Achieving Noise Suppression of Multichannel Lock-In Amplifier Based on Signal Segmentation

This study proposes a multichannel signal-segmented lock-in amplifier (LIA) for suppressing co-frequency noise and processing signals containing local anomalies. Signal-segmented LIAs can counteract the interference caused by co-frequency noise in different signal segments, preventing the superimposition of co-frequency noise at different times in the input signal, which ultimately reduces its impact on the output. A module for detecting localized anomalous signals was incorporated into the LIA, enhancing its anti-interference capability. Through simulation, the mean square error (MSE) of the outputs from the signal-segmented LIA and the conventional software-implemented LIA (SILIA) were compared, validating the effectiveness of signal segmentation. Furthermore, an eddy current magneto-optic (ECMO) nondestructive testing (NDT) system with signal-segmented LIA was developed for detecting defects in weakly conductive materials carbon fiber-reinforced polymer (CFRP). Comprehensive experiments demonstrated that the signal-segmented LIA significantly suppressed co-frequency noise interference in its output compared to the traditional SILIA and exhibited superior anti-interference to localized anomalous signals. These results further validate the efficacy of the signal segmentation approach.