Invisible problem

By Delfos
Nov 16, 2008
Topic Status:
Not open for further replies.
  1. Hello everyone. I've been reading this forums for a while, whenever I've a problem that I do not know about I come here to check, this is a great service you're all doing. But now it's my turn to tell a story, I can't figure this on my own.

    My PC specs are on the sig, any additional questions just feel free to ask. I practically bought this computer by parts, taking my old boards and cards out of the case. I know my share on hardware, specially electrical stuff.

    Problem: The Computer started to freeze and crash in BF2 with a multi-colored screen bars everywhere, it would just crash the game and come to desktop. I thought it was because I was testing the limits of my graphics, started to go on ease on it and it was fine. Not more than a week ago my computer started do the same thing in any kind of game. I thoguht it would only do that on heavy graphical games but started to do that on low graphic ones too, yesterday the frequency of doing that was high enough not to open any game, specially because this time the whole computer would freeze on multi-colored screen, and even more weird, the sound would start looping, so gfx and sound freeze, also freezing my controls (mouse and kb), which led to manual reboots.

    After testing alot from any forum I would find, even through this forum, I couldn't find any solution. Today I've unplugged everything and plugged everything back, the problem still exists but now it doesn't even freeze, just power crashes the computer, which makes me think this cannot be a software problem.

    This is what I know: This is not a overheating problem, I installed a new fan under my gfx card because my old one was doing some noise and stopping sometimes, I was afraid the overheating would have led the gfx card or sound card (in each side of this fan) to melt or burn something, but they're fine.
    I've ran CPU+RAM stress tests, graphic stress tests and reinstalled gfx driver and dx 9.0c (same config as I ever had).

    For crashing both gfx and sound, I thought it was maybe the power supply, I've opened it and there's nothing melted or burnt, there could still be an over-power problem causing short-circuit, which I doubt it's happening or else it would crash without games.

    Also maybe the CPU, which is always mysterious, besides cleaning, is there anything I should look for signs of CPU malfunction?

    My biggest guess is that my Corsair RAM cards are malfunctioning, would explain why low graphics games would crash after a while, and why games using heavy single player bots would crash not even a minute after starting the game, but doesn't explain how now it power crashes the whole computer, does it?

    Could always be my ASRock motherboard, which never gave me reasons of problems, so I've no idea how this problem could be related to it.

    I've read somewhere about leaking capacitors, but I don't see anything wrong anywhere.
    There were a series of power cuts in my town in the last months, I've checked everything and there's no signs those power cuts affected the computer, but who knows? Any tip? Anyway I can figure where this problem is coming?
  2. kimsland

    kimsland Ex-TechSpotter Posts: 18,353

    Well your first post was sure a big one :) It may have made up for all the other times of just viewing (but that's ok)

    You will need to do the basics

    Confirm all drivers are updated
    Even confirm Windows security updates are done (SP3 would be good ;) )
    Run Memtest on your Ram
    Remove any dust internally, if possible

    That will help for a start
  3. Delfos

    Delfos Newcomer, in training Topic Starter

    All drivers are updated, they were already, there were no software/hardware changes since the time it was working fine.

    I don't like Windows stuff much, but I'll try the security updates if you advise it so much, although I doubt it'll fix the problem.

    I tried Prime95 instead of Memtest, but I can do it if you think it'll get a different result. I'm avoiding it because I only have DVDRs and no floppy drive, I don't want to waste a DVD but if it's what it takes to find the problem, I can either do that or go buy a CD.

    I always keep my PC dust free, I leave the computer open to look inside, modify, feel stuff if they're overheating and other reasons, everything is dust free right now, when I unplugged everything and plugged again I cleaned every bit, even inside powersupply.

    I'm sorry for the long post, but I rather avoid wasting people's time asking stuff that I already though and tested. Thank you for the advices, keep them coming.
  4. Delfos

    Delfos Newcomer, in training Topic Starter

    I've ran the memtest, it power-crashed after a few minutes it started. I've also tested both slots and cards individually, same result, plus every time I had to wait like a minute or so to successfully start the computer again, but I've felt the cards and there's no sign of overheating. If it's not definitly both Corsair RAM cards problem, we're back at where I was, although I'm almost sure it's not a power supply problem. Couldn't it still be the PS, CPU or MoBo?

    I'm really afraid it's the CPU, they're expensive and I don't want to waste my planned xmas money on one :(
  5. kimsland

    kimsland Ex-TechSpotter Posts: 18,353

    I think Benchtest may be a good idea, if it crashed whilst trying to run a Memtest (or you could just try swapping parts in (not that I'm suggesting that you have copious amounts of parts lying about!)

