Hmm, sounds like a problem I have.... I am using RS232 drivers/receivers to poll a 10X10 matrix of buttons. The each button is installed in an apartment up to 150 feet away. They are all connected with 24 gauge twisted pair telco cable. I use the drivers for noise immunity as driving the lines straight out of the PIC didn't even go 10 ft. I put 1 output bit high and check the 10 inputs one at a time. Then I move on to the next output bit. I scan all 100 switches about 3 times a second. It worked great in the lab, but didn't work at all at the site. Several things happened. I had to slow the program down quite a bit to compensate for the increased capacitance on the driver lines. I put a scope on the driver output and the signal was more like a sawtooth than a square wave. I had to wait for the line to settle before I could start checking the inputs. A bunch of NOPs here and there took care of that, but I was surprised at how much the slew rate changed (I don't remember exactly what it changed to). I didn't think 150 ft of wire would do that much.. Second thing that happened (which was the whole point to this post) is that I was blowing the drivers every other week. I was using Motorola MC1488 and 1489s. They are supposed to be current limited. They would get VERY hot during operation. I called Motorola on this and the input current draw was operating within spec. Putting heatsinks on the chips and a fan in the cabinet greatly improved the reliability to almost 100%. I cant tell exactly how good it is because I'm constantly dealing with people ripping the buttons off the wall, grounding them out or connecting phone lines to them. (The multi-cable is also has the phone wires in it) It seems the drivers just cant handle the extra capacitance on the line created by long runs. Just to let you know, the whole purpose of the system was to scan the 100 switches, if it finds one pressed, it displays the apartment number on an LCD in the guard booth. It also stores the last 25 apartments that pressed their buttons. Thanks for the bandwidth. ----------------------------------------- Mike Montalvo G-FORCE Motorsports http://www.g-forcemotorsports.com 516-794-0858 -----Original Message----- From: John Date: Saturday, May 29, 1999 10:08 AM Subject: [OT]RE: RS485/RS422 >A recent post mentioned running RS422 full-duplex with a 5-wire cable. >Is the 5th line for ground? I've assumed (uh oh!) one of the benefits of >RS422 and RS485 was the absence of a common ground running a couple >thousand feet (ground loops, noise, etc). > >A friend working as a network consultant has a client who recently moved >into a new and much larger facility. A PC communicates via RS232 to some >uP-based I/O boxes out in the warehouse (cable run was <50ft. now is >~150ft.). The PC's serial port has blown several times since the move. >The PC and the remote I/O boxes are on separate breakers, but >the electrician says all is OK with AC service and grounding. Their plan >is to convert to RS485. My serial comm experience has been with RS232, >so I can't offer anything beyond the MAXIM 485 data sheets, etc. >Has anyone dealt sucessfully with this symptom by replacing RS232 with >485? > >John >___________________________________________________________________________ ___ > >John Leonard, instrument technician Electron Hacker >College of Science & Engineering Bit Banger >Saginaw Valley State University Bassist >University Center, MI USA 48710 Dad >___________________________________________________________________________ __ > ----------------------------------------- Mike Montalvo G-FORCE Motorsports http://www.g-forcemotorsports.com 516-794-0858