Please title this page. (ds1620-txt-rs232.htm)

include p16f628.inc list p=16f628 __config (_CP_OFF & _DATA_CP_OFF & _PWRTE_ON & _WDT_OFF & _INTRC_OSC_CLKOUT & _LVP_OFF & _BODEN_OFF) #define TXBUF_PTR 0x20 ; Current TX buffer pointer #define OP_BUF 0x21 ; Output temp location #define OP_LOOP 0x22 ; Output loop var #define TEMP1 0x23 #define DIV_QUOT 0x25 ; Division quotient #define DIV_RET 0x26 ; Division return #define DELAY1 0x28 #define DELAY2 0x29 #define CEL_BAK 0x2A #define LZERO_FLG 0x2B ; Flag indicating if we still drop leading zeros #define CEL_INT 0x2C #define CEL_FLT 0x2D #define CEL_REMAIN 0x2E #define CEL_PER_C 0x2F #define FDIV_TMP 0x3A #define FDIV_NUM 0x3B #define FDIV_DEN 0x3C #define FDIV_RES_W 0x3D #define FDIV_RES_F 0x3E #define FDIV_LOOP 0x3F #define TXBUF_START 0x40 ; Beginning of TX buffer #define FR_W_1 0x50 ; Fraction conversion stuff #define FR_F_1 0x51 #define FR_W_2 0x52 #define FR_F_2 0x53 #define R_TMP1 0x54 #define R_TMP2 0x55 #define INT_TMP1 0x60 org 0 goto progstart org 4 interrupt: swapf STATUS, W retfie progstart: call bank1 bsf PCON, 3 ; 32kHz (0) or 4 Mhz (1) internal clock movlw b'11110111' ; Bit5 as output andwf TRISB, F movlw b'11110001' ; Bits 3:1 as outputs andwf TRISA, F call bank0 movlw 0x07 movwf CMCON ; Comparators off call tx_init clrf PORTA clrf PORTB movlw 0x32 call delay_w call deg_write_conf ; Initialise DS1620 movlw 0x32 call delay_w testloop1: call deg_read call buffer_cel_val call tx_next_byte mainloop: btfsc PIR1, TXIF call tx_next_byte movf TXBUF_PTR, W movwf FSR movf INDF, W btfss STATUS, Z goto mainloop goto testloop1 tx_init: call bank1 bsf TRISB, 1 bsf TRISB, 2 bcf TXSTA, TX9 ; 8-bit transmission bsf TXSTA, TXEN ; Transmission enabled bcf TXSTA, SYNC ; Asynchronous mode bsf TXSTA, BRGH ; High speed baud rates movlw 0x19 movwf SPBRG ; 9600 baud at 4Mhz call bank0 bsf RCSTA, SPEN ; Enable USART bcf RCSTA, RX9 ; 8-bit receive bcf RCSTA, CREN ; Non-continuous return ; Should be called to put the next byte for transmission into the ; output register. tx_next_byte: bcf PIR1, TXIF ; Clear TX flag movf TXBUF_PTR, W movwf FSR movf INDF, W iorlw 0x00 btfsc STATUS, Z ; If next byte is 0, finish goto tx_finished movwf TXREG incf TXBUF_PTR, F return tx_finished: ; Clears the buffer and resets movlw 0x00 ; the pointer value. movwf TXBUF_START movlw TXBUF_START movwf TXBUF_PTR return bank0: bcf STATUS, RP1 bcf STATUS, RP0 return bank1: bcf STATUS, RP1 bsf STATUS, RP0 return output_w: movwf OP_BUF movlw 8 movwf OP_LOOP output_loop: movlw '0' rlf OP_BUF, F btfsc STATUS, C movlw '1' call buffer_w decfsz OP_LOOP, F goto output_loop movlw '\n' call buffer_w return ; Put the the temperature into the buffer. The value in ; CEL_INT and CEL_FLT are destroyed by this routine, so copies must ; be kept if they are still needed. buffer_cel_val: clrf LZERO_FLG ; Drop leading zeros clrf TXBUF_START ; Empty buffer movlw TXBUF_START movwf TXBUF_PTR btfss CEL_INT, 7 goto positive movlw '-' call buffer_w movlw 0xFF xorwf CEL_INT, F ; Get 2's complement incf