No LF & CR

[AtomSoft]

Im trying to write a ROM string out to UART at high speed (even at low speed same results)

My Speed is 115.2kbps (using a 14.7456MHz Crystal)

NO PLL ...

rom char welcome[23] = "AtomSoftTech Hey\xD\xA";

//... Init UART
init_uart();

//...SEND ROM STRING
UartRomStr(welcome);

//... Functions
void init_uart(void){
    TRISC  = 0b10000000;
    TXSTAbits.SYNC = 0;
    BAUDCTLbits.BRG16 = 0;
    TXSTAbits.BRGH = 1;
    SPBRG = 7;        //115,200bps (115.2Kbps) for 14.7456 MHz

    RCONbits.IPEN = 1;
    IPR1bits.RCIP = 1; //PRIORITY

    TXSTAbits.TXEN = 1;  //ENABLE TX
    RCSTAbits.CREN = 1;  //ENABLE RX
    RCSTAbits.SPEN = 1;  //ENABLE SERIAL PORT and PIN Config

    //Disable All Interrupts
    PIE1 = 0; 
    PIR1   = 0;
    INTCON = 0;

    Delay10TCYx(100);
}

void UartRomStr(rom char *string){
    char temp = 0;

    while(*string)
    {
        temp = *string++;
        while(!PIR1bits.TXIF);     //wait until TXIF is high
        TXREG = temp;    //put byte into Transmit Register
    }
}

The code works as i get the message on my terminal but it does not create a Line Feed...

I tried:
\n\r --------and-------\r\n
\xA\xD --------and-------\xD\xA

ALSO TRIED:
welcome[]....

NO LUCK! You think its my terminal ? Im using Putty (0.62) (windows 7)

Thanks in advance guys (and some gals)

 

[AtomSoft]

Update:

I changed the While to :

while(!TXSTAbits.TRMT);

It does a CR now but LF is still a no go

 

[AtomSoft]

Ok i got it to work like:

rom char welcome[] = "AtomSoftTech Hey\r\n \0";

Has to have a space...

EDIT: Fixed

rom char welcome[24] = "AtomSoftTech Hey\r\n\0";

 

[Ian Rogers]

Jason... I found that some esc chars don't work... When you simulate does the string (in memory) contain the two characters or is there a '0' (null) stuck in there.

 

[Ian Rogers]

Never mind

 

[AtomSoft]

Thanks for trying tho Ian

 

[Ian Rogers]

I love the way you talk (scribble) out loud...

 

[AtomSoft]

****Remove this post please****

 

[AtomSoft]

Heh thanks my scribble is understandable tho lol

 

[AtomSoft]

Ok switched to a 10Mhz crystal so i can use PLL... need more speed..

To my calculations:
10mhz x 4xPLL = 40 Mhz

40Mhz / 115200 = 347.2222222222222 / 16 (BRGH = 1) = 21.70138888888889 - 1 = 20.70138888888889

I rounded it up to 21 which would be: FOSC/(16 ([SPBRGH:SPBRG] + 1))

40Mhz / (16 x (21 + 1)) = 40Mhz / (16 x (22)) = 40MHz / (352) = 113636.3636363636

Which means i get a error of : ((113636.3636363636 - 115200)*100) / 115200 = -1.35%

Is that too much ? I know its way off but it seems to work. I just dont want to run into a problem later on

 

[Gobbledok]

Hi Atom,

Theoretically, if you have a start bit, 8 data bits and a stop bit, then you can allow for around 5% frequency mismatch before you start reading the last bit wrong.

Be aware that's 5% total, so each end may have 2.5% or one end may have 5%.


My thinking goes:

The UART samples in the middle of the bit therefore you can allow your frequency to be off by half a bit before you start getting errors. In total there are 20 half-bits. 1/20 = 5%

 

[AtomSoft]

Ah cool thanks

 

[Pommie]

I've found RS232 to be quite tolerant of frequency mismatches. If you do get problems then switch to two stop bits and that normally fixes it.

Mike.