Accelerometer

Choose by project: tilt, rotation, heading, Klipper, or impact

Start with the measurement task. Static tilt and gravity reading need a different sensor family than rotation control, heading, or impact logging. ADXL345 is an accelerometer only, while MPU6050 combines an accelerometer and gyroscope; if you need rotation sensing as its own category, see these gyro sensor modules. For heading-capable builds, this example shows how a 9-axis IMU handles heading and orientation.

“10DOF” only means a board adds a barometer such as BMP180 or BMP280; it does not improve motion sensing by itself.

ADXL345 vs MPU6050 vs ADXL335: the fastest way to avoid the wrong first buy

For the digital accelerometer route, there is solid code support for using a digital ADXL345 with Arduino. Raspberry Pi users should stay with digital I2C or SPI sensors and pair them with compatible Raspberry Pi boards instead of starting from an analog board.

Compatibility checks before you buy: host, voltage, address, and mounting

Digital modules are the simpler path here. Analog modules change the wiring and the extra parts you need.

• Digital modules generally fit Arduino, ESP32, Raspberry Pi, and Pico projects through I2C or SPI at 3.3 V logic, so they are the default choice when you want the simplest host path. If you still need a controller for one of these sensors, start with compatible Arduino boards.
• Raspberry Pi has no analog inputs, so the ADXL335 analog accelerometer board needs an ADC such as MCP3008 or ADS1115. A digital board like the ADXL345 digital accelerometer or GY-521 MPU6050 accelerometer and gyroscope avoids that extra step.
• Most sensor chips here are 3.3 V devices, but breakout boards vary in whether they add a regulator or level shifting. Do not assume every module is universally 5 V-safe just because it plugs into an Arduino.
• The GY-521 MPU6050 accelerometer and gyroscope specifically lists an onboard voltage regulator, which reduces risk on common hobby hosts, but it is still worth verifying the exact board wiring before applying 5 V signals.
• MPU6050 uses selectable I2C address 0x68 or 0x69, and ADXL345 commonly appears at 0x53 or 0x1D. Those addresses matter if you are sharing a bus or troubleshooting a module that is not detected.
• Klipper results depend on rigid mounting and correct axis orientation, not just buying the right chip. A loosely mounted ADXL345 can produce poor resonance data even when the electrical connection is correct.
• The X, Y, and Z arrows on the PCB show the chip’s axis orientation, and those markings affect both physical mounting and axes_map or sign conventions in code.
• The ADXL362 ultra-low power accelerometer mentions SPI on its product card and does not mention I2C, so low-power buyers should confirm interface fit before treating it like a direct MPU6050 or ADXL345 substitute.

For a quick pre-purchase wiring check, this walkthrough covers wiring and addressing an MPU6050 on Arduino.

Feature and spec matrix for the modules on this page

Once you have the right sensor family, this matrix is the quickest way to compare the features that separate one board from another.

If you are comparing these boards mainly for the BMP180 or BMP280 side, that is as much a pressure and altitude sensor decision as a motion-sensor one. Buyers who want direct serial angle output can also review using a serial angle-output IMU.

When not to buy an accelerometer module

Accelerometers cannot reliably measure distance or position because double integration of noisy acceleration makes error grow very quickly, often within seconds. If the real goal is position or distance, use GPS, wheel encoders, or optical flow instead, and choose a capable controller such as an ESP32 board when the project also needs fusion or logging.

If your real goal is orientation output rather than raw sensor data, fused IMUs such as BNO055 or BNO085 are often considered because they output orientation directly. The tradeoff is 5–10× price, plus a real community split: convenience is the appeal, but calibration behavior and raw-noise quality are often criticized compared with MPU6050-style raw-data workflows. For heading, a 6-axis MPU6050 alone will still drift in yaw; use a magnetometer-based 9-axis option such as the MPU9250 9-axis IMU module or see this example of building a heading-capable setup with accel, gyro, and magnetometer.

FAQs on Accelerometer Modules for Arduino, Raspberry Pi & Robotics

Glossary

Accelerometer

A sensor that measures acceleration and gravity on three axes, which makes it useful for tilt, vibration, and impact detection but not reliable position tracking.

Gyroscope

A sensor that measures rotation rate, and it is the deciding difference between accel-only boards like ADXL345 and IMUs like MPU6050.

Magnetometer

A digital compass sensor that provides the reference needed for drift-corrected heading.

DOF / axis

On this page, 3-axis means accelerometer only, 6-axis adds a gyroscope, 9-axis adds a magnetometer, and 10DOF adds a barometer.

g-range

The maximum acceleration the sensor can measure before clipping, such as ±3.6g on ADXL335, ±16g on ADXL345, or 200g on ADXL375.

Drift

The slow wandering in gyro-based readings that is inherent to MEMS sensors and managed through calibration and sensor fusion.

SPI

A faster digital interface than I2C, and the one that matters for Klipper input shaping and some higher-rate logging setups.

ADC

An analog-to-digital converter, required when you want to use an analog-output sensor like ADXL335 with a host that has no analog inputs, such as Raspberry Pi.