Python - IMU Brick

This is the description of the Python API bindings for the IMU Brick. General information and technical specifications for the IMU Brick are summarized in its hardware description.

An installation guide for the Python API bindings is part of their general description.

Examples

The example code below is Public Domain (CC0 1.0).

Simple

Download (example_simple.py)

 1
 2
 3
 4
 5
 6
 7
 8
 9
10
11
12
13
14
15
16
17
18
19
20
21
22
23
24
#!/usr/bin/env python
# -*- coding: utf-8 -*-  

HOST = "localhost"
PORT = 4223
UID = "6QGRFq" # Change to your UID

from tinkerforge.ip_connection import IPConnection
from tinkerforge.brick_imu import IMU

if __name__ == "__main__":
    ipcon = IPConnection() # Create IP connection
    imu = IMU(UID, ipcon) # Create device object

    ipcon.connect(HOST, PORT) # Connect to brickd
    # Don't use device before ipcon is connected

    # Get current quaternion
    x, y, z, w = imu.get_quaternion()

    print("x: " + str(x) + "\ny: " + str(y) + "\nz: " + str(z) + "\nw: " + str(w))

    raw_input('Press key to exit\n') # Use input() in Python 3
    ipcon.disconnect()

Callback

Download (example_callback.py)

 1
 2
 3
 4
 5
 6
 7
 8
 9
10
11
12
13
14
15
16
17
18
19
20
21
22
23
24
25
26
27
28
29
#!/usr/bin/env python
# -*- coding: utf-8 -*-  

HOST = "localhost"
PORT = 4223
UID = "ayQskyoNrCW" # Change to your UID

from tinkerforge.ip_connection import IPConnection
from tinkerforge.brick_imu import IMU

# Quaternion callback
def cb_quaternion(x, y, z, w):
    print("x: " + str(x) + "\ny: " + str(y) + "\nz: " + str(z) + "\nw: " + str(w) + "\n")

if __name__ == "__main__":
    ipcon = IPConnection() # Create IP connection
    imu = IMU(UID, ipcon) # Create device object

    ipcon.connect(HOST, PORT) # Connect to brickd
    # Don't use device before ipcon is connected

    # Set period for quaternion callback to 1s
    imu.set_quaternion_period(1000)

    # Register quaternion callback
    imu.register_callback(imu.CALLBACK_QUATERNION, cb_quaternion)

    raw_input('Press key to exit\n') # Use input() in Python 3
    ipcon.disconnect()

API

Generally, every method of the Python bindings can throw an tinkerforge.ip_connection.Error exception that has a value and a description property. value can have different values:

  • Error.TIMEOUT = -1
  • Error.ALREADY_CONNECTED = -7
  • Error.NOT_CONNECTED = -8
  • Error.INVALID_PARAMETER = -9
  • Error.NOT_SUPPORTED = -10
  • Error.UNKNOWN_ERROR_CODE = -11

All methods listed below are thread-safe.

Basic Functions

IMU(uid, ipcon)
Parameters:
  • uid -- string
  • ipcon -- IPConnection

Creates an object with the unique device ID uid:

imu = IMU("YOUR_DEVICE_UID", ipcon)

This object can then be used after the IP Connection is connected (see examples above).

IMU.get_orientation()
Return type:(int, int, int)

Returns the current orientation (roll, pitch, yaw) of the IMU Brick as Euler angles in one-hundredth degree. Note that Euler angles always experience a gimbal lock.

We recommend that you use quaternions instead.

The order to sequence in which the orientation values should be applied is roll, yaw, pitch.

If you want to get the orientation periodically, it is recommended to use the callback CALLBACK_ORIENTATION and set the period with set_orientation_period().

The returned namedtuple has the variables roll, pitch and yaw.

IMU.get_quaternion()
Return type:(float, float, float, float)

Returns the current orientation (x, y, z, w) of the IMU as quaternions.

You can go from quaternions to Euler angles with the following formula:

xAngle = atan2(2*y*w - 2*x*z, 1 - 2*y*y - 2*z*z)
yAngle = atan2(2*x*w - 2*y*z, 1 - 2*x*x - 2*z*z)
zAngle =  asin(2*x*y + 2*z*w)

This process is not reversible, because of the gimbal lock.

