overheating possibly leading to short circuit?

Discussion in 'Abit' started by kaiser, Jan 1, 2006.

  1. kaiser

    kaiser Guest

    My computer (which I have been building over the last year) has had no
    issues with any hardware problems until two nights ago...

    I have a P4 3.E processor (appr. 3.01 to .02 GHz) w/ HT. Abit IC7-G
    motherboard. ATI AGP Radeon 9550XL 256mb video card. 600w Ultra power
    supply.

    I was playing AGE III with one of my flat mates. About an hour and
    half into the game my computer's alarm started going off. After waking
    everyone in the room I turned off the game and the computer stopped
    beeping. I immediately thought that it was overheating because I can't
    think of another reason why it would beep like that. I was surprised
    thought because I have a total of 9 fans in my computer. Also, my
    thermal reader said that my CPU was only about 122 deg. F. The thermal
    reader at that time was only loosely connected to the heat sink
    (actually barely hanging on with the thermal tape). So I figure that
    the CPU would be about 20 deg. more or so. My alarm goes off at 185
    deg. and shuts down at 195 and I didn't think there could be anyway it
    could be that warm. When the game was off the thermal reader dropped
    to fluctuating temps, between 113 and 119. I grabbed a fan and open up
    the side panel and the fan dropped the temp another 10 deg. We tried
    starting the game over again. This time the alarm went off around 132
    deg. and it only took about 45 min to get that warm. After that we
    just quit playing AGE. I turned off my computer this time. When my
    computer came back on the CPU was only running at about 85 deg (with my
    fan is still blowing into the machine) with minimal applications. I
    then changed games and played Diablo II by myself since I didn't think
    that it would get as hot with that game. I played for a few hours with
    no problem. The temp was between 119 and 127 with no alarm. Then
    yesterday I played AGE by myself and I was able to play a larger game
    for longer but then the alarm eventually came on. When I went to save
    the game the alarm stopped. I played so more and eventually it went
    off again. This time I save and quit. My flat mate wanted to play AGE
    again but we decided to play Diablo instead. I had no problems once
    again except for a couple of time he had a phone call or had to take a
    smoke break I would tab out of the game and do something else and
    eventually the alarm would come on again. I would save the game and
    close out and the alarm would stop. My flat mate eventually wanted to
    play AGE again so I decided to open the box and see if I can figure out
    what is wrong. I noticed the heat sink has dust caked on it. I take
    it off and grab a can of air. I clean that and blow out the rest of my
    computer too. I also fixed all of my thermal readers so that they were
    better attached to the things that they were reading (3 of 4 had become
    loose or fallen off). When I went to put the heat sink back on I
    realized I just broke the glue seal that was mounting the heat sink to
    the processor chip. Now there is only tried glue smeared on both the
    heat sink and the processor. I didn't think that would make much of a
    difference (someone can confirm if I am wrong) so I put it back on.
    When I turned the computer back on there was an error when searching
    for the CPU and said it had been changed. I then went into the BIOS to
    see if maybe a setting was messed up. Everything looked fine. I did
    change the temp for the alarm from 185 to 175. I saved and exited and
    when it restarted it restarted just fine. However, when windows began
    to load the alarm went off again. When I got to the welcome screen
    instead of logging into my user I just turned off the computer again.
    This time when I opened it up I took out the processor chip to make
    sure there was nothing wrong looking with it and it looked fine. Here
    is the kicker. After that I have never been able to boot my computer
    back up. It powers up just fine. All of the lights work on the mother
    board, for the wireless card, in the power supply, the optical drives
    open and close, the thermal reader reads, however no single is sent to
    the monitor. The light on the monitor stays orange. I thought maybe I
    screwed up my VGA card when I was doing all the messing around so I put
    in my old one and still the same thing.

