I went digging through some papers and found some information Ann David
from TAMU faxed me last June (when the backplane problem first came
up). Although the manufacturer of the backplane, Schroffs, was unwilling
to send schematics (claiming proprietary). They did send Ann a
multi-lingual data sheet. I just reread it and noticed something I must
have overlooked before.
** The terminators are switchable from active to passive termination. **
Switching termination from active to passive is done using jumpers
labelled P-A located at both ends of the board. Switching the jumpers
from A to P changes the termination to "passive".
Reminder: the VME data signals are terminated at +3V. I presume this can
be done passively with a simple divider off the +5V supply as well as
with the L272M op-amp circuit which provided an active level. The
advantage of the active level is that it is roughly 1A lower current
draw. Considering how much we already, this is no big savings. (Although
we are drawing a fair fraction of the maximum capacity of the +5V
supply... that is another story).
Anyway, the bottom line is: next time you are debugging backplane
troubles, consider moving the jumpers to the passive setting. In fact,
if you trace it to a blown L272M op-amp, do not desolder the chip and
replace. Instead, just convert that backplane to PASSIVE. However, I
see no reason to change the configuration of the backplanes now in use
(until they show signs of trouble).
Knowledge is power,
Ed