Waveshare MLX90640 IR Array Thermal Imaging Camera, 32×24 Pixels, 110° FOV

The module is broadly compatible with any platform supporting I2C communication. Waveshare provides specific example code and libraries for Raspberry Pi (using Python), Arduino IDE (for boards like Uno or Mega), ESP32 (via the Arduino framework), and STM32. The onboard voltage translator ensures safe operation with both 3.3V and 5V logic levels, making integration straightforward across these common development ecosystems.

Connect the module's 4-pin PH2.0 connector to your host board: VCC to 3.3V or 5V, GND to ground, SDA to the I2C data line, and SCL to the I2C clock line. Ensure the I2C pull-up resistors are present on your host board or are added externally. The module draws less than 23mA during operation. For initial testing, use the provided example code to scan for the device at I2C address 0x33 to confirm communication.

The MLX90640 supports a programmable refresh rate from 0.5 Hz to 64 Hz. You configure this via specific I2C commands sent to the sensor's control registers. A lower refresh rate (e.g., 0.5Hz or 2Hz) reduces power consumption and data bandwidth, suitable for periodic monitoring. Higher rates (16Hz, 32Hz, 64Hz) are used for capturing faster thermal changes. Refer to the MLX90640 datasheet and Waveshare's examples for the exact register write sequences to set the desired rate.

Yes, the module outputs a grid of 768 temperature values (32 columns by 24 rows). To create a visual thermal image, you need to read these values via I2C and map them to a color palette (like jet or grayscale) in your software. Waveshare's Raspberry Pi example includes Python scripts that do this, displaying the image on screen. For Arduino or ESP32, you would typically send the processed data to a display or transmit it over serial/USB to a PC application for visualization.