Ok, here's a preliminary inspection. The ROM tray and the ROM
modules use the pin-and-through hole connection scheme that was
used by HP at the time; this type of interconnection between boards
can be seen in other HP machines such as the HP97 and the HP75.
While the pins are somewhat delicate, this system seems to provide
a good connection, especially since surfaces are gold-plated.
Now enter the HP85 ROM modules. The pc board that they are mounted on
is of the laminated-fiberglass-looking variety. Looking inside
the through-hole, it seems that the miriad layers in the pc board
induce some rugosity in the walls of the through-hole, and these
imperfections can be seen also in the gold plating of the walls.
Indeed, it doesn't seem like the surface of the walls is very
appropriate for plating; I bet that chunks of plating can be scraped off
easily with just a little friction. I think that I need to look inside
the holes with a microscope and a good light; proper illumination
is going to be a problem, though.
If I had a tin through-hole riveter I would try sleeving the holes,
but I don't have one. Anyway, the goal right now is to rule out
bad contacts.
carlos.
Joe wrote:
>
> At 06:19 PM 7/10/00 -0400, Carlos wrote:
> >
> > By the way, this weekend
> >my I/O ROM started to go flaky; on power up, it reports an error 112,
> >which is a checksum failure. I remember that this happened before and
> >reseating fixed it, but I haven't had luck this time. I tried other
> >slots in the ROM drawer too. Guess I need to pry open the ROM capsule
> >and have a look. This is weird, because both the pin and the through-hole
> >that make the contacts are gold-plated; they were made to last. I am
> >afraid that this is a case of bit rot :-( .
>
> Keep me posted about what you find. I've only seen bit rot in the HP
> 41 plus in ROMs. I've never seen it in other HP machines.
>
> Joe
--
Carlos Murillo-Sanchez email: cem14_at_cornell.edu
428 Phillips Hall, Electrical Engineering Department
Cornell University, Ithaca, NY 14853
Received on Tue Jul 11 2000 - 10:11:16 BST