A Nano- g MEMS Accelerometer With Electrostatic Anti-Spring Based on SOI Technology

A hybrid comb-finger MEMS accelerometer is presented to achieve nano-$g$-level noise performance by enhancing acceleration sensitivity through lowering the resonant frequency of its mechanical structure. Gap-varying comb fingers are employed to realize electrostatic anti-spring tuning for mechanical spring softening, while area-varying comb fingers are used for capacitive displacement sensing or electrostatic actuation. The device is fabricated using a standard SOI process within a footprint of$8\times 12$mm2, achieving fabrication simplicity and structural miniaturization. Experimental results show that the resonant frequency can be tuned from 78 Hz to 12 Hz, resulting in an approximately 40-fold enhancement in acceleration sensitivity and a fourfold reduction in bias instability. A noise floor of 8.5 n$g$/$\surd $Hz at 1.2 Hz is achieved, and micro seismic signals are successfully detected. While achieving nano-$g$precision under open-loop conditions, the proposed architecture decouples negative stiffness generation from the actuation path, thereby alleviating the pull-in instability and paving the way to stable closed-loop operation at near-zero stiffness.[2025-0214]