Feasibility Considerations on an Inkjet-Printed Capacitive Position Sensor for Electrostatically Actuated Resonant MEMS-Mirror Systems

We present a systematic approach to enable an inkjet-printed capacitive position sensor for a resonant microelectromechanical systems (MEMS)-mirror system in a smart packaging solution. We adopt systematic noise analysis and employ a stochastic (Bayesian) estimator for position reconstruction. In this way, we exploit prior knowledge of the system and its dynamics. We can thus demonstrate a way to achieve very high position resolutions of res
_pos
<; 50 nm. The considered arrangement is a part of a Michelson interferometer setup and, as such, subject to concise specifications regarding resolution and accuracy. The measurement system's analog front-end is designed in a carrier frequency setup. It enables a flexible choice of bandwidth and provides a measurement signal at a frequency well separated from that of the mirror actuation. The wide measurement range of r
_m
= 1000 μm, at an offset of d
₀
= 1000 μm, and the high required position resolution of res
_pos
<; 50 nm present major challenges for the inkjet-printed capacitive sensor. To overcome noise limitations inflicted by the measurement and system setups, and to enable nanometer resolution independently of the measurement bandwidth extended Kalman filtering as Bayesian approach to statistical signal processing is suggested. Noise analysis and stochastic position estimation are applied to the MEMSmirror system. Evaluations demonstrate the high achievable resolutions using extended Kalman filtering.