    Here's the hardware most likely to least likely

    Conflicts (including addon cards, even USB devices plugged in)
    Cmos settings (not exactly hardware, but resetting is always good)
    Ram
    PSU
    M/b
    CPU

    It can also even be data cabling that type of thing
    I have all this type of info in my No Post guide, so maybe have a look there, or h e r e rather :)
  6. Delfos

    Delfos Newcomer, in training Topic Starter

    About that CMOS, I turn the PC off, (I've a switch) switch PS off, put a jumper cap on the thingies for 5 seconds and take it off and start the PC back on? What about the battery? Do I also remove the battery while shorting the pins? and where do I find the cap? :S
  7. kimsland

    kimsland Ex-TechSpotter Posts: 18,353

    Damn how time flies, I've been away talking to others about computer stuff (how nerdy hey) and just noticed your reply

    Anyway yes it's usually battery out (we are talking PC not laptop here obviously)
    But actually the manual (your free manual) should give exact steps on resetting cmos settings (as per the guide above)
  8. Delfos

    Delfos Newcomer, in training Topic Starter

    CMOS reset didn't fix the problem, tested and it power-crashed
    ...but you know what? I remembered how I always had problems installing the CPU heatsink, this time it wouldn't be sort-of fixed, I forced the heatsink out and two of the 4 pins weren't properly attached, so I forced the heatsink back and I was able to force-fix the heatsink back to the CPU.

    I tested a game I'm using for test, Red Orchestra, using high graphics and bots and all, I was so releaf that I finally fixed the problem, but after some minutes I've got the multi-colored screen freeze and PC rebooted right after. So did I fix the power-crash problem and got back to the multi-colored screen problem or is it the same problem and I just fixed 1 part of it?
    Maybe the CPU is really overheating but the heatsink even if force-attached isn't able to "heatsink" which might explain why it's cold... tell me your recomendation for CPU temperature test please.

    I doubt there are conflicts, as I said above, I've made no changes from when it was working fine and there's nothing important connected to the PC, printer is off, unconnected firewire cable, Logitech Joystick, mouse and kb, unconnected USB cord, CRT cord, sound and mic cords and that's it. Plus I've this config for long time, 2 years next summer, last summer I added the firewire card but it was fine until now.

    RAM, what else can I do about it? just try to get some lent RAM cards to see if it were corrupted RAM cards?

    What else?

    EDIT: Maybe I should try the memtest again and see if those crashes were just because of the CPU heatsink?
  9. kimsland

    kimsland Ex-TechSpotter Posts: 18,353

  10. Delfos

    Delfos Newcomer, in training Topic Starter

    CPU is fine, no more power-crashes as I said above,

    either way, I've ran memtest again, found an error on test 3 pass 1, found some more errors but I guess it was the same error from the other passes since it wasn't reported, anyway I didn't know how to stop it and see the full report, was there a way I could tell where the fault come from? Because I don't want another half a day memtest on each RAM card to see if either only one or both are faulty.

    Oh, and does this definitly mean it's a RAM card problem?
  11. kimsland

    kimsland Ex-TechSpotter Posts: 18,353

    Hmm, then I better not tell you the answer :rolleyes:

    Yes, I'd replace both, that way, you don't need to test again ;)
     
  12. Delfos

    Delfos Newcomer, in training Topic Starter

    I'll test them both eventually, but I'll replace both most likely.

    Do you have any advices on brands? I thought Corsair wouldn't give me any troubles...

    I'm looking at this ones:
    KINGSTON DDR2 2048MB 667MHZ (24.50€)
    OEM - 2Gb DDR2 PC667 (21.69€)
  13. kimsland

    kimsland Ex-TechSpotter Posts: 18,353

    Hmm, damn I'm the only one replying in this thread
    The problem is I don't like getting caught up in sales questions
    I've quoted before, then someone has come along and said, "no this is better"
    So therefore I'm best not to say anything!

    I can say this though, you should go to the manufacturer's website and confirm the brands that are compatible with your system (because that's a technical answer ;) )
  14. Delfos

    Delfos Newcomer, in training Topic Starter

  15. kimsland

    kimsland Ex-TechSpotter Posts: 18,353

    Well according to that link page

    Memory
    Therefore Yes (but don't quote me ;) )
  16. Delfos

    Delfos Newcomer, in training Topic Starter

    I eventually found which one was faulty, I'm runing with one only, I'm ordering 2 of those 2GB Kingstons.

    Thanks for the support, basically:
    - Power-Crash - was the result of a not properly attached CPU HeatSink.
    - Multi-Colored Screen/Freeze/Crash - was the result of a faulty RAM card.

    Mission Accomplished.
  17. kimsland

    kimsland Ex-TechSpotter Posts: 18,353

    Thanks for the update :grinthumb
Topic Status:
Not open for further replies.


Add New Comment

TechSpot Members
Login or sign up for free,
it takes about 30 seconds.
You may also...


Get complete access to the TechSpot community. Join thousands of technology enthusiasts that contribute and share knowledge in our forum. Get a private inbox, upload your own photo gallery and more.