement
  1. seconds

    seconds Thread Starter

    Joined:
    Apr 27, 2009
    Messages:
    16
    Long story short. On my old computer I had this problem that my computer restarts itself randomly when doing stuff (most often just surfing the web). About 6 months ago I got new computer and got all new parts except psu and gpu.

    I'm still getting random problems. But little different. On my old pc it was either pc restarting or bluescreen. On my new pc the pc just goes into vegetative state and I can hear buzzing sound from the audio speakers. Only way to get it back up is pressing physical reset button on the pc case. The pc basically never crashes when playing games. The only pattern I have figured out is that the computer is on when it crashes (trying to be funny).

    I don't have any minidumps. I'm not getting any error messages. I've searched all device managers and logs and there is nothing to suggest anything even happened. Other problem is that the problem is super random. Sometimes it can be months between crashes and then some day the pc crashes 6 times in one day (like today).

    Today it was different though. The pc crashed. I reset. Computer restarts again. But this time after just few seconds the computer freezed again on the desktop. Mouse just stopped and I heard short buzz from the speakers. Restarted again. On 3rd try windows shows some kind of error message about windows not starting/closing (not sure which it was) properly. It offers selection of couple different startup methods (normal, and some kind of safestart). I choose normal. Again crash. I reboot and once the computer starts I immediatelly open task explorer to see if some program is crashing the computer. This time it did not crash and here we are. But in all previous times all I had to do was to reboot and continue where I left off. Crashing on start is something new.

    Because the problem is kinda similar with the problem I had with my old computer I'm thinking it is either the gpu or psu. I'm totally broke and I don't have any friends to borrow any parts. So I don't want to get new psu just to find out it is the gpu that is the problem. Memtest runs fine. Temperatures are fine.

    The two old parts are the psu and gpu:
    psu: corsair 620hx. Bought in 2008 october.
    gpu: powercolor ati radeon 6950 1gb. Has 3rd party heatsink. Bought in december 2011.


    And tsg info:
    Tech Support Guy System Info Utility version 1.0.0.2
    OS Version: Microsoft Windows 8.1, 64 bit
    Processor: Intel(R) Core(TM) i5-4440 CPU @ 3.10GHz, Intel64 Family 6 Model 60 Stepping 3
    Processor Count: 4
    RAM: 8120 Mb
    Graphics Card: AMD Radeon HD 6900 Series, 1024 Mb
    Hard Drives: C: Total - 228407 MB, Free - 137941 MB; E: Total - 1897854 MB, Free - 655776 MB;
    Motherboard: MSI, H87M-E33 (MS-7817)
    Antivirus: Bitdefender Antivirus Free Edition, Updated: Yes, On-Demand Scanner: Enabled
     
  2. Tanis

    Tanis

    Joined:
    May 29, 2006
    Messages:
    3,969
    First Name:
    Shane
    What sort of temps are you getting that are 'OK' and what software are you using to check them?

    I assume your not overclocking anything? If you are change everything back to stock speeds for testing.

    I would certainly be looking at the PSU and GPU as possible culprits as they are older parts. The PSU is a good quality unit but it is 6 years old now and they all degrade (by varying degrees) over time.
     
  3. seconds

    seconds Thread Starter

    Joined:
    Apr 27, 2009
    Messages:
    16
    I don't actually know the exact temps. When I built the computer I did run some benchmarks to see if it overheats and it didn't. Haven't bothered with it since.

    As for the actual parts. There is no overclock. The gpu has scythe musashi heatsink and has heatsinks for the vrms. The heatsink is older than the gpu as it was on my previous gpu as well. The cpu is stock i5 4440 and has Katana 4 cooler which should be enough as well. Whole thing sits in antec p182 case and the case has 4 120mm fans. Wire management is tidy but I'm not ocd about it. Overheating would be really really odd especially considering that the pc doesn't never crash when playing games.
     
  4. Tanis

    Tanis

    Joined:
    May 29, 2006
    Messages:
    3,969
    First Name:
    Shane
    Have a look in the BIOS as find the +12v, +5.5v and +3v readings, post them all in a reply here. Can't tell you where they would be as it varies from one BIOS to the next. You may also find some temperature info in there also.

    Just because something didn't overheat when you built it doesn't mean it isn't overheating now, there are so many variables that can impact it. Having said that, I would lean away from temps a little if it doesn't crash when gaming as you said.

    Have you confirmed that all your drivers are fully compatible with Windows 8.1?
     
  5. crjdriver

    crjdriver Moderator

    Joined:
    Jan 2, 2001
    Messages:
    37,749
    I looked at the msi site and that board does support win 8.1. It appears that there is some monitoring software on the support page for that board. Download that and install; post the requested info; ie temps and voltages.

    The corsair hx series are fine pw supplies however it is now 6yrs old and it has pretty much reached its useful life. Most pw supplies last 5~7yrs at most.
     
  6. seconds

    seconds Thread Starter

    Joined:
    Apr 27, 2009
    Messages:
    16
    Uhh, which one do you want me to install? :) I did not find anything that was any kind of monitoring app on the msi site.

    It would be strange if the board did not support windows 8.1 when in the bios it has specific options for "windows8/windows 8.1" (both are mentioned) for the energy saving thingies. My psu does not support those so I have turned them off (I think they were off when I got the board already).

