Path: utzoo!mnetor!tmsoft!torsqnt!news-server.csri.toronto.edu!cs.utexas.edu!uunet!nuchat!steve
From: steve@nuchat.sccsi.com (Steve Nuchia)
Newsgroups: comp.unix.sysv386
Subject: Re: Doin' it right (was Re: Comtrol)
Message-ID: <1991Feb14.071814.26648@nuchat.sccsi.com>
Date: 14 Feb 91 07:18:14 GMT
References: <9@equinox.UUCP> <6801@rsiatl.Dixie.Com> <1844@chinacat.Unicom.COM>
Organization: South Coast Computing Services, Inc.  Houston
Lines: 48

In article <1844@chinacat.Unicom.COM> chip@chinacat.Unicom.COM (Chip Rosenthal) writes:
>This has already been done.  Steve Nuchia has put together (from what I
>hear) a nice little four-port dumb 16550 card.  It is his project, but
>net.folks have had input and have played guinea pig for it.  That might
>be a starting point.  Steve - it's been a while since we've heard from
>you.  How about a status report/availability info?

Whimper.  Whine.   Leave me alone!                       :-)

What's happening is I'm trying to graduate in May.  I made 12 cards
and sold them all, as far as I know they're all working.  I don't want
to sell any more of the original design for several reasons, and the
redesign is held up waiting for me to get time to play with it.  I do
have a spring break coming up, and if all else fails it will be my
first project starting 19 April.

All my hardware designs for the forseeable future will be sold
as kits.  I'm trying to meet the needs of the Unix and post-unix
hobbyists: low cost, featurefull and completely documented hardware.

The serial board is a 4-port design using the NS16550A chips and
the rj45 type jacks with 3B2 pinout.  Status port, addressing matrix,
and IRQ sharing matrix is all the logic on the board, but there
are a boucoup of traces to make all that go.  Rev A used a handful
of standard TTL and some PALs, rev B looks like it will be all pals
(except for uarts and receiver/drivers).  Its most exciting feature,
other than working well with FAS and telebits, is that it has access
to the high-numbered IRQ lines, relieving crowding on the low ones.

After that will be a parallel port with DMA support, configurable
as a centronics output (or input, for that matter), bidirectional
or bussed packet exchanger, or whatever.  Should be just the thing
for some plotters, and it would fill a niche for high throughput
low cost data exchange and hardware hacking.  It also happens that
I know somebody trying to drive a nearly-but-not-quite centronics
plotter in real time and the handshaking is eating way too much cpu.

After that come a system monitor board, some power management kits,
and a logic analyser.  Suggestions are welcome too.  Also if anybody
wants to get prereleases of the logic analyser specs to start developing
the software (this is going to be a group effort, guys) let me know and
I'll start a mailing list.

-- 
Steve Nuchia	      South Coast Computing Services      (713) 964-2462
	"Innocence is a splendid thing, only it has the misfortune
	 not to keep very well and to be easily misled."
	    --- Immanuel Kant,  Groundwork of the Metaphysic of Morals