Wanted: M9312 RX01 bootstrap ROM

From: Ethan Dicks <ethan.dicks_at_gmail.com>
Date: Thu Dec 16 17:15:12 2004

On Thu, 16 Dec 2004 13:19:12 -0500, Ashley Carder <wacarder_at_usit.net> wrote:
> > > The fun wears off after the second or third time
> > > of toggling in that code!
> >
> > Wouldn't VTserver also be a way out here?
> >
> > John A.
>
> Yeah, I could load up the bootstrap code in a PC file and run
> VTServer and have it load and run the bootstrap file. But then
> I'd have to have my PC hooked up as the console. I'm trying to
> run my configuration using all 1970s vintage hardware and
> peripherals.

You could do what we did at Software Results in the late 1970s through
the 1980s...
we needed to downline feed a PDP-11 test bed with our product in it
for test. We had a number of diagnostic programs, both for the
embedded 68000 and the PDP-11 itself (tests like the DMA test required
code running on both CPUs simultaneously). Note that this PDP-11 had
*no* peripherals other than the console and our board under test, and
no operating system, just ODT (an 11/04 for Unibus cards, an 11/03 for
Qbus cards). We had this special console switch that one of the
engineers built that would let us bank in a real VT100, the PDP-11
console, and a port on our VAX. Under normal use, one would have the
switches thrown such that the VT100 was either a PDP-11 console or a
VAX terminal. One would go to the PDP-11, type a few ODT commands,
switch to the VAX, run a special in-house FORTRAN program that would
take a binary and convert it to ODT commands, pausing to let one flip
switches, then... we would pick a _third_ setting of the switches to
gate the VAX output to the PDP-11 input while gating the PDP-11 output
to the VT100, allowing us to watch what was going on. So the VAX
would 'toggle in' the test programs via ODT, saving us the hassle.

AFAIK, the switch was pretty simple - common ground for all 3 serial
ports, with a single-pole/double-throw switch on each Rx and Tx, with
the common lead going to the PDP-11. Up/Up to connect the VT100 to
the PDP-11, Down/Down to connect the VT100 to the VAX, and Up/Down to
enter "load mode". We had two flavors of it - one was hand-made with
switches from Radio Shack. A later version was two AB switch boxes,
internally re-wired, but they used RJ-11 jacks, the same as all of our
serial comms gear (we had Nevada Western patch panels, telco-cable
adapters, DB25 hoods, Hydras, etc.) It doesn't matter _how_ you rig
the switches mechanically, as long as they swap round Tx and Rx
properly.

This way, you can have a cable to a PC that could be used for booting
(you could even run simh on it and feed it from a vintage OS ;-) and
also to a real VT100 or VT52, etc.

One could probably do something similar with 20mA, but with active and
passive devices, it might take a bit of fiddling and an active circuit
or two to keep two devices (the PDP-11 and the (VAX/PC/vtserver host)
from both trying to drive the loop. If you are looking for a PCish
20mA interface, the original IBM 8-bit ISA serial board has both EIA
drivers and a 20mA current loop.

But with all that said, I understand about trying to keep it all
vintage. The right boot PROM would solve a lot of hassles. I have an
ancient Data I/O burner that we used to burn 6309s in (256x8) - I
would be surprised if it couldn't handle the PROMs for an M9312, but
a) I'm n ot home yet, and b) I wouldn't have any blanks. If you are
still looking three or four months from now, and you can locate a
blank or two, I'd be happy to burn one for you for free (you pay S&H
both ways...) I can try to locate blanks for you if you like, but I
couldn't guarantee I'd have any better luck finding them than you
will.

Cheers,

-ethan

P.S. - I may have just found a use for the extra serial ports on my
SBC6120 - a PDP-11 bootstrap server... :-) Set the switches on the
FP6120, throw the console switch on the -11 to 'load mode', then give
the SBC6120 a kick and let it emit the necessary ODT commands...
Would need a trivial app for OS/8, but I _have_ one written in VAX
FORTRAN I could start with. Not sure how to read the switch register
in FORTRAN, though. Might have to use the VT100 I/O stream.
Received on Thu Dec 16 2004 - 17:15:12 GMT