Atmega Tutorial PDF

An Introduction to programming an Atmega
microcontroller
Benjamin Reh
November 14, 2013
Contents
1. Preface 3
2. Compilation and transfer 3

2.1. Preparation . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . 3
2.1.1. Basis . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . 3
2.1.2. First steps . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . 3
2.1.3. Fuse-bits . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . 4
3. The basics 4
3.1. Pins and Ports . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . 5
4. First steps 5
4.1. Controlling a LED . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . 5
4.1.1. Plugging in the LED . . . . . . . . . . . . . . . . . . . . . . . . . . . . . 5
4.1.2. Switching it on . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . 6
4.1.3. Making the LED blink . . . . . . . . . . . . . . . . . . . . . . . . . . . . 7
4.1.4. Making the LED blink (advanced) . . . . . . . . . . . . . . . . . . . . . 8
4.2. Reading an input . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . 8
4.2.1. Detecting an edge . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . 9
4.2.2. Providing a clean signal . . . . . . . . . . . . . . . . . . . . . . . . . . . 10
5. Timer/Counter 11
5.1. Hardware timer . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . 11
5.2. Servo control . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . 11
5.2.1. About servos . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . 11
5.2.2. Servos in basis . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . 12
5.3. Timer for software interrupt . . . . . . . . . . . . . . . . . . . . . . . . . . . . . 12
6. Analogue-Digital-Converter (ADC) 13
6.1. ADC in basis . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . 13
1
Version 1.1 IWR - robotics lab
7. Serial connection(UART) 13
7.1. UART in basis . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . 14
7.1.1. Example . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . 14
7.2. The terminal . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . 14
A. Appendix 15
A.1. Algebra with bits . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . 15
A.1.1. Shifting . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . 15
A.1.2. Bit-wise logic . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . 16
A.1.3. Setting a bit . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . 16
A.1.4. Clearing a bit . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . 16
A.2. Pin out of the Atmega 168 . . . . . . . . . . . . . . . . . . . . . . . . . . . . . 17
B. References 18
2
1. Preface
This document an introduction into the programming of an Atmega microcontroller. It is
separated into the first part guiding like a tutorial for beginners and a second part which is a
reference book to the functions provided in the basis.
The examples and explanations provided are neither exhaustive nor complete. The only aim
of this document is to lower the burden of getting started. Only a basic knowledge in C is
required.
2. Compilation and transfer

2.1. Preparation
Before starting to write own programs, it is advisable to first familiarize with the tool-chain
which will be used.
Prerequisites
To compile your programs and transfer them to the microcontroller, a few tools have to be
installed. If you are working on a PC in the Lab, then you can skip this section.
First, I assume you are working with Linux, so I also assume you already have your favorite
editor. If you do not, I recommend geany, which is a light, but powerful editor using gnome.
For the Atmega you need the avr port of the gcc, a program for uploading the code to the
device (avrdude) and make for running Makefiles.
Everything can be installed on Debian/Ubuntu systems with the following command
sudo apt-get install make avr-libc avrdude binutils-avr gcc-avr gdb-avr
For other distributions please take a look inside your documentation how to install new packages
and how to find their names.
2.1.1. Basis
For a clean start you should get a fresh copy of the basic programming environment we call
basis. You can either check it out from the svn-repository with
svn checkout http://ornella.iwr.uni-heidelberg.de/svn/basis
or download it from http://roboter.uni-hd.de. After downloading it, you will need to

extract the files from the archive of course.
Change the directory (cd basis) and start with your first steps.
2.1.2. First steps

Just type
make
to compile and get an output similar to this
3
Compiling main.c ...

