Searching \ for 'Experiences of data transmission on transformer co' in subject line. ()
Make payments with PayPal - it's fast, free and secure! Help us get a faster server
FAQ page: techref.massmind.org/techref/mems.htm?key=data
Search entire site for: 'Experiences of data transmission on transformer co'.

Truncated match.
PICList Thread
'Experiences of data transmission on transformer co'
1999\04\06@065412 by Russell McMahon

picon face
I'm using PICs for this at present and PIC multi-processor networks
seem to have been of general interest to many here in the past so
presumably no-one will complain with religious fervor about me not
OTing this :-)

I have been busy re-inventing the wheel over the last month or so and
should be able to share some interesting results with you sometime
soon (patent applications (not mine) are being processed covering
some aspects of this).

I am connecting up to (hopefully :-)) about 400 PICs on a physically
short network (about 100 metres). Each PIC is *essentially*
transformer coupled to the network. Similar physically to but, for
technical reasons, definitely not the same as eg ethernet.

QUESTION:

Has anyone here done this or know about systems which work this way.
Use of a transformer leads to different treatment of data "pulses"
and edges than for eg RS232 or RS422.

Do any of the known small system/home automation networks connect
this way and how do they handle the physical level signals. My way
works well but it would be interesting to know what else is being
done.


regards


       Russell McMahon

{{  PS In case you wonder - there were constraints in this case which
mandated doing it how it's been done. All shall be revealed. I'm not
doing the research at the wrong end of the project:-)  }}

1999\04\06@141457 by Lee Jones

flavicon
face
> I'm using PICs for [a] multi-processor networks
>
> I am connecting up to (hopefully :-)) about 400 PICs on a physically
> short network (about 100 metres). Each PIC is *essentially*
> transformer coupled to the network. Similar physically to but, for
> technical reasons, definitely not the same as eg ethernet.
>
> QUESTION:
>
> Has anyone here done this or know about systems which work this way.
> Use of a transformer leads to different treatment of data "pulses"
> and edges than for eg RS232 or RS422.

A very well known and widely installed system is Apple's
Localtalk network.

Originally, the protocol stack and the physical interface
were called Appletalk.  Later, Apple redefined Appletalk to
describe the logical layers, Ethertalk for the physical layer
when on Ethernet hardware, and Localtalk using the original
transformer coupled 230.4Kbps(*) physical network.

>From the first Macintosh in 1984, the serial port hardware
supported networking.  Signaling used MFM encoding (I think)
with hardware support from the Zilog 8530 SCC (Serial
Communications Controller).  Good description of the physical
layer in Zilog's data manual for that chip.

Apple's original coupler design used a special Belden sheilded
twisted pair cable and miniDIN-3 connectors.  The marketplace
revised this to use standard unshielded phone cable and RJ11
modular phone jacks/plugs.  Only real difference was the
termination resistors & transformers varied due to the
impedence change between the cable types.

Physical and logical layers were documented in Apple's "Inside
Macintosh" series, various tech notes, and the definitive book
"Inside Appletalk" by Gursharon S Sidhu, Alan B Oppenheimer,
& Richard Andrews, Addison Wesley, 06/1990, 500 pages.

                                               Lee Jones

(*) Yes, this may start another flame war about it being
   230.4 Kelvin bits per second.  On a more useful note,
   the rate was an integer multiple of 9600, 19200, and
   57600 and based on a 7.37 or 3.684 MHz crystal.

More... (looser matching)
- Last day of these posts
- In 1999 , 2000 only
- Today
- New search...