Strange problem with HP 9100A

From: Tony Duell <ard_at_p850ug1.demon.co.uk>
Date: Fri Jun 20 16:35:31 2003

> Hello,
>
> I just acquired an HP 9100A in great condition with no signs of abuse but it
> has a strange problem: although the numeric input keys and four functions
> work fine, the transcendentals and other keys on the leftmost pad do
> unexpected things.
>
> When in the radian mode, many of the keys (like sin x, cos x, tan x, e^x)
> perform an x<>y, but even more strangely, in the degree mode pressing these
> same keys blanks the display and hangs the machine. After such a hang, when
> I cycle power the display lights up with 9.999999999 99 in the x register.

This is a classic case of ROM address failure (or at least that's what
I'd go for on a 9100B -- I've not been lucky enough to find a -A yet).
Take the processor chassis out, flip it over. On the bottom, there are 4
PCBs soldered together. Across the back is the address decoder, in the
middle is the ROM iteslf, down each side are the sense amplifiers.

Most likely the problem is one of the transistors on the ROM PCB. To get
to it, remove the sideboards that link the ROM to the gating PCB, then
undo the screws and remove the complete ROM assembly. You can now get to
the transistor connections to test them

Take great care when working on the ROM PCB. It's a 16 layer (or so) PCB,
and the layout is critical (tracks looping or not looping each other
define the contents of the ROM). So if you rip out a via, there is
probably no way to repair it!

Note those transistors seem to be HP-selected so you may have to
experiment to find something that works...

If those transistors are all good, start investigating the circuitry on
the address decoder PCB.

-tony
Received on Fri Jun 20 2003 - 16:35:31 BST