Compiling adc.c ...
Compiling timer.c ...
Compiling pwm.c ...
Compiling uart.c ...
Compiling servo.c ...
Linking basis.elf ...
Creating basis.hex ...
Creating basis.eep ...
avr-objcopy: --change-section-lma .eeprom=0x0000000000000000 never used
If you encounter any problems or errors with this, you should verify your installation of the
toolchain.
The make command produces a binary output file called basis.elf which now needs to be
transferred to the microcontroller. You should supply power to the board and plug in the
programming device.
By typing
make prog
a program called avrdude will be started to initiate the transfer.
Since there is more than just one possible programmer, maybe the Makefile has to be adjusted
to the one you are using.
...
#select programmer
prog: prog-usb
#prog: prog-ser
#prog: prog-funk
...
In this example the usb-programmer is chosen because it is the only line that does not start
with a #.
If you encounter problems that look like a problem with permissions you can try executing
the command as root with for example sudo make prog. For being able to program as a
normal unprivileged user, you can modify your udev-rules accordingly. In the lab this has
already been done. This problem should therefor not occur.
2.1.3. Fuse-bits
A really essential part of initializing a new controller is setting the so called fuse-bits. You
should take a look at the board’s documentation how to do that in detail or visit http:
//www.engbedded.com/fusecal.
From this point of view it is sufficient to just call
make fuse
once every time you put a new controller chip in your board. If your are in doubt whether
your controller has already been fused, just it again. Fusing more than once doesn’t hurt.
3. The basics
In this chapter you learn what registers and ports are and how to set a pin on a certain port.
4
3.1. Pins and Ports

Pins
The word pin can be referred to as the metal connectors on the package of the controller chip.
While some of those pins can be freely used within your programs, some have fixed purposes
for example to supply voltage. There are also some pins, which are both: They normally have
a distinctive purpose, but in some applications can also be used freely.1
The electric connector pins that are not configurable cannot be accessed from within the
software.
So whenever the word pin is used in the following text, the ones that are accessible from inside
the software are meant.
Generally every (software) pin can be used as a digital input or output. This means that it
can either ”listen” to the signal that is supplied to the pin or provide a signal itself. These two
behaviors are called the Data Direction.
Ports
As the reader may recall from lectures in computer engineering2 , a processor has a set of
registers which contain data or control the peripheral devices. A microcontroller typically is
nothing more than a microprocessor with memory, some special hardware (counter, ADCs,. . . )
and direct access to IO.
The Atmega series we are using has a 8 bit architecture. It can calculate values of the size of
1 byte at normally one clock cycle.
For this reason 8 pins are combined to 1 port, which simply is represented by one register in
the CPU.
Ports are enumerated with capital letters (A,B,C,. . . ), pins with number from 0 . . . 7. For
example PIND6 represents the pin no. 6 on port D
Each bit in a register for IO represents a (physical) pin. This makes it necessary to know how
to manipulate just several bits at a time while leaving the others untouched.
If your are not familiar with bit shifting and bit algebra in expressions like a |= (1<<5); you
can take a look at A.1 to learn the basics.
4. First steps
This chapter elaborates about setting an output, reading an input and programming simple
time depending tasks. If you are not familiar with the bit expressions used in this section feel
free to take a look at A.1.
4.1. Controlling a LED

4.1.1. Plugging in the LED
In the lab we provide simple LEDs with integrated resistors soldered on a connector to be
directly plugged onto the board.
1
An example is the reset pin which can also be used for normal IO. This tutorial keeps on the usage that is
recommended by an unchanged board with unchanged software.
2
Technische Informatik, typically Informatik II
5
3 2 1
+ -
(a) Hold LED against (b)
light to see the polar-
ity
Figure 1: LED and board
To find out the polarity of the LED, it can simply he held against the light, see 1(a). The
smaller piece inside the plastic housing is the anode (+). It will be plugged on the board on
the pin next to the microcontroller. The controller will provide its signal there.
The other pin, the cathode (-), is plugged to the pin which is the most far away from the
controller. This pin is on ground. If you do not have a board, take a look at 1(b).
The LED will be placed on PORTC5, which is located in the corner of the MC at pin no 28.
Take a look at A.2.
4.1.2. Switching it on
In the previous paragraph the LED was placed in PORTC5. In the file main.c where the main-
loop is located the following changes need to be made:
1. Set pin 5 in PORTC to output mode with the Data Direction Register for the port DDRC.
2. Switch pin 5 in PORTC on
3. Stop the controller

A program on a microcontroller should never leave its main loop. This normally causes
the controller to start from the beginning of the programs. But not all hardware is reset
resulting in an undefined state in not properly initialized hardware. To prevent this, normally
the controller is kept in an infinite loop at the end of a program.
...
int main(void) {
//Set the Data Direction Register to output
DDRC |= (1<<5);
//Set the signal to high
PORTC |= (1<<5);
6
//Stop in an infinite loop

while (1);
}
...
You can now type make to see if the program compiles.
After connecting the board to power and plugging in the programmer, the binary can be
transferred with the command make prog.
After the transfer the LED should now emit light. If it does not, check for possible errors during
the transfer on the screen. Or maybe the LED is placed with the wrong polarity. Simply turn
it around for a try, wrong polarity does not damage the LED.
4.1.3. Making the LED blink

