I2C devices, when powering up, can be accidentally placed inside a READ
or WRITE sequence due to its detecting a false start.

I always issue a STOP then a START then a STOP after powerup. It fixes
ALL I2C chips on the bus.

--Bob



Dmitriy Fitisov wrote:
> I think you are talking about START byte - see I2C specification
> 10.1.2 for version 2.1
> Basically it used for slaves which are software-coded, and in some cases,
> yes, slave remembers last address it was addressed with - see 10-bit address
> READ mode - it sets address in WRITE mode, after that send repeated START,
> 2-nd byte of the address with READ byte.
>
> Dmitriy
>
> -----Original Message-----
> From: pic microcontroller discussion list
> [mailto:PICLIST@MITVMA.MIT.EDU]On Behalf Of Nick Veys
> Sent: March 7, 2004 4:50 PM
> To: PICLIST@MITVMA.MIT.EDU
> Subject: [PIC]: I2C Guru's, Help!
>
>
> Ok, this has to be a quirk of I2C that I'm not aware of, this is odd.
>
> I'm using a PIC18F2320 to talk to a Matrix Orbital GLK12232-25SM LCD, It
> has a 16C87 on it FYI.
>
> I was having trouble talking to it so I thought I had the address wrong,
> I wrote a loop to go though all possible addresses, and bingo it worked.
>   So I binary searched til I found it, decimal 80, which was what I
> thought it was in the first place.  So I took the loop out and tried to
> talk to 80, no luck.  What the hell I say to myself. :)
>
> So I play around a bit and find that if I try to send data to it without
> first having tried to send data beforehand, ANY data, it won't work.
> But if I just send 0x00 to address 0x00 (or anything to anything) it
> works fine.
>
> So I'm guessing it's some kind of reset condition on the I2C bus?  I set
> SDA/SCL high before starting anything and that didn't seem to do
> anything, maybe something in the PIC that gets setup after the first
> run?  Seems odd but I'm out of ideas here! :)
>
> Here's my code to illustrate if I'm doing something absurd:
>
> --------- SNIP ------------
> void init() {
>    TRISC = 0b10000000; // See pinout
>
>    SMP = 1; // Slew Rate Standard (100KHz - 1MHz)
>    SSPADD = 19; // I2C clock = 8M/(4*(19+1)) -> 100KHz
>    SSPCON1 = 0x08; // Master, clock from SSPADD
>    SSPEN = 1; // Synch. Serial Port Enabled
> }
> void sendI2C(unsigned char addr, char* data, unsigned char len) {
>    unsigned char i;
>
>    SSPIF = 0; // Clear SSP flag
>    SEN = 1; // Start Enable
>    while(SEN); // Wait to finish
>    SSPIF = 0; // Clear SSP flag
>    SSPBUF = addr; // Send address
>    while(!SSPIF); // Wait for SSPIF flag
>    SSPIF = 0; // Clear SSP flag
>    for(i = 0; i < len; i++) {
>      SSPBUF = data[i]; // Send data
>      while(!SSPIF); // Wait for SSPIF flag
>      SSPIF = 0; // Clear SSP flag
>    }
>    PEN = 1;
>    while(PEN); // Wait to finish
>    SSPIF = 0; // Clear SSP flag
>
>    delayMilliseconds(1);
> }
>
> void sendStuff() {
>    char data[] = { 0xFE, 0x62, 2, 0, 0 };
>    char len = 5;
>
>    sendI2C(80, data, len);
>    sendI2C(80, data, len);
> }
>
> void main() {
>    init(); // setup ports, registers, etc...
>
>    delaySeconds(2); // wait for everyone to wake up
>    sendStuff();
>
>    while(1);
> }
> ------------- SNIP --------------
>
> This works.  If I change sendStuff() to have just 1 sendI2C, it doesn't
> work.  Any suggestions are appreciated on the code as well of course.
>
> Thanks!
>
> Nick
>
> --
> http://www.piclist.com hint: PICList Posts must start with ONE topic:
> [PIC]:,[SX]:,[AVR]: ->uP ONLY! [EE]:,[OT]: ->Other [BUY]:,[AD]: ->Ads
>
> --
> http://www.piclist.com hint: PICList Posts must start with ONE topic:
> [PIC]:,[SX]:,[AVR]: ->uP ONLY! [EE]:,[OT]: ->Other [BUY]:,[AD]: ->Ads
>


--

  Replies: NOTE-Script, EXE,BAT and COM
     files will be rejected by server
              --------------
                Bob Axtell
        PIC Hardware & Firmware Dev
          http://beam.to/baxtell
              1-520-219-2363

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