Note: this page refers to a product that is retired.

Arduino 101 CurieIMU Orientation Visualiser

This tutorial demonstrates how to make use the 6-axis IMU to read the X, Y, and Z values of both the accelerometer and the gyroscope.

This tutorial demonstrates how to make use the Arduino 101's onboard 6-axis accelerometer/gyro to read the X, Y, and Z values of both the accelerometer and the gyroscope. While the gyroscope is able to determine the orientation of the board, the accelerometer measures the angular velocity of the board. Together, the accelerometer and the gyroscope form an Inertial Monitoring Unit (IMU) which can be used to precisely identify the orientation of the board. Madgwick's filter algorithm is used in this example to calculate quarternions from the 6 axes' values. The quarternions are then used to calculate Euler angles Pitch, Yaw, and Roll, which are received by Processing and used to control the rotation of an object around the X, Y and Z axes.

Hardware Required

The CurieIMU library uses the IMU (accelerometer + gyroscope) built into the Arduino 101.

Instructions

  1. Set up the Arduino software (IDE) as described in Getting Started with Arduino 101.

  2. Connect the 101 to your computer.

  3. Launch the Arduino software (IDE) and select Arduino 101 from the Tools > Board menu.

  4. Install the Madgwick library from library manager. To do this, open the Arduino Software (IDE), go to Sketch > Include Library > Manage Libraries. There you can search 'Madgwick' and install the library directly from there.

  5. Download and Launch the Processing software and create a file with the Processing code shown below.

  6. Change the Serial port to the one that your 101 is using (see "Processing Sketch" section).

  7. Upload the example contained in the Madgwik library called Visualizer101 to your 101, making sure that the board is flat and stationery so it can perform the calibration accurately.

  8. After a few seconds, run the Processing sketch, adjust the orientation of your board, and watch as the Processing sketch gives a visualization of your board. The Processing is contained in the "extras" folder of the Madgwick library.

The Circuit

genuino101fzz

image developed using Fritzing.

How it works

The Madgwick filter algorithm is open-source and is well documented in Madgwick's information and reports. The Madgwick filter algorithm was developed by Sebastian Madgwick during his Ph.D. in 2010 and is designed to be computationally inexpensive and efficient even at low sampling rates. The algorithm takes raw values from a gyroscope and accelerometer, and uses them to return four quaternions, which are 4-dimensional numbers which contain x, y, and z values to represent the axis around which rotation occurs, as well as a ω value which represents the value of rotation which occurs around the same axis. These quaternions can be used to calculate the Euler angles pitch, yaw, and roll; three angles used to describe the orientation of a rigid body in terms of x,y, and z as presented by Leonhard Euler in the 1700s. The equations (7) (8) (9) in Madgwick's Report are used to calculate the values for pitch, roll, and yaw, and their functions are included within the library.

We can create a 3D representation of the Arduino 101's onboard IMU in Processing, which will move as the board does. This is achieved with the values for Euler angles pitch, roll and yaw obtained by the Madgwick filter algorithm. These values can then be sent via Serial to Processing and used as angle arguments for Processing's to compute the position of the Arduino 3D model using the methods: applyMatrix(), pushMatrix(), and popMatrix() functions.

Arduino Sketch

The sketch uses functions inside the CurieIMU library to get the data from the accelerometer/gyro.

In order to see a 3D representation in Processing, the Arduino sketch must incorporate two main functionalities; using the IMU data and algorithm to calculate yaw, pitch, and roll values, and enabling serial communication in a handshake fashion in order to send those values to Processing.

First, we must create a Madgwick object to access the functions from the Madgwick class in the library. Here, we call it filter:

1Madgwick filter;

In the

setup()
function we perform a preliminary configuration of the CurieIMU, by setting the sample rate of the accelerometer and the gyro and the filter to 25Hz:

1CurieIMU.begin();
2
3CurieIMU.setGyroRate(25);
4
5CurieIMU.setAccelerometerRate(25);
6
7filter.begin(25);

Then we set the accelerometer range to 2g and the gyro range to 250 °/s:

1CurieIMU.setAccelerometerRange(2);
2
3CurieIMU.setGyroRange(250);

In the

loop()
function we will need to send a sample according to the sample rate we set for the CurieIMU reading:

1microsPerReading = 1000000 / 25;

We can then 'get' accelerometer and gyroscope data using the following functions from CurieIMU library and convert the raw data to acceleration (g) and angular velocity (°/s):

1CurieIMU.readMotionSensor(aix, aiy, aiz, gix, giy, giz);
2
3// convert from raw data to gravity and degrees/second units
4
5ax = convertRawAcceleration(aix);
6
7ay = convertRawAcceleration(aiy);
8
9az = convertRawAcceleration(aiz);
10
11gx = convertRawGyro(gix);
12
13gy = convertRawGyro(giy);
14
15gz = convertRawGyro(giz);

We can then use the function updateIMU() from the Madgwick library.

