Probably not a huge amount. The big thing that the Decsystem20/20 uses
tape for is installing operating systems and microcode. The trick though
was that there were two different types of TM03 formatters, one could
handle 16 bit data and the other had a "bit fiddler" that could handle
18 bit data. I know you could not use a 16 bit TU45/TU77 on a 2020 but I
don't recall if an 18 bit TU45/77 would work on a pdp11 or Vax.
If I get that desperate in my life though I could always sit down and
figure out a way to re-write the console microcode to just boot from a
TU58 by packing 144 bit "words" (4 36 bit) into 9 16 bit words and be
done with it. :-)
CZ
On 10/5/26 22:57, Jim Bender wrote:
Is there much demand for tape support? It probably wouldn't be that
difficult to add it, but seems a small number of use cases for it...
Using it of course requires a Massbus controller configured for tape
and a
separate emulator if you want to run both disk and tape.
Jim
On 10/5/26 21:25, Chris Zach via cctalk wrote:
Hi Jim!
Just watched the video. Impressive no doubt.
Fitting it on the 3 5904 board would be perfect, Then it just sits
there in the RH11/RH70/Whatever and you don't need to deal with the
ribbon cables or a nice big MASSBUS bulkhead and cable.
I would love to keep an eye on this, thank you for doing it. Tape
emulation, specifically a TU45 with a TM03 with the 18 bit fiddler
would be optimal, but just getting the drive going would be a major
step. We can load the microcode off an RM03.
Chris
On 10/5/26 21:14, Jim Bender wrote:
Hi Chris:
Don't see why it wouldn't work on a 2020, but someone with a 36-bit
machine
will need to test it.
I think your expectations of the possible drive shrinkage factor are
a bit
off... :-)
While a solution that fits inside a terminator would be convenient,
it just
is not possible. You could, however, run ribbon cables directly
from the
Massbus transceivers (usually M5904s) to the emulator and hide the
emulator
inside the rack somewhere.
My eventual goal is to reduce the emulator to a three-board
"sandwich" that
takes the place of the M5904 trio on systems that use it. That
eliminates
the differential drivers and receivers and simplifies the
electronics quite
a bit. Also solves the problem of "where do I put the emulator at?"
:-)
Something like this:
https://www.dmv.net/mbe/RS06_Setasi.pdf
Jim
On 10/5/26 19:57, Chris Zach via cctalk wrote:
Jim:
Nicely done. If it runs on a 2020 this will finally be a
game-changer. :-)
Can it be built into a Massbus terminator plug? That would be a
proper mounting....
Chris
On 10/5/26 18:00, Jim Bender via cctalk wrote:
It's been a while since my initial posting about this project but
it is time
for an update.
This has been mostly a spare time project, and you know how life
tends gets in
the way sometimes...
The work itself has been a bit of a "maze of twisty little
passages, all
alike" with many potholes and obstacles, but progress has been made.
The video below tells the tale:
https://youtu.be/KwaNschJg6w
Synopsis:
A real PDP-11/70 boots RSTS/E V10.1 through it's unmodified RH70
controller
with the data coming from the Massbus emulator.
Operation is rock solid and so far have not been able to break it.
It also
passes the ERHAE1 XXDP diagnostic without errors.
We now have a viable option for modern storage for Massbus systems.
Still to do is performance optimizations, hardening against some
edge cases,
and a large amount of code cleanup.
I only have a PDP-11/70 to test with, but it *should* work with
any Massbus
system. Adherence to the DEC Massbus Specification was strict and
every
attempt made to make it behave as a proper Massbus device rather
than just
coding things to make the RH70 happy.
I am kicking around the idea of putting together a blog
chronicling the
various trials, tribulations and milestones encountered along the
way if
anyone is interested in reading the gory details....
Cheers!
Jim
