Abstract

This paper proposes a clock strategy to improve the precision of power-efficient readout circuit for capacitive sensor. To achieve high power efficiency, capacitive sensor such as MEMS accelerometer is designed with open-loop architecture rather than close-loop one. In the open-loop architecture, the capacitance variation of sensing element is limited to femto-farad level in order to overcome nonlinearity problem. However, due to this limitation, the signal charge from sensing element is weak and the interference charge due to coupling capacitance between clock wires and sensing electrodes becomes a significant issue. Therefore, split clock bus is employed to meet this challenge, but the split clock bus introduces the timing mismatch which causes the charge injection sensitive to fabrication process and circuit operating voltage. As a result, the input offset, the variation of the offset and the nonlinearity of readout circuit will increase, and this will lead to reduction of precision. In this work, a clock scheme named “multi-nested clocks” is proposed to address the charge injection problem. The multi-nested clocks are demonstrated in a readout circuit fabricated using a $0.18-\mu \text{m}$ BCD process. The measurement results show that compared to the readout circuit using traditional clock, the readout circuit using the multi-nested clocks significantly reduces the equivalent input offset from 1.66fF to 0.25fF, the offset variation from 1.4fF to 0.2fF and the nonlinearity from 5.5% to 0.9%.

Full Text
Published version (Free)

Talk to us

Join us for a 30 min session where you can share your feedback and ask us any queries you have

Schedule a call