SCSI options for PDP 11/23

From: Tom Leffingwell <tom_at_sba.miami.edu>
Date: Wed Feb 27 23:52:07 2002

On Wed, 27 Feb 2002, Zane H. Healy wrote:

> OK, what I see that concerns me is the fact you're using a 11/03 box.
> Unless I'm mistaken the 11/03 was only Q16, which has me wondering how
> you're even running a 11/23 CPU, which should be Q18, in it (it's possible
> someone upgraded the backplane to be 18-bit).

It did have the M8186 11/23 CPU in the 11/03 chassis when I got it, not
that that means anything. I did change the memory module to a 128KW
MSV11-LK.

There's an older Q-bus VAX system at the University that will probably be
taken out of service soon, and there's a chance I can grab its SCSI
controller (I'm not sure exactly what it has, but I know it has one).
 
> I took a quick look and it looks like CMD made SCSI controllers that would
> work in either a Q18 or Q22 system. It also looks like Viking controllers
> only work in a Q22 system.
>
> As for your Fortran programs, I would think they would work on V5.x, but am
> not sure. The best people on the list to know about that would be Megan or
> Jerome.
>
> Is this a hobbyist system, or is it a production system?

Its mostly a hobbyist system, in that I paid for it personally, and if it
weren't for it being a hobby, I would have given up a long time ago. :)

Its kind of hard to explain, but I'll try. I work for a university on a
National Science Foundation funded project that provides satellite
communication to scientists at the geographic south pole, and a few other
assorted places in the Antarctic. Normal commericial satellite
communication isn't affordable to the scientific community, so the
operation uses old satellites (ranging in launch year from 1967 to 1977)
that are no longer usable for their original purpose, such as imager
failures in the case of the older GOES satellites. We also use ATS-3, the
oldest continously operating satellite in the world. It took the first
color pictures of the earth from space in 1967. Also, being low on
station keeping fuel, which causes the orbit to drift north and south of
the equator, so that it its no longer geostationary is another reason.
That's good for us, because at geostationary orbit, you can only see down
to about 80 degrees south. With the drift, you can see the poles for
about 6 hours a day.

In order to perform maneuvers (fire the jets) on a spin-stabilized
satellite, you have to synchronize your commanding with a reference point
(usually the sun). If you were in the 1970s and wanted to design a
reliable "portable" computing system for each launch site required around
the world, what would you use? You guessed it, a PDP-11. The software
was on an RX02 along with RT-11.

Anyway, we built our own command equipment awhile ago with a small
microcontroller. This was intended for commanding where the timing isn't
critical, such as turning things on and off, switching sensors, etc. The
attitude of a spinning satellite precesses slowly over time, and you
correct it by firing jets in a synchronized manner, which we can't do with
our equipment. The attitude starting to be off enough to where it needs
to be corrected. Lacking the funding to develop something new, combined
with everyone assuming I wouldn't be able to revive this system, I decided
to rebuilt it on my own. (We only had a few pieces of the original system
that attached to the DMA and PIO interfaces, which didn't work, and none
of the PDP equipment). It was working pretty well until the RX02
incident. I was hoping I could scrape up a SCSI controller from somewhere
around the university, so I could have a longer-term storage option. So
in the next couple of months, it will get used for a "production" purpose,
but then it won't be needed for quite awhile.
Received on Wed Feb 27 2002 - 23:52:07 GMT

This archive was generated by hypermail 2.3.0 : Fri Oct 10 2014 - 23:34:49 BST