Performance Evaluation of MMA7260QT and ADXL345 on Self Balancing Robot
Abstract: A self balancing
robot (SBR) controller needs to detect platform inclination. For this purpose,
an accelerometer is used. From various types of accelerometer, we can divide
into digital and analog ones. The problem is how to select the right type for
the SBR. This paper evaluates the performance of the ADXL345, 3-axis digital
output accelerometer and the MMA7260QT, 3-axis analog output accelerometer. The
Arduino is used to read data from the sensor and send it to PC for plotting.
Both sensors use the lowest sensitivity. The sensors are evaluated with three
criteria, i.e. stationary, dynamical response and collaborating with ITG3200
3-axis gyroscope for Kalman filter fusion. For stationary criterion, the
ADXL345 is better than the other sensor for all stationary position. For
dynamical response, both sensors suffer from the noise due to acceleration of
the platform. The sensors do not only sense the gravity but also the
acceleration of the platform when it is moved. But the noise level for the
ADXL345 is lower than the other. Using Kalman filter makes both sensors show
good performance for a SBR application. If three criteria are combined with
hardware aspect, then the authors recommend using the ADXL345. Besides, it has
several useful features to handle abrupt acceleration.
Author: Hany Ferdinando,
Handry Khoswanto, Djoko Purwanto
Journal Code: jptkomputergg130018