1filter.updateIMU(gx, gy, gz, ax, ay, az);

After that, we are ready to obtain from the filter the roll, pith and yaw values:

1roll = filter.getRoll();
2
3pitch = filter.getPitch();
4
5heading = filter.getYaw();

These values are sent 25 times per second over the serial port to the Processing application.

As seen in the code, the gyroscope values have been scaled down by a variable factor so that they fit into a range which works well with the algorithm. Without this scaling, the values which are inputted to the function are too high and the visualization of the movement of the board becomes very sensitive to small changes of the 101's position, interpreting a slight change as a great change and causing the 'virtual' board to spin.

The full code can be found at the bottom of the page.

Note that the serial prints for gx,gy,gz,ax,az,ay are left in loop in comments for debugging and must be commented whilst communicating with Processing.

Processing Sketch

If you haven't already, the first thing to do is to download the latest version of Processing from processing.org. Processing is a language similar to Arduino which allows the user to draw dynamic imagery in the familiar

void setup()
and
void loop()
structure. For more information on using Processing, please visit their Getting Started guide.

The processing code receives incoming data from the serial port which is parsed and assigned to floats

yaw
,
pitch
, and
roll
, which are then used to compute the transformation matrix which moves the 3D model of the Arduino board.

To enable Processing to read from the same port that Arduino is sending to, myPort needs to be changed to your serial port's name. In setup(), this is the Second parameter of Serial.

1myPort = new Serial(this, Serial.list()[0], 9600);

The correct port can be found by using the list() function from the Serial class. The number inside the square brackets refers to the number of the serial port and will be 0, 1, 2, etc. The sketch works if you have just one COM port active on your machine. As an alternative, you could specify directly the "COMx" port of the 101 board - the one used to program from Arduino Software (IDE) the 101 board - commenting this line

1myPort = new Serial(this, Serial.list()[0], 9600);

and uncomment of the lines corresponding to your OS:

1myPort = new Serial(this, "COM5:", 9600); // Windows
2
3myPort = new Serial(this, "/dev/ttyACM0", 9600); // Linux
4
5myPort = new Serial(this, "/dev/cu.usbmodem1217321", 9600); // Mac

You must replace the port string with the correct name for your COM port.

If in doubt, you can print a list of your available serial ports in a separate sketch to determine this name.

The function serialEvent() is then used to receive and parse data.

1void serialEvent()
2{
3
4 int newLine = 13; // new line character in ASCII
5
6 String message;
7
8 do {
9
10 message = myPort.readStringUntil(newLine); // read from port until new line
11
12 if (message != null) {
13
14 String[] list = split(trim(message), " ");
15
16 if (list.length >= 4 && list[0].equals("Orientation:")) {
17
18 yaw = float(list[1]); // convert to float yaw
19
20 pitch = float(list[2]); // convert to float pitch
21
22 roll = float(list[3]); // convert to float roll
23
24 }
25
26 }
27
28 } while (message != null);
29}

This reads from the serial port until ASCII character 13 (new line) and then uses the split() function to separate the values using the comma character. Since we know that we sent from Arduino in the order yaw, pitch, roll, we can then convert each string to a float and assign them to the first three values in String array list[]. The strings are then converted into floats and stored in float variables. The full Arduino and Processing sketches can be seen below.

Code

Arduino Code

1#include <CurieIMU.h>
2#include <MadgwickAHRS.h>
3
4Madgwick filter;
5unsigned long microsPerReading, microsPrevious;
6float accelScale, gyroScale;
7
8void setup() {
9
10 Serial.begin(9600);
11
12 // start the IMU and filter
13
14 CurieIMU.begin();
15
16 CurieIMU.setGyroRate(25);
17
18 CurieIMU.setAccelerometerRate(25);
19
20 filter.begin(25);
21
22 // Set the accelerometer range to 2G
23
24 CurieIMU.setAccelerometerRange(2);
25
26 // Set the gyroscope range to 250 degrees/second
27
28 CurieIMU.setGyroRange(250);
29
30 // initialize variables to pace updates to correct rate
31
32 microsPerReading = 1000000 / 25;
33
34 microsPrevious = micros();
35}
36
37void loop() {
38
39 int aix, aiy, aiz;
40
41 int gix, giy, giz;
42
43 float ax, ay, az;
44
45 float gx, gy, gz;
46
47 float roll, pitch, heading;
48
49 unsigned long microsNow;
50
51 // check if it's time to read data and update the filter
52
53 microsNow = micros();
54
55 if (microsNow - microsPrevious >= microsPerReading) {
56
57 // read raw data from CurieIMU
58
59 CurieIMU.readMotionSensor(aix, aiy, aiz, gix, giy, giz);
60
61 // convert from raw data to gravity and degrees/second units
62
63 ax = convertRawAcceleration(aix);
64
65 ay = convertRawAcceleration(aiy);
66
67 az = convertRawAcceleration(aiz);
68
69 gx = convertRawGyro(gix);
70
71 gy = convertRawGyro(giy);
72
73 gz = convertRawGyro(giz);
74
75 // update the filter, which computes orientation
76
77 filter.updateIMU(gx, gy, gz, ax, ay, az);
78
79 // print the heading, pitch and roll
80
81 roll = filter.getRoll();
82
83 pitch = filter.getPitch();
84
85 heading = filter.getYaw();
86
87 Serial.print("Orientation: ");
88
89 Serial.print(heading);
90
91 Serial.print(" ");
92
93 Serial.print(pitch);
94
95 Serial.print(" ");
96
97 Serial.println(roll);
98
99 // increment previous time, so we keep proper pace
100
101 microsPrevious = microsPrevious + microsPerReading;
102
103 }
104}
105
106float convertRawAcceleration(int aRaw) {
107
108 // since we are using 2G range
109
110 // -2g maps to a raw value of -32768
111
112 // +2g maps to a raw value of 32767
113
114
115
116 float a = (aRaw * 2.0) / 32768.0;
117
118 return a;
119}
120
121float convertRawGyro(int gRaw) {
122
123 // since we are using 250 degrees/seconds range
124
125 // -250 maps to a raw value of -32768
126
127 // +250 maps to a raw value of 32767
128
129
130
131 float g = (gRaw * 250.0) / 32768.0;
132
133 return g;
134}

