Improved Processing Circuit for MEMS Accelerometer as a Respiration Sensor

Capturing respiration signal using a microelectromechanical (MEMS) accelerometer is a challenging task. The design of the interface circuit of the sensor must consider the characteristics of the respiration signal resulting from abdominal movement. It is mainly characterized by very low magnitude and low-frequency contents. Hence, any DC offset or drift in the sensor's response is a key design factor of the circuit. In addition, the OnChip processing circuit of the sensor itself may introduce DC offset and limit the lower range of the frequency response. This paper presents a design and implementation of a signal processing interface circuit that can be used to process MEMS accelerometer output to capture respiration signals. The circuit consists of a low pass filter that captures the DC component of the signal, then a differential amplifier that subtracts the estimated DC component from the signal followed by amplification. The circuit is self-calibrated by the position of the sensor. The circuit diagram, evaluation results, and validation of the circuit are presented in this paper.