Computer turns off intermittently

This is a discussion on Computer turns off intermittently within the Tech Board forums, part of the Community Boards category; This just started last night. I was debugging some camera code related to a shaking issue and on my last ...

  1. #1
    Super Moderator VirtualAce's Avatar
    Join Date
    Aug 2001
    Posts
    9,596

    Computer turns off intermittently

    This just started last night. I was debugging some camera code related to a shaking issue and on my last debug attempt the entire computer just shut off. I went to bed thinking that somehow I triple faulted the CPU and all was ok.

    When I got home from work I turned the computer on and wasn't paying much attention to it. When I looked back up my computer was off. Booted again and it blue screened. Went to safe mode and removed some startup programs via msconfig and set the system so it would not reboot if it encountered a system error. Viewed the event logs...nothing. Rebooted. Everything was fine until I went into Mercenaries 2. The minute it went it to the actual 3D the computer shut down.

    So I'm thinking power issues or video card or virus. Scanned for malware and viruses with Spybot S&D and Antivir. Nothing. Tested Direct3D with dxdiag thinking that if the card was prob. drawing more power in 3D which might make a weak PSU just shut down that dxdiag might cause the system to shut down during a test. All tests passed. Fired up Mercenaries 2 and played for hours and hours - no temp issues (GPU 70 C max) and nothing crazy.

    So....what the heck is going on? Is it possible I have something in the PSU that is going out? If so I thought more heat (IE: more up time) would reveal the problem as opposed to it revealing itself when its cold.

    The PSU is a 1000W Antec and it's less than a year old.
    Last edited by VirtualAce; 12-15-2009 at 10:08 PM.

  2. #2
    (?<!re)tired Mario F.'s Avatar
    Join Date
    May 2006
    Location
    Portugal
    Posts
    7,487
    Just a stab in the dark, but I suppose that's what you expect us to do
    I'm inclined towards your video card. Surely seems like it.

    - Do you overclock the GPU? Could be some power issue over that.
    - Would pay to open the box, remove the card and firmly plant it again (think I read about it somewhere, someday).
    - Looking back at all the GPU settings could reveal something out of the ordinary.
    - Stress testing your GPU with FurMark might help you discard the videocard as the problem. Or confirm it.
    The programmer’s wife tells him: “Run to the store and pick up a loaf of bread. If they have eggs, get a dozen.”
    The programmer comes home with 12 loaves of bread.


    Originally Posted by brewbuck:
    Reimplementing a large system in another language to get a 25% performance boost is nonsense. It would be cheaper to just get a computer which is 25% faster.

  3. #3
    Registered User
    Join Date
    Dec 2006
    Location
    Canada
    Posts
    3,183
    Failing memory? have you tried memtest?

  4. #4
    Super Moderator VirtualAce's Avatar
    Join Date
    Aug 2001
    Posts
    9,596
    • Card is overclocked from the factory - it's a BFG Tech - however I'm only running one on a 1000W power supply. I should have heaps of power waiting to be used with my system configured as it is.
    • Re-seating the card definitely could not hurt anything - it is a beast and I'm not using screws so maybe the tension on the fastener is dying out on me.
    • GPU settings are fairly normal - I don't mess with them too often - no aftermarket O/C or any other custom settings
    • Stress testing sounds like a good plan

    Happened again tonight while playing Silent Hunter 4. Was firing torps at a nice fat destroyer and the computer just shut off. I turned the power switch on the PSU off and let it sit for a bit just in case some circuit inside got into a weird state (if even possible). So far so good but I'm not feeling confident about the system right now.

    My PSU has a 5-year warranty on it and is supposed to be able to run 2 to 3 graphics cards on it. I'm running one card, a DVD RW, two hard drives, a 3.5 floppy drive, some internal fans, and that's about it. At times it is powering a usb flight stick, gamepad, or wheel but usually never all at the same time.

    My 5V reading is sitting at 5.11 which seems a bit off since the 12V is sitting at about 12.03 which seems more normal. What is the accepted differential between the actual and the desired voltages?

    Voltages as reported by PC Probe 2:
    12V - 12.03
    5V - 5.11
    DDR2 - 2.16
    HT - 1.26
    SB - 1.25
    BR 1.22
    VDDA - 2.53
    DDR2 Term. - 1.06
    VDDNB - 1.55 - 1.57 (Probe has flagged this since day one but I think it's b/c Probe doesn't quite understand what VDDNB should be for the AMD Phenom 940 BE)

    CPU: 44c
    MB: 42c
    Case: 28c
    CPU fan speed: 5869 RPM

    Interesting note is that my PSU comes with:
    • OCP - Over current protection
    • OVP - Over voltage protection
    • SCP - Short circuit protection
    • UVP - Under voltage protection


    With all of this protection it seems to me that the PSU may be revealing another problem in the system rather than it being defective. Of course anything is possible. But I really spared no expense on the components in this new system. The only thing I did not get was an Intel I7 and instead went with AMD's Phenom and the ASUS Crossfire 2 mobo.