Now that we know how to switch on a LED, we also want to switch it off again after a certain
time. We do this by the most simple way: a delay function.
The AVR library provides a very accurate delay function. Make sure your file has the line
#include <util/delay.h> to include the delay functions.
Now the program looks like this:
...
#include <util/delay.h>
...
int main(void) {
DDRC |= (1<<5);
while (1) {
//Set the signal to high
PORTC |= (1<<5);
//wait 0.5 sec
_delay_ms(500);
//Set the signal to low
PORTC &= ~(1<<5);
//wait 0.5 sec
_delay_ms(500);
}
}
...
After compiling and transferring the program with make prog the led should blink in a fre-
quency of 1 sec.
If it is blinking faster or slower then you maybe have forgotten to fuse the controller. Try make
fuse and see if it changes.
This program does what it is supposed to, but it can be written shorter and with a reduction
of redundant code. First the bit is set and later cleared, this can also be done with the toggle
function which is the equivalent of the XOR operator (see A.1):
...
#include <util/delay.h>
...
7
int main(void) {
DDRC |= (1<<5);
while (1) {
//Toggle the signal
PORTC ^= (1<<5);
//wait 0.5 sec
_delay_ms(500);
}
}
...
4.1.4. Making the LED blink (advanced)

The previous program is nice but the delay function blocks the whole main loop from working
on other tasks. It seems wise to find a method of doing actions in a certain time but without
blocking the controller. Such situations are normally resolved using a state machine.
In the basis-code a function called getMsTimer() is provided. It returns the time in millisec-
1
onds ( 1000 s) since resetting the controller. Beware that the return value of this function is a
uint32 t, when storing it. The variable ms timer is deprecated and cannot be used in newer
revisions.
...
//Time between switching the led on or off
#define delay 500
int main(void) {
init();
DDRC |= (1<<5);
uint32_t next_ms_timer=getMsTimer()+delay;
while (1) {
//Is it time to do smth.?
if (getMsTimer()>next_ms_timer) {
//Toggle the signal
PORTC ^= (1<<5);
//don’t forget to set the next "alarm"
next_ms_timer+=delay;
}
//plenty of free time to do other stuff
}
}
...
4.2. Reading an input

If the bit in the DDR (Data Direction Register) for the corresponding pin is 0, it is configured
as input. A voltage provided at that pin is interpreted as either 0 or 1 and stored in the
PIN-Register (e.g. PIN).
8
Example
In the following example a switch is placed on pin C3. While the switch is pushed, the LED
(still placed at C5) is turned off. In both examples the so called internal pull-up resistor is
switched on. Please refer to section 4.2.2 for its meaning.
...
int main(void) {
init();
//Set the Data Direction Register for the LED to output
DDRC |= (1<<5);
//Set the Data Direction Register for the switch to input
DDRC &= ~(1<<3);
//Enable the pullup resistor for the switch
PORTC|=(1<<3);
while (1) {
if (PINC & (1<<3))
PORTC|=(1<<5);
else
PORTC&=~(1<<5);
}
}
...
4.2.1. Detecting an edge

Most of the time it is requiered to detect a change of a pin. This is done by storing the old
value and comparing it to the current one.
...
int main(void) {
init();
//Set the Data Direction Register for the LED to output
DDRC |= (1<<5);
//Set the Data Direction Register for the switch to input
DDRC &= ~(1<<3);
//Enable the pullup resistor for the switch
PORTC|=(1<<3);
//save old state
uint8_t oldc=PINC;
while (1) {
if ((oldc & (1<<3)) != (PINC & (1<<3)) ) {
//Switch LED
PORTC^=(1<<5);
//save new state
oldc=PINC;
}
}
}
...
9
VCC
VCC
100K
R1
JP1 1 JP2 JP3
1 2 1 1
3
TL36YO
TL36YO
S1
3
TL36YO
S2
S3
2
2
GND GND GND
Figure 2: Differently connected switches: (a) between 5V and GND, (b) between pin and GND
with an external pull-up resistor, (c) as (b) but with an internal pull-up
4.2.2. Providing a clean signal

