AVR Programming

in C
LECTURE# 08
MICROPROCESSOR SYSTEMS AND INTERFACING

Saad Arslan COMSATS UNIVERSITY ISLAMABAD 1

Loop Revision
How it works
How many times the loop will iterate?
How many clock cycles each instruction needs?
How many times each instruction is repeated?

;this program adds value 3 to the R20 ten times

.INCLUDE "M32DEF.INC"
LDI R16, 10 ;R16 = 10 (decimal) for counter
LDI R20, 0 ;R20 = 0
LDI R21, 3 ;R21 = 3
MUST be together, AGAIN: ADD R20, R21 ;add 03 to R20 (R20 = sum)
to make decision on DEC R16 ;decrement R16 (counter)
DEC R16 result BRNE AGAIN ;repeat until COUNT = 0
OUT PORTB, R20 ;send sum to PORTB

Saad Arslan COMSATS UNIVERSITY ISLAMABAD 2

Quiz 1b
Get ready for quiz
About to start

Saad Arslan COMSATS UNIVERSITY ISLAMABAD 3

Quiz 1b
Q1. What is the value of the C and Z flags, upon last execution of DEC R16
◦ For the program given on the right LDI R17, 3
LDI R18, 10
LDI R16, 6 ;loop variable
loop: ADD R18, R17
Q2. Find the number of clock cycles needed DEC R16
◦ To execute the program given on the right BRNE loop

◦ Also find the duration when running on 2 MHz oscillator

Q3. Write a simple assembly code to load value 20 in R20.

◦ Increment R20 NNN times using a loop (where NNN is your reg#)
◦ Write incremented value to PORTD in each loop iteration
◦ PORTD must be declared output, in the start
◦ It is suggested to use BRNE conditional branching instruction

Saad Arslan COMSATS UNIVERSITY ISLAMABAD 4

Last Lecture
I/Ports in AVR
◦ I/O Port registers (DDR, PORT and PIN registers)

Inputting data
◦ Role of DDR and PORT register
◦ Synchronizer delay

Dual Roles of Ports and Pins

I/O port bit-addressability
◦ Output using SBI and CBI instructions
◦ Input using SBIS and SBIC in instructions
◦ Examples

Saad Arslan COMSATS UNIVERSITY ISLAMABAD 5

Assembly vs. C language
Hex files are downloaded into AVR Flash
Assembly language code is used to generate hex files
Assembly language is often tedious and time consuming
◦ C language is easier
◦ C language provide higher level programming
However, the generated hex files from C language may be large
◦ First C language code is compiled to get Assembly language code
◦ The generated assembly code is assembled to get hex file
The added process of compiling may not be as accurate as writing
assembly code
◦ There is a trade-off between convenience in coding and size/accuracy of
coding

Saad Arslan COMSATS UNIVERSITY ISLAMABAD 6

Assembly vs. C language cont.
C language code can be ported easily to other devices
C language code libraries can be used
There are several third-party C compilers are available for AVR

Saad Arslan COMSATS UNIVERSITY ISLAMABAD 7

C data types for AVR
Data types define the size required for the variable in the memory
AVR has limited registers and to generate smaller hex file
◦ We need to accurately declare variable types, how?

AVR is an 8-bit microcontroller

◦ So natural choice of C data type will be char (or unsigned char)

Storage of large variable (16- or 32-bit)

◦ Require additional space in the AVR (more than one register)
◦ Arithmetic operations on such variables require multiple iterations

Saad Arslan COMSATS UNIVERSITY ISLAMABAD 8

C data types for AVR
unsigned char
Most proper choice, since AVR is an 8-bit microcontroller
The variable can store values 0-255 (0x00 - 0xFF)
This variable type should be first choice
◦ Unless the values to be stored are out of range

See Example 7-1

◦ Write a C program to send values 0x00 - 0xFF to Port B
◦ Will the loop ever break? #include <avr/io.h>
◦ No, z will always be int main (void)
{
◦ less than or equal to 255 unsigned char z; //loop count variable
◦ Because, z cannot be 256 DDRB = 0xFF;
for (z=0; z<=255; z++)
◦ Incrementing after 255 will overflow to 0 PORTB=z;
return 0;
}

