I’m using coyote-2500 for doing some not so complex tasks: to read from a KB, to display on a lcd20x4, and processing a comm protocol under rs485. In order to make the three tasks to work together and at the same time I used the multiprocess facilities for the rabbit (“costates”).
When the debug cable is attached to the board (running in debug mode) everything works really nice (in specific the protocol), but when unpluging the debug cable (running stand-alone), ups, the protocol starts to unbehave, and it’s not reliable anymore. The baud rate is 9600.
Anybody has some ideas what’s going on?? Or moreover, can you tell me how I can make it to work as it must without the debug cable (stand-alone mode)?? Thanks in advanced =)
I have one idea on this without getting more information. Do you have a common ground between the coyote and the RS-485 port it is communicating with? If not, try adding one and see what happens. It could be that when connected to the PC, the coyote was grounded through that connection causing things to work. When unplugged from the debug cable, that path to ground is then gone.
No, there isn’t common ground between rabbit and the 485. What I wanted to say is that while in debug mode the protocol works nice, but when running in stand-alone mode the protocol unbehaves. Perhaps the fact that the common ground is missing may be causing that unbehavior. I also thought that the timing may be the problem, because in debug mode the applications run slower than in stand-alone because the debug overhead. I set some delays among sent and received bytes in order to see what’s going on. That helped a little bit, but not at all. Let me see what happens when attaching a common ground.