CEL_INT, F positive: movf CEL_INT, W movwf DIV_QUOT movlw d'100' ; Do hundreds first call divide addlw '0' call buffer_w_nz movlw d'10' ; Do tens call divide addlw '0' call buffer_w_nz movf DIV_QUOT, W ; Remainder is units addlw '0' call buffer_w movlw '.' call buffer_w ; That concludes the integer part; now do the floating point bit. fraction_loop: movf CEL_FLT, W call mult_10 addlw '0' call buffer_w movf FR_F_1, W ; Get remaining fraction movwf CEL_FLT ; and copy to celcius float btfss STATUS, Z goto fraction_loop movlw '\n' call buffer_w movlw TXBUF_START ; Reset pointer to start of buffer movwf TXBUF_PTR return divide: ; Returns the result of DIV_QUOT / W as an int (in W) ; The result is returned in W and the remainder is ; left in DIV_QUOT clrf DIV_RET div_loop: ; Division loop subwf DIV_QUOT, F btfss STATUS, C goto div_return incf DIV_RET, F goto div_loop div_return: addwf DIV_QUOT, F ; Undo last subtract movf DIV_RET, W return mult_10: ; Multiply 8-bit binary fraction by 10 clrf FR_W_1 ; returning the whole part in W and movwf FR_F_1 ; the fraction in FR_F_1 clrf FR_W_2 movwf FR_F_2 movlw FR_W_1 ; Multiply FR_x_1 by 8 movwf FSR call shift_l_fsr call shift_l_fsr call shift_l_fsr movlw FR_W_2 ; Multiply FR_x_1 by 2 movwf FSR call shift_l_fsr movf FR_F_2, W ; Add the two results addwf FR_F_1, F btfsc STATUS, C incf FR_W_1, F ; Including carry if needed movf FR_W_2, W addwf FR_W_1, F movf FR_W_1, W return shift_l_fsr: ; Does a 16-bit shift of the two memory incf FSR, F ; locations pointed to by FSR (MSB first) rlf INDF, F bcf INDF, 0 decf FSR, F rlf INDF, F return ; Places the contents of W into the buffer area buffer_w_nz: ; Only does buffering if char is not 0 movf LZERO_FLG, F btfss STATUS, Z goto buffer_w ; Not first char, so don't drop sublw '0' btfsc STATUS, Z return sublw '0' ; Undo subtraction incf LZERO_FLG, F ; Indicates drop no more zeros until reset buffer_w: movwf TEMP1 movf TXBUF_PTR, W movwf FSR ; Point FSR at next buffer location movf TEMP1, W movwf INDF ; Put W into buffer location incf FSR, F clrf INDF ; Zero terminate string incf TXBUF_PTR, F ; Update buffer pointer return delay_w: ; Delays W * DELAY1 (= 983ms for W=0) movwf DELAY2 delay_w_loop: call delay1 decfsz DELAY2, F goto delay_w_loop return delay1: ; Delays approx 15 * 256 cycles = 3.84ms @ 4Mhz clrf DELAY1 ; movlw 0x02 ; Use 2 as test value if running at 32kHz ; movwf DELAY1 delay1_loop: nop ; Needs to take 15 cycles nop nop nop nop nop nop nop nop nop nop nop decfsz DELAY1, F goto delay1_loop return ; DS1620 reading routines. #define DEG_DQ 1 ; DS1620 data pin #define DEG_CLK 2 ; Clock pin #define DEG_RST 3 ; !Reset pin ; **** DS1620 Subroutines deg_read: bsf PORTA, DEG_CLK ; Clock starts off high bsf PORTA, DEG_RST ; Take out of reset mode movlw 0xEE call deg_tx_w bcf PORTA, DEG_RST ; Reset movlw d'150' ; Delays about 1sec while the call delay_w ; conversion is performed movlw d'150' call delay_w movlw 0xAA ; AA=Read Temp, A0=Counter, A9=Slope call ds1620_send_read movwf CEL_INT clrf CEL_FLT movlw 0x40 ; 0.25 in