Robust estimation of magnetic Barkhausen noise based on a numerical approach

This paper deals with a multifeature numerical instrument for nondestructive testing of ferromagnetic materials, based on the Barkhausen noise. Starting from traditional instruments based on the root mean square value of the band-pass filtered measured signal, the proposed method is based on the simultaneous analysis of the excitation voltage, the excitation current, and the induced signal. A good repeatability improvement is achieved by monitoring the excitation impedance, while synchronizing the rms evaluation to the excitation current signal increases the robustness with respect to the environmental electromagnetic noise. Moreover, experimental results encourage the use of shape-based features to better characterize the Barkhausen signal.