🚀 go-pugleaf

In article <1990Mar5.0...@cs.rochester.edu> k...@cs.rochester.edu writes:
>I wonder how many other people used a UART as I once did.

>                     But the board didn't have a timer chip.
>So I wrote characters to the
>serial chip. The serial data just fell off the output pin, of course.

I've had to do things like that.

(Actually, now that you remind me, I think that's how I discovered the Altos
586 serial port's 50 bps rate is actually 5 bps - I was looking for the
longest time interval with the fewest output characters!)

The most interesting application I've thought of so far for a UART is as an
8-channel multiplexor:

serial input channels	(mind you, this scheme would need to clock at
  A B C D E F G H	something like 384K bps for eight 9600 bps lines...)
  | | | | | | | |
+-+-+-+-+-+-+-+-+-+                           +-----------------+
|     U A R T     +---------------------------+     U A R T     |
+-----------------+   multiplexed data line   +-+-+-+-+-+-+-+-+-+
                                                | | | | | | | |
(and no, I haven't done it.  Yet.)		A B C D E F G H
					      serial output channels
-- 
"Logic is a little bird that sings in the meadow."	-Mr. Spock
cha...@C3.COM	{decuac.dec.com,cucstud}!c3pe!charles	ex::!echo Boo:

In article <8445@c3pe.UUCP>, charles@c3pe.UUCP (Charles Green) writes:
> serial input channels	(mind you, this scheme would need to clock at
>   A B C D E F G H	something like 384K bps for eight 9600 bps lines...)
>   | | | | | | | |
> +-+-+-+-+-+-+-+-+-+                           +-----------------+
> |     U A R T     +---------------------------+     U A R T     |
> +-----------------+   multiplexed data line   +-+-+-+-+-+-+-+-+-+
>                                                 | | | | | | | |
> (and no, I haven't done it.  Yet.)		A B C D E F G H
> 					      serial output channels

  It's been done. Consider the Gandalf PACX IV data switch - 256 terminal
ports to 128 computer ports with _NO_ UARTS in the data path (true, there
are 4 for the "connection engine" and 2 for the console and status ports.
All the serial lines end up as bits on parallel ports which are then scan-
ned at 38.4 KBPS and copied (more or less sucessfully) to their destina-
tions.

  Here's my entry in the "silly UART trick" contest:

  The Volker-Craig VC4404 "Chat" terminal was a microprocessor-driven
replacement for the VC404 terminal. The 4404 had 2 UARTS - one for the
host's serial interface and one to receive data from the keyboard. The
KB UART had no use for the transmitted data pin. During development,
it was determined that there wasn't enough CPU power (the terminal
only had an effective 1/2 6502 CPU, as alternate cycles were given to
the display) to beep the beeper. So, they programmed the transmit side
of the KB UART for 1200 BPS and wrote x'55' characters to it to pulse
the feeper element!

	Terry Kennedy		Operations Manager, Academic Computing
	terry@spcvxa.bitnet	St. Peter's College, US
	te...@spcvxa.spc.edu	(201) 915-9381

> 
> The most interesting application I've thought of so far for a UART is as an
> 8-channel multiplexor:
> 
> serial input channels	(mind you, this scheme would need to clock at
>   A B C D E F G H	something like 384K bps for eight 9600 bps lines...)
>   | | | | | | | |
> +-+-+-+-+-+-+-+-+-+                           +-----------------+
> |     U A R T     +---------------------------+     U A R T     |
> +-----------------+   multiplexed data line   +-+-+-+-+-+-+-+-+-+
>                                                 | | | | | | | |
> (and no, I haven't done it.  Yet.)		A B C D E F G H
> 					      serial output channels

Around 7 years ago Solana Electronics sold a multiplexor that did just that.
(They may still be around.)  There was this large chip on the board which
usually had the lettering removed, but once, after a warranty repair, we
got one back with the lettering intact.  It turned out to be a UART.  The
multiplexed data line was run with balanced RS422 drivers up to 1000 or 2000
feet.  At $400 a box, it was a cheap high-speed (at the time) connection 
method for long in-building runs, and included manual port sharing within an
office.  We had about 24 in service, and there is still one next to my desk for
sharing a test printer between 2 mainframes.

> -- 
> "Logic is a little bird that sings in the meadow."	-Mr. Spock
> cha...@C3.COM	{decuac.dec.com,cucstud}!c3pe!charles	ex::!echo Boo:
-- 
-------------------------------------------------------------------------
Steve Roseman
Lehigh University Computing Center
LU...@VAX1.CC.Lehigh.EDU
 

How about this (I don't know if the PROM is available, but something
equivalent can be cooked up:

(----) 1 bit data path
(====) 8 bit data path

	    +---------------+   +------+
----<->-----|  UART         |=<=|      |
	    +---------------+   |      |
	     ||                 |      |---- transducer.
	    +---------------+   | ADC  |
	    | 256 by 1 PROM |->-|      |
	    +---------------+   +------+

With an appropriate delay logic to trigger the uart when the value back from
the ADC is valid. You write a byte out that matches the only valid '1' in the
PROM, and that triggers the ADC and sends back the value after a short delay.
With some simple coding, you could put (say) 128 7-bit ADCs off a single RS232
port, so long as you don't have to poll them too fast. They should be pretty
cheap to build in quantity, too.
-- 
 _--_|\  Peter da Silva <pe...@sugar.hackercorp.com>.
/      \
\_.--._/ I haven't lost my mind, it's backed up on tape somewhere!
      v  "Have you hugged your wolf today?" `-_-'