Some words on a well defined signal for the input should be said. The input stage of the
controller has an intrinsic capacitor that keeps its electric charge for a certain time. If you
connect a switch, it should always switch the pin between 5V and GND (0V) to allow it to get
properly (dis-)charged. This is shown on 2 (a).
If a switch can not change between two connectors but just opens and closes a single line, a
so called pull-up resistor must be used. This can be an external resistor with a value between
100KΩ and 1M Ω (2 (b)).
The most elegant way is to use the internal pull-up already installed inside the Atmega. To
activate them the PORT is set to 1, while the DDR remains set to 0.
...
DDRC &= ~(1<<3);
PORTC = (1<<3);
if (PINC & (1<<3)) {
//do smth.
}
...
Be aware of the fact that by using a pull-up the logic of the switch is inverted!
10
5. Timer/Counter
This is the end of the basic tutorial section of this document. The following sections describe
the usage of functions provided in the basis-code.
5.1. Hardware timer

The Atmega controllers provide hardware counters. Those
counters are registers that are incremented normally by a sig-
nal from the oscillator which also drives the Atmega. The
oscillator is not connected directly but via a variable prescaler
to run slower.
To every counter there is at least one compare reg-
ister. When the counter value equals the value of
the compare register, a certain action can be trig-
gered. This can be a reset of the timer, throw-
ing an interrupt (like in 5.3) or toggling a pin as
used in PWM or for servo control (5.2). Actions
can also be triggered when the timer value reaches
an overflow. Please take a look at the data sheet
[1] for more information, this is just a brief introduc-
tion.
Example calculation
As an example we calculate the frequency in which Timer1
overflows. Timer1 is a 16-bit timer.
foscillator = 16M Hz = 16 · 106 Hz = 16000000Hz

prescaler = 8
foscillator 16000000
foverf low = = Hz ≈ 30.5Hz
prescaler · 216 8 · 65536
1
toverf low = ≈ 0.0327s ≈ 33ms
foverf low
As we see the timer would overflow about every 33ms that
equals a frequency of about 31Hz.
Figure 3: Interior of a
servo. Picture 5.2. Servo control
from wikipedia[2]
5.2.1. About servos
Servos are motors with an integrated controller unit. They normally have a range of motion
of about 180◦ , most servos also a little more. A signal is applied to tell the servo the desired
position and the controller inside then tries to actuate the motor to reach it. For that reason
servos are very handy because of their easy usage and very compact design.
11
Unfortunately there is no way of asking the servo about its position. If for example the
counter momentum is too large and the motor is too weak to reach the desired position, there
is no way of knowing it on the microcontroller. So there is no guarantee that a servo actually
reaches a position.
A servo normally has three connectors. One for
Ground, the middle for supply voltage (typically 2
≈ 4.8V . . . 6V ), and the last for the signal. The

signal is a high-pulse of 1 . . . 2ms every 20ms. The
duration of the signal is proportional to the angle
of the position.
1
5.2.2. Servos in basis

The basis contains simple functions to provide this
signal for two servos. Those servos should be con-
nected to PORTB1 and PORTB2 where is output pins 0
for Timer1 are located. They are called OC1A and 0 2 4 6 8 10 12 14 16 18 20 22
OC1B (see A.2).

After initializing the timer with function Figure 4: Signal form for a servo: 1 . . . 2ms
servoInit() the function setServo(uint8 t nr, high every 20ms
uint16 t us) is the only interface needed. nr is
the servo number, i.e. 0 or 1. The second param-
1
eter is the signal length in 1µs = 1000 ms. The
range is therefore 1000 . . . 2000.
Some servos allow a wider range which can be used safely. For safety reasons they are not
allowed per default in the basis. If you want to use them you need to make changes in the
servo.h file.
But be careful not to stall the servo over a longer period of time! It can overheat and destroy
the electronic inside!
Example
This example sets the first servo to one boundary position.
...
int main(void) {
servoInit();
setServo(0, 1000);
while (1) { }
}
...
5.3. Timer for software interrupt

