| View previous topic :: View next topic |
| Author |
Message |
Beowulf
Joined: 10 Feb 2012
Posts: 7
Location: washington
|
| PIC16F1824 INT_RDA does not work |
 Posted: Fri Feb 10, 2012 3:18 pm |
 |
|
I am using a PIC16F1824 with compiler Version 4.128. I cannot get the INT_RDA routine to fire when data arrives on the RS232 line. I am using a PC to communicate through a MAX3232. I am able to communicate over the RS232 in both directions....but the INT_RDA never fires.
Here is the code that I am testing.
| Code: | #include <16F1824.h>
#device adc=16
#FUSES NOWDT //No Watch Dog Timer
#FUSES INTRC_IO //Internal RC Osc, no CLKOUT
#FUSES WDT_SW //No Watch Dog Timer, enabled in Software
#FUSES NOBROWNOUT //No brownout reset
#FUSES PLL_SW //4X HW PLL disabled, 4X PLL enabled/disabled in software
#FUSES NOLVP //No low voltage prgming, B3(PIC16) or B5(PIC18) used for I/O
#use delay(int=16000000)
//#use rs232(baud=9600,parity=N,xmit=PIN_A0,rcv=PIN_A1,bits=8,stream=PORT1) GENERATED BY COMPILER
//#use rs232(baud=9600,parity=N,UART1,bits=8,stream=PORT1) //doesn't work
#use rs232(baud=9600,parity=N,UART1,bits=8,stream=PORT1,ERRORS) //doesn't work
#int_RDA
void RDA_isr(void)
{
fputc('Z',PORT1);
}
void main()
{
enable_interrupts(INT_RDA);
enable_interrupts(GLOBAL);
setup_oscillator(OSC_16MHZ|OSC_INTRC|OSC_PLL_OFF,0);
for(;;)
{
output_bit(PIN_C4,1); //this flashes an LED to verify the chip is alive
delay_ms(500);
output_bit(PIN_C4,0);
delay_ms(500);
}
} |
most of this code has been generated by the compiler through the wizard.
fputc(fgetc()) works for any of the #useRS232 statements above.
other interrupts on this chip seem to work ok....INT_EXT works for instance.
I have tried most of the advice given in this forum...stumped |
|
 |
PCM programmer
Joined: 06 Sep 2003
Posts: 21708
|
|
| Posted: Fri Feb 10, 2012 3:33 pm |
|
|
You have to call fgetc() inside the #int_rda routine. You have to read the
char to prevent the PIC from interrupting over and over again. Add an
fgetc() statement at the start of the isr. |
|
|
Beowulf
Joined: 10 Feb 2012
Posts: 7
Location: washington
|
|
| Posted: Fri Feb 10, 2012 3:39 pm |
|
|
I have tried various version of this
| Code: |
#int_RDA
void RDA_isr(void)
{
char temp;
temp = fgetc(PORT1);
fputc('Z',PORT1);
}
|
the processor doesn't appear enter this block..... |
|
|
PCM programmer
Joined: 06 Sep 2003
Posts: 21708
|
|
| Posted: Fri Feb 10, 2012 4:02 pm |
|
|
What pins are you using for Tx and Rx ? This PIC has default and alternate
pins for those signals. |
|
|
Beowulf
Joined: 10 Feb 2012
Posts: 7
Location: washington
|
|
| Posted: Fri Feb 10, 2012 5:46 pm |
|
|
If I give the command
| Code: |
#use rs232(baud=9600,UART1,ERRORS)
|
the compiler chooses A0 for tx and A1 for rx...so that would seem to be the default. The data sheet allows C4 for tx and C5 for rx to be selected through APFCON as well. The chip only has one UART. |
|
|
PCM programmer
Joined: 06 Sep 2003
Posts: 21708
|
|
| Posted: Fri Feb 10, 2012 6:01 pm |
|
|
I don't think it defaults to A0 and A1 (for Tx and Rx). I made the
following program and tested it just now in hardware, and I see pulses
on pin C4. Also, in the Overview section of the 16F1824 data sheet,
where it lists the pins, it says for pins C4 and C5:
| Quote: |
2: Default function location.
|
I think you probably need to use APFCON if you want to use A0 and A1.
