The original note also mentioned localtalk, this uses a 'carrier' (actually
FM-0, aka biphase space) on the bus while packets are transmitted, the Zilog
SCC can be programmed to generate this signalling and uses SDLC framing with
CRC for packet encoding. Signalling is at 230.4Kbits per second with one
transition within the cell for a one and two transitions in a cell to
represent zero. High-impedance driver buffers are switched on only while
transmitting.

Roger Brown

----------
>From: Chas Williams <chas@cmf.nrl.navy.mil>
>To: Bill Studenmund <wrstuden@nas.nasa.gov>
>Subject: Re: Does mack68k NetBSD support actual RS422 or just RS232 
>Date: Thu, Jan 6, 2000, 2:04 pm
>

>In message <Pine.SOL.3.96.1000105162334.25095D-100000@marcy.nas.nasa.gov>,Bill 
>Studenmund writes:
>>It's been a while since I played with RS442/485, but I think the signaling
>>is different for them. I'm not sure if just a cable will work. RS485 is
>>the differential version of RS442.
>
>rs422 is just differential rs232.  in the case of the macintosh most cables
>just connect the tx+ and rx+ to ground and drive the tx- and rx- 
>with the rs232 tx and rx in order to connect an rs232 device to the
>mac's rs422 port.  if the other device uses an 8-pin mini-din you just
>need a plain macintosh serial cable and you should be fine.
>
>[on a side note, some of the larger sgi's have a special ioctl for
>opening their serial ports in order do rs422.  since these boxes
>have db9s, you need some way of saying enable the tx+ and rx+ pins
>on the connectors]
>
>btw, i thought rs485 was a bus version of rs422.  it uses differential coding
>like rs422 but you can have multiple senders and receivers on the same
>set of wires.  i believe rs485 specifies some way of determining when it
>is clear to access the 'bus'
>
>