In this paper, three different types of acoustic emission sensors have been applied to detect the outer race defect of angular contact ball bearings. The sensors include R6a, which is a general-purpose, WSa, wideb-and sensor and Pico, miniature sensor. All three sensors differ in size and operational frequency range and sensitivity. In this research, using signal analysis methods, different features and parameters in time-domain have been extracted from the signal and different four states (healthy and three faulty states) have been identified based on them. An artificial neural network used for separating, detection and classification of defect sizes. In the first comparison of the signals, the results showed that R6a was better for absorbing acoustic emission energies and for data acquisition and processing than the other two sensors. But by creating the neural network and selecting different parameters, therefore all three sensors could be useful in diagnosing the faults. (C) 2020 Published by Elsevier Ltd.
