Brad Parker wrote:
But I think the language misses the point. It
doesn't matter what
language it's written in. The important question is what will it run
on. There are still pdp-11's out there running applications every day
and it doesn't matter what they were written in. It does matter if the
11/44 power supply blows up, however :-)
Not much though. How long would it take to bodge up a working power
supply for a PDP-11 of any flavour? It's not like it's incredibly
critical, unlike modern PC motherboards.
I suspect the original mention of 11/44 was done for a reason. The
PDP11/44 poser supply is very complicated inside, it's really 3 separate
SMPSUs running off the stame 400V DC supply (rectified mains), with a
board of control logic at one end to do things like turn off the CPU and
Unibus power, but beep the memory alive. Oh yes, there's an optional
battery backup unit for the memory supplies. And even a full-H driver for
the special cooling fans (and that is also supplied by the battery unit
if fitted).
However, DEC schematics are pretty clear, and there's nothing that odd in
the PSU. I think I could keep one going, and repair it in less time than
it would take to bodge in some modern, less well made and less well
documented replacement.
However, if you're the sort of peron who'd afraid of coke-can sized
capacitors charged to 200V, then there's noting that odd about the
outputs of the 11/44 PSU,. There's a 5V rail (120A or so) for the logic
and +/-15V rails for the Unibus slots. Another +5V and +/-12V rails for
the memory (there's no good reason why the 2 5V rials couldn't be linked
if you're not keeping the battery backup). Maybe a -5V.
You'd lose the battery backup ability, you'd lose the ability to power
down the CPU and Unibus but keep the memory on, and you'd better change
the fans for something more convenitonal (12V DC would be the obvious
choice).
But yes, you could bodge in another PSU without too much hassle.
-tony