Xref: utzoo comp.unix.xenix:2711 comp.periphs:1071
Path: utzoo!attcan!uunet!lll-winken!lll-tis!ames!hc!lanl!cmcl2!phri!cooper!gene
From: gene@cooper.cooper.EDU (Gene (the Spook) )
Newsgroups: comp.unix.xenix,comp.periphs
Subject: Re: serial to parallel
Message-ID: <1354@cooper.cooper.EDU>
Date: 19 Jul 88 15:42:19 GMT
References: <5403@ecsvax.uncecs.edu>
Organization: The Cooper Union (NY, NY)
Lines: 44

in article <5403@ecsvax.uncecs.edu>, cjl@ecsvax.uncecs.edu (Charles Lord) says:
> Xref: cooper comp.unix.xenix:2108 comp.periphs:767
> 
> If you want to hack it yourself, there is an ingenious little 
> circuit in the current issue of EDN (special Designs issue)
> that seems to do a quickie job of attaching a Centronics-interface
> printer to a serial output.

I'm not quite familiar with the referenced EDN article, but I *can*
recommend a nifty (almost)one chip solution to your problem. The reason
I put in the "(almost)" is that I'm not including the 1488/1489s and
the appropriate baud rate generator.

What I'm talking about is the General Instruments' AY-3-1015D UART.
All you need is the chip itself and the baud-rate generator, along with
the 1488/1489s to get a fully operational serial-to-parallel or parallel-
to-serial converter. All options can be selected in hardware (ie, no
programming the UART) by means of dipswitches or just plain old jumpers.
The UART is all CMOS and draws a ridiculously low 30 mA or so (about as
much as a bright LED), and can probably be powered right off the serial
line.

What's so nifty about the circuit is that it works both ways, as S-P and
P-S, with the appropriate buffers (1488/1489s on the serial end; 'LS373
or 'LS374 on the parallel end). For an even nicer twist, get hold of a
Maxim MAX238 which has an on-board 12V converter (5V only operation), and
both input and output line drivers.

Oh, yeah. The baud rate generator can be an AY-[35]-8116, if memory serves
me correctly. If you don't need complete and total accuracy, try a CMOS
555 (ie, a 7555) with good quality components. This might not be exactly to
specification for clock accuracy, but it will be cheaper. Aw, hell, stick
with a crystal and the BRG!

Anyway, I'm in a rush now, and I'm starting to babble, but think about the
1015D/8116/238 solution. It's so simple all you'd have to do is hook them
all together like in the data sheets and it'll work! Have fun!

						Spookfully yours,
						Gene


	"Bring the little ones unto me, and I will get a good price
	 for them."
					- Dr. Bertram X. Fegg