Nonlinearity of a Closed-Loop Micro-accelerometer

The sensing mass of the MEMS sensor working in the closed-loop system can be driven not only by inertial force due to external acceleration, but also by elastic restoring force and the electrostatic driving force. As all the forces are included in the loop, the errors they bring in separately could be interrelated with each other. Driven by undistinguishable forces with initial biases due to the manufacturing procession of the sensor, the MEMS sensor consequently works with biases, which could be correlated with the errors of the accelerometer, such as nonlinearity. In order to make clear how nonlinearity comes into being, a model of MEMS sensor with initial biases is proposed, and the formula, which approximately shows the relationship between the nonlinearity of the accelerometer and the initial biases, is deduced. The model, which expatiates the tradeoff between nonlinearity and zero bias of the accelerometer, and helps to quantitatively determine how to adjust the proof bias as well as how much the mechanical manufacturing error can be accepted, could be useful while designing and debugging.