It is also possible to calculate independent angles. You can calculate yaw, pitch and roll in a right-handed vehicle coordinate system according to DIN70000 with:

yaw   =  atan2(2*x*y + 2*w*z, w*w + x*x - y*y - z*z)
pitch = -asin(2*w*y - 2*x*z)
roll  = -atan2(2*y*z + 2*w*x, -w*w + x*x + y*y - z*z))

Converting the quaternions to an OpenGL transformation matrix is possible with the following formula:

matrix = [[1 - 2*(y*y + z*z),     2*(x*y - w*z),     2*(x*z + w*y), 0],
          [    2*(x*y + w*z), 1 - 2*(x*x + z*z),     2*(y*z - w*x), 0],
          [    2*(x*z - w*y),     2*(y*z + w*x), 1 - 2*(x*x + y*y), 0],
          [                0,                 0,                 0, 1]]

If you want to get the quaternions periodically, it is recommended to use the callback CALLBACK_QUATERNION and set the period with set_quaternion_period().

The returned namedtuple has the variables x, y, z and w.

IMU.leds_on()
Return type:None

Turns the orientation and direction LEDs of the IMU Brick on.

IMU.leds_off()
Return type:None

Turns the orientation and direction LEDs of the IMU Brick off.

IMU.are_leds_on()
Return type:bool

Returns true if the orientation and direction LEDs of the IMU Brick are on, false otherwise.

IMU.set_convergence_speed(speed)
Parameters:speed -- int
Return type:None

Sets the convergence speed of the IMU Brick in °/s. The convergence speed determines how the different sensor measurements are fused.

If the orientation of the IMU Brick is off by 10° and the convergence speed is set to 20°/s, it will take 0.5s until the orientation is corrected. However, if the correct orientation is reached and the convergence speed is too high, the orientation will fluctuate with the fluctuations of the accelerometer and the magnetometer.

If you set the convergence speed to 0, practically only the gyroscope is used to calculate the orientation. This gives very smooth movements, but errors of the gyroscope will not be corrected. If you set the convergence speed to something above 500, practically only the magnetometer and the accelerometer are used to calculate the orientation. In this case the movements are abrupt and the values will fluctuate, but there won't be any errors that accumulate over time.

In an application with high angular velocities, we recommend a high convergence speed, so the errors of the gyroscope can be corrected fast. In applications with only slow movements we recommend a low convergence speed. You can change the convergence speed on the fly. So it is possible (and recommended) to increase the convergence speed before an abrupt movement and decrease it afterwards again.

You might want to play around with the convergence speed in the Brick Viewer to get a feeling for a good value for your application.

The default value is 30.

IMU.get_convergence_speed()
Return type:int

Returns the convergence speed as set by set_convergence_speed().

Advanced Functions

IMU.get_acceleration()
Return type:(int, int, int)

Returns the calibrated acceleration from the accelerometer for the x, y and z axis in mG (G/1000, 1G = 9.80605m/s²).

If you want to get the acceleration periodically, it is recommended to use the callback CALLBACK_ACCELERATION and set the period with set_acceleration_period().

The returned namedtuple has the variables x, y and z.

IMU.get_magnetic_field()
Return type:(int, int, int)

Returns the calibrated magnetic field from the magnetometer for the x, y and z axis in mG (Milligauss or Nanotesla).

If you want to get the magnetic field periodically, it is recommended to use the callback CALLBACK_MAGNETIC_FIELD and set the period with set_magnetic_field_period().

The returned namedtuple has the variables x, y and z.

IMU.get_angular_velocity()
Return type:(int, int, int)

Returns the calibrated angular velocity from the gyroscope for the x, y and z axis in °/14.375s (you have to divide by 14.375 to get the value in °/s).

If you want to get the angular velocity periodically, it is recommended to use the callback CALLBACK_ANGULAR_VELOCITY and set the period with set_angular_velocity_period().

The returned namedtuple has the variables x, y and z.

IMU.get_all_data()
Return type:(int, int, int, int, int, int, int, int, int, int)

Returns the data from get_acceleration(), get_magnetic_field() and get_angular_velocity() as well as the temperature of the IMU Brick.

The temperature is given in °C/100.

If you want to get the data periodically, it is recommended to use the callback CALLBACK_ALL_DATA and set the period with set_all_data_period().

The returned namedtuple has the variables acc_x, acc_y, acc_z, mag_x, mag_y, mag_z, ang_x, ang_y, ang_z and temperature.

IMU.get_imu_temperature()
Return type:int

Returns the temperature of the IMU Brick. The temperature is given in °C/100.

IMU.set_acceleration_range(range)
Parameters:range -- int
Return type:None

Not implemented yet.

IMU.get_acceleration_range()
Return type:int

Not implemented yet.

IMU.set_magnetometer_range(range)
Parameters:range -- int
Return type:None

Not implemented yet.

