[GreenKeys] 1 or 1.5 stop bits?

[Jim Haynes]

I've given this lecture a few times before, so if you are tired of it
just delete.

The principle of start-stop synchronization, which makes teletypewriter
operation simple at the expense of a little slowness, is that the
receiving mechanism must come to rest in between characters.  Western
Electric built some machinery using electromechanical sending and
receiving distributors, and put the sending and receiving distributors
on the same shaft.  This doesn't allow any time for the receiving
distributor to stop between characters, so they added a relay to slow
down the transmitting distributor trip a little bit.

Morkrum, ancestor of Teletype, ran the receiving distributor a little
faster than the transmitting, so the receiving distributor always got to
tne end of its rotation before the transmitting distributor finished
sending a single unit stop pulse.

Western Electric measured the time delay introduced by the relay and found
it to be 0.42 as long as a normal pulse.  Thus to have their machines
interoperable with Teletype they demanded that the Teletype transmit an
elongated stop pulse 1.42 times as long as the normal pulses.  The
source of this is a very obscure old document that someone made available
to me.

Teletype equipment has always been capable of operating with 7.00 unit
code because of the different sending and receiving shaft speeds, but
the elongated stop pulse demanded by the Bell System became the U.S.
standard.  Long after the Western Electric equipment was no longer in
use so the need for the elongated stop pulse had ceased.

Western Union didn't like the delay added by the longer stop pulse, so
they tended to operate machines set up for 7.00 unit code.  For Teletype
equipment this means no change to the receiving selector shaft speed, but
changing the gearing and camming on the keyboards and tape transmitters
to eliminate the extra-long stop pulse.  I you happen to get old W.U.
equipment you may find the keyboard gears and cams different from the
ones in non-WU machines.

Somehow, and I don't know how, the standard speed for Teletype equipment
in the U.S. was nominally 60 wpm, with a 22 millisecond signal pulse
length, speed of 368 operations per minute.  22 x 7.42 gives 163.24
milliseconds as the minimum length of a character, This gives 6.12595
characters per second, or 367.55 characters per minute.  And at 6 
characters per word (including the space between words) that's 61.26
wpm.

Keeping 22 ms pulse length for the sake of interoperability, the 7.0
code gives 154 milliseconds as the minimum length of a character.
That gives 6.5 characters per second, or 65 wpm.

Now in Europe they use 50 baud, or 20 ms per signal pulse.  I have no
idea if the designers of European teleprinters decided they needed an
elongated stop pulse and if so settled on 1.5 unit length.  50 baud
and 7.5 unit code means 150 milliseconds per character, or 66 wpm.

>From this you can see why it's not a good idea to throw around words
per minute casually, as 7.00 code 45.45 baud is almost the same wpm
as 7.50 code 50 baud, yet the two are hardly interoperable.

When we get into electronics it's obvious that it's a lot easier to
generate 7.50 code than 7.42.  There's no problem receiving 7.00
code.

So if you are talking to Teletype equipment it should not matter whether
you generate 7.00, 7.42 or 7.50 code.  If you are talking to other 
equipment I will not offer an opinion whether the elongated stop pulse]
is needed.

 	---

 	"Ya can argue all ya wanna, but it's dif'rent than it was."
 	"No it ain't! No it ain't!  But ya gotta know the territory."
 		Meredith Willson, The Music Man