Weak Signal Detection Method Based On Lock-in Amplifier

ABSTRACT:Weak signal detection is a discipline which develops with the engineering application. In recent years, the related research of weak signal detection has become a hot research field, such as the exploring of weak signal detection method, the design of weak signal detection system and the research and development of weak signal testing instrument. This paper mainly studied the principle and implementation method of weak signal detection which uses the phase-locked amplifier to extract useful signals.Firstly, the basic theory of weak signal detection and several popular detection methods are introduced in this paper, where the principle of digital phase-locked amplifier signal detection is mainly discussed; on this basis, the overall design scheme of digital phase-locked amplifier is given combining with the related detection principle of digital phase-locked amplifier. The design of digital phase-locked amplifier is emphatically discussed using the principle of correlation detection algorithm and phase shift algorithm.This paper focuses on the research of the relationship between sampling frequency and the related calculation results, and MATLAB is used to simulate the algorithm in the process of design. The simulation results showed that there exist inherent error when the reference signal turns into a square wave and there are a certain relationship between sampling frequency and signal frequency. In order to reduce the error, this paper proposed a method which dynamic sampling rate is introduced into the design of digital phase-locked amplifier. The result of dynamic sampling operation found that the more the number of sampling frequency is, the less error of the singularity caused. This method effectively solves the singularity problem.Finally, LabVIEW is used to simulate our proposed system. The test results show that it can effectively detect the signal below the frequency of500kHz when the signal amplitude of digital lock in amplifier is5v and the noise standard deviation is no more than50(SNR=34.04dB), and the relative error of the system test results and the theoretical calculation value is no more than2%.