My microcontroller cant do maths!

Heres my code:

Code:

#include <htc.h>
#include <stdint.h>

#define _XTAL_FREQ 8000000

__CONFIG (FOSC_XT & WDTE_OFF & PWRTE_ON & MCLRE_ON & CP_OFF & CPD_OFF & BOREN_OFF & IESO_OFF & FCMEN_OFF & LVP_OFF & DEBUG_OFF);

void initDigitalOut() {
	PORTA = PORTB = PORTC = PORTD = 0;
	TRISA = TRISB = TRISC = TRISD = 0;
	ANSEL = ANSELH = 0;
}

unsigned int math() {
	float revolutionTime_S = 3000/31250;
	float horizontalDelay_US = revolutionTime_S/276;
	float horizontalDelay_US = horizontalDelay_US*1000000;
	return (unsigned int) horizontalDelay_US;
}

void main() {
	initDigitalOut();
	while(1) {
	PORTB = math();
	PORTD = math() >> 8;
	}
}

I seperated horizontal delay into 2 sections because revolutionTime_S ranges from about 0.032 to 2, and I dont think float would like to deal with numbers like 2,000,000? Anyway, I get no output on the ports at all. When I return a constant of say 350, the ports light up nicely. The value the function should be outputting ranges from about 350 to about 7500. All within the range of a 16 bit number. Can anyone shed light onto what's happening here?

Thanks!!!

> float revolutionTime_S = 3000/31250;
This will always evaluate to 0, since the division is done in integer math.

Try
float revolutionTime_S = 3000.0/31250.0;

Or like this

Code:

float revolutionTime_S = (float) 3000/31250;

you sir have just given me the solution to a much bigger problem. Thanks alot!!!!

You both*

Code:

float revolutionTime_S = 3000/31250;

"This will always evaluate to 0, since the division is done in integer math."

Salem, would you care to explain what is going on in this statement?

The numbers 3000 and 31250 both look like integers, so the compiler treats them as such and performs an integer division, which truncates any fractional part and has an integer result. When you use 3000.0 and 31250.0 instead, both numbers look like floating point types. As long as at least one operand of a division is a floating point type, the compiler does floating point division.

One way'd be to scale the output so the fractional part is insignificant:

Code:

unsigned int math()
{
    return ((3000 * 1000000) / (276 * 31250));
}

And since 3000/31250 is all fractional, the output is 0. I was getting confused thinking that any integer maths when assigned to a float would come out as 0. Thanks for the help everyone.