    Do you think that I killed my processor and if so wouldn't I still be
    able to at least get to the first screen of the BIOS for it to tell me
    my CPU is unreadable?
     
    kaiser, Jan 1, 2006
    #1
    1. Advertisements

  2. kaiser

    Dylan C Guest

    I'm not much of an Intel person, but here goes...
    First off, too many case fans just add more dust and noise to the
    system. From my experience, dust kills more components than heat.
    Since your temps weren't at alarm levels, are you sure it was a temp
    alarm and not a voltage alarm or something similar? I'm really not
    familiar with Pentium temps, though, and this is just a guess on my part.
    Ahah...Dust! I bet your PSU is/was even worse than your CPU cooler.
    Could also lead to an overheated power supply with irregular voltages.
    Its not glue. At least not on an AMD CPU. Its thermal compund that
    helps pass heat from the CPU to the heatsink. You REALLY should have
    cleaned both surfaces and reapplied fresh compound. You can use the
    cheapest stuff you can find at any PC parts store. You can pay for more
    expensive stuff if you want, but you wont get much if any better
    performance for your money. Until you replace this compound, you really
    shouldn't run the computer.

    Now that you don't even POST, you need to start at square one and swap
    all critical components with known working ones. I would start with the
    Power Supply. Then the CPU, memory, graphics and mainboard. be sure to
    disconnect all other devices (Drives, PCI cards, extra fans, etc.)
    before starting. If you dont have spare parts, you can pull the
    motherboard, memory and Processor, and most PC repair places will
    individually test each for a reasonable fee.
    No, a bad CPU is one of many faults that will cause a computer to behave
    exactly as you described.

    -Dylan
     
    Dylan C, Jan 1, 2006
    #2
    1. Advertisements

  3. kaiser

    TomG Guest

    we usually talk in Celsius temperatures so I would have to do a conversion
    to see how hot your temperatures were... as a test, go into the bios and
    disable the alarms for CPU fans and such and see what happens... it is
    possible that your CPU cooling fan is getting noisy RPM feedback or other
    such problem although, I have a problem with seeing that be an issue only
    under a load of gaming.

    try running the system with the side cover off for the duration of a game to
    see if you still get the alarm... P4 processors have a built in thermal
    protection and they cycle themselves down if they start to go overtemp so I
    don't think you are hurting the CPU but anything is possible...
     
    TomG, Jan 1, 2006
    #3
  4. kaiser

    TomG Guest

    also, don't forget to check for dust build-up in the cooling fins, fan
    blades, etc.
     
    TomG, Jan 1, 2006
    #4
  5. kaiser

    kaiser Guest

    Actually there are 10 fans in my comp but they aren't all case fanse.
    Technically I only have 5 case fans. Two on the front blowing in and
    two on the back blowing out (these four are the only ones I added
    because the case was built with fan mounts that I just mounted fans
    to). There is on the side panel blowing out (it came with the case I
    didn't add it). then I have a fan on the CPU heat sink and on the
    chipset heatsink (it came with the mother board I didn't add it). Then
    my power supply came with two fans. One blowing into the PSU inside
    the computer case and then one on the other side blowing out. Then my
    thermal reader has a little fan which is on on the fron of the case.
    The thermal reader sits between my two optical drives and the two case
    fans (that I added) are at the bottom of the case.
    I am not sure if it was a temp alarm vs. a voltage alarm. I only
    assumed that because I don't have enough stuff in my computer to max
    out my 600w PSU. Also, the alarm stopped when I closed the game. I am
    also not sure what the actually temp was of the CPU because my thermal
    reader only reads the heat sink and not the chip itself and I am not
    sure how much hotter the chip is than the outside of the heat sink.
    Ok how do I clean out my PSU. The side panel of my PSU is clear and I
    see some dust in it but it isn't really that bad. About as much in it
    as on my book case (I should go dust that too). I took my PSU out and
    looked at the other side and it is clean. There is only a little dust
    on the side with the fans. I blow throw it with the can of air and
    some of the loose dust came out. Only enough to sneeze but not enought
    to make it look any different inside.
    No you are right it is a thermal compound even with the the Pentiums.
    I can go buy some more. Is there also some sort of special cleaning
    angent to remove the old stuff?

    Also, without the compound do think my CPU could have been damaged just
    from turning it on?
    I was afraid of this. I strip it down and start over to see what I
    get. I will let you know what happens
     
    kaiser, Jan 1, 2006
    #5
  6. kaiser

    kaiser Guest

    I can't go to the BIOS now because my computer doesn't send a signal to
    the monitor so I can't read anything. I am not sure if the computer is
    running not. All of the lights and fans come on. The optical drives
    work and thermal reader works. Everything is getting power except my
    monitor doesn't turn on.

