RS-232 headaches INT_EXT versus INT_RDA

[petedd]

I have set up a board, running at 20MHz, with two RS232 ports: one on the UART (C7/C6) and one to use the EXT interrupt (B0/B1). I am using CCS sample code basically. Receive data on one port, buffer, and retransmit on the same port. By changing the DEFINE at the top of the program, one can select between the UART and the S/W port.

The UART version works great. The S/W version does not. The output is jbberish.

By setting a breakpoint and watching the buffer data, I can see that the corruption of data seems to be happening upon receipt. The S/W output seems to mimic whatever is in the buffer.

I need both ports running at 4800 baud (GPS application). Eventually, once I solve this problem, I intend to use two RS-232 ports at once. The processor is the 16F737 and it is running on a crystal at 20MHz. (At least I think I have told the compiler to use the crystal)...

Can anyone see what is wrong here? THANKS!!!

/////////////////////////////////////////////////////////////////////////
//// Buffered Serial Input/output ////
//// ////
//// pick between software (IND_EXT) B0/B1 or UART (INT_RDA)C7,C7 ////
//// ////
//// CHARACTERS BEING CORRUPTED UPON RECEIPT DURING INT_EXT ////
//// RECEIPT -- WORKS PERFECTLY USING INT_RDA ////
//// ////
/////////////////////////////////////////////////////////////////////////


//&&&&&&&&&&&&&&&&&&&&&&&&&&&&&&&&&&&&&&&&&&&&&&&&&&&&&&&&&&&&&&&&&&&&&&&&&&&&&
//&&&&&&&&&&&&&&&&&&&&&&&&&&&&&&&&&&&&&&&&&&&&&&&&&&&&&&&&&&&&&&&&&&&&&&&&&&&&&
// CHANGE MODES BY COMMENTING IN OR OUT THE FOLLOWING LINE
#define USE_EXT // use S/W RS232, comment out to use RDA (UART) port instead
//
//&&&&&&&&&&&&&&&&&&&&&&&&&&&&&&&&&&&&&&&&&&&&&&&&&&&&&&&&&&&&&&&&&&&&&&&&&&&&&
//&&&&&&&&&&&&&&&&&&&&&&&&&&&&&&&&&&&&&&&&&&&&&&&&&&&&&&&&&&&&&&&&&&&&&&&&&&&&&

#include <16F737.h>

#fuses HS,NOWDT,NOPROTECT
#use delay(clock=20000000)

#if defined(USE_EXT)
#use rs232(baud=4800, xmit=PIN_B1, rcv=PIN_B0, parity=N, BITS = 8, stop=1) //, disable_ints)
#else // use RDA
#use rs232(baud=4800, xmit=PIN_C6, rcv=PIN_C7, parity=N, BITS = 8, stop=1)
#endif

// Initialize Buffer
#define BUFFER_SIZE 64
BYTE buffer[BUFFER_SIZE];
BYTE next_in = 0;
BYTE next_out = 0;

// Set up serial interrupt
#if defined (USE_EXT)
#int_ext
#else
#int_rda
#endif

void serial_isr() {
int t,x;

buffer[next_in]=getc();
t=next_in;
if (t==32) // just for breakpoint
x=t;
next_in=(next_in+1) % BUFFER_SIZE;
if(next_in==next_out)
next_in=t; // Buffer full !!
}

#define bkbhit (next_in!=next_out)

BYTE bgetc() {
BYTE c;

while(!bkbhit) ;
c=buffer[next_out];
next_out=(next_out+1) % BUFFER_SIZE;
return(c);
}

void main() {

#if defined (USE_EXT)
enable_interrupts(int_ext);
#else
enable_interrupts(int_rda);
#endif

enable_interrupts(global);

printf("\r\n\Running...\r\n");

do {
if(bkbhit) {
putc(bgetc());
}
} while (TRUE);
}

[peteD]

Board: Olimex PIC-28 schematic is here: http://www.olimex.com/dev/images/PIC/PIC-P28-sch.gif (I am using two extra ports of MAX232 for the B0/B1 RS-232 port).

Compiler: 4.064

IDE: MPLAB 8.00.000

[peteD]

If I remove the PIC from the board and jumper B1 and B0 pins together, the data moves through and the output is clean. This eliminates the MAX232 as a possible problem. All it could be in my mind is timing, but how come I can't get even the simplest of software rs232 repeaters to run without having the input corrupted???

such as this:

#include <16F737.h>
#fuses HS,NOWDT
#use delay(clock=20000000)
#use rs232(baud=4800, xmit=PIN_b1, rcv=PIN_b0, ERRORS)

//================================
void main(void)
{
char c;

while(1)
{
c = getc();
putc(c);
}

}

[Guest]

46 people have viewed this problem but not one person has a clue on the problem?

[SET]

[PCM programmer]

[SET]