# Design and Implementation of a Low-Noise Analog Front-End Circuit for MEMS Capacitive Accelerometers

## TL;DR

This paper describes a low-noise circuit for MEMS accelerometers used in optical image stabilization, achieving high sensitivity and low noise.

## Contribution

A novel low-noise analog front-end circuit with chopper stabilization for MEMS accelerometers is designed and implemented.

## Key findings

- The AFE circuit uses chopper stabilization to reduce low-frequency noise and parasitic coupling.
- The packaged system achieves 342 mV/g sensitivity and 14 μg/Hz noise floor.
- The MEMS accelerometer has a small footprint of 1000 μm × 950 μm.

## Abstract

This paper presents a low-noise analog front-end (AFE) integrated circuit (IC) circuit for capacitive micro-electromechanical system (MEMS) accelerometers that can be used for optical image stabilization (OIS) in various optical imaging systems. The AFE circuit design features a fully differential chopper stabilization technique that efficiently minimizes low-frequency 1/f noise and parasitic coupling. The AFE circuit chip is fabricated in a 0.18 μm complementary metal-oxide-semiconductor (CMOS) technology and co-packaged with an x-axis capacitive MEMS accelerometer based on a silicon-on-glass (SOG) process. The SOG accelerometer has a footprint of 1000 μm × 950 μm. The packaged system demonstrates a sensitivity of 342 mV/g and a nonlinearity of 1.1% between −1 g and +1 g, a dynamic range of 88 dB, and an equivalent noise floor of 14 μg/Hz.