A software error compensation method for equivalent-time sampling oscilloscopes

When an equivalent sampling oscilloscope measures signals, it will inevitably introduce noise and jitter, affecting the accuracy of the signals. The traditional processing method is to design an excellent circuit to reduce the generation of errors, which is relatively complex. This paper proposes a software method based on the orthogonal distance regression algorithm to compensate for the generated errors. Compared to error compensation implemented through hardware circuits, the implementation via software algorithms is significantly more convenient. Moreover, the signal error post software compensation is substantially reduced, achieving precision at the fs scale. At the same time, by studying the orthogonal distance regression (ODR) algorithm in comparison with the ordinary least squares (OLS) method and the moving least squares (MLS) method, it was concluded that the ODR algorithm exhibits the best performance, with a relative error of less than 0.01. By optimizing the ODR algorithm from nonlinear to linear, the computation was significantly reduced. The underlying causes of significant parameter errors in the algorithm were systematically analysed to offer insights and guidance on its applicability.