??? 06/09/04 05:34 Read: times |
#72135 - RE: RS 485 looping-problem Responding to: ???'s previous message |
Santanu,
if you use this concept of having a bus length of much more than 1000meters, this will only work, 1. with absolutely stable earth potentials at all recievers, 2. with absolutely perfect match between termination and characeristic impedance of cable, and 3. when using slew rate limted drivers. to 1. If you route your bus system over a wide area, you will experience extra high common mode interference. Think of thunderstorms, solar activity, indirect air discharge, etc. Your whole system at least would need some surge protection. But having this extreme wide bus sytem running without any shielded cable, make me wonder that this application works at all! Michael Neary is totally right: Common mode interference IS an issue! Please tell us, what local power supplies are used. Are they floating? What connection is made between local signal ground and protection earth? to 2. It does not mainly matter what kind of cable you use. It matters how perfect you match termination with characteristic impedance! So, even the best cat-5 cable will fail, if termination does not exactly match characteristic impedance. I have seen, that some manufacturers of cat-5 cables specify characteristic impedance to 100Ohm +-15%. This is an unacceptable tolerance!! RG179U is specified to 75Ohm +-4%, for instance. This is much better. Also, your termination network will not show exactly 100Ohm: Your both 1kOhm resistors are in parallel to 100Ohm, so you get only 95Ohm. This is 5% too low. Finally, you may have a mismatch of up to 20%! This will not be suited to have a such large bus system running, especially if a non slew rate limited driver is used. to 3. Slew rate limited drivers were not only invented, because then a higher mismatch between termination and characteristic impedance can be tolerated, but also to keep radiation at acceptable limits. I highly believe, that your bus system will heavily violate CE conformity. Finally, your system should at least have some very simple software handshaking. Otherwise, you will never know, what actual data has arrived at these many recievers. A remedy could be to use optical links, or RS485 transcievers providing optical isolation. I cannot believe, that with non galvanic isolation such a large area bus system will ever work reliably, especially if routed in rough industrial environment, and especially if not using any shielded cable... Kai |
Topic | Author | Date |
RS 485 looping-problem | 01/01/70 00:00 | |
RE: RS 485 looping-problem | 01/01/70 00:00 | |
RE: RS 485 looping-problem | 01/01/70 00:00 | |
RE: RS 485 looping-problem | 01/01/70 00:00 | |
RE: RS 485 looping-problem | 01/01/70 00:00 | |
RE: RS 485 looping-problem | 01/01/70 00:00 | |
RE: RS 485 looping-problem | 01/01/70 00:00 | |
RE: RS 485 looping-problem | 01/01/70 00:00 | |
RE: RS 485 looping-problem | 01/01/70 00:00 | |
RE: RS 485 looping-problem | 01/01/70 00:00 | |
RE: RS 485 looping-problem | 01/01/70 00:00 | |
RE: RS 485 looping-problem | 01/01/70 00:00 | |
RE: RS 485 looping-problem | 01/01/70 00:00 | |
RE: RS 485 looping-problem | 01/01/70 00:00 | |
RE: RS 485 looping-problem | 01/01/70 00:00 | |
RE: RS 485 looping-problem | 01/01/70 00:00 | |
RE: RS 485 looping-problem | 01/01/70 00:00 | |
RE: RS 485 looping-problem![]() | 01/01/70 00:00 |