"Failure Imminent" should i continue?

Discussion in 'Asus' started by Lindsay, Jul 10, 2004.

  1. Lindsay

    Lindsay Guest

    Ok, a few weeks ago, my BIOS chip died, so I ordered a new one.
    I have a ASUS P4S533 motherboard (it's like 1 1/2 years old). I
    started the comp, everything worked, except for the CD and DVD ROMS.
    Turns out, before I found out it was the BIOS that was dead, I had
    taken it to a computer store to have it looked at. Without even asking
    if they could start work on it, they switched a bunch of wires around
    and never pulgged the power cords back into the CD drives but they DID
    plug in the floppy, HD, and MB (with the 4 pin rectangular
    So I tried to plug them back in. I switched them around to try and get
    them all to fit into the 2 CD drives, the floppy, the HD, and the MB
    and couldn't. I then called ASUS and they said that the motherboard
    had something like EASY power or something (Im not sure) and that I
    didn't need to have the 4 pin plugged into the MB. So i unplugged it
    and reset the RAM just incase.
    I started the comp and got a message that read SMART predicts that a
    failure is imminent blah blah blah.
    Now that didnt show up before i switch the wires. Could that be the
    cause? Should I just hit "f1" and continue? or should I do something
    else? I WAS fetting a different error before that. it read "failed to
    write ESCD." So im not really sure what to do here. I've got alot of
    important crap on there.
    Lindsay, Jul 10, 2004
    1. Advertisements

  2. Lindsay

    Outback Jon Guest

    Backup the hard drive IMMEDIATELY. I've had a couple of drives indicate
    SMART failures. One continued to work for months. One created more and
    more errors as I continued to use it. And one died COMPLETELY a few weeks
    after getting the error.

    It is doubtful that the wires are causing the problem.

    "Outback" Jon Gould | Let those who RIDE,
    2003 Kawasaki Concours (wreck)|
    1976 Honda CB750F (needs work)| DECIDE.
    1972 Yamaha DS7 (project) |
    CQ CQ CQ de KC2BNE |
    AMD XP 2400+ @ 2180 MHz 768M PC2100 DDR @ 290 MHz
    Asus A7N8X-E Deluxe 120GB and 40GB Maxtor HD
    Asus GeForceFX 5600 (330/650) Cornerstone 21" Monitor
    Sony 52x40x52 CDRW NEC MultiSync 17" Monitor
    Sony 8X DVD+-RW Nostromo n50 Speedpad
    http://folding.stanford.edu - got folding? Team 33432
    Outback Jon, Jul 10, 2004
    1. Advertisements

  3. Lindsay

    Paul Guest

    You are probably working on two or more failures at the same time.
    This complicates life.

    When you say the BIOS chip died, was it just innocently sitting there
    and one day the computer wouldn't POST ? Or, did you attempt a flash
    upgrade, and it died on the spot ? The difference is important, as
    the "failed to write ESCD" is another BIOS chip problem.

    The BIOS chip has the following areas - at least the ones I've
    run into.

    Main BIOS code - Basic Input/Output Drivers for hardware. BIOS setup
    screens and the like.
    Boot Block code - A subset of basic I/O, usually just enough to aid
    in reflashing the BIOS, if there is a problem with
    the main BIOS code.
    DMI block - Some kind of management information. Like inventory records
    suitable for harvest by a central database in a big company.
    This block is writable and there are some kinds of tools
    for changing the contents.
    ESCD block - Again, some kind of inventory info, but in this case
    used by the BIOS during enumeration of the hardware.
    If new hardware is detected, the ESCD block will be
    written by the BIOS. This is what the failure message is
    telling you, that during POST, the BIOS is attempting
    to do an update. It has to do at least one update, after
    getting a new BIOS chip, as flashing the BIOS can erase
    this area, giving the processor a chance to write a clean
    Microcode cache - On a P4 board, when you plug in a new P4 processor,
    a 2KB code segment from the BIOS is copied into a
    temporary area. There is room for two segments,
    before they are reused. If the cache fails to be
    written, there might not be any message, as the
    BIOS still has the CPUCODE.exe code segment to get
    any of the microcode patches it needs. Microcode
    patches correct errata in Intel processors.

    So, now maybe you understand why the circumstances of your flash failure
    are important. If the BIOS chip died on its own before, it sounds
    like it is still having trouble, implying something else is wrong
    with the computer. For example, if the power supply voltages were out
    of spec, that might be enough to do it. But, if that were true, there
    might be some other symptoms as well.

    Now comes the SMART error message from the disk. A disk can have trouble
    too, if the voltages from the power supply aren't correct. That could
    be the source of your SMART message.

    You cannot afford to ignore the SMART message. This is how I would
    proceed - your response may differ due to budget concerns.

