[conspire] RS-232-C (was: Re: DE-9, not DB-9 (was: conspire list hacked?))

[Michael Paoli]

> From: "Rick Moen" <rick at linuxmafia.com>
> Subject: Re: [conspire] DE-9, not DB-9 (was: conspire list hacked?)
> Date: Mon, 4 Mar 2019 16:33:22 -0800

> have locked the serial port to 115,200 bps rather than 38,400.  (I don't
> know if any RS-232C ports supported 224,000 BPS, and no longer care.)

Another bit of trivia for RS-232-C spec.
Max speed one could do without bending(/breaking) spec. was "38.4"(k),
or more specifically 38,400 - at least of the standard
speed doublings up from 300.  And why?  Had to do with the maximum
rise/fall times on the voltages plus the minimum thresholds of level
(and hold time probably also) for the voltages, to have a legitimate
to-specification signal.

"Of course" the specification got, uh, bent a lot, and higher speeds
were pretty common (at least on the wire) in fairly quick* order.
Much of the earlier stuff only had top wire speed of 9600, if I recall
correctly - and I think that was due to spec limitations on the common
(TTL to RS-232-C) level conversion ICs that were most commonly used
at that time.  Or maybe(?) it was 2 or 4 x that, but no higher, until
subsequent faster speed capable chip sets came along.
There are also successor standards that support higher speeds, and
are mostly (at least signal-wise) backwards compatible with
RS-232-C.

*quick ... well, if one thinks from about 1969 to about 1985 as quick order.

Oh, other bit I forgot ... Electronics Industry Association (EIA) - wrote
and copyrighted the RS-232-C specification.  That's also why at least some
of the equipment also had "EIA" marked by their RS-232-C ports - to help
identify what they were.

And just because it's a DB-25 doesn't mean it's RS-232-C.  Could be
"Centronics compatible" (signal levels) printer port, or SCSI!  So,
generally best to never presume.  Some would also do funky things on
DB-25 connectors.  E.g. I recall I had a UPS ... as for RS-232-C on that
port (and yes, as well documented in the manual) ... yes, for the
customary signal ground, data transmit and receive lines (on primary channel).
But others used for very different things - e.g. EPO signal to UPS (N.O.,
close contacts, UPS shuts down cold immediately).  I think it had some other
non-RS-232-C signaling on there (UPS alarm and/or on UPS power).
So, case like that, one wouldn't want to just connect a full 25-line straight
through cable from UPS to terminal - in fact the manual very specifically
spelled out to NOT do that.