The program below outputs a signal on pin C4:
| Code: |
#include <16F1824.h>
#fuses INTRC_IO, NOWDT, BROWNOUT, PLL_SW, NOLVP
#use delay(clock=4M)
#use rs232(baud=9600, UART1, ERRORS)
//========================================
void main()
{
while(1)
{
putc(0x55);
}
} |
|
|
|
Beowulf
Joined: 10 Feb 2012
Posts: 7
Location: washington
|
|
| Posted: Fri Feb 10, 2012 6:41 pm |
|
|
after compiling your code ...the LST file shows
| Code: |
0056: MOVLB 02
0057: BSF 1D.2
0058: BSF 1D.7
|
bank 2 offset 0x1D is the APFCON0 register
setting these bits selects A0 and A1.
Adding this
| Code: |
#byte APFCON0 = getenv("SFR:APFCON0")
#bit TXSTA = APFCON0.2
#bit RXSTA = APFCON0.7
TXSTA =0;
RXSTA =0;
|
compiles to a LST file that has
| Code: |
.................... TXSTA =0;
0065: BCF 1D.2
.................... RXSTA =0;
0066: BCF 1D.7
|
which clears the bits in APFCON0
but I still get RS232 on A0 & A1 with no INT_RDA |
|
|
PCM programmer
Joined: 06 Sep 2003
Posts: 21708
|
|
| Posted: Fri Feb 10, 2012 7:34 pm |
|
|
The PIC data sheet says you have to set the APFCON0 bits to 1,
in order to select the alternate pins. I did that and it works:
| Code: |
#include <16F1824.h>
#fuses INTRC_IO, NOWDT, BROWNOUT, PLL_SW, NOLVP
#use delay(clock=4M)
#use rs232(baud=9600, UART1, ERRORS)
#byte APFCON0 = getenv("SFR:APFCON0")
#bit TXSEL = APFCON0.2
#bit RXSEL = APFCON0.7
#int_RDA
void RDA_isr(void)
{
char c;
c = getc();
putc(c);
}
//========================================
void main()
{
TXSEL = 1;
RXSEL = 1;
clear_interrupt(INT_RDA);
enable_interrupts(INT_RDA);
enable_interrupts(GLOBAL);
while(1);
} |
|
|
|
Beowulf
Joined: 10 Feb 2012
Posts: 7
Location: washington
|
|
| Posted: Fri Feb 10, 2012 8:53 pm |
|
|
just updated to the 4.129 software version.
this no longer appears in the LST file in the default case(tx=C4,rx=C5)
| Code: |
0056: MOVLB 02
0057: BSF 1D.2
0058: BSF 1D.7
|
I am no longer at my work bench ...so I can't load this onto the PIC to be sure...but It seems as though the 4.128 version has a bug when compiling for this processor. If I have time this weekend I will try this out to see if it works on the PIC. |
|
|
Beowulf
Joined: 10 Feb 2012
Posts: 7
Location: washington
|
| (Solved) |
| Posted: Mon Feb 13, 2012 9:15 am |
|
|
I listed my compiler version incorrectly in the problem statement. The version that has this issue is:
| Code: |
CCS PCM C Compiler, Version 4.124
|
I am convinced that this was a compiler bug. I upgraded to 4.129 and the problem has gone away. Th tx/rx pins are correctly assigned now and the interrupts occur as expected.
**always check for updates** |
|
|
Ttelmah
Joined: 11 Mar 2010
Posts: 19537
|
| Re: (Solved) |
| Posted: Mon Feb 13, 2012 9:23 am |
|
|
| Beowulf wrote: | I listed my compiler version incorrectly in the problem statement. The version that has this issue is:
| Code: |
CCS PCM C Compiler, Version 4.124
|
I am convinced that this was a compiler bug. I upgraded to 4.129 and the problem has gone away. Th tx/rx pins are correctly assigned now and the interrupts occur as expected.
**always check for updates** |
No!.
Not if the compiler works for what you want.....
Key here is that the 1824 is a pretty new chip, and it always seems to take CCS perhaps a dozen versions to get these working.
However CCS's 'updates', generally are not reliable till they have been tested for a while, so rule has to be 'check for updates _if_ you have something not working right'. Otherwise you may well be asking for problems.....
Best Wishes |
|
|
Beowulf
Joined: 10 Feb 2012
Posts: 7
Location: washington
|
|
| Posted: Mon Feb 13, 2012 11:27 am |
|
|
| you are correct....updates are only useful when there is an actual problem to solve... |
|
|
|
FAQ 
Search
Register
Profile 
Log in to check your private messages 
Log in
