Comparative Analysis of High Frequencies for the Broken Bar Fault Diagnosis Using MCSA and Park’s Vector Demodulation

Manufacturing companies intend to find reliability in their processes and procedures, reasons for which faults must be predicted in order to generate actions in a timely manner so that the operation of the electric machine is reliable. The present work shows a procedure for detecting faults in bars of a three-phase induction motor, based on upper lateral bands near the fundamental and up to 900 [Hz], focused on the technique of the Park transformation, a method that demodulates the currents resulting in the associated components of torque and magnetization which will be analyzed by Fast Fourier Transform (FFT). The spectra near the fundamental of the current presented an average frequency coincidence with respect to the theoretical equation of 98.6% for bands 1× and 7×, and an average frequency coincidence far from the fundamental in 99.4% in bands 10×, 17×, 20×, and 77×. The tests were performed using a laboratory test bench for electric motors where it is observed that the Park transform provides wide and clean results over the entire frequency range, eliminating the uncertainties and/or errors that may arise in the event of a diagnosis of broken bar fault compared to MCSA results.