1. This forum section is a read-only archive which contains old newsgroup posts. If you wish to post a query, please do so in one of our main forum sections (here). This way you will get a faster, better response from the members on Motherboard Point.

revive motherBoard hit by lightning?

Discussion in 'Motherboards' started by Samuel, Sep 17, 2003.

  1. Samuel

    Samuel Guest

    hi folks,

    my p4 is dead following storm.dead as in silent,not even fans. i think
    it got hit via usb adsl modem (as i have no dial tone either now) and
    exited into my home LAN via the network card as when i remove this
    card and switch on the atx PSU fan turn on for 5 sec then shuts down.i
    have other pcs on this lan (without modems) which wont turn on either
    but i am not worried about them as they are obsolete.

    does this shutdown after 5 sec indicate a short on the board?

    how to i proceed to diagnose the short?would it be near the USB
    port?anyone ever manage to revive his board?thanks

    Sam
     
    Samuel, Sep 17, 2003
    #1
    1. Advertisements

  2. Samuel

    Paul Guest

    About all I can do for you is explain how an ATX PSU and motherboard
    work together.

    The ATX PSU has two parts - the +5VSB supply runs as long as the switch
    on the back of the PSU is switched on. This should cause a green power
    LED on the surface of your motherboard to light up. This LED indicates
    your motherboard has standby power operating, and usually the power
    supply can provide an amp or two on that output. That power can be used
    to operate a keyboard, mouse, LAN card etc, or any other device that
    provides a "wakeup" function for the computer.

    The +5VSB is essential to supporting the "turn on sequence" for the
    computer. Without +5VSB, the computer won't turn on. The +5VSB powers
    chips on your motherboard, and one of those chips controls a signal
    called PS_ON#. PS_ON# is pulled up with a resistor to logic 1, and
    when the chip on your board grounds the PS_ON# signal, the power supply
    then turns on the rest of the supply voltages +3.3V, +5V (main),
    +12V, -5V, -12V.

    A number of functions can be tied into the PS_ON# logic.

    1) AGP burnout protection. Modern P4 Asus motherboards use 1.5V AGP
    and have a small transistor circuit, that prevents the power supply
    from starting, if a 3.3V AGP card is detected. Sometimes this
    circuit gets confused, or if the video card is damaged, sometimes
    this circuit stops the computer from operating. (So this won't
    result in the board running for 4 seconds - it'll just be dead.)
    2) Thermal protection. Some motherboards have protection against
    overheat. That might be tied into killing the PS_ON# signal,
    and may take a few seconds to trip.
    3) Software controlled wakeup state. If I plug in my P4B based
    computer, it operates for a couple of seconds - basically long
    enough for the BIOS to figure out that the machine should not
    be running, and then it shuts down again.
    4) The power switch has a 4 second time constant. Pressing the
    power switch for more than 4 seconds or less than 4 seconds is
    detected by the motherboard chip, and can be used to commence
    suspend or shutdown functions. A stuck power switch might
    result in behaviors with a relatively short time constant.
    5) The voltage regulator(s) on the motherboard contribute to a
    signal called Power_Good. Generally, a processor will not run
    unless Power_Good is valid in a relatively short time after
    power up. What I don't know, is if this is tied into the PS_ON#
    logic or not - it doesn't have to be - the motherboard could
    safely sit there dead with no Power_Good, and with the power
    supply still running happily.

    To debug your problem, you need to simplify as much as possible.
    You didn't say what model it is, but if the board has vocal POST
    messages, you could strip everything off the motherboard, and
    see if the vocal POST reports CPU missing or memory missing etc.
    Then, add components back, noting whether the latest added component
    stops it from starting properly.

    If the motherboard and power supply by themselves don't seem to
    be doing anything, you can test the power supply by connecting
    PS_ON# to COM. Page 19 of this document shows the pinout of the
    ATX 20 pin connector on the end of the PS cable harness.

    http://www.formfactors.org/developer/specs/atx/atx2_1.pdf

    Some people don't recommend running a power supply without a load.
    The reason for this can be that the outputs can achieve higher than
    normal voltages when no load is present. Check your power supply
    label, to see if any output has a non-zero minimum load value
    specified. I got some 5 watt resistors and pins that plug into
    an ATX connector from my local electronics store, to draw anywhere
    from 1/2 amp (for the +5VSB, -5V, -12V outputs) to 1 amp for the
    main outputs. Loading all the outputs helps ensure the PS isn't
    damaged when I connect PS_ON# to COM and test my supply. Then,
    using a voltmeter, I can check that all the outputs are sane, and
    leave the PS running for a while to see if it is stable and isn't
    overheating due to some kind of internal problem. (Stay out of
    the inside of the PS, because there are two big caps in there
    with enough juice in them to stop your heart.)

    Depending on your home insurance policy, you may want to just
    junk the thing, and get the computer replaced at the same time
    as any other appliances that got blown. I don't know how a
    home insurance policy would handle the construction of a "DIY"
    computer :) Maybe have a local computer shop assemble one
    from the parts list you specify - that way you get the parts
    you want, and the insurance company gets a nice itemized bill
    for what was done.

    In terms of what damage the lightning can do, it can range from
    damaging just the chip connected to an external cable (in this
    case that could be the Southbridge, that hosts the USB interface),
    all the way to blowing the tops off all the chips on the board.
    A dead Southbridge would be fatal (which suggests that in future,
    use a separate PCI LAN card to connect to the ADSL modem, on the
    off chance that just the LAN card gets blown).

    HTH,
    Paul
     
    Paul, Sep 17, 2003
    #2
    1. Advertisements

  3. Samuel

    rstlne Guest

    The +5VSB is essential to supporting the "turn on sequence" for the
    logic 0
    ps_on is a active low
     
    rstlne, Sep 17, 2003
    #3
  4. Samuel

    Paul Guest

    Where did I say otherwise ? I said a resistor is used to pull it to
    the inactive state. That is why, if you sit the PS on a tabletop and
    plug it in, it doesn't run. Grounding the PS_ON# pin (as you say
    correctly, logic 0) causes the PS to operate all outputs. Or, for the
    newbies out there, connecting PS_ON# to one of the adjacent COM pins,
    because that is less ambiguous. Basically the PS_ON# pin is open
    collector capable, but not necessarily always driven that way,
    so don't assume it is safe to short PS_ON# to ground when the
    motherboard is connected to the PS. Some older motherboards use a
    74 series totem pole driver to drive PS_ON#.

    Paul
     
    Paul, Sep 17, 2003
    #4
  5. Samuel

    rstlne Guest

    "PS_ON# is pulled up with a resistor to logic 1,"
    Not wanting a fight.. was just pointing out
     
    rstlne, Sep 17, 2003
    #5
    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.