binary subwf CEL_FLT, F ; TODO - fix for negative values decf CEL_INT, F ; Next, do the calculation to calculate the exact temperature. ; See the DS1620 datasheet for details of what happens here. movlw 0xA0 call ds1620_send_read movwf CEL_REMAIN movlw 0xA9 call ds1620_send_read movwf CEL_PER_C movwf FDIV_NUM movwf FDIV_DEN movf CEL_REMAIN, W subwf FDIV_NUM, F call fdivide movf FDIV_RES_F, W addwf CEL_FLT, F btfsc STATUS, C incf CEL_INT, F return ds1620_send_read: ; Send command in W and read response bsf PORTA, DEG_RST call deg_tx_w call bank1 bsf TRISA, DEG_DQ ; Make DQ an input call bank0 movlw 0x09 ; Read 9 bits movwf R_TMP1 deg_read_loop: rrf R_TMP2, F bcf PORTA, DEG_CLK ; Clock goes low causing data bit nop nop btfss PORTA, DEG_DQ bcf R_TMP2, 7 btfsc PORTA, DEG_DQ bsf R_TMP2, 7 bsf PORTA, DEG_CLK nop nop decfsz R_TMP1, F goto deg_read_loop bcf PORTA, DEG_RST ; Reset call bank1 bcf TRISA, DEG_DQ ; Make DQ output call bank0 movf R_TMP2, W return deg_write_conf: bsf PORTA, DEG_CLK ; Clock starts off high bsf PORTA, DEG_RST ; Take out of reset mode movlw 0x0C ; W = Write config command 0C call deg_tx_w movlw 0x03 ; One shot and CPU only mode call deg_tx_w movlw d'50' ; Allow time for EEPROM write (max 50ms) call delay_w bcf PORTA, DEG_RST ; Reset return ; Transmit the contents of W. The protocol is that the data must ; be valid during the low->high transition of the clock cycle. This ; is done by setting the data then cycling the clock ->low->high. ; Data is sent LSB first. deg_tx_w: movwf R_TMP1 movlw 0x08 movwf R_TMP2 ; Bits left to send deg_tx_loop: btfss R_TMP1, 0 ; Set DQ line as LSB bcf PORTA, DEG_DQ ; of the data byte btfsc R_TMP1, 0 bsf PORTA, DEG_DQ bcf PORTA, DEG_CLK ; Cycle clock nop nop nop bsf PORTA, DEG_CLK rrf R_TMP1, F ; Shift out the bit just sent decfsz R_TMP2, F goto deg_tx_loop return ; *** TEST CODE *** fdiv_test: movlw d'21' movwf FDIV_NUM movlw d'8' movwf FDIV_DEN call fdivide ; 21/8 = 2.625 goto fdiv_test ; ***************** ; Divides one 8-bit number by another, including floating point ; calculation to 8 binary digits. The denominator must be less ; than 128. fdivide: call idivide movwf FDIV_RES_W ; Whole part is easy movlw 0x08 movwf FDIV_LOOP fdiv_loop: rlf FDIV_NUM, F ; Multiply by 2 bcf FDIV_NUM, 0 call idivide ; Result will be 0 or 1 rrf FDIV_TMP, F ; Rotate result into floating rlf FDIV_RES_F, F ; part of result. decfsz FDIV_LOOP, F goto fdiv_loop rlf FDIV_NUM, F ; Round off result bcf FDIV_NUM, 0 call idivide btfss FDIV_TMP, 0 return ; No rounding needed incf FDIV_RES_F, F btfsc STATUS, Z incf FDIV_RES_W, F return ; Integer division section. idivide: ; Returns the result of FDIV_NUM / FDIV_DEN as an int ; The result is returned in W and the remainder is ; left in FDIV_NUM. The result is also found in FDIV_TMP clrf FDIV_TMP movf FDIV_DEN, W idiv_loop: ; Division loop subwf FDIV_NUM, F btfss STATUS, C goto idiv_return incf FDIV_TMP, F goto idiv_loop idiv_return: addwf FDIV_NUM, F ; Undo last subtract movf FDIV_TMP, W return end

Welcome to www.piclist.com!