Software UART at 300 Baud corrupt data

[Hans Wedemeyer]

My circuit has two chips linked by CMOS level;
Chip#1 ---- Chip#2
Tx1(C2) to Rx2(A3)
Rx1(G4) to Tx1(A0)

As this link only needs to send 2 bytes and recieve 2 bytes I figured software
UART at 300 Baud should do the trick.

#use rs232(baud=300,parity=N,xmit=PIN_C2,rcv=PIN_G4,bits=8,FORCE_SW)

Chip #1 transmits a request byte and chip #2 replies. However the reply data is corrupt.

I tried using DISABLE_INTS in the #use RS232 statements, but the complier complains (PCH V3.249)
I tried to disable the interrupts, that did not help.

In the following code snippets, chip#1 sends a command byte 'b', I assume this has been received correctly by chip#2 because it replies everytime, the assumption is based upon if ( D == 'b' ) which must be true to ever send a reply, and it never fails to do that.

Speed is not the issue here, infact I could take the Baud rate down to 110 Baud or lower !

I have never liked software UARTS so I expected trouble !

What am I doing wrong here... ?

Thanks.

//////////////////////////////
Code for chips #1


#use rs232(baud=300,parity=N,xmit=PIN_C2,rcv=PIN_G4,bits=8,FORCE_SW)

int8 V1,V2;

putchar(0XFF); // use it as the wake up char for Chip #2
delay_ms(25); // may not be needed
putchar( 'b' ); // the command char request

V1 = getchar(); // will wait for next char
V2 = getchar();

/////////////////////////////
Code for chip #2

#use rs232(baud=300,parity=N,xmit=PIN_A0,rcv=PIN_A3,bits=8)

int8 D;

while(1)
{

if (kbhit() )
{
D = getchar(); // wait for command char

if ( D == 'b' )
{
putchar(0X55); // return test values
delay_ms(16);
putchar(0XAA);
}
D=0;
}

}

[PCM programmer]

What PICs are you using ?

What is the crystal frequency on each board ?

Have you tried having one PIC transmit at 300 baud to a terminal
window on a PC ? Does the PC receive characters correctly ?

[Hans Wedemeyer]

Thanks PCM programmer,

Q) What PICs are you using ?
A) #1 PIC16F6720 and PIC12F629

Q) What is the crystal frequency on each board ?
A) Chip #1 4.194304Mz using 4XPLL, and Chip #2 2MHz resonator (industrial 0.5%)

Q) Have you tried having one PIC transmit at 300 baud to a terminal
window on a PC ?
A) No because it's difficult. I put the code in a loop and transmited 0X55 the square wave indicates 301 Baud !

Q) Does the PC receive characters correctly ?
A) Not tested but may have to cut and jumper in a MAXnnn level shifter and try this.

Unless I'm mistaken the PIC12F629 8 pin chip does not allow debugging, so I'm kind of stuck on that side of it. I'd like to connect my second ICD-U40.

I may have to wire up another PIC device to test/debug the code, and then port it to the 12F629... !

I'm curious why the compiler complains about DISABLE_INTS ! I thought that item was intended to assit the use of software UARTs.

If this does not work out, I may have to go to some of two wire bit banger comm. that I have from way back.

[Ttelmah]

You say 'by CMOS level'. What voltage is each chip actually running at?. How are they actually joined?.
With 3.249, 'DISABLE_INTS' in the RS232 line, works fine for me with this chip. What compiler version are you running?.

Best Wishes

[Hans Wedemeyer]

[Ttelmah]

The phrase 'puzzled', does leap out on this one.
Post the actual 'header' (the #use delay, fuses, and #use RS232), from each program. What is 'odd', is that:

[Hans Wedemeyer]

[PCM programmer]

There is no OR symbol. It all has to be done with commas.

[Hans Wedemeyer]