    1) Buy a new disk drive. A person who builds computers for themselves
    should always keep a spare disk drive handy, for emergencies. They
    are cheap, and very useful.
    2) Find a second known working and stable computer. Take the drive
    with the SMART errors and the new disk drive, and try to make
    a copy of the drive, to the new disk. Before doing the copy,
    you could do a surface scan of the disk, looking for reports of
    CRC errors coming from the disk. If any errors are reported, a drive
    to drive copy is going to lose the intelligence in those sectors,
    but there is little you can do about it now. A CRC error happens
    when the drive has run out of spare sectors, to correct the CRC
    errors, so the sector you are reading from, is a bad sector that
    the drive cannot automatically replace.

    I copy drives with Partition Magic, and when I plug them in, they
    seem to boot OK, so I guess all the necessary info is getting
    copied across. You would think the MBR (Master Boot Record) wouldn't
    be copied, but I don't seem to have any problems with the method.
    3) Armed with your clone copy of the boot disk, you are now ready for
    more hardware debugging, secure in the knowledge you have two copies
    of your valuable info.

    When working with the disk drive(s), I assume you know about master/slave
    or Cable Select jumpering for the disks.

    Now, back to the duff computer. If you can get into the BIOS, check
    the power monitor screen that lists the voltage coming from the power
    supply. Check to see if any of the listed voltages are more than 5%
    off from the nominal +3.3, +5, +12, -5, and -12. The last two aren't
    that important, but if they are malfunctioning, it can still indicate
    the need to change out the power supply.

    It is possible, if you replace the power supply, that at least the BIOS
    error message will go away.

    Now, assuming the power supply swap gets rid of the "ESCD write failed"
    message, you are still left with a drive that has SMART stats filled
    to the roof. I don't know if reformatting a drive will erase the SMART
    stats or not. If you can get the computer to run with the clone copy
    of the drive, and all your info is intact, you might consider formatting
    the old drive, and see if the SMART error message goes away. Don't
    format the drive, until you are sure all your info is safe on one or
    more backup devices, such as the clone drive or drives.

    The EZplug power connector (called AUX12V1 in your manual), was placed
    on Asus boards during the transition period during which ATX power
    supplies didn't have the necessary 2x2 square ATX12V connector. The 2x2
    connector has two 12V pins on it, and they are capable of carrying at
    least the 10 amps of current that a high end processor might need.
    A disk drive connector, has one +12V pin on it, and that is rated
    for 8 amps. Either the 2x2 connector, or the EZplug, or both can
    be used, as the more connections in parallel, the more current carrying
    capacity, and the less voltage drop. As the manual states, your power
    supply has to have enough current output at +12V, to meet the needs
    of the processor. The difference between an older, midrange processor,
    and a high end processor, might be having a power supply with say
    10A to 15A of current on +12V. Many supplies now will be able to
    source the second of those two numbers without complaint.

    With cabling, you have to think about where to put the drives in
    the case, and how to route the cables, pretty carefully. Some supplies
    now, don't have enough drive cables on them, so every connector counts.
    Plan the cables, so each cable is fully used, with the exception of
    your EZplug cable. If you are using that cable, instead of the 2x2,
    then don't power a drive from the same cable. Better yet, use the 2x2
    connector, as the wire on it powers only the 2x2 connector. Disk drive
    cables make a poor power supply to a disk drive, if one of the
    connectors on the cable is connected to a "current hog", such as
    the EZplug, an ATI9800 or a FX5900 type high end video card. It is
    OK for two or more drives to share a drive power cable, but don't
    have a disk drive sharing with a high power toy, as there may be too
    much voltage drop for the disk drive to tolerate.

    In summary, I would:

    1) Back up the current disk drive. Use a known stable computer.
    Make sure your info is safe first. Only use one of the two
    drives in the duff computer until it is fixed.
    2) Try a power supply swap, if the symptoms at least hint at power
    not being stable. The fact the drive is in a mess, and the BIOS
    flash chip cannot update itself makes me suspicious. Many of the
    problems in this news group are traced to supply, so it is the
    first suspect.
    3) Plan drive cabling carefully, using all connectors on a cable if
    only drives are being powered on that cable. Save some drive
    cables for private powering of high power loads, like high end
    video cards and the like.

    Paul, Jul 10, 2004
  4. Lindsay

    Greysky Guest

    Usually, the POST message is separate from the BIOS stuff, although as Paul
    so brilliantly put in his excellent post, it may be just a bad power supply.
    If it *is* your hard drive, and it's out of warranty, I always recommend the
    user take it apart and sand the disk platters using Coarse sandpaper even it
    they previously wiped it... can't be too careful. Also, there are two very
    powerful flat magnets housed in the arm assembly that you could carefully
    pry out and have a lot of fun with! So even a dead drive can be a source of
    amusement and fun :)
    Greysky, Jul 11, 2004
    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.