S2466- broken cap, bad voltages, high T- what else?

Discussion in 'Tyan' started by Mike, Aug 22, 2003.

  1. Mike

    Mike Guest

    The saga of my broken S2466-based pc continues...
    It returned from the manufacturer a couple of days ago with a leaky
    capacitor on the S2466. The repair rep says they tested it for 2 days,
    and it's ok, and they're not replacing it. So today I decide to run
    SiSoftware Sandra on it just to see what it says. Here's the interesting
    Temperature Sensor(s)
    Board Temperature : 77.0°C / 170.6°F
    CPU1 Temperature : 77.5°C / 171.5°F
    CPU2 Temperature : 74.5°C / 166.1°F

    Cooling Device(s)
    Auto Fan Speed Control : No
    CPU1 Fan Speed : 4688rpm
    CPU2 Fan Speed : 4927rpm

    Voltage Sensor(s)
    CPU1 Core Voltage : 1.73V
    CPU2 Core Voltage : 1.73V
    +3.3V Voltage : 3.36V
    +5V Voltage : 5.03V
    +12V Voltage : 9.61V
    -12V Voltage : -12.61V
    -5V Voltage : 0.23V
    +5V Standby Voltage : 4.92V
    Ok, the fan speeds aren't interesting but they are between what is-
    namely, the really high temperatures and the wacky voltages. Clearly
    this board is toast. What I'm wondering is if the CPUs or other components
    might have been damaged too. Everything seems to be operating ok, but
    I don't see that as a long-term condition. What else should I demand they


    p.s. yes, I turned it off now.
    Mike, Aug 22, 2003
    1. Advertisements

  2. Is the machine really that hot or is Sandra just not properly
    calibrating the sensors? Does the BIOS show the same speed? Are the
    chips/board hot to the touch? Has the machine been running long enough to
    get that hot (assuming the heat sinks are attached properly, otherwise CPUs
    can heat up to 170F in seconds.)

    David Schwartz, Aug 23, 2003
    1. Advertisements

  3. Mike

    Arno Wagner Guest

    Does not matter. Obvious damage is damage, even if some
    tests succeed. Send it back and tell them they should start
    acting like they are requred by law.


    Arno Wagner, Aug 23, 2003
  4. Mike

    Mike Guest

    Hi David,
    The machine had been running for a couple of hours at 100% cpu usage.
    It still shouldn't be that hot. I wouldn't be surprised if the heatsinks
    had been installed improperly. (I just turned it on and started Sandra
    right away- the T was only 2-3 C lower.)
    I don't know if Sandra was improperly calibrated, but it had shown
    reasonable T values when the machine was not broken. I downloaded the
    latest version, and it came up with the same values. The BIOS does not
    show the T or other info anywhere I can find- it's the version dated
    03/27/02. Will a newer BIOS tell me more info?

    Mike, Aug 25, 2003
  5. Mike

    Brian Guest

    Mike wrote:

    The board is NOT that hot. There is a bug in the chips or something
    that when the first time the temp is read, it returns temps in the 70s.
    Subsequent reads will return the correct temp. If you go to the temp
    monitor in the BIOS, you will see it show 70 the first time, then it
    will show the real temp a few seconds later.

    I recommend using a tool like MBM5 (http://mbm.livewiredev.com/) to
    monitor your temps and other things.

    Brian, Aug 25, 2003
  6. Mike

    Mike Guest

    I installed MBM5. Initially, it did show the same temperatures (even after
    a refresh). But when I RTFM, I discovered that the settings were wrong.
    When I set the sensors to the correct settings for the cpus (HF/2D combo
    5 diode + HF/2D combo 6 diode) it shows the CPU temps as 44°C, 51°C, and
    43°C for the board. Under load, they all went up 3-5°C. The voltages are
    still screwy, but I won't worry about something frying (other than my
    temper- the customer (dis)service rep has not returned any of my e-mails
    since Thursday! Probably busy trying to think up another lame excuse about
    why this board is ok).


    Mike, Aug 25, 2003
  7. Mike

    pahis Guest

    that is more their style nowadays.
    when they are out of excuses they simply stop responding.
    happend to me when i was teasing them about new BIOS.

    now it seems that bankcropcy is just behind one motherboard
    replacement. sad end of story to former industry leader...:(

    pahis, Aug 26, 2003
  8. Mike

    pahis Guest


    capacitors stop leaking when they are empty. now all of it's job
    is done with other capacitors in same group. ie. i have one MSI-board
    wich worked when 3 of 4 capacitors were and had only little start-up
    problems. it depend how much there is oversize in design. in Tyan, i think
    (and hope) there is pretty lot. after all, them are mostly serverboards and
    they are not watched every day for visual errors. but when something like
    this is found, replace as soon as possible is only choice.

    pahis, Sep 3, 2003
  9. Mike

    ewy Guest

    Tell tyan that you have somthing like a K7X Pro or one of their more
    expensive line boards and see how they act.
    do a little bait and switch on them.
    ewy, Sep 30, 2003
    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.