IMU.get_magnetometer_range()
Return type:int

Not implemented yet.

IMU.set_calibration(typ, data)
Parameters:
  • typ -- int
  • data -- [int, int, int, int, int, int, int, int, int, int]
Return type:

None

There are several different types that can be calibrated:

Type Description Values
0 Accelerometer Gain [mul x, mul y, mul z, div x, div y, div z, 0, 0, 0, 0]
1 Accelerometer Bias [bias x, bias y, bias z, 0, 0, 0, 0, 0, 0, 0]
2 Magnetometer Gain [mul x, mul y, mul z, div x, div y, div z, 0, 0, 0, 0]
3 Magnetometer Bias [bias x, bias y, bias z, 0, 0, 0, 0, 0, 0, 0]
4 Gyroscope Gain [mul x, mul y, mul z, div x, div y, div z, 0, 0, 0, 0]
5 Gyroscope Bias [bias xl, bias yl, bias zl, temp l, bias xh, bias yh, bias zh, temp h, 0, 0]

The calibration via gain and bias is done with the following formula:

new_value = (bias + orig_value) * gain_mul / gain_div

If you really want to write your own calibration software, please keep in mind that you first have to undo the old calibration (set bias to 0 and gain to 1/1) and that you have to average over several thousand values to obtain a usable result in the end.

The gyroscope bias is highly dependent on the temperature, so you have to calibrate the bias two times with different temperatures. The values xl, yl, zl and temp l are the bias for x, y, z and the corresponding temperature for a low temperature. The values xh, yh, zh and temp h are the same for a high temperatures. The temperature difference should be at least 5°C. If you have a temperature where the IMU Brick is mostly used, you should use this temperature for one of the sampling points.

Note

We highly recommend that you use the Brick Viewer to calibrate your IMU Brick.

The following constants are available for this function:

  • IMU.CALIBRATION_TYPE_ACCELEROMETER_GAIN = 0
  • IMU.CALIBRATION_TYPE_ACCELEROMETER_BIAS = 1
  • IMU.CALIBRATION_TYPE_MAGNETOMETER_GAIN = 2
  • IMU.CALIBRATION_TYPE_MAGNETOMETER_BIAS = 3
  • IMU.CALIBRATION_TYPE_GYROSCOPE_GAIN = 4
  • IMU.CALIBRATION_TYPE_GYROSCOPE_BIAS = 5
IMU.get_calibration(typ)
Parameters:typ -- int
Return type:[int, int, int, int, int, int, int, int, int, int]

Returns the calibration for a given type as set by set_calibration().

The following constants are available for this function:

  • IMU.CALIBRATION_TYPE_ACCELEROMETER_GAIN = 0
  • IMU.CALIBRATION_TYPE_ACCELEROMETER_BIAS = 1
  • IMU.CALIBRATION_TYPE_MAGNETOMETER_GAIN = 2
  • IMU.CALIBRATION_TYPE_MAGNETOMETER_BIAS = 3
  • IMU.CALIBRATION_TYPE_GYROSCOPE_GAIN = 4
  • IMU.CALIBRATION_TYPE_GYROSCOPE_BIAS = 5
IMU.orientation_calculation_on()
Return type:None

Turns the orientation calculation of the IMU Brick on.

As default the calculation is on.

New in version 2.0.2 (Firmware).

IMU.orientation_calculation_off()
Return type:None

Turns the orientation calculation of the IMU Brick off.

If the calculation is off, get_orientation() will return the last calculated value until the calculation is turned on again.

The trigonometric functions that are needed to calculate the orientation are very expensive. We recommend to turn the orientation calculation off if the orientation is not needed, to free calculation time for the sensor fusion algorithm.

As default the calculation is on.

New in version 2.0.2 (Firmware).

IMU.is_orientation_calculation_on()
Return type:bool

Returns true if the orientation calculation of the IMU Brick is on, false otherwise.

New in version 2.0.2 (Firmware).

IMU.get_api_version()
Return type:[int, int, int]

Returns the version of the API definition (major, minor, revision) implemented by this API bindings. This is neither the release version of this API bindings nor does it tell you anything about the represented Brick or Bricklet.

IMU.get_response_expected(function_id)
Parameters:function_id -- int
Return type:bool

Returns the response expected flag for the function specified by the function ID parameter. It is true if the function is expected to send a response, false otherwise.

For getter functions this is enabled by default and cannot be disabled, because those functions will always send a response. For callback configuration functions it is enabled by default too, but can be disabled by set_response_expected(). For setter functions it is disabled by default and can be enabled.