    When the computer was working I did try playing with the cover off and
    I brought another fan over. BTW, my room is about 15 deg. Celsius so I
    was blowing some cool air into the case.
     
    kaiser, Jan 1, 2006
    #6
  7. kaiser

    kaiser Guest

    yep I took my can of air to all of my fans and blew them off.
     
    kaiser, Jan 1, 2006
    #7
  8. kaiser

    kaiser Guest

    ok I have stripped everything down so that the only thing plugged in is
    the motherboard which powers the CPU/heat sink fan, chipset heat sink
    fan, one case system fan (the system fan), and VGA card. And I can't
    the same thing. Everything powers on just fine but my monitor still
    receives no signal and the light remains orange.
     
    kaiser, Jan 1, 2006
    #8
  9. kaiser

    kaiser Guest

    I even unplugged my CPU hoping that I would leas get a warning beep
    letting me know that there is no CPU but still nothing. My computer
    right now works just as good without a processor as it does with one.
     
    kaiser, Jan 1, 2006
    #9
  10. kaiser

    w_tom Guest

    Posts that quickly blame problem on heat usually means the
    poster did not know where to start or learned why electronics
    fail. IOW heat is not an issue with only one PSU fan. Each
    fan after that should cause only single digit or tenths of
    degree temperature change - totally irrelevant. Too many fans
    create other problems. Dust should never be a problem in any
    machine. But with too many fans, even an extremely rare
    excessive dust problem can occur.

    Still this does not say (yet) where to start. For example,
    posted are classic symptoms of a failed power supply. Lights
    can illuminate, fans spin, and power supply is still 100%
    defective. First thing to measure is power supply voltages
    with a 3.5 digit multimeter. Never make a single change.
    First measure. I did not say motherboard voltage monitor for
    good reason. That meter is required.

    Confirm the system is constructed on a foundation that has
    not crumbled. Use the meter. Purple wire (power supply to
    motherboard), with power off, must measure above 4.87 volts.
    Green wire must measure more than 2 volts. Then when power
    switch is pressed, green wire voltage must drop to less than
    0.8 volts (and purple wire voltage stay in limits). Gray wire
    must be much more than 2.4 volts when power switch is
    pressed. Then yellow, orange and red wires must measure more
    than 11.7, 3.22, and 4.87.

    Do all these voltages exist? Then power supply is OK. Move
    on to other suspects. And still nothing has been removed.

    With problems announced, you should have immediately
    consulted the system (event) logs and device manager to read
    what problems existed before a system crash.

    Now try to boot (maybe from floppy) to run comprehensive
    hardware diagnostics. These diagnostics are provided for free
    by all responsible computer manufacturers. Otherwise you most
    locate individual diagnostics from component manufacturers or
    third parties. This assumes you can boot the machine without
    Windows. IOW does machine appear to read floppy drive or
    CD-ROM? If not, then we move on - and still have not changed
    anything.

    It appears you tried to shotgun a solution. It may or may
    not have complicated the problem. Strip system down to only
    motherboard, power supply, CPU, power switch, and speaker. No
    keyboard, peripheral cards, mouse, video, drives, or DRAM
    strips. That's right. Not even memory. Now power up
    system. If that minimal hardware is working, then speaker
    beeps. If not, you have the short list of defective suspects
    - minus the power supply that you know is working because you
    used a 3.5 digit multimeter.

    Of course, plug power cord was disconnected from wall before
    removing anything. That right. No power switch is even
    sufficient. Power cord must have been removed. Otherwise you
    may have made more damage. Also you were using proper static
    electric protection including a room with more than 40%
    humidity.

    Meanwhile, eliminate at least seven fans. The CPU has a
    fan. Power supply has a fan. And just in case the power
    supply fan (which is all any computer case needs) fails,
    install one more so that airflow is maintained if its power
    supply fan fails. Only hype would promote so many fans.

