Hello again,

Continuing the thread about movement- or position-sensitive computers.

>  I believe Bruce's comment was about a post that is so loose that it
>contacts other traces/pins on the board.  A properly installed post
>will only contact the board immediately around the mounting hole, and
>many board designs RELY on this as a ground connection to the case
>itself.

>David G

Sorry I wasn't clear on that point. Yes, many cases are so poorly
designed and poorly manufactured that the motherboard "hangs" at least
on one side or end. What's more, there are often metal bits sticking up
here and there (used for mounting other-sized motherboards) and these
have been known to short the bottom of the the board. This is
particularly the case when the board is not securely mounted, which
sometimes is not even possible without drilling holes and getting
involved in things one would better avoid. In one other case that had me
stumped for a long time, the board actually touched a bit of the case
ABOVE it when the machine was put into its normal postion. The one
mounting screw was very loose.

For years now we have been clipping the bottom bit off of nylon mounting
posts to use as standoffs to keep the board from touching the case. I
have also used epoxy to glue a threaded brass coupling onto the chassis
so that I could have two screws to fasten in the motherboard. Also note
that the screws should not be tightened down until the slot cards
(graphic, sound, network, etc. cards) have been put in place the first
time. Physical stress is a killer to the board and the cards.

Beware however that too much medicine can kill the patient. We had a
mechanic in the shop who was so disgusted at the quality of the cases
that he "remanufactured" them, putting screw-down posts at all mounting
points on the motherboards. Predictably, this resulted in a number of
the motherboards developing hairline cracks because of stress that
occurred as the temperature in the case changed. Two screws near the
keyboard connector (point of major stress) are adequate as long as nylon
or brass standoffs are used on the rest of the board. Circuit ground for
the board is by the power connector. Witness, a board will run fine on a
workbench not connected to any metal. I don't know whether there is any
additional "antistatic" effect of the posthole grounding or if the
postholes really ground the board at all.

Thanks to David for pointing out my inadequate description of the
problem.

Best wishes,

Bruce

--
------------------------------------------------------------------
        Bruce Boschek - Haus Christiansruh - Nidda, Germany
[log in to unmask] - CompuServe ID 100331,2566
       Using IBM OS/2 Warp Connect and MR/2ice (Registered)
   "Scitum est inter caecos luscum regnare posse."
(It is well known, that among the blind the one-eyed man is king.)
                          - Gerard Didier Erasmus (c. 1465-1536)
------------------------------------------------------------------

                                  -----
       **Need help with PCBUILD mailing list? Send an Email to:**
        Bob Wright<[log in to unmask]> or Drew Dunn<[log in to unmask]>