Enabling the response expected flag for a setter function allows to detect timeouts and other error conditions calls of this setter as well. The device will then send a response for this purpose. If this flag is disabled for a setter function then no response is send and errors are silently ignored, because they cannot be detected.

See set_response_expected() for the list of function ID constants available for this function.

IMU.set_response_expected(function_id, response_expected)
Parameters:
  • function_id -- int
  • response_expected -- bool
Return type:

None

Changes the response expected flag of the function specified by the function ID parameter. This flag can only be changed for setter (default value: false) and callback configuration functions (default value: true). For getter functions it is always enabled and callbacks it is always disabled.

Enabling the response expected flag for a setter function allows to detect timeouts and other error conditions calls of this setter as well. The device will then send a response for this purpose. If this flag is disabled for a setter function then no response is send and errors are silently ignored, because they cannot be detected.

The following function ID constants are available for this function:

  • IMU.FUNCTION_LEDS_ON = 8
  • IMU.FUNCTION_LEDS_OFF = 9
  • IMU.FUNCTION_SET_ACCELERATION_RANGE = 11
  • IMU.FUNCTION_SET_MAGNETOMETER_RANGE = 13
  • IMU.FUNCTION_SET_CONVERGENCE_SPEED = 15
  • IMU.FUNCTION_SET_CALIBRATION = 17
  • IMU.FUNCTION_SET_ACCELERATION_PERIOD = 19
  • IMU.FUNCTION_SET_MAGNETIC_FIELD_PERIOD = 21
  • IMU.FUNCTION_SET_ANGULAR_VELOCITY_PERIOD = 23
  • IMU.FUNCTION_SET_ALL_DATA_PERIOD = 25
  • IMU.FUNCTION_SET_ORIENTATION_PERIOD = 27
  • IMU.FUNCTION_SET_QUATERNION_PERIOD = 29
  • IMU.FUNCTION_ORIENTATION_CALCULATION_ON = 37
  • IMU.FUNCTION_ORIENTATION_CALCULATION_OFF = 38
  • IMU.FUNCTION_ENABLE_STATUS_LED = 238
  • IMU.FUNCTION_DISABLE_STATUS_LED = 239
  • IMU.FUNCTION_RESET = 243
IMU.set_response_expected_all(response_expected)
Parameters:response_expected -- bool
Return type:None

Changes the response expected flag for all setter and callback configuration functions of this device at once.

IMU.enable_status_led()
Return type:None

Enables the status LED.

The status LED is the blue LED next to the USB connector. If enabled is is on and it flickers if data is transfered. If disabled it is always off.

The default state is enabled.

New in version 2.3.1 (Firmware).

IMU.disable_status_led()
Return type:None

Disables the status LED.

The status LED is the blue LED next to the USB connector. If enabled is is on and it flickers if data is transfered. If disabled it is always off.

The default state is enabled.

New in version 2.3.1 (Firmware).

IMU.is_status_led_enabled()
Return type:bool

Returns true if the status LED is enabled, false otherwise.

New in version 2.3.1 (Firmware).

IMU.get_protocol1_bricklet_name(port)
Parameters:port -- chr
Return type:(int, [int, int, int], str)

Returns the firmware and protocol version and the name of the Bricklet for a given port.

This functions sole purpose is to allow automatic flashing of v1.x.y Bricklet plugins.

The returned namedtuple has the variables protocol_version, firmware_version and name.

IMU.get_chip_temperature()
Return type:int

Returns the temperature in °C/10 as measured inside the microcontroller. The value returned is not the ambient temperature!

The temperature is only proportional to the real temperature and it has an accuracy of +-15%. Practically it is only useful as an indicator for temperature changes.

IMU.reset()
Return type:None

Calling this function will reset the Brick. Calling this function on a Brick inside of a stack will reset the whole stack.

After a reset you have to create new device objects, calling functions on the existing ones will result in undefined behavior!

IMU.get_identity()
Return type:(str, str, chr, [int, int, int], [int, int, int], int)

Returns the UID, the UID where the Brick is connected to, the position, the hardware and firmware version as well as the device identifier.

The position can be '0'-'8' (stack position).

The device identifier numbers can be found here. There is also a constant for the device identifier of this Brick.

The returned namedtuple has the variables uid, connected_uid, position, hardware_version, firmware_version and device_identifier.

Callback Configuration Functions

IMU.register_callback(id, callback)
Parameters:
  • id -- int
  • callback -- callable
Return type:

None

Registers a callback with ID id to the function callback. The available IDs with corresponding function signatures are listed below.

