Study On The Readout Technique For Femto-farad MEMS Capacitive Sensor

The CMOS process for IC manufacturing and the MEMS process for sensor manufacturing are continuously scaling down,in order to meet the demands of low-cost,low power and small space from the Internet of Things?IoT?applications.However,the MEMS capacitive sensor meets a huge performance bottleneck when it is scaled to micron-meter level(10^-6 m).This is because the capacitance variation of the sensor decreases at a 5^th order rate?-100dB/dec?with the process scaling down,while the value of the parasitic capacitance of the sensor decreases at a 2^nd order rate?-40dB/dec?with the process scaling down.As a result,the accuracy of the traditional switched-capacitor based readout circuit is seriously reduced by the parasitic capacitance when the MEMS sensor is scaled to femto-farad level(10^-15 F).And the traditional readout circuit cannot meet the performance requirements on accuracy and power consumption of the IoT application any more.What's more,the reduction of signal dynamic range,the decreasing of intrinsic gain and the increasing of leakage current caused by the scaling down of modern CMOS process further add to this challenge.In this paper,the effects of scaling down on the performance of femto-farad MEMS capacitive sensor system are studied in detail,and the corresponding solutions are proposed.The main innovative achievements of this paper are as follows.1).An iterative readout technique named"Oversampling Successive Approximation?OSA?"is proposed to improve the readout accuracy of capacitive sensors.The main works are described as follows.Firstly,the deterioration of gain error caused by parasitic capacitance in femto-farad MEMS capacitance sensor system is analyzed.Secondly,the concept of OSA technique is proposed and the Capacitor-to-Voltage Converter?CVC?and the integrator based on OSA technique are implemented.Thirdly,the accuracy and noise performance of the circuit based on OSA are analyzed in detail.Finally,OSA-CVC is analyzed in detail on the charge injection,and a dedicated clock driving strategy is given to further improve the accuracy.2).An open-loop readout circuit for MEMS capacitive accelerometer based on OSA-CVC is proposed.The main works are described as follows.Firstly,the five key non-ideal characteristics of OSA-CVC that reduces system accuracy and settling time in practical application are analyzed.They are holding error,recovery degradation,increment degradation,rising edge degradation and charge injection.The solutions to these problems are given.Secondly,the proposed readout circuit is manufactured by commercial0.18-micron BCD process,and the traditional readout circuit is reproduced and manufactured by the same process for comparison.Finally,the experiment is conducted and the results show that compared with the traditional open-loop readout circuit,the proposed open-loop readout circuit based on OSA-CVC reduces the affect of common-mode parasitic capacitance on accuracy by 23.8 dB,reduces the power consumption by 69.3%,and reduces the chip area by 50%.This research is published in the academic journal IEEE JSSC.3).An closed-loop third-order Sigma-Delta readout circuit for capacitive accelerometer based on OSA-CVC and OSA integrator is proposed.The main works are described as follows.Firstly,the force feedback close-loop readout circuit for accelerometer is analyzed in detail.Secondly,the OSA circuit is employed to reduce the gain requirement of the amplifier in close-loop readout circuit,and thus the power consumption and the chip area are significantly reduced.Thirdly,the charge injection and leakage current interference which the close-loop readout circuit is very sensitive to are analyzed in detail and the interference suppressing solutions are given.Finally,the proposed closed-loop readout circuit is manufactured by commercial 0.18um BCD process.The experimental results show that the power consumption of the readout circuit chip is 0.5mW,the sampling rate is9MHz,the gain error is 0.07%,and the background noise is 24ug/rtHz.This research is published in the academic journal IEEE TCAS-I.4).The MEFS-OSA touch sensing technique with low sensitivity to water drop is proposed.The main works are described as follows.Firstly,the method to distinguish the water drop and human finger via physical dimension is proposed.The proposed system distinguishes the interference from the water drop based on physical size.Since the height of the finger?height>10mm?is much higher than the height of the water drop?height<1mm?,the electric field of the sensing cell is projected at high altitude?height>10mm?instead of gathering at low altitude?height<1mm?,so that the sensing cell can distinguish the finger from the water drop.Secondly,in order to implement this electric field projection,Mutual-Coupling Electrical Field Shaping?MEFS?technique is employed to build the sensing cells.MEFS sensing cell converts the interference of water drop to common-mode parasitic capacitance.Thirdly,the OSA readout technique is employed to suppress the interference from the common-mode parasitic capacitance.Finally,the experiments are conducted and the results show that for the proposed MEFS-OSA touch sensing system,when the sensing cell is covered by water drop and touched by fingers,the system output is1.11V and 1.23V,respectively.For the traditional touch sensing system,when the sensing cell is covered by water droplets and touched by fingers,the system output is 1.32V and1.33V,respectively.This results show that the MEFS-OSA touch sensing system can resist the interference of water drop better than the traditional sensing system does.This research is published in the academic journal IEEE TCAS-I.To summarize,the main innovation of this dissertation is the proposing of the OSA technique.The OSA technique is proved to be able to significantly reduce the gain error,the power consumption and the sensitivity to common-mode parasitic capacitance in femto-farad MEMS capacitive sensor system.