>I don't think so, remember they didn't have fifos, internal baudrate
>gens  etc. in those days... All you need is a shiftreg, some clock
>dividers and a pretty simple state machine - rather less than even a
>4004. I'm sure you could do it with less than a dozen 74TTLs

>There was also a Texas Instruments chip that was pin compatible to these.
It had
>a problem operating at high speed (as in 9600B was high speed) and would lock
>up. The system we had these in was connected to a Burroughs machine which ran a
>polled network, so in went a timer that reset the chip if there were no polls
>received over a certain period.... Then we got some 6402 chips and the problem
>was solved properly.


Well, I did build my own UART and it took about 10 TTL chips.  It
had a bug, too, because I jumped from the parity bit to the stop
bit without waiting.  The state machine was just a counter.  The
baud rate was done with a 555 and a 10-turn pot.  It was NOT
stable but I could "tune" it each time until it worked.

I had seen these all-in-one chips called "UART"s at swap meets
but they were rare and I didn't get to the swap meets all that
often.   Finally I decided to get one (like $6) and so made
myself a pc board and assembled the thing.  It worked, too.
Until one day I got +5 across one of the outputs and then,
no more chip!  I was sunk.  I'd ruined my board.  Lo and
behold, a friend got me some datasheets for this chip and we
discovered, as Alan said, that it (the TI 6011) was indeed
pin-for-pin compatible to the AY-5-1013.  I was saved!

I still have all that stuff.  One day I'll get a digital
camera and...

Barry

--
http://www.piclist.com hint: The PICList is archived three different
ways.  See http://www.piclist.com/#archives for details.