Sorry if you had something to do with the design of the DZ11. It does smack of the
"We have to sell this as a timesharing system, so what's the least we can get away
with doing?" It's pretty much all the joy of 8 DL11s stuck together on a board, with the
added pleasure of "which port is that interrupt for?"
This wasn't limited to multiplexors. The RB730 was the disk equivalent - "We have to
sell this VAX with disk storage and a way to back it up, what's the cheapest thing we
can come up with?" and you get the RB730 which talks to only one R80 (not RM80,
not RA80, just R80) and only one RL02.
I looked an an original DH11 and "ran away screaming" - but that was substantially later
(1982-ish). Having designed 8-port multiplexors with FIFOs and shared memory (instead
of DMA) on an S-100 card by then, the low density and high complexity of the DH11 ap-
peared to be incredibly primitive. But the technology had advanced by orders of magni-
tued in the intervening years.
I ran some Emulex DH emulators on some of my systems. IIRC, each distribution panel
was "smart" and just made more ports appear on the host adapter board.
You're correct that the DHU/DHV require rather different drivers that the original DH11 -
I wasn't aware of that - I figured they were single-board shrinks of the DH11 like the
clones. I guess DEC took the opportunity to eliminate some of the DH11 quirks like the
silo alarm and changed the driver interface.
I have a Camintonn DHV11 clone (8 ports on a dual Q-bus card) in my BA123 11/93
system, so I gravitated toward the DHU11 as the emulated multiplexor of choice in the
version of simh that the PiDP-11 runs. As I understand it, the Blinkenbone mods can't
be accepted upstream, so PiDP-11 owners are stuck with that version (or doing a lot
of work to get them into their own fork of Open-Simh).
Once you get into VAX space, all sorts of odd multiplexors appear - the DMF32 is quite
nice as a multifunction board (8 serial ports, an LP11 port, and a synchronous port
that often served as a BITNET link). I assume the reason it never got drivers for any
PDP-11 operating systems was simply that by the time it came out, most development
was for VAX and the PDP-11 operating systems were slowly at the start of "maintenance
mode". Although maybe there is something VAX-specific to the hardware.
The DMZ32 is a true oddball - what we have there is a mostly (but not completely) func-
tional T1 controller re-purposed as a multiplexor. Transformer-coupled at both ends,
with a powered distribution panel that can be located up to a mile away using 2-pair
cable. Excellent for environments like factories where the computer room is some dis-
tance away from where the serial ports are needed, with complete isolation between
the two.
ᐧ