Processing Code

1import processing.serial.*;
2Serial myPort;
3
4float yaw = 0.0;
5float pitch = 0.0;
6float roll = 0.0;
7
8void setup()
9{
10
11 size(600, 500, P3D);
12
13 // if you have only ONE serial port active
14
15 myPort = new Serial(this, Serial.list()[0], 9600); // if you have only ONE serial port active
16
17 // if you know the serial port name
18
19 //myPort = new Serial(this, "COM5:", 9600); // Windows
20
21 //myPort = new Serial(this, "/dev/ttyACM0", 9600); // Linux
22
23 //myPort = new Serial(this, "/dev/cu.usbmodem1217321", 9600); // Mac
24
25 textSize(16); // set text size
26
27 textMode(SHAPE); // set text mode to shape
28}
29
30void draw()
31{
32
33 serialEvent(); // read and parse incoming serial message
34
35 background(255); // set background to white
36
37 lights();
38
39 translate(width/2, height/2); // set position to centre
40
41 pushMatrix(); // begin object
42
43 float c1 = cos(radians(roll));
44
45 float s1 = sin(radians(roll));
46
47 float c2 = cos(radians(pitch));
48
49 float s2 = sin(radians(pitch));
50
51 float c3 = cos(radians(yaw));
52
53 float s3 = sin(radians(yaw));
54
55 applyMatrix( c2*c3, s1*s3+c1*c3*s2, c3*s1*s2-c1*s3, 0,
56
57 -s2, c1*c2, c2*s1, 0,
58
59 c2*s3, c1*s2*s3-c3*s1, c1*c3+s1*s2*s3, 0,
60
61 0, 0, 0, 1);
62
63 drawArduino();
64
65 popMatrix(); // end of object
66
67 // Print values to console
68
69 print(roll);
70
71 print("\t");
72
73 print(pitch);
74
75 print("\t");
76
77 print(yaw);
78
79 println();
80}
81
82void serialEvent()
83{
84
85 int newLine = 13; // new line character in ASCII
86
87 String message;
88
89 do {
90
91 message = myPort.readStringUntil(newLine); // read from port until new line
92
93 if (message != null) {
94
95 String[] list = split(trim(message), " ");
96
97 if (list.length >= 4 && list[0].equals("Orientation:")) {
98
99 yaw = float(list[1]); // convert to float yaw
100
101 pitch = float(list[2]); // convert to float pitch
102
103 roll = float(list[3]); // convert to float roll
104
105 }
106
107 }
108
109 } while (message != null);
110}
111
112void drawArduino()
113{
114
115 /* function contains shape(s) that are rotated with the IMU */
116
117 stroke(0, 90, 90); // set outline colour to darker teal
118
119 fill(0, 130, 130); // set fill colour to lighter teal
120
121 box(300, 10, 200); // draw Arduino board base shape
122
123 stroke(0); // set outline colour to black
124
125 fill(80); // set fill colour to dark grey
126
127 translate(60, -10, 90); // set position to edge of Arduino box
128
129 box(170, 20, 10); // draw pin header as box
130
131 translate(-20, 0, -180); // set position to other edge of Arduino box
132
133 box(210, 20, 10); // draw other pin header as box
134}

Suggest changes

The content on docs.arduino.cc is facilitated through a public GitHub repository. If you see anything wrong, you can edit this page here.

License

The Arduino documentation is licensed under the Creative Commons Attribution-Share Alike 4.0 license.