MEM11 Status Update

Guy Sotomayor ggs at
Wed Jan 21 22:20:08 CST 2015

On 1/21/15 9:29 AM, Noel Chiappa wrote:
>      > From: Guy Sotomayor
>      > I'm building a "LED" board that attaches to the MEM11. ... I couldn't
>      > resist more blinkin lights.
> Oh frabjous day! Das blinken-litz!! Excellent!!!
> Now I'm _really_ going to want a stack of these... :-)
>      > The LED board will display all of the indicators for the RF11
>      > controller in a layout similar to what was on a real RF11. ...
>      > It's optional and will connect to the MEM11 with a small cable (I plan
>      > on using RS485 drivers for the cable) so the LED board can be a fair
>      > distance away.
> Hmm. Looking at the original RF11 display panel, some of it may be impossible
> to really emulate; e.g. it displays the contents of the shift register used
> to assemble words from the bit stream being read from the disk; it's probably
> not feasible to emulate that... And of course, many of the error bits only
> make sense with an actual disk (e.g. the various Timing Track errors).
No, it'll be close and look similar but won't have all of the bits that 
are on the original RF11.
But it will have other stuff (such as the memory address) so I think the 
same number of bits
will actually be used.
> How much bandwidth will you have from the board to the display panel anyway?
> I assume there will be latches on the display panel, and one sends commands
> down the link to turn given latches on/off?
No, the way that it works is that 144 bits will be shifted out and then 
all latched at once.
> Looking at the original display panel, although it had the extended memory
> bits, it didn't show the memory address (it does have the disk address).
> Maybe we could put that where the shift register display used to be?
Yea, I'll do something like that as I mentioned above.
> (And speaking of the original display, I couldn't find a picture of one - does
> someone have one, or, better yet, an actual RF11, so as to copy the look of it
> as closely as possible?)
There's a pretty good diagram in the RF11/RS11 System Manual (on 
bitsavers of course) that shows
the layout.
>      > (I haven't looked at doing an overlay to make it look pretty)
> Oh, yes, please do!!!
Eventually.  ;-)
>      > it'll be somewhat generic (4 rows of 36 LEDs).
> I counted up (no doubt you did the same :-), and on the longest row (at the
> top), there are 30 lights, plus 4 empty spaces for the gaps between blocks,
> so 36 is good. Will you be grouping them in 3's, with slightly larger gaps
> between each group of 3? That would really maximize the ability to look just
> like the old RF11 display panel.
No, it'll be just like the DEC panels.  All of the LEDs (in this case) 
will be evenly spaced.  The
overlay masked the bulbs that weren't used.  The original display panel 
that DEC produced was
generic and used for a bunch of different controllers.  It was what it 
was plugged into (parallel
connections) and the overlay that determined what it was.
> And warm white LED's, please (of course :-).
I'll probably make that an option.  ;-)
> I'm really going to have to get going on my UNIBUS PDP-11's. I've mostly been
> working on the QBUS ones, but if I can have blinking lights... :-)
Blinking lights are *good*.  ;-)
> I'm already planning on how I can put /, /bin and the pipe device on separate
> RF platters, so I can watch the lights and see what the machine is up to.. :-)

TTFN - Guy

More information about the cctalk mailing list