    Did heat damage the Intel chip? Not even possible. Even
    back in the 486 days, any Intel chip that might overheat,
    instead, protects itself automatically. This being different
    from AMDs that could self-destruct. Just another reason why
    heat could not damage your CPU.

    Another fact to collect before ever making changes. Did you
    examine every motherboard electrolytic capacitor for bulging -
    a problem created by overseas counterfeiting? Last thing on
    your list of possible damaged components was the Intel
    processor - that is until you removed it. Now even static
    electricity could have caused damage. Just another reason why
    the problem is identified before trying to fix anything.

    Numbers and other information collected should be posted
    here for numerous reason. One reason is that voltage numbers
    also report other facts.

    You posted a temperature of 195 somewhere. If on CPU, then
    CPU shutdown to protect itself. No fans were going to solve
    this - a human created problem - probably due to mismounting
    of CPU heatsink or failure of CPU fan. Not a heat problem.
    Instead a human assembly created problem. Ignore the hype
    about heat. That is only promoted by those who never first
    learn numbers. 195F number would be useful information.

    One final point. Another diagnostic tool is heat. We heat
    computers to locate hardware problems. Standard test of any
    minimally acceptable PC is to run it in a 100 degree F room.
    A system heated with a hairdryer on high is in 'pigs heaven' -
    works just fine - without all those fans - if hardware is
    working. Hardware that appears to work at 70F and fails under
    a hair dryer is 100% defective. Heat being a tool to locate
    defective hardware. Those who never learned, instead, cure
    defective hardware with more fans - which is why Tim Allen
    uses that as a joke, "More Power".
     
    w_tom, Jan 3, 2006
    #10
  11. kaiser

    Ed Forsythe Guest

    Hi K.
    First you must*thoroughly* clean the mating surfaces of the CPU and the
    heatsink with either 99% denatured alcohol (preferred) or Acetone. While you
    have the heatsink/fan assembly off thoroughly clean all the dust of heat
    sink fins, fan blades, *everything*!. Excessive dust buildup mayl seriously
    affect operating temps. Blowing canned air into the fans and allowing them
    to spin will not usually do the job. Buy a package of Q-Tips and hold the
    blades while you carefully swab them off. Once everything is cleaned
    *properly* apply the thermal compound and re-assemble and mount the
    CPU/HS/Fan to the MB. Now we can begin the trouble shooting process. In my
    experience (PC Power & Cooling and other premium PS, they either work or
    not. Sinvce ypou system [owers up i'm inclined to think that the PS is OK.
    You speak of alarms and since you don't describe them I'll throw this in the
    pot. The IC7 Max 3 has the followin beep codes:
    1. Long Beep: Memory problem
    Explanation: There is a failure of some sort related to the system memory.
    Diagnosis: The first bank of memory probably has a failure of some sort;
    this is usually just a physical problem such as an incorrectly inserted
    module, but may also mean a bad memory chip in a module. It is possible that
    there is a failure related to the motherboard or a system device as well.

    2. Long-Short-Short (-..) = Video problem.
    Explanation: The BIOS is unable to access the video system in order to write
    any error messages to the screen.
    Diagnosis: This is usually caused by a problem with the video card, or the
    memory on the video card. It can also be a motherboard issue.

    If you're gettin the long-short-short beep code and your monitor is not
    firering, I'd strongly suspect a video problem. Remove the card, clean the
    contacts with a paper towel and denatured alcohol and re-insert it.

    Just for kicks remove your memory sticks and clean the contacts. Now turn on
    your computer *before* you turn the monitor on.
    If your monitor fires up go into the BIOS and insure that all settings are
    OK especially the one that specifies AGP first (assuming that you have an
    AGP vid card). What happened?
     
    Ed Forsythe, Jan 3, 2006
    #11
  12. kaiser

    Ed Forsythe Guest

    Oops, Sorry about the spelling errors. My spell checker must have konked
    out??
    --
    TallyHo!
    Ed
     
    Ed Forsythe, Jan 3, 2006
    #12
    1. Advertisements

Ask a Question

Want to reply to this thread or ask your own question?

You'll need to choose a username for the site, which only take a couple of moments (here). After that, you can post your question and our members will help you out.