division include floating point... - Electronic Circuits Projects Diagrams Free

meera83 · 14th June 2007, 04:30 AM

i m using PIC16F877, Assembly language..

if i want to do math division, how should i do to make sure the LCD will also show floating point like 1/4=0.25(two decimal point)...

i had a division program, but the LCD didn't show result in two decimal point..1/4=0..

thanks..

ericgibbs · 14th June 2007, 07:47 AM

hi meera,
Are you using the microchip.com floating point subroutines?

Can you post the code that you are using, maths and LCD driver/conversion subr?

**Nigel Goodwin** · 14th June 2007, 03:36 PM

You also need to consider if you really need floating point?, most applications work fine with integers - much faster, much smaller, and more accurate - simply insert the decimal point in the display routines.

For example - don't use dollars and cents ($4.23), use just cents instead (423 cents) - this is how spreadsheets do maths for money, and for the same reasons.

Mike, K8LH · 15th June 2007, 09:07 PM

I agree with Nigel. Scale your values and stick with Integer arithmetic. It's simpler. Then insert the decimal point manually.

Take 12-bit two's compliment sign extended Dallas 18B20 Temperature Sensor data, for example. The 12-bit binary data represents °Celcius in 0.0625° increments, or put another way, °Celcius*16. By scaling the formulas you can get away with a simple 8x16 multiply routine, a simple divide-by-16 routine, and a simple binary-to-bcd routine to provide decimal output to 1 decimal place (you still have to insert the decimal point manually).

Code:

;******************************************************************
;
;  BC2DC - convert binary °C*16 to decimal °C*10 in D3:D0 vars
;
;  formula is: °C*10 = (10*(Celsius*16))/16
;                    = (10*(TempH:TempL))/16
;
;  D3:D0 result '-250' to '1250' (°C*10)
;
BC2DC
        movlw   d'10'           ; multiplier 10                   |B0
        call    Multiply        ; do 10 x TempH:TempL             |B0
        goto    Div16           ;                                 |B0

;******************************************************************
;
;  BC2DF - convert binary °C*16 to decimal °F*10 in D3:D0 vars
;
;  formula is: °F*10 = (18*(Celsius*16)+(320*16))/16
;                    = (18*(TempH:TempL)+( 5120))/16
;
;  D3:D0 result '-670' to '2570' (°F*10)
;
BC2DF
        movlw   d'18'           ; multiplier 18 (9/5*10)          |B0
        call    Multiply        ; do 18 x TempH:TempL             |B0
;
;  add (32 * 10 * 16) to result
;
        movlw   low (.320*.16)  ; add (320 * 16) to Product       |B0
        addwf   ProdL,F         ;                                 |B0
        skpnc                   ;                                 |B0
        incf    ProdH,F         ;                                 |B0
        movlw   high (.320*.16) ;                                 |B0
        addwf   ProdH,F         ;                                 |B0
        skpnc                   ;                                 |B0
        incf    ProdU,F         ;                                 |B0
;
;  divide result by 16
;
Div16
        movlw   d'4'            ;                                 |B0
        movwf   Count           ;                                 |B0
DivNext
        clrc                    ; clear C                         |B0
        rrf     ProdH,F         ;                                 |B0
        rrf     ProdL,F         ;                                 |B0
        decfsz  Count,F         ; all done?                       |B0
        goto    DivNext         ; no, branch, else                |B0
;
;  check for negative number result
;
        bcf     TempH,7         ; clear "negative" flag           |B0
        btfss   ProdH,3         ; negative result?                |B0
        goto    Bin2Dec         ; no, branch, else                |B0
;
;  convert it to a postive number
;
        bsf     TempH,7         ; set "negative" flag             |B0
        movf    ProdL,W         ; temperature low                 |B0
        sublw   h'00'           ; two's complement it             |B0
        movwf   ProdL           ; ^ is that a verb (grin)?        |B0
        movf    ProdH,W         ; temperature high in W           |B0
        skpc                    ; borrow?  no, skip, else         |B0
        incf    ProdH,W         ; temperature high + 1 in W       |B0
        sublw   h'00'           ; two's complement it             |B0
        movwf   ProdH           ;                                 |B0
;
;  Peter Hemsley's 12-bit Bin_to_BCD (ProdH:ProdL binary input
;  and D3:D0 unpacked BCD output)
;

Hero999 · 17th June 2007, 05:21 PM

Used fixed point binary numbers, look it up on Wikipedia!

Thread Tools
Show Printable Version Email this Page
Display Modes
Linear Mode Switch to Hybrid Mode Switch to Threaded Mode

Similar Threads
Title	Starter	Forum	Replies	Latest
Looks simple enough. But.....	hjl4	Micro Controllers	3	29th January 2007 12:17 PM
SIMPLE ADC PROBLEM	neelam29	Micro Controllers	2	9th March 2006 09:46 AM
Can PIC16F877 do floating point math?	stevenlkz	Micro Controllers	5	12th June 2005 06:00 PM
ya	mari_69	Micro Controllers	4	7th June 2004 02:29 PM

14th June 2007, 04:30 AM	(permalink)
meera83	division include floating point... i m using PIC16F877, Assembly language.. if i want to do math division, how should i do to make sure the LCD will also show floating point like 1/4=0.25(two decimal point)... i had a division program, but the LCD didn't show result in two decimal point..1/4=0.. thanks..

14th June 2007, 07:47 AM	(permalink)
ericgibbs	hi meera, Are you using the microchip.com floating point subroutines? Can you post the code that you are using, maths and LCD driver/conversion subr? __________________ Eric "Good enough is Perfect" PIC tutorials: Gramo's: www.digital-diy.net/ Bill's: www.blueroomelectronics.com/

14th June 2007, 03:36 PM	(permalink)
Nigel Goodwin	You also need to consider if you really need floating point?, most applications work fine with integers - much faster, much smaller, and more accurate - simply insert the decimal point in the display routines. For example - don't use dollars and cents ($4.23), use just cents instead (423 cents) - this is how spreadsheets do maths for money, and for the same reasons. __________________ PIC programmer software, and PIC Tutorials at: http://www.winpicprog.co.uk

17th June 2007, 05:21 PM	(permalink)
Hero999	Used fixed point binary numbers, look it up on Wikipedia! __________________ I also post at the following sites: http://www.stop-microsoft.org http://www.heated-debates.com Screen name: Aloone_Jonez And http://www.silicontronics.com, same screen name as here.