+3.3v problem

Discussion in 'General Hardware' started by Deuslunea, Oct 17, 2008.

  1. Deuslunea

    Deuslunea Geek Trainee

    Likes Received:
    0
    Trophy Points:
    0
    Hey folks.

    I'm experiencing a total system shutdown whenever I stress it for a little while. I might be posting this in a wrong forum, but I'm pretty much at a loss of what is causing it.

    I've recently formatted the system and I think I've installed the same software and drivers I usually operate with. The only difference is that I upgraded the OS from XP Pro SP2 to XP Pro SP3 at the same time, but I doubt that it is what's causing the crashes.

    The nature of the crash itself (black screen, the reboot) lead me to suspect that the system overheated. I checked the fans, but they seem to be running as intended. Then I tried using PC Probe 2 to monitor the system, which came up with the +3.3v malfunction. I have no idea what this means, but it goes as low as 2.8v which is when the program notifies me that something is wrong.

    As I said earlier, the crashes only occur when the system is being stressed. If I shut down whatever I am doing when the notification sounds and return to the desktop, the voltage slowly returns to a stable 3.3v and thus avoid crashing.

    Here are some of my specs. If more are needed, let me know. I'm no expert on these things, but I'll do my very best.

    Motherboard: ASUSTeK Computer INC. - P5WD2
    Physical memory: 2gb SDRAM
    Processor: Pentium 4 3GHz
    Hard drive: Maxtor 6L200M0
    Graphics card: Radeon X1800 Series

    Again, I'm sorry if I'm posting in the wrong forum or anything. And I really appreciate any help or advice you have to offer. Thanks in advance.
     
  2. Swansen

    Swansen The Ninj

    Likes Received:
    0
    Trophy Points:
    36
    To me it seems like your 3.3 rail is just dieing, how old is your computer/power supply.
     
  3. Big B

    Big B HWF Godfather

    Likes Received:
    145
    Trophy Points:
    63
    Well, I wouldn't worry about the PSU just yet. Software monitoring is notoriously inaccurate because of the way it reads your voltages. It reads them off a sensor on the motherboard. This sensor isn't right when the voltage runs into the motherboard, and as the juice flows through it, the readings drop after running through all the resistors and other components. If you want an accurate reading, you'll need to measure it straight from the PSU itself with a multimeter. I'm not saying software is worthless, but don't put a ton of stock in it and sweat over it immediately. Let's investigate the system.

    Where I would start is looking at the CPU temps. Asus PCProbe isn't that great when it comes to monitoring, and as with voltage monitoring, software monitoring only gets you so far. However, it might help clue you in to a potential heating issue.
    If it's too high, then you'll want to first make sure the CPU heatsink is properly installed and that the case airflow is enough and installed properly.
    Heat rises, so you want your intake at the bottom (and sides if available) and exhaust on the back (and top, if available).

    Next, you'll want to test your memory for errors. Memtest86 is one thing that's great to run. You'll need to download it for a floppy and let it create a bootable floppy, or a CD .ISO file and burn it to disc. When you're done with that, boot off said disc. Let it run for a minimum of one hour, but preferably overnight. You want 0 errors. If there are, you may have a bad stick of RAM. If you do, you'll want to isolate the problem and test each stick individually.
     
  4. Deuslunea

    Deuslunea Geek Trainee

    Likes Received:
    0
    Trophy Points:
    0
    Sorry about the respond time, my ISP has pulled my leg once again. Anyways...

    @Swansen:
    Most of the system is a few years old, but the power supply in about six months.

    The hardware and tower setup hadn't been altered when the problem first occurred. I've checked that everything was in its place and cleaned the heat sink and fans. I also removed the side covers on the tower, but even with the extra intake, the difference was hardly noticeable.

    In addition to the other observations, it would seem that the ram also overheats. At least that's what PC-Probe tells me. I've only seen this once though. I'm going to run the memtest86 program tonight and post the result tomorrow.

    I'm not entirely sure about this, but I seem to remember that my CPU fan used to crank up the RPM when I stressed the systems. In any case, it always runs at the same speed as it is now. Might that indicate a driver issue? I'm pretty sure I've installed every driver there is for my motherboard, except for the BIOS updaters.

    Thanks for the input so far. Best regards, Deuslunea
     
  5. Deuslunea

    Deuslunea Geek Trainee

    Likes Received:
    0
    Trophy Points:
    0
    The Memtest found no errors.
     
  6. Big B

    Big B HWF Godfather

    Likes Received:
    145
    Trophy Points:
    63
    BIOS isn't the same as drivers, though sometimes it does affect stability. As a general rule, if the BIOS doesn't address an issue you have or add support (typically CPU microcode updates) for new hardware, then I would pass.

    You mentioned the RAM overheating. If it were a problem, you should see some errors in Memtest, especially if the +3.3V line is dropping, since the RAM feeds of that rail.

    What PSU do you currently have?
     
  7. Deuslunea

    Deuslunea Geek Trainee

    Likes Received:
    0
    Trophy Points:
    0
    Upon closer inspection it would seem that the RAM aren't what's overheating, it's the motherboard. Sorry. Also, the Memtest ran for quite a long time, so it should have made a complete test.

    I have a 400w NorthQ model 4001EXT-USB PSU. It's only been in use for about half a year, so I doubt that's what causing it.

    Thank you for the input. I hope we can resolve this.
     
  8. Big B

    Big B HWF Godfather

    Likes Received:
    145
    Trophy Points:
    63
    That's not a sound argument. Having something a short while isn't a good indicator of it being a solid product. You're most likely to have problems with something early on due to a manufacturing and/or design defect or at the end of the products life when it's been well worn.

    Right now, we're going to need to play part roulette in order to figure out some issues.
    Our suspected products are the PSU and RAM, so that's where we'll focus on. You'll need to borrow another PSU that is known good and sufficient for your power needs. Secondly, you'll need some known good RAM. You'll need to swap out each of these components one at a time and run them for awhile to see if they've addressed the problem. If they still check out, then, yeah I'd agree with the motherboard. However, before going through the trouble of taking out the motherboard, try swapping out the PSU and RAM first.

    Given that you've opened up the case and tried to adjust the cooling, it's probably safe to say that there's not an overheating issue causing problems.
     

Share This Page