Bob McConnell wrote:
James Knott wrote:
NoOp wrote:
On 05/07/2008 11:59 AM, James Knott wrote:
mike scott wrote:
On 6 May 2008 at 17:15, James Knott wrote:
Jerry Feldman wrote:
...
It was 1.5 stop bits back in 1976 when we were integrating a
1200bps
modem into our POS system. I'm not talking about Baudot code. In
any
...
It was never 1.5 stop bits for ASCII devices. It was either 2
stop bits at 110 b/s or 1 at 300 & above. It wouldn't hurt
anything, other than
I'm not sure that's correct. I'm pretty sure it was an option (1,
1.5, 2) on VAX serial ports.
The option may have been available, as it was on the 8250 UART used
in
I've a few of those, and 8251 USART's out in the garage... I know the
8251 could be configure to allow 1.5 stop bits, can't recall if the
8250
also had this option. However, 1.5 was only used for characters with 5
information bits, so it wouldn't apply to standard/modern modems
anyway.
On the 8250 and later, you specified the number of data bits and stop
bits separately, so it was possible to have any combination of data
bits and stop bits. However, in all my years of experience with
working with this stuff, I have never, not once, seen 1.5 stop bits
used with 8 bit codes. It's always been either 1 or 2 stop bits,
though 1.42 and 1.5 were common with 5 bit codes. I've also never
seen a UART that could do 1.42 stop bits, which is found only on
mechanical devices. A UART would be configured for 1.5 stop bits,
which isn't much of a difference from 1.42. So, while it is
technically possible to configure for 8 data bits and 1.5 stop bits,
it's never, to my knowledge, ever been done in production systems, as
it doesn't comply with any of the standard codes. As I mentioned in
another note, this is one area, having worked for 36 years in data
communications and computers, that I have a *LOT* of experience. In
fact, when I designed & built that 8 port serial card I mentioned
earlier, I found a bug in the 8250 that National Semiconductor didn't
know about. I even got a nice letter from them for finding it.
The fractionally longer stop bit was a carryover from the mechanical
teleprinter days. Before synchronous motors became common, shaft
speeds were controlled by a governor. No matter how carefully they
were adjusted, there was always some variation between units. So the
stop bit was lengthened to allow the slower device to stay in sync
with each character. I believe this applied to all devices up through
the early Model 28 equipment, somewhere in the mid 1950's. The
extended bit was added to the transmit side of early electronic
interfaces as well, just in case there was a mechanical device at the
other end. Variations included 1.42, 1.5 and 2.0 bits.
For more information about mechanical teletypes, here are some web
sites and an active mailing list for former TTY repairmen like myself.
I am also a former TTY repairman. As I mentioned in another note, I
started in the datacom business, some 36 years ago, overhauling Teletype
models 32, 33, 28 & 35 (I could overhaul 2 M32/33 machines per day or 1
M28/35). Two stop bits were used for ASCII machines at 110B and 1 at
300B and above. The newest mechanical one I saw, though never worked
on, was the Model 37, which IIRC could do 150B and also used 1 stop
bit. However, that was an unusual speed, which I never saw elsewhere.
I also worked with Texas Instruments Silent 700 terminals, which ran at
300B. I worked with modems, computer serial ports and many other
devices, over the years.
--
Use OpenOffice.org <http://www.openoffice.org>
---------------------------------------------------------------------
To unsubscribe, e-mail: [EMAIL PROTECTED]
For additional commands, e-mail: [EMAIL PROTECTED]
