Multi-Sensor Condition-Monitoring Dataset of a Brushed DC Servo Motor

Dataset Short Description This dataset contains raw multi-sensor recordings from a brushed PM DC servo motor (3PI12.12) operated under multiple conditions and load levels. It includes four sensor modalities: armature current waveforms (BIN) vibrometer waveform audio (WAV) smartphone audio (M4A) vibrometer spot measurements (XLS) The dataset is organized into four main condition families: normal operation, loose foundation, suboptimal speed-regulator tuning, and suboptimal speed-regulator tuning with RT (current-regulator) coefficient variation. Each family is provided in two variants: without reversal (constant rotation direction) and with reversal (rotation direction reversed every 4 seconds). The files are organized by condition and sensor, with metadata in metadata.csv, and are intended for condition monitoring research such as fault classification, load estimation, and phone-vs-instrument benchmarking.