Arduino Nano 33 BLE Sense Rev2: IMU & Environmental Data Logger Build & Debug Guide

The Arduino Nano 33 BLE Sense Rev2 (Part: ABX00069) is a powerhouse for edge computing, packing a 9-axis IMU, environmental sensors, and a microphone into a 45x18mm footprint. However, the transition from Rev1 to Rev2 changed the underlying sensor silicon, breaking countless legacy tutorials. This guide targets the Rev2 variant specifically, walking you through building a standalone inertial and environmental data logger with onboard MicroSD storage.

Parts List and Board Specifications

Before wiring, verify your board revision. The Rev2 board swaps the STMicroelectronics sensors used in Rev1 for Bosch and Renesas alternatives. If you attempt to compile Rev1 libraries on a Rev2 board, you will hit immediate I2C initialization faults.

Required Hardware

• Microcontroller: Arduino Nano 33 BLE Sense Rev2 (ABX00069) - ~$65.00
• Storage: Adafruit MicroSD Card Breakout Board+ (Product ID: 254) - ~$7.50
• Power: 3.7V LiPo Battery (500mAh minimum) with JST-PH 2.0 connector - ~$8.00
• Media: Class 10 MicroSD Card (8GB to 32GB, FAT32 formatted)
• Wiring: 24 AWG solid core jumper wires, half-size breadboard

Sensor Silicon: Rev1 vs Rev2 Comparison

Pin Mapping and Wiring Guide

The Nano 33 BLE Sense operates strictly at 3.3V logic. Feeding 5V into any digital I/O pin will permanently damage the nRF52840 SoC. The Adafruit MicroSD breakout includes a 3.3V voltage regulator and logic level shifters, making it safe to interface with the 3.3V SPI bus.

Note: While the nRF52840 supports multiple SPI instances, using the default hardware SPI pins (D11, D12, D13) ensures maximum transfer speeds and compatibility with the standard Arduino SD.h library.

Complete Data Logger Firmware

The following code targets the Arduino Nano 33 BLE Sense Rev2. It initializes the BMI270 IMU and HS3003 environmental sensor, then logs the data to the MicroSD card at 10Hz. Error handling is built into the setup() loop to halt execution and report faults via the Serial Monitor if hardware initialization fails.

#include <Arduino_BMI270_BMM150.h>
#include <Arduino_HS300x.h>
#include <SD.h>

// Pin Definitions
const int SD_CS_PIN = 10;
const int LED_ERROR_PIN = LED_BUILTIN;

// Data logging interval (milliseconds)
const unsigned long LOG_INTERVAL = 100; 
unsigned long lastLogTime = 0;

File dataFile;

void setup() {
  Serial.begin(115200);
  while (!Serial) {
    ; // Wait for serial port to connect (needed for native USB)
  }

  pinMode(LED_ERROR_PIN, OUTPUT);

  // Initialize IMU (BMI270 + BMM150)
  if (!IMU.begin()) {
    Serial.println("ERROR: Failed to initialize IMU!");
    blinkError(3);
    while (1); // Halt execution
  }
  Serial.println("IMU initialized successfully.");

  // Initialize Environmental Sensor (HS3003)
  if (!HS300x.begin()) {
    Serial.println("ERROR: Failed to initialize HS3003 Humidity/Temp sensor!");
    blinkError(4);
    while (1);
  }
  Serial.println("HS3003 initialized successfully.");

  // Initialize SD Card
  if (!SD.begin(SD_CS_PIN)) {
    Serial.println("ERROR: SD card initialization failed! Check wiring and FAT32 format.");
    blinkError(5);
    while (1);
  }
  Serial.println("SD card initialized.");

  // Open file for appending
  dataFile = SD.open("datalog.csv", FILE_WRITE);
  if (dataFile) {
    // Write header if file is empty
    if (dataFile.size() == 0) {
      dataFile.println("Timestamp_ms,Accel_X,Accel_Y,Accel_Z,Gyro_X,Gyro_Y,Gyro_Z,Temp_C,Humidity_Pct");
    }
    dataFile.close();
  } else {
    Serial.println("ERROR: Could not open datalog.csv for writing.");
    while (1);
  }
}

void loop() {
  unsigned long currentTime = millis();

  if (currentTime - lastLogTime >= LOG_INTERVAL) {
    lastLogTime = currentTime;

    float ax, ay, az, gx, gy, gz;
    float temp, humidity;

    // Read IMU data
    if (IMU.accelerationAvailable() && IMU.gyroscopeAvailable()) {
      IMU.readAcceleration(ax, ay, az);
      IMU.readGyroscope(gx, gy, gz);

      // Read Environmental data
      temp = HS300x.readTemperature();
      humidity = HS300x.readHumidity();

      // Write to SD
      dataFile = SD.open("datalog.csv", FILE_WRITE);
      if (dataFile) {
        dataFile.print(currentTime);
        dataFile.print(","); dataFile.print(ax);
        dataFile.print(","); dataFile.print(ay);
        dataFile.print(","); dataFile.print(az);
        dataFile.print(","); dataFile.print(gx);
        dataFile.print(","); dataFile.print(gy);
        dataFile.print(","); dataFile.print(gz);
        dataFile.print(","); dataFile.print(temp);
        dataFile.print(","); dataFile.println(humidity);
        dataFile.close();
      }
    }
  }
}

// Helper function to blink LED on fatal errors
void blinkError(int count) {
  for (int i = 0; i < count; i++) {
    digitalWrite(LED_ERROR_PIN, HIGH);
    delay(250);
    digitalWrite(LED_ERROR_PIN, LOW);
    delay(250);
  }
}

Debugging: Resolving IMU and I2C Initialization Faults

When working with the Nano 33 family, I2C bus lockups and library mismatches are the most common points of failure. If your Serial Monitor outputs ERROR: Failed to initialize IMU!, follow this diagnostic path.

The First Three Things to Check

1. Verify the Board Revision and Library: Open the Arduino IDE Boards Manager. Ensure you have the Arduino_BMI270_BMM150 library installed. If you are using Arduino_LSM9DS1, it will compile but fail at runtime on a Rev2 board.
2. Check the I2C Pull-up Resistors: The Nano 33 BLE Sense routes its internal sensors to the external I2C header (pins A4/A5). If you have external sensors connected that are dragging the bus low, or if the internal pull-ups are disabled via software, the internal IMU will fail to enumerate.
3. Inspect the 3.3V Rail: The onboard IMU requires a clean 3.3V supply. If you are powering the board via a weak USB hub or a damaged cable, the voltage may droop below 3.0V during IMU startup, causing the BMI270 to brown out and reject I2C addressing.

Ranked Causes for 'Failed to initialize IMU!'

Extending and Simplifying the Build

Depending on your project constraints, you may need to alter the architecture of this data logger.

How to Simplify (Drop the SD Card)

If you don't need long-term standalone storage, remove the MicroSD breakout and stream the data over Bluetooth Low Energy (BLE). Use the ArduinoBLE library to create a custom BLE service with characteristics for Acceleration and Temperature. This reduces power consumption by roughly 40% and eliminates SPI bus overhead, allowing you to push the IMU sampling rate to 100Hz+.

How to Extend (Edge Machine Learning)

The nRF52840 features 1MB of Flash and 256KB of RAM, which is sufficient for running TinyML models. Use Edge Impulse to capture gesture data via the BMI270, train a neural network, and export it as an Arduino C++ library. Replace the SD logging loop with an inference loop that triggers a digital pin when a specific motion signature (like a 'wave' or 'drop') is detected.

Frequently Asked Questions