Grinnel framestore. This seems to be a digitiser
as well (there's
what looks like an ADC on one of the boards), and is mostly
hand-wired, not PCB. The Unibus interface is a DR11-B. I know noting
about this unit
Oh, that takes me back.
Back in the '80s, the lab I worked at had a Grinnell display (I'm
pretty sure that's how it was spelled). I don't recall the hardware of
You're probably right about the spelling. The unit I have is stuck in a
corner of my workshop, and I am not going to climb over assorted other
machines to read the nameplate :-)
its interface, despite having written a BSD driver for
it. It was
connected to a VAX-11/780 (and I think later to a 750).
Mine came off an 11/34 machine. When I got it, there were 2 cables with
it, with a round 'milspec' connectors on one end that fit connectors on
the back of the framewstore, and DEC card-edge paddleboard things on the
other end. Also with it was a DR11-B. On the grounds those paddleboards
would fit the user device connectors on the DR11-B, I've assumed that's
how it was connected.
Ours didn't have a digitizer, as far as I can recall. It was 256x256
I'm nor _sure_ mine does. It's just that one of the boards seems to have
a lot of fast (for the time) analogue comparators on it, which look like
a semi-flash ADC. I don't have schematics, and haven't had the time to
trace them out yet!
(framebuffer memory) or 256x240 (displayed), with a
joystick-and-button
input interface (which might be what your ADCish stuff is all about).
It had capabilities I've yet to see anywhere else. The closest X term
for its display capabilities is DirectColor, but it could do things X
has no way to describe. In particular, if you describe the three major
pieces as three (6-bit-deep) memory planes, three (6 bits in and I
think 8 bits out) lookup tables, and three DACs to drive the three
primaries on the monitor, there was a way to control which memory plane
fed each lookup table and which lookup table fed each DAC, and for each
switching layer, each switch was independent of the others, meaning
that all 27 setting combinations were achievable.
Interesting...
I will have to dig out the docs on my I2S image processor/display
systems. Those machines are stuffed with programmable look-up tables.
From what I remember, the outputs of each byteplane
feed lookup tables,
the outputs of those are added (full adder), the output of the
full adder
feeds more lookup tables, etc. The manual suggests programming 2's
complement tables to do subtraction, log/antilog tables to do
multiplication, and so on.
-tony