    Failing memory? have you tried memtest?
    Also a distinct possibility and it has crossed my mind. I have not ran memtest on it yet but I will here soon if it continues. I've had bad sticks before but they normally cause other issues like random faults in systray apps, etc, etc. I have 4 Gig and I'm only ever using about 50% of it and that's when inside a huge game world. It could be that the game reaches that one area of one stick and the data comes back incorrect which would cause all types of problems. The RAM is Corsair Dominator which is also O/C from the factory. Not likely that RAM that has had almost a year of power cycles on it would suddenly fail....but it could happen.

    The blue-screen may be unrelated to the issue or might have been caused by improper shutdown. If it happens again I'll post the error codes here.
    Last edited by VirtualAce; 12-16-2009 at 12:07 AM.

  5. #5
    Registered User
    Join Date
    Dec 2006
    Location
    Canada
    Posts
    3,183
    I THINK the voltage tolerance for ATX is 5% for all the lines, so they should all be fine.

    The readings may not be accurate, though. It would be best if you can check with a multimeter. The CPU power plug (4 pins thing) and the ATX power plug (20/24 pins) should all have exposed metal contacts on top. But then I think the motherboard does its own voltage regulation for the CPU (vcore), so there's probably no way to check that.

    There could be ripple, too, but you'll need an oscilloscope to check that... (I've always wanted one)

    That said, the Antec PS is probably close to the last thing I'll suspect. The power rating is WAY more than enough, too.

    I run my Core 2 Duo + 9600 GT (both HEAVILY overclocked) on an Antec 380W. Still running fine after 2 years. Good power supplies won't give you problems when you try to draw too much current, too. They should just shut down.

    USB devices can't draw too much current. They are limited to 500mA at 5V, that's just 2.5W. Most don't draw more than 100mA.

    CPU fan speed: 5869 RPM
    Wow. That's one fast fan.

  6. #6
    Super Moderator VirtualAce's Avatar
    Join Date
    Aug 2001
    Posts
    9,596
    Wow. That's one fast fan.
    Indeed and it looked odd to me as well. Perhaps it's time for a yearly cleaning.

    They should just shut down.
    It is I just don't know why. I should have plenty of power left over to cover any spikes. I prob have in the neighborhood of over 500W left over.
    It's a very high dollar well made PSU so I really don't think that is it but I am going to call Antec and see what they think. After that it's on to BFG and then last but not least...ASUS. I really hate talking to ASUS. They used to play the blame game a lot. Hopefully that has changed.

    Regardless of what it is all this stuff is under at least 3-year warranties so one or more of the manufacturers owe me some new hardware.
    Last edited by VirtualAce; 12-16-2009 at 12:29 AM.

  7. #7
    Registered User
    Join Date
    Dec 2006
    Location
    Canada
    Posts
    3,183
    The warranties don't include diagnosing, so unless you can isolate the problem yourself to one component, they probably won't replace it. It can be very hard, especially if the problem is intermittent. I'm in the middle of trying to figure one out myself (I get memory errors, which may be caused by memory, northbridge, or CPU cache, intermittently, and seemingly only at night, when there's a full moon). Took me days already. I can't use my warranty until I can isolate the problem. Talking to manufacturers probably won't help since they will just be blaming each other.

    In your case, I would suspect memory first, though. Are you using Crucial memory by any chance? They make a lot of high end memory and the last batch has a VERY high failure rate. If you turn off the automatic restart thing, the bluescreens should give you some clue, too (if it's the same driver every time, or memory access violations, etc).

  8. #8
    Super Moderator VirtualAce's Avatar
    Join Date
    Aug 2001
    Posts
    9,596
    I'm using Corsair memory. I've had a high failure rate among some of their modules but I did test these right after I installed them. Another memory test is in order for sure. First thing I will do is open the case and make sure nothing has been jostled around and that everything is still seated correctly.

    I'll let you know more details as I investigate further.

    The warranties don't include diagnosing, so unless you can isolate the problem yourself to one component, they probably won't replace it.
    Most companies will let you RMA the component and they will test it. If it is not faulty then they will send it back and will not replace it.

    EDIT:
    Noticed I've been getting bugcheck 0x50 quite often in my event viewer. It says the computer has been rebooted due to bugcheck 0x000000050 - which is page fault in non paged area.

    Here is a partially corrupted dump: Most of the time the system does not have time to do a crash dump. It's looking like a memory issue.
    Last edited by VirtualAce; 03-12-2011 at 10:41 AM.

