A Design Of Acquisition System And Noise Reduction Method For MEMS Inertial Sensors Parallel Test Data

MEMS inertial elements,which are extensively applied in industry,military industry and consumer electronics industry,is of great market demand.The test efficiency improvement and cost reduction in manufacture process are both cutting-edge topics.This paper designs a parallel MEMS test data collection scheme to improve the existing single-chip MEMS test system.The scheme is mainly based on hardware design,and denoise the collected MEMS signal from two aspects namely hardware optimization and algorithm design.This paper first introduce the design method of parallel test system structure.The test data is collected and transmitted after output from MEMS.The data is collected by 8 8-channel ADCs to realize multi-channel analog signal synchronous acquisition,and is received via FPGA in parallel.The high speed data transmission among stator and rotor of turntable is based on FPGA,USB2.0protocol chip and electric slip ring.This paper also analyzes the noise reduction of twofold,namely the instrumental noise of analog components of data collection circuit and the MEMS measurement signal in real scene.The instrumental noise mainly includes the resistance thermal noise,the component1/f noise,the power ripple noise,as well as the phase noise,switching noise and the quantization noise of ADCs,etc.Models of the above noises and detailed mathematical expressions of which are provided.Suggestions and methods regarding noise reduction are given from the point of view of hardware design and practical operation.For the MEMS measurement signal noise reduction,the Mallat algorithm and the threshold denoising method are applied based on the principle of wavelet analysis,and decent results are obtained.