> > I am assuming the UART is used only for communicating
> > between the 2 PICs.  In other words, pic 1-TX is
> > connected to pic2-RX; pic1-RX is connected to pic2-TX.
> > If that is the case, put a pull-up resistor on one of
> > the 2 lines and put a pull-down resistor on the other
> > line.

> > At start-up, all lines are inputs.  Read the state of
> > the RX & TX lines.
> > Pic 1 will have TX hi & RX low
> > Pic 2 will have TX lo & RX hi
> >
> So, for example, the PIC that sees TX hi & RX low would
> be the master, and the PIC that sees TX lo & RX hi
> would be the slave.

I just wanted to add two more points now that I have
implemented this scheme:

1. The master and slave must be synchronized before
initializing the UART (in other words, even before
transmission commences).  This is because UART
initialization causes a 1 to be driven on the Tx line (idle state).

2. When the slave sends to the master, the slave MUST wait
one bit time after TRMT=1 (to allow the stop bit to be
transmitted) before "shutting down" (disabling the UART).
Otherwise, the pull-down on the slave Tx line will
interrupt the stop bit.

   I bring this last point up because there was a thread
("[PIC]: TRMT vs TXIF ..." on 04/30/03) debating under
which conditions waiting for this bit time was necessary.

Rick

--
http://www.piclist.com hint: PICList Posts must start with ONE topic:
[PIC]:,[SX]:,[AVR]: ->uP ONLY! [EE]:,[OT]: ->Other [BUY]:,[AD]: ->Ads