Experimental Feature-Based Diagnosis of Switching Transients and Phase Loss in Low-Power Induction Motors

Abstract

Low-power induction motors used in pumps, conveyors, ventilation units, agricultural machinery, and auxiliary industrial drives operate under frequent switching, long feeder cables, supply asymmetry, and occasional phase loss, all of which accelerate insulation stress and reliability degradation. Switching transients and phase-loss events are usually handled as separate protection problems, yet their electrical responses also carry quantifiable diagnostic information. This study presents an original experimental basis and a reliability-oriented, feature-based diagnostic framework that unifies both phenomena for low-power induction motor drives. Approximately 7000 switching-overvoltage oscillograms recorded on 4A and AIR series motors rated up to 3 kW were processed into diagnostic features, namely overvoltage multiplicity, impulse number, front duration, total transient duration, and waveform class, and their statistical distributions were identified. The maximum de-energizing overvoltage multiplicity increased from 4.02 to 7.61 when the feeder-cable length increased from 4 to 80 m. A two-level three-factor experiment further quantified the effects of starter size, protective capacitance, and feeder length and enabled design-oriented capacitance selection. Phase-loss operation was derived in closed form using symmetrical components. Unlike studies that treat switching stress, motor-fault signals, or single phasing separately, the proposed method integrates experimentally quantified transient features, a protection-design model, and analytical phase-loss severity within one four-level decision framework with an emergency override.