I don't think it is easy (or even possible) to implement CSMA/CD in an RS-485 bus. I tried it once without success, each node can only see its only transmission. It seems that because the RS-485 transceivers drive both levels actively, and together with the bus resistance makes the distorted levels undetectable. Best regards, Isaac Em 07/07/2013 10:11, Charles Craft escreveu: > ">I will be building a small network of 4 clients to send one of the=20 >> house's 4 heating/cooling zones temperature and humidity to a central=20 >> processor." > Will the clients be sending asynchronously or will you poll the clients? > If you allow the clients to send won't there be a lot of extra code for C= SMA/CD? > > thanks > chuckc > > > > -----Original Message----- >> From: Carl Denk >> Sent: Jul 6, 2013 3:42 PM >> To: PICLIST >> Subject: [PIC] RS-485 star network ?? >> >> I will be building a small network of 4 clients to send one of the=20 >> house's 4 heating/cooling zones temperature and humidity to a central=20 >> processor. The client likely will be 18F1320, and the processor 18F4320= =20 >> since I have a few laying around, but could consider something=20 >> different. The client will also have a LCD display. The processor will=20 >> based on temperature and humidity sent the house HVAC zone controller=20 >> and humidifier signals (24VAC) via relays based on current set condition= s. >> >> The 4 zone lengths from the processor range from 10 to 50 feet and=20 >> planning to use cat-5 cable. 2 pair for power to the client, and one=20 >> pair for RS-485 signal. The network speed can be very slow, even down to= =20 >> 300 baud, since 1 message from each client in 1second and a message=20 >> length of maybe 20 characters at most would be a lot. >> >> The question is: Will a star arrangement work, and what should the=20 >> terminating resistors be? >> >> As an alternative, to each client, I could use 1 pair of the cat-5=20 >> outbound and 1 pair inbound, and make it one long (less than 200')=20 >> network with jumpers at each client to select the continuity to the next= =20 >> client, and the usual terminating resistors at the end. >> >> >> Trying to keep this very simple and any advice appreciated. >> --=20 >> http://www.piclist.com/techref/piclist PIC/SX FAQ & list archive >> View/change your membership options at >> http://mailman.mit.edu/mailman/listinfo/piclist > --=20 http://www.piclist.com/techref/piclist PIC/SX FAQ & list archive View/change your membership options at http://mailman.mit.edu/mailman/listinfo/piclist .