Hi all, I'm having issues with the RS232 ports on a PAC unit and thought I would share my experiences and ask a couple of questions. This issue has been ongoing for a week or so and the testing I have carried out is fairly extensive, so please bear with me... I am working on an installation where a PAC will control a PA system. I understand the protocol to be used and have written code that executes flawlessly on the simulator that gives correct behaviour when connected to the PA. So far so good! However, when I come to transfer the code to a PAC I hit problems. No matter what I do I cannot not get the unit to send out data over the RS232 line. I quickly moved to testing the PAC with a PC, rather than with the PA. The PAC receives data OK over the line (and behaves as expected in response to that data), but I never see any return data (which should definitely be being sent as verified in the simulator). I have read the documentation thoroughly and understand that either a separate 24V AC supply is required, or the RTS/DTR lines can be oppositely biased. Initially I was using the RTS/DTR lines. This certainly appears to work, as the PAC receives data (although I actually now suspect that power is the issue - see below). Initially I suspected that my serial cable was wired incorrectly. However, I verified this pin-to-pin and am 100% sure it is correct and electrically sound. (Even going so far as to rig up a reverse test cable with a Cat5 socket to simulate the PAC using another PC). I then suspected some flaw in the code, so simplified it to a single 'writeserial' on each loop iteration. Again, in the simulator this works as expected and again on the unit is does not work. I then suspected that the RTS/DTR lines where not providing sufficient power to the unit to generate RS232 compliant output voltages and so connected a 24VAC power supply. I still saw the same behaviour: able to receive OK, but not able to send data. I then suspected that I had a faulty unit and so began experimenting with a second brand new PAC. Again, I started by powering the unit from the RTS/DTR lines (confirming the voltages on a meter) and again saw the same behaviour. At this point I began to wonder if the unit was in fact sending data, but that the signal levels where below thresholds and so broke out my oscilloscope. This confirmed that hypothosis - the unit is generating a signal, but at much reduced signal levels - only -3V to +0.8V, well outside the RS232 specification, which explains why I am unable to receive data on a connected terminal device. I then went back to the first unit to test in the same way and in fact saw no signal at all. My supposition is that either the unit was shipped faulty or that the transmit circuitry was damaged when I connected the 24VAC supply. I went back and checked that PSU and it is giving an output of about 27V RMS - slightly above the stated voltage. I am obviously somewhat hesitant to connect this PSU to the new PAC for fear of damaging it. My questions then are as below: 1) What is the required input voltage on the RTS/DTR lines to give correct behviour? I am seeing approx -9V and + 9V... 2) What is the actual limit on the input voltage for the 24V connection? Would 27V damage the PAC (and specifically just damage the transmit side of the serial circuitry)? 3) Is there any current limiting circuitry on the 24V input? The PSU I am using is rated to 3.5Amps - way over the stated 30mA nominal current draw. Presumably the unit would not pull more current than it is rated for ? (Although I saw a post saying these units were easily damaged by short circuits on the serial side?) 4) Is there a known good 24VAC PSU? A DIN mounted unit would be preferable as I am trying to keep everything in a single board and having an external plug style PSU would not be ideal (and may give issues from a building regulations point of view, as my reading is that all cabling that enters a board in which there are mains voltages should have insulation rated to mains voltage, which a simple wall plug-style power adapter may not have).... One separate observation as well, if I may. I understand the unit uses a separate power supply as it is internally optically isolated. However, 24VAC PSUs seem to be fairly difficult to obtain, particularly in a DIN form at anything like appropriate current ratings, and it is a unique power requirement, not shared with any other CBus kit, meaning more PSUs in a board... I wonder if there is any reason that some form of 5V powered RS232 driver IC is not used to drive the serial lines instead? Something like the Maxim Max220 family could be used which only require a single 5V supply and generate the -13V/+13V voltages using an internal charge pump. This could be coupled with a simple voltage regulator and then accept any voltage say 5V - 12V, making it compatible with other CBUS devices that are not bus powered. Just a thought! Thanks for any input - as ever the community's help is much appreciated. Kind regards Jon
