I guess that if the 485 line is terminated at each end then a 'directional coupler' could detect collisions reliably by detecting the line impedance deviating from 60 ohms. Then I thought: This might even be reflected in the supply current of the line driver? Maybe a simple window comparator could detect this, contention would increase the supply current, both drivers in the same polarity should reduce it (half?). Oliver. ----- Original Message ----- From: Howard Cripe To: Sent: Monday, October 02, 2000 3:21 PM Subject: Re: [EE]: Multiple serial devices > Olin has the right idea for a single line RS/485 network. The collision > avoidance/detection usually will not work on RS/485 with standard chips. > The problem we had was that when sending, you receive the data you are > sending (since rx and tx are tied together) as well as other's data. The > levels on your own transmitter override any other unit's data, so in some > cases, you can't detect a collision. Then you have to implement > time-out/retry software, and it all gets very messy. If you need faster > response time than with a polled system like Olin describes, go to a > bi-directional bus (requires 4 wires plus ground). This way the master unit > such as a PC will be the only one sending on one line, and all the slave > units respond on the other line. Again, though, you need to poll the units > or you will have collisions. You won't have any turn around time, though. > > Howard > > -- > http://www.piclist.com#nomail Going offline? Don't AutoReply us! > use mailto:listserv@mitvma.mit.edu?body=SET%20PICList%20DIGEST > > > > -- http://www.piclist.com#nomail Going offline? Don't AutoReply us! use mailto:listserv@mitvma.mit.edu?body=SET%20PICList%20DIGEST