IMU

An Inertial Measurement Unit (IMU) is a sensor device that measures and reports motion, orientation, and gravitational forces. IMUs combine multiple sensors to provide comprehensive motion tracking capabilities, making them essential for applications such as robotics, drone stabilization, motion sensing, orientation tracking, and navigation systems. Modern IMUs typically integrate a 3-axis accelerometer (measuring linear acceleration), a 3-axis gyroscope (measuring angular velocity), and often a 3-axis magnetometer (measuring magnetic field) to create a complete 9-axis motion tracking solution.

1. ICM-20948 IMU

Your board is equipped with the InvenSense ICM-20948, a high-performance 9-axis motion tracking device. The sensor combines a 3-axis gyroscope, 3-axis accelerometer, and 3-axis magnetometer in a single package.

1.1. Key Features

3-Axis Gyroscope with programmable full-scale ranges: ±250, ±500, ±1000, ±2000 DPS
3-Axis Accelerometer with programmable full-scale ranges: ±2g, ±4g, ±8g, ±16g
3-Axis Magnetometer (AK09916) for compass functionality
Digital-output temperature sensor for temperature compensation
User-programmable interrupts for motion detection and data ready signals
Low-power modes for battery-powered applications
Digital Low Pass Filter (DLPF) for noise reduction

1.2. Hardware Connection

The ICM-20948 on your board is connected via SPI interface at /dev/spidev0.3.

2. Example Application

imu project included in t3gemstone/examples repository provides a complete C library and test application to interact with the IMU sensor. The library handles all low-level communication and provides an API for reading sensor data.

2.1. Test Application

The test application performs the following sequence:

Initialization - Opens the SPI device at /dev/spidev0.3 and initializes the sensor
Configuration - Sets accelerometer to ±8g and gyroscope to ±2000 DPS range
Information - Prints configuration values and temperature at start
Continuous monitoring - Reads accelerometer and gyroscope data every 10ms and prints calculated angles
Angle calculation - Applies Kalman filtering to compute orientation angles
Auto-recovery - Monitors sensor connectivity and reconfigures if needed

The Kalman filter implementation provides smooth angle estimates by fusing accelerometer and gyroscope data, reducing noise while maintaining responsiveness to real motion.

gemstone@t3-gem-o1:examples$ ./build/examples/imu/imu
# ICM-20948 IMU Accel/Gyro/Temp Test
# ==================================
# 
# icm20948 SPI bus and device initialized successfully
# icm20948 Device ID:0xEA
# icm20948 Device id correct!
# 
# Temperature:            47.31
# Accel Sensitivity:      4096.00
# Accel Full Scale Range: 8g (+/-)
# Gyro Sensitivity:       16.40
# Gyro Full Scale Range:  2000DPS (+/-)
# 
# Continuous test will begin shortly. Press Ctrl+C to exit.
# Angle: x=   0.01, y= 179.22, z=   0.00
# Angle: x=   0.03, y= 179.23, z=   0.00
# Angle: x=   0.05, y= 179.23, z=   0.00
# Angle: x=   0.06, y= 179.23, z=   0.00
# Angle: x=   0.08, y= 179.23, z=  -0.00
# Angle: x=   0.09, y= 179.23, z=  -0.00
# Angle: x=   0.11, y= 179.23, z=  -0.00
# Angle: x=   0.12, y= 179.23, z=  -0.00

2.2. API Reference

The icm20948 library provides:

Bus initialization for both I2C and SPI interfaces
Sensor configuration with customizable full-scale ranges
Raw data reading for accelerometer, gyroscope, and temperature
Angle calculation using Kalman filtering for smooth orientation tracking
Auto-recovery with online status checking and reconfiguration
Flexible data access through a structured data interface

2.2.1. Initialization

icm20948_handle_t icm20948_create(icm20948_data_t* data, char* tag);
int icm20948_spi_bus_init(icm20948_handle_t sensor, const char* dev_path);
int icm20948_configure(icm20948_handle_t sensor, 
                       icm20948_acce_fs_t acce_fs, 
                       icm20948_gyro_fs_t gyro_fs);

2.2.2. Data Reading

int icm20948_get_acce(icm20948_handle_t sensor);
int icm20948_get_gyro(icm20948_handle_t sensor);
int icm20948_get_temp(icm20948_handle_t sensor);
void icm20948_get_angle(icm20948_handle_t sensor, float dt);

2.2.3. Configuration

int icm20948_set_acce_fs(icm20948_handle_t sensor, icm20948_acce_fs_t acce_fs);
int icm20948_set_gyro_fs(icm20948_handle_t sensor, icm20948_gyro_fs_t gyro_fs);
float icm20948_get_acce_sensitivity(icm20948_handle_t sensor);
float icm20948_get_gyro_sensitivity(icm20948_handle_t sensor);

2.2.4. Utility

int icm20948_check_online(icm20948_handle_t sensor);
void icm20948_delete(icm20948_handle_t sensor);

2.2.5. Data Structure

The sensor data is accessed through the icm20948_data_t structure:

typedef struct {
    int16_t ax_raw, ay_raw, az_raw; // Raw accelerometer data
    int16_t gx_raw, gy_raw, gz_raw; // Raw gyroscope data
    float ax, ay, az;               // Accelerometer (g)
    float gx, gy, gz;               // Gyroscope (deg/s)
    float anglex, angley, anglez;   // Computed angles (deg)
    float temp;                     // Temperature (°C)
} icm20948_data_t;

Getting Started

Development Boards

Operating System

Image Writer

Projects

Sdk

Guides

Other

1. ICM-20948 IMU

1.1. Key Features

1.2. Hardware Connection

2. Example Application

2.1. Test Application

2.2. API Reference

2.2.1. Initialization

2.2.2. Data Reading

2.2.3. Configuration

2.2.4. Utility

2.2.5. Data Structure

​1. ICM-20948 IMU

​1.1. Key Features

​1.2. Hardware Connection

​2. Example Application

​2.1. Test Application

​2.2. API Reference

​2.2.1. Initialization

​2.2.2. Data Reading

​2.2.3. Configuration

​2.2.4. Utility

​2.2.5. Data Structure

1. ICM-20948 IMU

1.1. Key Features

1.2. Hardware Connection

2. Example Application

2.1. Test Application

2.2. API Reference

2.2.1. Initialization

2.2.2. Data Reading

2.2.3. Configuration

2.2.4. Utility

2.2.5. Data Structure