The hardware Timer2 is used to provide a software interrupt at 1kHz. Every time the interrupt
is triggered, the global variable ms timer is incremented.
Therefor in every part of the code the variable has the actual time since in ms since the last
reset of the controller.
12
To get the actual time the function uint32 t getMsTimer() must be used. Be aware of the
fact that the return value is a uint32 t. Storing it in a container of smaller size will probably
not work. Even if a 32-bit variable seems quite large it will overflow every 49.7 days. For
long-time applications another way of storing the time must be developed.
6. Analogue-Digital-Converter (ADC)
An input pin in default mode is only able to distinguish between a high and a low signal. Some
pins can be configured as ADCs. They can be read in a resolution of up to 10 bit (210 = 1024).
The controller has only one ADC but uses a multiplexer or mux to switch between the pins.
On an Atmega168 those are C0 to C5.
6.1. ADC in basis

There are numerous ways to configure the ADC. For the basis a compromise between speed and
accuracy was chosen that should fit to most purposes. The ADC is initialized with ADCInit().
A value can be read with getADCValue(uint8 t channel), where channel specifies the pin
to read from.
example
...
int main(void) {
ADCInit(0);
uint16_t value;
value = getADCValue(2);
while (1) { }
}
...
7. Serial connection(UART)
The Atmega controller has a so called Universal Asynchronous Receiver Transmitter (UART)
which can be used to communicate with any other RS-232 device like a PC. The only hardware
required is a signal converter that connects the 0 . . . 5V signal of the microcontroller to the
−12 . . . 12V of the PC. We provide such converters in the lab.
When programming uart one should always be aware of two pitfalls:
1. Buffersize: The hardware buffer for input and output on the Controller only hold 1 byte.
That means
a) If the program does not read fast enough, it loses bytes.
b) It is only possible to send one byte at a time.
To prevent data loss and blocking calls for sending it is possible to use interrupts. Those
functions are currently not implemented in the basis.
13
2. Speed: The baud (bandwidth) has to be generated by the controller’s oscillator. An

internal prescaler is applied to generate the baud. That limits the baud usable with a
16MHz oscillator to 38.400kbit/s.
7.1. UART in basis

First the baud has to be specified in uart.h. After initializing with uartInit() the following
functions can be used:
• uart putc(unsigned char c) sends the char stored in c. It blocks if the send buffer is
not empty.
• uart puts(char *s) sends a string (terminated with ’\0’) stored in s.
• uart puts pgm (const prog char * str) sends a string like uart puts(). The only
difference is, that this string is stored in flash memory and therefor does not waste
valueable SRAM.
• uart puti (int16 t i) sends an integer. Be careful that cannot be longer than 16 bit.
• unsigned char uart getc() receives a char. It blocks until one is in the in buffer.
• uart data waiting() determines if the input buffer stores a valid byte. This can be used
as a test before calling uart getc() because uart data waiting() is non-blocking.
7.1.1. Example
...
int main(void) {
uartInit();
//This string wastes RAM, because it’s static and could be stored in flash
uart_puts("Hello World\n\r");
//This string is stored in program memory
uart_puts_pgm(PSTR("Hello World!\r\n");
if (uart_data_waiting())
uart_puts_pgm(PSTR("There is already data in the input buffer!\r\n");
while (1) {
//get a character (blocking)
uint8_t c = uart_getc();
//send it back, aka echo
uart_putc(c);
}
}
...
7.2. The terminal

On the side of the PC a terminal program is needed to directly communicate with the controller.
There is for example minicom which is old and text based or gtkterm.
Settings:
14
• Baud as you selected in uart.h, i.e. 1200, 9600, 38400 . . .
• 8 bit, no parity, 1 stop bit, a.k.a. 8N1
• No hardware flow control (RTS/CTS)
• No software flow control (Xon/Xoff)
A. Appendix
A.1. Algebra with bits
This section shortly introduces how to set, unset and toggle3 a single or multiple bits in a 8bit
register or variable in your program memory.
There are of course more things that can be done in bit-manipulation. If you are interested,
feel free to search the web for it. The following part is focusing on the basics.
There are multiple ways to define an unsigned integer value in C:

// decimal
a = 123;
// hexadecimal
b = 0x7b;
// octal
c = 0173;
But there is no binary representation in standard C.

