Orthogonal modulation based light beam induced current method for anti-noise defect detection in photovoltaic cells

In this paper, orthogonal modulation theory is introduced into light beam-induced current (LBIC) scanning to develop an orthogonal modulation-based LBIC (OMLBIC) method that is proposed for anti-noise defect detection applications in photovoltaic cells. In the proposed method, the response signal from the scanning area is treated similarly to the characteristics of a communication channel; first, orthogonal structured light patterns are generated to illuminate the scanning area surface, the combined current output of the entire device is then measured for each pattern, and demodulation is finally performed to extract the signal precisely from the noise. The feasibility of the proposed method and its detection performance were evaluated. The experimental results indicate that the proposed method can cause a maximum mean square error (MSE) reduction of 16,384 times with 128 × 128 scanning points, and it provides an enhanced direct link between the inherent spatial non-uniformities and the overall current map for the sample. When compared with the conventional LBIC method, the OMLBIC method provided increased consistency during noisy scanning, greater detail in defect detection, and a reduction in the sensitivity to noise, without requiring any time-consuming increase in sampling or hardware upgrades. The association between the noise elimination effect and the size of the structured light pattern was also studied to provide a superior system configuration. The proposed method shows major potential for improvement in the defect detection accuracy and classification of photovoltaic cells in industrial applications.