Re: Network

From: Ethan Dicks (erd_at_infinet.com)
Date: 1998-04-15 18:09:39

 
Ruud writes to Andre:
> I want to react on your "Network page":
> 
> > The asynchrounous communication will be derived from the RS232
> > specification, in that normal UART 16550 or ACIA 6551 chips can be used.
> 
> Very nice. (I always used the 8250 and 16450 with 65xx-systems)

Are there any odd interfacing issues?  (R/W vs /IOR and /IOW kinds of
stuff)
 
> > The physical layer could be derived from the RS485 specifications.
> > The original RS485 specification uses a terminated bus, and allows
> > one master to send at any time only. Otherwise drivers could be damaged.
> 
> Can you give me in a few lines some more specifications, please? How many
> lines does it use? Why can the drivers be damaged? What ICs does it use?

Years ago, I put together a proposed network interface for the Amiga with
a guy at Zenith, Jeff Easton.  We never finished the projects, but the
schematics (in HPGL and PCL) are still kicking around on Aminet 
(ftp://wuarchive.wustl.edu/pub/aminet/hard/hack/pocket.lzh).  Someone at
Commodore (Dave Haynie, IIRC), kept putting a similar circuit on new Amiga
designs, and management kept takingit out, but it was RS-485 based (thus
the point of mentioning it).

We used the COM20010 (or COM20020) integrated ARCnet chip and an SN75176
RS-485 driver.  Yes, the network was terminated, but I don't remember if
there was a prohibition about multiple simultaneous transmitters; it's
been too long.  I'd have to check the 75176 data sheets (it's made by
National Semiconductor and TI, if anybody want to dig for themselves).

The COM20010 implemented a token passing bus, if using twisted-pair
drivers.  SMC had plenty of application notes, etc, in 1992; perhaps
they still do if the chip is still in production.  I'd recommend those
docs for anyone planning on implementing a -485 token based network.

The only reason why I wouldn't recommend the COM20010 chip today is
availability.  16550 chips are going to be around for a long, long
time, plus, you can code around them at the physical layer, should
something better come along.  The COM20010 only had one descendent
that I knew of, the COM20020, and the only difference appeared to be
how much onboard SRAM there was for buffers.  On the other hand, you
could implement a full ARCnet adapter with it, and 6 years ago, there
were ARCnet networks to hook into.

For a physical connection, Commodore was going to use a switched RCA
jack like the kind at the front of most modern VCRs for external
inputs.  When empty, the switch would close, allowing the network to
self-terminate.  RJ-11's or RJ-45's are fine, too, but you've got
to remember to stick in a terminator, like with Farallon's PhoneNet
stuff.  I suppose you could even use yellow/black on phone wire, as
long as you didn't also have a PhoneNet network on the same cable.

Cheers,

-ethan

Archive generated by hypermail 2.1.1.