IMU.set_acceleration_period(period)
Parameters:period -- int
Return type:None

Sets the period in ms with which the CALLBACK_ACCELERATION callback is triggered periodically. A value of 0 turns the callback off.

The default value is 0.

IMU.get_acceleration_period()
Return type:int

Returns the period as set by set_acceleration_period().

IMU.set_magnetic_field_period(period)
Parameters:period -- int
Return type:None

Sets the period in ms with which the CALLBACK_MAGNETIC_FIELD callback is triggered periodically. A value of 0 turns the callback off.

IMU.get_magnetic_field_period()
Return type:int

Returns the period as set by set_magnetic_field_period().

IMU.set_angular_velocity_period(period)
Parameters:period -- int
Return type:None

Sets the period in ms with which the CALLBACK_ANGULAR_VELOCITY callback is triggered periodically. A value of 0 turns the callback off.

IMU.get_angular_velocity_period()
Return type:int

Returns the period as set by set_angular_velocity_period().

IMU.set_all_data_period(period)
Parameters:period -- int
Return type:None

Sets the period in ms with which the CALLBACK_ALL_DATA callback is triggered periodically. A value of 0 turns the callback off.

IMU.get_all_data_period()
Return type:int

Returns the period as set by set_all_data_period().

IMU.set_orientation_period(period)
Parameters:period -- int
Return type:None

Sets the period in ms with which the CALLBACK_ORIENTATION callback is triggered periodically. A value of 0 turns the callback off.

IMU.get_orientation_period()
Return type:int

Returns the period as set by set_orientation_period().

IMU.set_quaternion_period(period)
Parameters:period -- int
Return type:None

Sets the period in ms with which the CALLBACK_QUATERNION callback is triggered periodically. A value of 0 turns the callback off.

IMU.get_quaternion_period()
Return type:int

Returns the period as set by set_quaternion_period().

Callbacks

Callbacks can be registered to receive time critical or recurring data from the device. The registration is done with the register_callback() function of the device object. The first parameter is the callback ID and the second parameter the callback function:

def my_callback(param):
    print(param)

imu.register_callback(IMU.CALLBACK_EXAMPLE, my_callback)

The available constants with inherent number and type of parameters are described below.

Note

Using callbacks for recurring events is always preferred compared to using getters. It will use less USB bandwidth and the latency will be a lot better, since there is no round trip time.

IMU.CALLBACK_ACCELERATION
Parameters:
  • x -- int
  • y -- int
  • z -- int

This callback is triggered periodically with the period that is set by set_acceleration_period(). The parameters are the acceleration for the x, y and z axis.

IMU.CALLBACK_MAGNETIC_FIELD
Parameters:
  • x -- int
  • y -- int
  • z -- int

This callback is triggered periodically with the period that is set by set_magnetic_field_period(). The parameters are the magnetic field for the x, y and z axis.

IMU.CALLBACK_ANGULAR_VELOCITY
Parameters:
  • x -- int
  • y -- int
  • z -- int

This callback is triggered periodically with the period that is set by set_angular_velocity_period(). The parameters are the angular velocity for the x, y and z axis.

IMU.CALLBACK_ALL_DATA
Parameters:
  • acc_x -- int
  • acc_y -- int
  • acc_z -- int
  • mag_x -- int
  • mag_y -- int
  • mag_z -- int
  • ang_x -- int
  • ang_y -- int
  • ang_z -- int
  • temperature -- int

This callback is triggered periodically with the period that is set by set_all_data_period(). The parameters are the acceleration, the magnetic field and the angular velocity for the x, y and z axis as well as the temperature of the IMU Brick.

IMU.CALLBACK_ORIENTATION
Parameters:
  • roll -- int
  • pitch -- int
  • yaw -- int

This callback is triggered periodically with the period that is set by set_orientation_period(). The parameters are the orientation (roll, pitch and yaw) of the IMU Brick in Euler angles. See get_orientation() for details.

IMU.CALLBACK_QUATERNION
Parameters:
  • x -- float
  • y -- float
  • z -- float
  • w -- float

This callback is triggered periodically with the period that is set by set_quaternion_period(). The parameters are the orientation (x, y, z, w) of the IMU Brick in quaternions. See get_quaternion() for details.

Constants

IMU.DEVICE_IDENTIFIER

This constant is used to identify a IMU Brick.

The get_identity() function and the CALLBACK_ENUMERATE callback of the IP Connection have a device_identifier parameter to specify the Brick's or Bricklet's type.

Creative Commons Licence The content of this page is licensed under Creative Commons Attribution 3.0 Unported License.