 |
 |
| View previous topic :: View next topic |
| Author |
Message |
evsource
Joined: 21 Nov 2006
Posts: 129
|
| spi_read on slave with simple code not working |
 Posted: Thu Jul 15, 2010 4:44 pm |
 |
|
Hi,
I'm using 4.092 PCH. Two different PIC18's trying to do some very simple communications with SPI just to get it working.
The master does get the expected value from the slave (0x08). However, the value that the slave reads is not correct (should be 0x15). If in my diagnostic line, if I enter 0x00, then I get a match.
In the master, I try putting spi_read(0), but then nothing works.
Any ideas?! I've tried every possible combination. I'm wondering if it's a compiler problem.
Master:
| Code: | #include <18f4685.h>
#fuses H4,WDT256,NOMCLR,NOWDT,NOBROWNOUT,BORV20,NOLVP
#fuses NOPROTECT,NOCPD,NOCPB,NOWRT,NOWRTD,NOWRTB,NOWRTC,NOEBTR,NOEBTRB
#use delay(clock=40000000)
// SPI mode definitions.
#define SPI_MODE_0 (SPI_L_TO_H | SPI_XMIT_L_TO_H)
#define SPI_MODE_1 (SPI_L_TO_H)
#define SPI_MODE_2 (SPI_H_TO_L)
#define SPI_MODE_3 (SPI_H_TO_L | SPI_XMIT_L_TO_H)
#define SLAVE_SELECT PIN_D7
void main() {
int8 return_value;
setup_spi(SPI_MASTER | SPI_MODE_3 | SPI_CLK_DIV_64);
while(1) {
output_low(SLAVE_SELECT);
spi_write(0x15);
output_high(SLAVE_SELECT);
output_low(SLAVE_SELECT);
return_value = spi_read();
output_high(SLAVE_SELECT);
if(return_value == 0x08) output_high(PIN_C7);
else output_low(PIN_C7);
delay_ms(1);
}
} |
Slave:
| Code: | #include <18f14k22.h>
#fuses INTRC,PLLEN,WDT256,NOMCLR,NOWDT,NOBROWNOUT,BORV22,NOLVP
#fuses NOPROTECT,NOCPD,NOCPB,NOWRT,NOWRTD,NOWRTB,NOWRTC,NOEBTR,NOEBTRB
#use delay(clock=64000000)
// SPI mode definitions.
#define SPI_MODE_0 (SPI_L_TO_H | SPI_XMIT_L_TO_H)
#define SPI_MODE_1 (SPI_L_TO_H)
#define SPI_MODE_2 (SPI_H_TO_L)
#define SPI_MODE_3 (SPI_H_TO_L | SPI_XMIT_L_TO_H)
int8 rx_byte;
void main() {
setup_oscillator(OSC_64MHZ);
setup_spi(SPI_SLAVE | SPI_MODE_3 );
while(1) {
output_low(PIN_C0); // diagnostics
if(spi_data_is_in()) {
output_high(PIN_C0); // diagnostics
rx_byte = spi_read();
spi_write(0x08);
if(rx_byte == 0x15) output_high(PIN_C1); // diagnostics
else output_low(PIN_C1); // diagnostics
}
}
} |
|
|
 |
evsource
Joined: 21 Nov 2006
Posts: 129
|
|
| Posted: Thu Jul 15, 2010 5:44 pm |
|
|
Just a couple more things I've discovered. Watching the signals on the scope, if send 0x0F from master to slave, I can see the four highest bits of 0x0F clocked in with the data line low, and then the last four bits clocked in with the data line low. The slave select line drops low just before the clock starts ticking. The slave select line goes high again right after the clock stops. I placed an output high to a pin right after the spi_data_is_in() call, and it goes high just a very small time after the slave select line goes high.
Why in the world is the slave reading zero instead of the value being sent??!! I looked at the list file to make sure it was reading from the correct register, which it is. I can't figure out where it would be getting the 0x00 from.
Last edited by evsource on Thu Jul 15, 2010 6:03 pm; edited 1 time in total |
|
|
evsource
Joined: 21 Nov 2006
Posts: 129
|
|
| Posted: Thu Jul 15, 2010 6:02 pm |
|
|
Just checked the errata for the 18F14K22, and there are SPI issues. However, they are all for master configuration. I'm using it in the slave mode.
The 18F4685's errata shows a small glitch that can happen on the SCK line. I don't think that's related to the slave receiving 0x00 when it's supposed to get 0x15. |
|
|
FvM
Joined: 27 Aug 2008
Posts: 2337
Location: Germany
|
|
| Posted: Thu Jul 15, 2010 11:34 pm |
|
|
| The receive event should be signalled after one SPI frame has been clocked in, independent of the SS pin transition. |
|
|
|
|
You cannot post new topics in this forum
You cannot reply to topics in this forum
You cannot edit your posts in this forum
You cannot delete your posts in this forum
You cannot vote in polls in this forum
|
Powered by phpBB © 2001, 2005 phpBB Group
|
FAQ 
Search
Register
Profile 
Log in to check your private messages 
Log in