  9. #9
    C++まいる!Cをこわせ! Elysia's Avatar
    Join Date
    Oct 2007
    Posts
    22,663
    This sounds more like some driver problem, a driver that somehow try to access some invalid type of memory or such. I don't know how you get a page fault in a non-paged area, though!
    But yes, if not a driver problem, it sounds like memory problem.
    Good luck in diagnosing that.
    Quote Originally Posted by Adak View Post
    io.h certainly IS included in some modern compilers. It is no longer part of the standard for C, but it is nevertheless, included in the very latest Pelles C versions.
    Quote Originally Posted by Salem View Post
    You mean it's included as a crutch to help ancient programmers limp along without them having to relearn too much.

    Outside of your DOS world, your header file is meaningless.

  10. #10
    and the hat of sweating
    Join Date
    Aug 2007
    Location
    Toronto, ON
    Posts
    3,545
    You could try either installing an OS in another partition or booting Linux from a bootable CD and see if you still have the problem. That would rule out viruses/driver problems.
    "I am probably the laziest programmer on the planet, a fact with which anyone who has ever seen my code will agree." - esbo, 11/15/2008

    "the internet is a scary place to be thats why i dont use it much." - billet, 03/17/2010

  11. #11
    Registered User
    Join Date
    Dec 2006
    Location
    Canada
    Posts
    3,183
    If the problem appears frequently enough, an easy way to test memory is to just take out one stick at a time. Or run memtest.

  12. #12
    Registered User
    Join Date
    Aug 2006
    Posts
    99
    Actually, I'd only RUN one stick at a time, and memtest on it. Prove you have all good sticks, then start trying them in combination to see if it's an actual bus timing problem

    It could still be the PSU though, no matter what PCProbe says... I had one fail after two years of flawless service, but only on one 12V rail (and I didn't know it was even a dual rail supply, as it was not advertised as such). That caused very odd shutdowns indeed.

  13. #13
    Super Moderator VirtualAce's Avatar
    Join Date
    Aug 2001
    Posts
    9,596
    Well I checked most of the memory but did not do a stick by stick check. Memtest86 was unable to test all of it since I have 4 GB. It said it could not address all of it which is odd since 4GB clearly fits into an unsigned int.

    Anyways all tests passed. I uninstalled my wireless N NIC and reinstalled it to see if the bugcheck errors in event viewer were due to the card. They were but I'm not sure I fixed them.

    So far after having shutoff the PSU last night and waiting for 30 secs to 1 min the system has not shut down by itself since I did that. Played some games tonight for several hours and did all kinds of debugging and programming and it never shut down.

    Maybe the PSU got into a flaky state due to some of the recent power outages we've had around here.

  14. #14
    Malum in se abachler's Avatar
    Join Date
    Apr 2007
    Posts
    3,189
    It's entirely possible that you just had some surface oxidation on one of the memory chips and by removing and reinserting it, you scratched it off. I've seen that problem several times before, it'll drive you nuts until you realize whats going on. Just use a super fine emery board to very lightly abrade the surface of the contacts and you will be fine. A QTip dipped in alcohol also works great, remember you want to remove as little material as possible. The layer of oxidation doesn't have to be visible for it to cause issues.
    Last edited by abachler; 12-17-2009 at 01:30 AM.
    Until you can build a working general purpose reprogrammable computer out of basic components from radio shack, you are not fit to call yourself a programmer in my presence. This is cwhizard, signing off.

  15. #15
    {Jaxom,Imriel,Liam}'s Dad Kennedy's Avatar
    Join Date
    Aug 2006
    Location
    Alabama
    Posts
    1,065
    An old (now dead) Radio Engineer told me the best way to clean contacts is with a pencil eraser. This has ALWAYS worked for me.

Page 1 of 2 12 LastLast
Popular pages Recent additions subscribe to a feed

Similar Threads

  1. Is anyone ever a computer scientist outside of college?
    By indigo0086 in forum A Brief History of Cprogramming.com
    Replies: 25
    Last Post: 08-31-2007, 09:26 AM
  2. Starting C++ & Computer Science Major
    By learning C++ in forum C++ Programming
    Replies: 8
    Last Post: 12-26-2003, 11:17 AM
  3. upgrade computer ?
    By teeyester in forum Tech Board
    Replies: 4
    Last Post: 10-11-2003, 09:51 PM
  4. Regarding Undergraduate Computer Majors
    By UnregdRegd in forum A Brief History of Cprogramming.com
    Replies: 11
    Last Post: 10-04-2003, 11:55 AM
  5. computer sex
    By Unregistered in forum A Brief History of Cprogramming.com
    Replies: 5
    Last Post: 09-17-2001, 07:09 PM

1 2 3 4 5 6 7 8 9 10 11 12 13 14 15 16 17 18 19 20 21