code:
 

 

                if(plen==0){ 

                lcd_clrscr(); 

				lcd_puts("Xacc Yacc Zacc"; 

				//select ADC1 and Use 3.3 Vcc as reference with cap at AREF pin 

				//Also set right adjusted in the result 

				ADMUX =0x41; //ADC1 

				delay_ms(1); 

				// start the conversion 

				ADCSRA |= 0x40; 

				while((ADCSRA&0x40) == 0x40){ 

				continue; 

				} 

				xacc = (ADC/25)-19; //put value into xacc 

				itoa(xacc,numstr,10); // convert integer to string 

				i=0; 

				while(numstr [ i ]!=0 && i<4) 

				{ 

				lcd_gotoxy(i,1); 

				lcd_putc(numstr[ i ]); 

				i++; 

				} 

//	ADC2 

				ADMUX =0x42; //ADC2 

				delay_ms(1); 

				// start the conversion 

				ADCSRA |= 0x40; 

				while((ADCSRA&0x40) == 0x40){ 

				continue; 

				} 

				yacc = 0; //put value into xacc 

				itoa(yacc,numstr,10); // convert integer to string			 

				i=0; 

				while(numstr [ i ]!=0 && i<4) 

				{ 

				lcd_gotoxy(i+5,1); 

				lcd_putc(numstr[ i ]); 

				i++; 

				} 

//	ADC3 

				ADMUX =0x43; //ADC3 

				delay_ms(1); 

				// start the conversion 

				ADCSRA |= 0x40; 

				while((ADCSRA&0x40) == 0x40){ 

				continue; 

				} 

				zacc = -5; //put value into xacc 

				itoa(zacc,numstr,10); // convert integer to string 

				i=0; 

				while(numstr [ i ]!=0 && i<4) 

				{ 

				lcd_gotoxy(i+10,1); 

				lcd_putc(numstr[ i ]); 

				i++; 

				} 

				delay_ms(500);				 

				// Ending accellerometer section here				 

				        continue; 

                } 

 

code:
 

 

Xacc  Yacc  Zacc 

0        0      -5 

 

or 

 

Xacc  Yacc  Zacc 

4       0      -5 

quote:

---

Originally posted by AdrianH

The maths will work if I use (ADC/2.5)-192 to give 0 to 100 display. So still to figure why (ADC-480)/25 will not work out.

---

quote:

---

Originally posted by MikeR
not sure if i'm talking down & apologise if i am. Just in case you're not aware ...



Signed integer - same as integer but the first half of the number is classed as negative. Eg 16 bit signed integer is -32000 to +32000 instead of 0 to 64000. If you started with the very first value and added 1 to it it would be classed as -31999, if you add 32000 to it, the computer thinks its actually 0.

---

quote:

---

Originally posted by MikeR
If you want to have a value with a decimal point it it, ie a real number you need to use floating point.

---

code:
 

while( ( numstr[ i ] !=0 ) && ( i<4 ) )  

{  

	lcd_gotoxy(i+10,1);  

	lcd_putc(numstr[ i ]);  

	i++;  

}  

 

 

lcd_gotoxy(10,1);  

while( ( numstr[ i ] !=0 ) && ( i<4 ) )  

{  

	lcd_putc(numstr[ i ]);  

	i++;  

}  

code:
 

/** 

 * Write a null terminated string to the LCD. 

 * Assumes cursor is at correct location prior to calling. 

 * @param str Pointer to null terminated string. 

 */ 

void lcd_puts( char *str ) 

{ 

    while( *str ) 

    { 

        lcd_putc( *str++ ); 

    } 

} 

 

 

/** 

 * Write a null terminated string to the LCD up to specified number of characters. 

 * Assumes cursor is at correct location prior to calling. 

 * @param str Pointer to null terminated string. 

 * @param len Number of characters to write. 

 */ 

void lcd_putsn( char *str, unsigned int len ) 

{ 

    while( *str && len ) 

    { 

        lcd_putc( *str++ ); 

        len--; 

    } 

} 

quote:

---

Originally posted by MikeR
eeek - its been a while. having to read up again now to relearn what i've obviously forgotten from my PC coding days. Sorry for the miss-information.

(i do vaguely remember at uni make up a adders circuit using logic gates - which i seem to recall needed two half adders which was a combination of AND gates plus something else).

---

quote:

---

Originally posted by MikeRJ

quote:

---

Originally posted by MikeR
If you want to have a value with a decimal point it it, ie a real number you need to use floating point.

---

You only need floating point maths when you are dealing with numbers that have a very large range, i.e. they can be very small or very large, and the floating point allows you to trade off resolution against range.

In the vast majority of embedded applications you can get away without using floating point maths, by using fixed point maths instead. Normally you would not scale intermediate values by factors of ten, but rather by factors of two since this gives you the most efficient/fastest calculations (though obvioulsy you need to scale the final result by a power of ten to display decimal places).

---

quote:

---

Originally posted by AdrianH
...
Anyone remember Nascom?

Adrian

---