Installation of PCI board resulted in meltdown of systemboard - Hardware

This is a discussion on Installation of PCI board resulted in meltdown of systemboard - Hardware ; Hi, Yesterday an embarrassing problem occured. I installed into a HP Proliant ML350 G4p system a bluecherry PV-153 video capture card. (We want to do video capturing with Linux). The card was seated well as far i could see. Turning ...

+ Reply to Thread
Results 1 to 6 of 6

Thread: Installation of PCI board resulted in meltdown of systemboard

  1. Installation of PCI board resulted in meltdown of systemboard

    Hi,

    Yesterday an embarrassing problem occured. I installed into a HP
    Proliant ML350 G4p system a bluecherry PV-153 video capture card. (We
    want to do video capturing with Linux). The card was seated well as far
    i could see. Turning on the system resulted into a short that caused
    both the system board and the card to be destroyed. The system
    developed strong fumes of burned plastics and the PCI slot showed black
    spots. The PCI slot on the card was burned on two places.

    Now the card and the system are destroyed, HP will replace the board
    for free, although they suspect that the card turns out the wrong
    voltage. The capture card was brand new, and PCI V2.1 compliant
    (although i do not know what that means).

    store.bluecherry.net, where we purchased the card, has some
    reservations if this situation is covered by warranty, and accuses HP
    to put too much voltage on the board. (The old game..).

    In the 20 yrs i work with computers, nothing like this has happened to
    me.

    Questions: Can somebody give some hints what could have caused the
    problem ?.

    Thanks in advance.


    Erik


  2. Re: Installation of PCI board resulted in meltdown of systemboard

    erikw_nl@yahoo.com writes:

    > Hi,
    >
    > Yesterday an embarrassing problem occured. I installed into a HP
    > Proliant ML350 G4p system a bluecherry PV-153 video capture card. (We
    > want to do video capturing with Linux). The card was seated well as far
    > i could see. Turning on the system resulted into a short that caused
    > both the system board and the card to be destroyed. The system
    > developed strong fumes of burned plastics and the PCI slot showed black
    > spots. The PCI slot on the card was burned on two places.
    >
    > Now the card and the system are destroyed, HP will replace the board
    > for free, although they suspect that the card turns out the wrong
    > voltage. The capture card was brand new, and PCI V2.1 compliant
    > (although i do not know what that means).
    >
    > store.bluecherry.net, where we purchased the card, has some
    > reservations if this situation is covered by warranty, and accuses HP
    > to put too much voltage on the board. (The old game..).
    >
    > In the 20 yrs i work with computers, nothing like this has happened to
    > me.
    >
    > Questions: Can somebody give some hints what could have caused the
    > problem ?.


    Faulty or badly designed mainboard or addin card or both. Perhaps the
    addin card was a 3.3V design incorrectly keyed as 5V-compatible. The
    photo on the web page shows a card keyed as universal 3.3/5V. What is
    the keying on the mainboard PCI slot?

    --
    Måns Rullgård
    mru@inprovide.com

  3. Re: Installation of PCI board resulted in meltdown of systemboard

    erikw_nl@yahoo.com writes:

    >Questions: Can somebody give some hints what could have caused the
    >problem ?.


    A PCI slot has 3 voltage rails: Fixed 5V, Fixed 3.3V and "VIO". VIO is meant to
    be used for the PCI signal drivers on the card and can be either 3.3V or 5V,
    depending on the chipset. There's a notch to differentiate the allowed VIOs:

    http://www.adexelec.com/faq.htm

    So either your board has 5V on VIO and the card works only with 3.3V, but has the
    wrong notch or the card internally connects VIO to 5V (I've seen that, too).
    That is a nice short between the 3.3 and 5V rail...

    --
    Georg Acher, acher@in.tum.de
    http://www.lrr.in.tum.de/~acher
    "Oh no, not again !" The bowl of petunias

  4. Re: Installation of PCI board resulted in meltdown of systemboard

    On Fri, 22 Dec 2006 11:11:52 +0000, Georg Acher wrote:

    > erikw_nl@yahoo.com writes:
    >
    >>Questions: Can somebody give some hints what could have caused the
    >>problem ?.

    >
    > A PCI slot has 3 voltage rails: Fixed 5V, Fixed 3.3V and "VIO". VIO is meant to
    > be used for the PCI signal drivers on the card and can be either 3.3V or 5V,
    > depending on the chipset. There's a notch to differentiate the allowed VIOs:
    >
    > http://www.adexelec.com/faq.htm
    >
    > So either your board has 5V on VIO and the card works only with 3.3V, but has the
    > wrong notch or the card internally connects VIO to 5V (I've seen that, too).
    > That is a nice short between the 3.3 and 5V rail...
    >


    I'm guessing that the PCI card had a short on it. It seems unlikely that
    anyone would be selling a card that was keyed wrong, it's much more likely
    that there was a manufacturing defect.

  5. Re: Installation of PCI board resulted in meltdown of systemboard

    General Schvantzkoph wrote:

    > On Fri, 22 Dec 2006 11:11:52 +0000, Georg Acher wrote:
    >
    >> erikw_nl@yahoo.com writes:
    >>
    >>>Questions: Can somebody give some hints what could have caused the
    >>>problem ?.

    >>
    >> A PCI slot has 3 voltage rails: Fixed 5V, Fixed 3.3V and "VIO". VIO is
    >> meant to be used for the PCI signal drivers on the card and can be either
    >> 3.3V or 5V, depending on the chipset. There's a notch to differentiate
    >> the allowed VIOs:
    >>
    >> http://www.adexelec.com/faq.htm
    >>
    >> So either your board has 5V on VIO and the card works only with 3.3V, but
    >> has the wrong notch or the card internally connects VIO to 5V (I've seen
    >> that, too). That is a nice short between the 3.3 and 5V rail...
    >>

    >
    > I'm guessing that the PCI card had a short on it. It seems unlikely that
    > anyone would be selling a card that was keyed wrong, it's much more likely
    > that there was a manufacturing defect.


    Agreed, the keying problem would show up in the required testing to produce
    the FCC required DOC.

    --
    JosephKK
    Gegen dummheit kampfen die Gotter Selbst, vergebens.Â*Â*
    --Schiller

  6. Re: Installation of PCI board resulted in meltdown of systemboard

    Hi,

    Thanks very much for the answers on this thread, i will be in touch
    with the supplier of this card about this unfortunate incident.

    Best wishes for 2007 !.

    Regards,

    Erik


+ Reply to Thread