Research On Resistance Bridge MEMS Sensor Interface Circuit

Based on the progress of micro-nano manufacturing technology,MEMS sensor realizes a significant development in recent years.Relying on its high precision,miniaturization,low cost,intelligence,integration,it has been applied in many fields such as automobile,mobile phone,aerospace,medical instrument and industrial monitoring.Resistance bridge MEMS sensor is a kind of sensor with wide application,which USES the pressure sensitivity and thermal sensitivity of resistive materials to detect external signals,such as pressure sensor,pressure resistance acceleration sensor and heat film flow sensor.Thermal film flow sensor is a typical resistive bridge principle sensor,and the digital thermal film air flow meter based on MEMS technology has the characteristics of higher measurement accuracy,faster dynamic response,stronger anti-interference ability and the recognition of the intake direction,which has become the mainstream of the development of vehicle flow sensor.Interface circuit is an important part of MEMS sensor.The design of a high-performance interface circuit is of great significance to improve the overall performance of MEMS sensor.This paper studies the resistance bridge MEMS flow sensor and designs a digital interface circuit chip for MEMS hot-film flow sensor.The flow sensor adopts the operating mode of constant temperature difference.Based on the closed-loop control of two constant current sources,the circuit realizes the constant temperature difference between inlet air temperature and heating temperature.2 order ?? traffic signals and temperature signals in real-time ?? ADC analog/digital conversion,CIC&IIR level cascade mode after filter down sampling and noise filtering.After the temperature signal and flow signal are processed by the temperature compensation algorithm,the compensated flow measurement value is finally output-through the.SPI communication interface.According to the design parameters of MEMS flow sensor,EA ADC was established of which simulation results showed the SNR was 103 dB while the OSR is 512.A 64 times down-sampling CIC filter and an IIR low-pass filter,were designed and prototype validated in FPGA.All circuit design and simulation were carried out in the Cadence,and the chip was manufactured by 0.18 ?m CMOS process.The test results of the chip show that the ENOB is 14bits,and the power consumption is less than 25mW,the stability(relative error)of 1h measurement is less than 770ppm.Compared with the traditional polynomial fitting method,the fitting error of the proposed piecewise fitting method is only 1/3 of the original one.The improved BP neural network temperature compensation algorithm improves the temperature adaptability of the sensor,and the flow detection accuracy of all sensors are less than 1%FS.