|
| |
Jallist Archive Dec 2003
RE: Ideas for an Asynchronous Net implementation.
From: Vasile Surducan <vasile(at)s3.itim-cj.ro>
Date: 01 Dec 2003 11:48:45 +0200 (EET)
On Fri, 28 Nov 2003, Wouter van Ooijen wrote:
> > The amount of PICs running in a wire it's mutch related to the hw
> implementation of the net: a low-cost rs-485 module can manage up to 32
> devices in a cable section (I hear about keep this value under 20). More
> devices means "repeaters".
>
> The Eur/Wouter network uses a repeater (a 74HC14 chip) in each node. The
> number of nodes that can be chained is limited by the varying delay of
> the edges caused by the 74HC14's and the RC networks used to suppress
> interference and reflections. (And also by the fact that power is also
> distributed along the chain, and will drop somethat due to connector
> contacts resistance). I am very interested in the actual number of nodes
> that can be used in a single chain, but that will only be known when Eur
> has a need for a realy long chain.
In RS485, there is a standard called 1/n, so even 256 devices may be
connected on the same cable section by decreasing the transmision current
and increasing the input impedance of device. Of course the cable lenght
can't be 1200m like in the 32/60mA standard. The numbers of repeaters in a
daisy chain RS485 may be as high as the user need.
Supplying such a chain on the one end is impossible, because of the large
voltage drop across the line.
Some friends of mine have informed me, about using multi-radial RS485
lines using repeaters in the distribution nodes. And the communication is
working ok.
|
|
|