RS232: fun, games, snafus.
Hedberg
Posts: 671
So, we've got this Polycom video conferencing system that I'm not familiar with.iPower 9000, it is. After setting everything up, I tried to communicate to it through the master using a telnet client without success. Unit didn't respond or return a string. Put in a short, 3-wire crossover jumper(2-3, 3-2, 5-5) and everything starts to work fine. All commands were successful and all responses from the Polycom were as advertised (almost). The unit also responded correctly to the IR remote keyboard that it came with.
The guy with the soldering iron came around and asked what that short cable was doing there. I explained and he said he'd fix it.
The guys with the wire, screwdrivers, and soldering irons had a bunch of stuff to do, so I didn't get back for a couple days. When I did, the Polycom was behaving very, very weirdly. It would respond to some commands but not all, and I was getting no return strings from it at all. The pan commands for the camera wouldn't work, and the tilt down command would send the camera all the way down and then the unit wouldn't respond to any RS232 commands. About half the time it would kill the IR keyboard response, too. Then, it would start to work again -- until I tried to tilt the camera down.
I thought that a cabling problem was unlikely, because the unit was receiving and properly responding to some commands. But, we were getting no response strings,and that could indicate a problem with the wire from Polycom 3 to AMX 2. So, I talked it over with the guy with the screwdriver and he opened up the hoods and found the cable wired 2-3, 3-2, 6-6!. Substituted a 3-wire crossover cable that I carry around and everything went back to working exactly as it should. Who would've known? Guy with the soldering iron shows up and asks what that cable is doing there . . .
Usually, when there is a cabling problem, the target device won't respond at all. Or, sometimes, I've seen the device not take any action but return a garbage string. I've never seen one before that would tease me into thinking that it was going to work and then crap out like this. Apparently, there was sufficient ground between the AMX and the Polycom to get it to work with only wires 2-3 and 3-2, but connecting 6-6 really fouled up the way the Polycom received the signals. Live and learn, they say.
The guy with the soldering iron came around and asked what that short cable was doing there. I explained and he said he'd fix it.
The guys with the wire, screwdrivers, and soldering irons had a bunch of stuff to do, so I didn't get back for a couple days. When I did, the Polycom was behaving very, very weirdly. It would respond to some commands but not all, and I was getting no return strings from it at all. The pan commands for the camera wouldn't work, and the tilt down command would send the camera all the way down and then the unit wouldn't respond to any RS232 commands. About half the time it would kill the IR keyboard response, too. Then, it would start to work again -- until I tried to tilt the camera down.
I thought that a cabling problem was unlikely, because the unit was receiving and properly responding to some commands. But, we were getting no response strings,and that could indicate a problem with the wire from Polycom 3 to AMX 2. So, I talked it over with the guy with the screwdriver and he opened up the hoods and found the cable wired 2-3, 3-2, 6-6!. Substituted a 3-wire crossover cable that I carry around and everything went back to working exactly as it should. Who would've known? Guy with the soldering iron shows up and asks what that cable is doing there . . .
Usually, when there is a cabling problem, the target device won't respond at all. Or, sometimes, I've seen the device not take any action but return a garbage string. I've never seen one before that would tease me into thinking that it was going to work and then crap out like this. Apparently, there was sufficient ground between the AMX and the Polycom to get it to work with only wires 2-3 and 3-2, but connecting 6-6 really fouled up the way the Polycom received the signals. Live and learn, they say.
0
Comments
I once came across some 20m serial runs to our projectors, where the D9 had been soldered up "mirrored" at the projector end. Instead of pins 2, 3 & 5 wired up, it was pins 4, 3 & 1. Soldering person must have been looking at the wrong side of the plug at the time.
The projector ran "fine" for 2 years where our control code was uni-directional, though sometimes you had to send the command to the projector several times before it received it.
When I updated the code to a bi-directional model, the bug showed up. My guess was that after a few strings had been sent down pin 3, there was sufficient line capacitance for the projector to somehow detect the message. Pain of a bug to try and find, especially since it had been "working" before my code change. For all new installs we now use pre-moulded serials. I've had no problems with runs up to 40m.
Roger McLean