    I did not find 3.3v. Here are the voltages I did find:
    cpu core: 1,728
    system/5V: 5,040
    system/12V 12,056

    I also had gpu-z running yesterday and logging all kinds of stuff when it crashed. Here is the last line of that log:
    Date 2014-10-26 12:23:56
    GPU Core Clock [MHz] 500.0
    GPU Memory Clock [MHz] 1250.0
    GPU Temperature [°C] 40.0
    Fan Speed (%) [%] 24
    Fan Speed (RPM) [RPM] -
    GPU Load [%] 0
    GPU Temp. #1 [°C] 41.0
    GPU Temp. #2 [°C] 47.5
    GPU Temp. #3 [°C] 46.5
    Memory Usage (Dedicated) [MB] 0
    Memory Usage (Dynamic) [MB] 152
    12V [V] 11.88
    VDDC [V] 1.000
    VDDC Current [A] 11.0
    VRM Temperature 1 [°C] 0
    VRM Temperature 2 [°C] 26

    There is no variation in those numbers during the whole time.

    Computer has been on now for 6 hours and I have done some excel work and 3d modelling. No crashes. In bios my cpu temp was 43celcius and motherboard/system temp is 33celcius. I can do some stress tests if necessary but I don't really see the point when the pc never crashes when playing games. Usually when it crashes it is when I'm browsing youtube, listening some music on the background and doing random stuff.

    My ssd disk temp is 25celsius and my hd is 28 celsius.
     
  7. Tanis

    Tanis

    Joined:
    May 29, 2006
    Messages:
    3,969
    First Name:
    Shane
  8. seconds

    seconds Thread Starter

    Joined:
    Apr 27, 2009
    Messages:
    16
    Hmm, I'm not sure if I want to install that stuff to be honest :). With all that cloud storage systems and online connectivity it smells more like spyware than any kind of useful program.
     
  9. seconds

    seconds Thread Starter

    Joined:
    Apr 27, 2009
    Messages:
    16
    The GPU temp maxes out at 69celcius under load according to gpu-z. Then there are gpu1, gpu2 and gpu3 which max out at 69, 76 and 75. I just spent some time playing the games I normally play and then checked the max values.
     
  10. crjdriver

    crjdriver Moderator

    Joined:
    Jan 2, 2001
    Messages:
    37,749
    OK, if you want to find the cause of your problem, you are going to have to install some type of monitoring software. MSI seems to have a program available on their site. I never build with msi boards so I really cannot tell you which one; YOU are going to have to read and find out. There should be a description in your mb manual under software.

    Up to you either install monitoring software [either from msi or you can try hwmonitor] or live with your problem.
     
  11. seconds

    seconds Thread Starter

    Joined:
    Apr 27, 2009
    Messages:
    16
    I just don't understand what is it exactly that you want me to monitor? What value do you want to know? How does mb monitor software help me with psu or gpu problem? Or are we trying to figure out if the mobo is the problem?

    I already have gpu-z, cpu-z, hwmonitor, crystaldisk and the tools from ati that show the gpu temps. I can go to bios to doublecheck those values. I can upgrade the bios (if it is not the latest version).
     
  12. crjdriver

    crjdriver Moderator

    Joined:
    Jan 2, 2001
    Messages:
    37,749
    You need to post the following; cpu temp, system temp, 12V, 5V, and 3.3V values. Post both a no load condition and with 100% load condition. Use one of the following to achieve a 100% load; prime95, OCCT, or IBT [all are free downloads]

    I like prime95 however any of the three will work.

    If your pw supply is failing, it will show a voltage drop on one or more rails when fully loaded.
     
  13. seconds

    seconds Thread Starter

    Joined:
    Apr 27, 2009
    Messages:
    16
    I'm sorry this is taking so long but I'm having a problem with the newer ati drivers. Some setting is off and it has made everything basically grayer and some things blurry. I'll get back to this once I have figured that out...
     
  14. seconds

    seconds Thread Starter

    Joined:
    Apr 27, 2009
    Messages:
    16
    Ok, did this with prime95. I logged the temps and voltages with different apps (gpu-z, hwmonitor). I did run the prime95 until I saw my cpu temp peak and then let it run for another 10 minutes.

    Screenshot:
    [​IMG]

    All I can see is that VCCIN drops from 1,728V to 1,712V which is about 1% change.
     
  15. crjdriver

    crjdriver Moderator

    Joined:
    Jan 2, 2001
    Messages:
    37,749
    Hwmonitor is almost useless; I really doubt your 12V rail is at 0 AND vdimm is running 2.2V. From what I can tell [which is not much since hwm is useless] Your temps are ok and the gpu monitoring app shows 12V as ok [again no one knows if that is accurate]

    At this point, I would say your video card is failing however no one can tell you for sure. This is why you were told to use the monitoring software from msi; it is designed for your board and will give much better readings.
     
  16. Sponsor

As Seen On
As Seen On...

Welcome to Tech Support Guy!

Are you looking for the solution to your computer problem? Join our site today to ask your question. This site is completely free -- paid for by advertisers and donations.

If you're not already familiar with forums, watch our Welcome Guide to get started.

Join over 733,556 other people just like you!

Thread Status:
Not open for further replies.

Short URL to this thread: https://techguy.org/1136157

  1. This site uses cookies to help personalise content, tailor your experience and to keep you logged in if you register.
    By continuing to use this site, you are consenting to our use of cookies.
    Dismiss Notice