It becomes necessary to find a way to circumvent this problem. One solution is bit-shifting.
A.1.1. Shifting
In C there are the shifting operators << and >>.
They must not be confused with the stream operators in C++!
Therefore it is strongly recommended to use brackets around them.
An expression like (a << b) will be evaluated as the binary representation of a will be shifted
b times to the left, the value is getting larger. (a >> b) shifts the to the right and therefor
decreases the value. Empty spaces are filled with 0.
The example demonstrates:
C code decimal value bit representation

1 1 00000001
3 3 00000011
4 4 00000100
(1 << 2) 4 00000100
(1 << 3) 8 00001000
(3 << 2) 12 00001100
(4 >> 1) 2 00000010
(3 >> 1) 1 00000001
(4 >> 4) 0 00000000
3
change a value to its inverse (1→0,0→1)
15
The obvious way to define bit no. n is using the expression. (1 << n).
A.1.2. Bit-wise logic

Assuming, the logic operators a well known. But for the sake of completeness here is another
overview:
A B AND OR XOR NOT(A)

0 0 0 0 0 1
0 1 0 1 1 1
1 0 0 1 1 0
1 1 1 1 0 0
The bit-wise logic needs to be distinguished from the Boolean logic.

In Boolean Logic the truth value of a whole expression is evaluated. If an expression equals 0
it is false, in any other case it is true. The representation are double symbols.
In bit-wise logic the those are single symbols.
Boolean bit-wise
AND && &
OR || |
XOR != ^
NOT ! ~
A.1.3. Setting a bit

Finally with bit-wise logic and bit shifting it is possible to set a certain bit while leaving the
other bits untouched.
For example bit no. 5 in variable a should be set to have the value 1:
// elaborately
a = a | (1 << 5);
// ...or shorter
a |= (1 << 5);
The OR-operator is used because x OR 1 = 1 and x OR 0 = x where x can be 0 or 1.

Therefor the previous value of this bit is irrelevant and is now set to 1.
For example:
100100102
| 001000002
= 101100102
A.1.4. Clearing a bit

Clearing a bit back to 0 works similar to setting it to 1. The only difference is that two
operators are needed, namely AND and NOT. The fact that x AND 0 = 0 and x AND 1 = x
is used. To achieve a 0 at the specific bit position and 1s at the others the NOT operator is
very helpful.
In this example bit no. 5 in variable a should be cleared.
16
// elaborately
ption a = a & ~(1 << 5);
// ...or shorter
40µA a &= ~(1 << 5);
5µA (including
ForOscillator)
example:
de: 101100102
& ~ 001000002
101100102
& 110111112
100100102
urations =
A.2. Pin out of the Atmega 168

ATmega48/88/168
PDIP
(PCINT14/RESET) PC6 1 28 PC5 (ADC5/SCL/PCINT13)

(PCINT16/RXD) PD0 2 27 PC4 (ADC4/SDA/PCINT12)
(PCINT17/TXD) PD1 3 26 PC3 (ADC3/PCINT11)
(PCINT18/INT0) PD2 4 25 PC2 (ADC2/PCINT10)
(PCINT19/OC2B/INT1) PD3 5 24 PC1 (ADC1/PCINT9)
(PCINT20/XCK/T0) PD4 6 23 PC0 (ADC0/PCINT8)
VCC 7 22 GND
GND 8 21 AREF
(PCINT6/XTAL1/TOSC1) PB6 9 20 AVCC
(PCINT7/XTAL2/TOSC2) PB7 10 19 PB5 (SCK/PCINT5)
(PCINT21/OC0B/T1) PD5 11 18 PB4 (MISO/PCINT4)
(PCINT22/OC0A/AIN0) PD6 12 17 PB3 (MOSI/OC2A/PCINT3)
(PCINT23/AIN1) PD7 13 16 PB2 (SS/OC1B/PCINT2)
(PCINT0/CLKO/ICP1) PB0 14 15 PB1 (OC1A/PCINT1)
TQFP TopFigure
View 5: Pin out of the controller excerpt from the data sheet[1].
MLFFor
Top View
orientation use the
PC4 (ADC4/SDA/PCINT12)
PC5 (ADC5/SCL/PCINT13)
PC4 (ADC4/SDA/PCINT12)
marking on top
PC5 (ADC5/SCL/PCINT13)
PC6 (RESET/PCINT14)
PC6 (RESET/PCINT14)
PC3 (ADC3/PCINT11)
PC2 (ADC2/PCINT10)
PC3 (ADC3/PCINT11)
PC2 (ADC2/PCINT10)
PD2 (INT0/PCINT18)
PD0 (RXD/PCINT16)
PD1 (TXD/PCINT17)
PD2 (INT0/PCINT18)
PD0 (RXD/PCINT16)
PD1 (TXD/PCINT17)
32
31
30
29
28
27
26
25
32
31
30
29
28
27
26
25
1 24 PC1 (ADC1/PCINT9) (PCINT19/OC2B/INT1) PD3 1 24 PC1 (ADC1/PCINT9)

