This dataset provides multi-domain monitoring data from a motor-driven electromechanical drivetrain operating under different health conditions and operating regimes. The experimental platform is instrumented to acquire three-phase stator currents, tri-axial vibration measurements at the motor and gearbox locations, and resolver-based rotational speed feedback provided by the drive. The dataset includes healthy operation, single-fault conditions affecting the bearing, rotor permanent magnet demagnetization and gearbox teeth, as well as combined fault scenarios involving gearbox degradation together with either bearing damage or demagnetization. Data were acquired under two load conditions and several operating profiles, including stationary speed points and non-stationary speed transitions, in order to represent operating variability commonly encountered in industrial actuation systems. By combining synchronized current, vibration and speed measurements with different load levels, stationary and non-stationary regimes, and both single and combined fault scenarios, the dataset provides a reusable benchmark for the development and validation of robust condition monitoring, fault diagnosis, signal processing and machine learning methods under changing operating conditions.