Saad Arslan COMSATS UNIVERSITY ISLAMABAD 9

C data types for AVR
unsigned char cont.
See Example 7-2
◦ Write a C program to send values 0,1,2,3,4,5,A,B,C,D to port B
◦ As ASCII characters
#include <avr/io.h>
int main (void)
{
unsigned char myList[] = "012345ABCD"; //character array

unsigned char z; //loop counter variable

DDRB = 0xFF; //port B as output

for (z=0; z<10; z++)

PORTB=myList[z]; //send characters to Port B

while(1); //the program stucks here

return 0;
}

Saad Arslan COMSATS UNIVERSITY ISLAMABAD 10

C data types for AVR
unsigned char cont.
See Example 7-3
◦ Write an AVR C program to toggle all the bits of Port B 200 times.
#include <avr/io.h> //standard AVR header
int main(void) //the code starts from here
{
DDRB = 0xFF; //PORTB is output

PORTB = 0xAA; //Initial PORTB value is 0b 1010 1010

unsigned char z; //loop counter variable

for(z=0; z < 200; z++) //run the next line 200 times
PORTB = ~PORTB; //toggle PORTB

while(1); //stay here forever

return 0;
}

Saad Arslan COMSATS UNIVERSITY ISLAMABAD 11

C data types for AVR
signed char
8-bit data type −𝟐𝟕 26 25 24 23 22 21 20
◦ Most significant bit represents sign D7 D6 D5 D4 D3 D2 D1 D0

The variable can store values -128 to +127

See Example 7-4 Value Hex
◦ Write an AVR C program to send values of -4 to +4 to Port B. -4 0xFC
#include <avr/io.h> //standard AVR header -3 0xFD
int main(void)
What binary
{ (hex) values -2 0xFE
char mynum[] = {-4,-3,-2,-1,0,+1,+2,+3,+4}; will be -1 0xFF
unsigned char z;
DDRB = 0xFF; //PORTB is output written on 0 0x00
for(z=0; z<=8; z++) //Nine iterations PortB? +1 0x01
PORTB = mynum[z];
while(1); //stay here forever +2 0x02
return 0;
+3 0x03
}
+4 0x04

Saad Arslan COMSATS UNIVERSITY ISLAMABAD 12

C data types for AVR
Integer (int)
unsigned int
◦ 16-bit data type
◦ The variable can store values 0 to 65,536 (0x0000 to 0xFFFF)

signed int
◦ 16-bit data type
◦ Most significant bit represents sign
◦ The variable can store values -32,768 to +32,767

Since AVR is 8-bit microcontroller, will take 2 locations for each variable
Multiple instructions will be needed for performing operations

Saad Arslan COMSATS UNIVERSITY ISLAMABAD 13

C data types for AVR
Other data types
If it is required to store larger number than 65536
◦ Or there is a need to store fractional numbers
◦ We may consider some other types

long (signed and unsigned) (32-bit)

float (for representing fractional numbers) (32-bit)
double (for representing fractional numbers) (32-bit)
Write a C program to toggle port B, 50,000 times or 100,000 times
◦ You may choose unsigned int for 50,000
◦ And unsigned long for 100,000

Saad Arslan COMSATS UNIVERSITY ISLAMABAD 14

Time Delay
There are three possible ways of achieving time delay in AVR C
1. A simple for loop, etc. See example 7-7 (on the next slide)
◦ Inaccurate, Unable to theoretically ensure the delay
◦ A loop iteration may be mapped to different number of assembly instructions
◦ An empty loop may be removed during the optimization process
◦ Also depends upon the crystal frequency, verified using oscilloscope etc.
2. Predefined Functions
◦ _delay_ms() and _delay_us() functions
◦ varies across compiler, no standard functions
◦ If wrapper function is used for compatibility it will also consume some cycles
3. AVR timers
◦ Accurate
◦ Using hardware module inside the AVR (will be studied in later lectures)

End of lecture

Saad Arslan COMSATS UNIVERSITY ISLAMABAD 15