[net.micro.cpm] need a break key

edens@ALMSA-1.ARPA (12/18/84)

I noticed someone made a request for information about a "break" key the
other day.  I also need this on my NEC APC.  I haven't found a work around
for it and would appreciate any information from you experts out there.

thanks in advance

Crede Edens

david@ukma.UUCP (David Herron, NPR Lover) (12/24/84)

You could do what the 4.2 tip program does.  For break it lowers the
baud rate to like 50 baud and sends characters out.  Gauranteed to make
framing errors on the receiver.  And it works too.

jeff@abnji.UUCP (jeff) (01/16/85)

[I'm new to USENET,so please be kind]

	Causing a BREAK without a break key is tricky since BREAK is not
a character but a line condition.  An active RS232 line (transmit or
receive data) is normally in the MARKING state (quiescent state)
(binary 1, voltage between -3 and -25 volts).  When a character is sent,
the line sends a start bit (goes SPACE, binary 0, voltage between +3 and
+25 volts), then the data bits (and parity if enabled), then a stop
bit (back to MARKING).

	A framing error is caused when the end bit isn't received
when it is expected, usually suggesting mismatched speeds.

	A break condition is when the line is put in the SPACE state
for the time it takes to transmit a character (including start and
stop bits) and thus cannot be confused with a character transmission.
Some serial communication chips will detect the start and end of
a break (since it may last a looooong time) and interrupt on both.
Others just give a framing error with 00h data.

	The UNIX Administrator's Manual under TERMIO(7) under IGNBRK
states that a break condition is a framing error with data 00h.  Non-null
data causing a framing error is considered a framing error.

	A break key causes the line to go spacing for about 1.1 to 2x
a character time length (usually a long time - around 200 milliseconds).

	Your trick of going to a slower speed may work if the software
treats framing error as a break (regardless if the data is 00h or not).
If they are treated differently, then you must transmit a NULL
character, usually control @.  This will be undistinguishable from a
break to all but the fussiest of receivers.  This also points out
that a framing error on a null character will be misinterpreted
as a break.

	I trust this definitively answers your question.  Ask anything
more technical and I will refer to the EIA Standard for RS-232-C
"Interface Between Data Terminal Equipment and Data Communication
Equipment Employing Serial Binary Data Interchange". So there!

			+-----------------------------------------------+
			|  Jeff 'oh my gawd - it's one of THOSE!' Skot	|
			|     at beautiful downtown Somerset NJ		|
			| 	      AT&T Info Systems			|
			|  		..!abnji!jeff			|
			+-----------------------------------------------+