Dear Ladies and Gentlemen,
   
   Thanks for the info on BA23! I have got another question that someone
here may have an answer to. I have previously written that whenever
MicroVAX and VAXstation twins used nearly the same system boards, they were
still different in that the MicroVAX version had the video circuitry
omitted. Well, among the stuff that Richard Parobek has given me on
Thursday there are two MV2000s. To test my theory, I have opened them up
and looked at the system boards. I was expecting to see signs of omitted
components (empty pads). I was very surprised to see all components in
place. There were no empty pads! I didn't have a board from a VS2000 in
front of me for comparison, but I have seen them before, and they looked
the same! Since my original theory turns out to be wrong, I'm now
constructing a new one. I have calculated that the frame buffer for a mono
(non-GPX) VAXstation should be only 128 KB, which is quite small. It is
plausible that BabyVAXen never have special memory for this purpose, but
subtract the frame buffer from the main memory instead. In this case there
would be no need to eliminate the video RAM on the MicroVAX version. There
still may be other components that can be omitted, like the video refresh
circuitry. Or maybe the video refresh is combined with DRAM refresh, just
like on BK0010 and BK0011 (non-standard Soviet PDP-11-like systems)? And
there is a question of why the MicroVAX version indicates the absense of
mono video by saying "F_..." during initialization. Well, that might be a
ROM difference. BTW, does that happen if you run the VAXstation version
with a serial console?
   I still have to take the 3100s into account, though. On these the
MicroVAX and VAXstation flavors differ slightly in speed, which suggests
the removal of some cache on the MicroVAX version. Also according to
Antonio Carlini the MicroVAX and VAXstation system boards have different
names, while on 2000s it's always KA410. This suggests that in this case
there are some differences. I'm currently thinking about some serial port
differences. I think that DEC was advertising MV3100 as having 1 console
port and 3 generic ports, just like MV2000. VAXstation versions, on the
other hand, reserve two ports for the keyboard and the mouse, leaving only
two generic ones. This problem is nicely solved on the 2000s: ports 0 and 1
are in the same DB15 as the video, port 2 is a standard RS-232, and port 3
is a separate DB9 with the same pinout as the MicroVAX II console
connector. On the MicroVAX version a special adapter screws into the DB15
and the DB9 and converts ports 0, 1, and 3 to MMJs. On the VAXstation
version ports 0 and 1 and converted to funny keyboard and mouse connectors
by the monitor cable, and port 3 is connected to a printer by a BCC05.
Everything works out nicely. On VS3100, on the other hand, something
strange is going on. Ports 2 and 3 come out of the system unit already as
MMJs, and ports 0 and 1 come out in two ways. First, they are accessible
through the monitor DB15, but there also are funny keyboard and mouse
connectors right on the system unit. On MV3100 all ports should be MMJs,
but I don't see how could they do it. My question to the club is this:
could someone with an MV3100 M10/10e/20/20e (but not M30+) please look at
its back panel and tell me what connectors does it (left to right or right
to left, doesn't matter)? I remember someone saying that he has such a box.
   
   Sincerely,
   Michael Sokolov
   Phone: 440-449-0299
   ARPA Internet SMTP mail: sokolov@alpha.ces.cwru.edu