A digitally controlled mems gyroscope with 3.2 DEG/HR stability

Abstract

The paper describes the implementation of digital feedback loops and digital signal conditioning for the control of vibrating MEMS gyroscopes. Compared to conventional analogue solutions, this implementation improves flexibility regarding algorithms and control parameters. The offset stability is also improved since mixing is performed using digital signals which are free from 1/f noise. The digital system is implemented in an FPGA that is connected to the gyro via an interface circuit. This circuit is a configurable analogue ASIC containing high performance capacitive readout amplifiers, 5/sup th/ order sigma-delta modulators and analogue feedback recovery circuits. The digital control architecture has been evaluated with a mass-produced silicon-micromachined butterfly gyroscope aimed for the automotive industry. The system shows a white noise-floor of 0.003 deg/(s/spl radic/Hz) and an Allan variance stability of 3.2 deg/hr.