2 23 PC0 (ADC0/PCINT8) (PCINT20/XCK/T0) PD4 2 23 PC0 (ADC0/PCINT8)
3 22 ADC7 GND 3 22 ADC7
4 21 GND VCC 4 21 GND
5 20 AREF GND 5 20 AREF
6 19 ADC6 VCC 6 19 ADC6
7 18 AVCC (PCINT6/XTAL1/TOSC1) PB6 7 18 AVCC
8 17 PB5 (SCK/PCINT5) (PCINT7/XTAL2/TOSC2) PB7 8 17 PB5 (SCK/PCINT5)
17
10
11
12
13
14
15
16
10
11
12
13
14
15
16
9
9
T21/OC0B/T1) PD5
22/OC0A/AIN0) PD6
PCINT23/AIN1) PD7
T0/CLKO/ICP1) PB0
PCINT1/OC1A) PB1
INT2/SS/OC1B) PB2
3/OC2A/MOSI) PB3
(PCINT4/MISO) PB4
T21/OC0B/T1) PD5
2/OC0A/AIN0) PD6
CINT23/AIN1) PD7
0/CLKO/ICP1) PB0
PCINT1/OC1A) PB1
NT2/SS/OC1B) PB2
3/OC2A/MOSI) PB3
PCINT4/MISO) PB4
NOTE: Bottom pad should be soldered to ground.

B. References
References
[1] Atmel Corporation. Datasheet to atmega48/88/168, 2005.
[2] Wikipedia. Servo — wikipedia, die freie enzyklopädie, 2011. [Online; Stand 23. November
2011].
Version
This document was created from git commit 2874ba28f297718d16661fa9a4aee6fd6ce3efb0
18

Atmega Tutorial PDF

Uploaded by

Atmega Tutorial PDF

Uploaded by

An Introduction to programming an Atmega

November 14, 2013

2. Compilation and transfer 3

2. Compilation and transfer

or download it from http://roboter.uni-hd.de. After downloading it, you will need to

2.1.2. First steps

to compile and get an output similar to this

Compiling main.c ...

3.1. Pins and Ports

4.1. Controlling a LED

Figure 1: LED and board

2. Switch pin 5 in PORTC on

3. Stop the controller

//Stop in an infinite loop

4.1.3. Making the LED blink

4.1.4. Making the LED blink (advanced)

4.2. Reading an input

4.2.1. Detecting an edge

4.2.2. Providing a clean signal

5.1. Hardware timer

foscillator = 16M Hz = 16 · 106 Hz = 16000000Hz

≈ 4.8V . . . 6V ), and the last for the signal. The

5.2.2. Servos in basis

for Timer1 are located. They are called OC1A and 0 2 4 6 8 10 12 14 16 18 20 22

OC1B (see A.2).

5.3. Timer for software interrupt

6.1. ADC in basis

2. Speed: The baud (bandwidth) has to be generated by the controller’s oscillator. An

7.1. UART in basis

• uart puts(char *s) sends a string (terminated with ’\0’) stored in s.

7.2. The terminal

• Baud as you selected in uart.h, i.e. 1200, 9600, 38400 . . .

• 8 bit, no parity, 1 stop bit, a.k.a. 8N1

• No hardware flow control (RTS/CTS)

• No software flow control (Xon/Xoff)

There are multiple ways to define an unsigned integer value in C:

But there is no binary representation in standard C.

C code decimal value bit representation

A.1.2. Bit-wise logic

A B AND OR XOR NOT(A)

The bit-wise logic needs to be distinguished from the Boolean logic.

A.1.3. Setting a bit

The OR-operator is used because x OR 1 = 1 and x OR 0 = x where x can be 0 or 1.

A.1.4. Clearing a bit

A.2. Pin out of the Atmega 168

(PCINT14/RESET) PC6 1 28 PC5 (ADC5/SCL/PCINT13)

1 24 PC1 (ADC1/PCINT9) (PCINT19/OC2B/INT1) PD3 1 24 PC1 (ADC1/PCINT9)

NOTE: Bottom pad should be soldered to ground.

You might also like