2009/5/4 Ruben J=F6nsson <ruben@pp.sbbs.se>:
> Because you are then building a star network as opposed to a linear bus n=
etwork
> (or daisy chained) which may be harder to terminate since you don't have a
> clear start and end of the bus. If the wires from your "switch" board to =
your
> node devices (temperature sensors or whatever) are short enough (with reg=
ard to
> the wavelength for the frequency of the signal) it might be OK.
>
> You can make the bus connectors in your nodes with two terminals for each
> signal (+, -, GND and perhaps shield) connected in parallel and then use =
one
> set as input and one as output (doesn't matter which). In the first node =
in the
> bus you can put the termination resistors in the input terminals and in t=
he
> output terminals in the last node.
>
> One thing I havn't seen mentioned in this thread is how you are intending=
 the
> communication to work. Are you going to have a master (a pc/plc perhaps) =
that
> continously polls the nodes in the network or are the nodes spontaneously=
 going
> to send data to a master or other nodes in the network? If it is the firs=
t one
> it is fairly easy to make a polled master/slave protocoll with RS485. If =
it is
> the last one I would recommend CAN since the message sending and collision
> detection between the nodes will be taken care of by the hardware.

I'll be using CAN.

Wow, this gets pretty complicated. If I use a star type network
topology, where would I put the terminating resistors?

-- =

http://www.piclist.com PIC/SX FAQ & list archive
View/change your membership options at
http://mailman.mit.edu/mailman/listinfo/piclist