Mechanical Time Constant Test Procedures: A Step-by-Step Experimental Guide for Electric Motors

The mechanical time constant is one of the most important dynamic performance parameters of an electric motor, defining how quickly the rotor accelerates from standstill toward its steady-state speed under an applied voltage. Accurate measurement of this parameter is essential for servo system design, control loop tuning, and motor quality verification in accordance with international standards such as IEC 60034 and GB/T 755. This article presents a complete, professional experimental procedure for determining the mechanical time constant, covering test samples, instrumentation, direct and indirect measurement methods, data processing, and frequently asked questions.

Technical Overview

The mechanical time constant, commonly denoted as tau-m, is defined as the time required for a motor shaft to reach 63.2 percent of its final steady-state no-load speed after a step voltage is applied to the terminals, starting from standstill. The value of 63.2 percent corresponds to the quantity (1 minus 1/e) that characterizes a first-order exponential system. For a brushed or brushless DC machine operated within its linear region, the speed response can be approximated by the first-order equation: n(t) equals n0 multiplied by (1 minus e raised to the power of minus t divided by tau-m), where n0 is the steady-state no-load speed.

In analytical form, the mechanical time constant is expressed as tau-m equals J multiplied by R, divided by the product of Ke and Kt, where J is the rotor moment of inertia in kg·m², R is the terminal winding resistance in ohms, Ke is the back-EMF constant in V·s/rad, and Kt is the torque constant in N·m/A. In consistent SI units, Ke and Kt are numerically identical for an ideal DC machine. A small mechanical time constant indicates a motor with high acceleration capability and wide achievable control bandwidth, which is critical in servo drives, scanning galvanometers, and high-dynamic positioning systems.

The mechanical time constant must be distinguished from the electrical time constant, tau-e, which equals the winding inductance divided by the resistance and describes how quickly the current rises in the armature. For the simple first-order extraction of tau-m to be valid, tau-e should be at least an order of magnitude smaller than tau-m. When the two time constants approach each other, a second-order model fit must be used. Test programs are typically structured in line with the general principles of IEC 60034-1 for rotating electrical machines, GB/T 755, and applicable micro-motor specifications such as GB/T 7345, and are commonly performed by accredited third-party laboratories or in-house motor test facilities.

Testing Samples

Brushed DC and Coreless Motors

Coreless (ironless-rotor) DC motors exhibit very low inductance and small rotor inertia, making them ideal candidates for direct step-response measurement of tau-m. Conventional brushed DC motors can also be tested directly, although brush contact noise and commutation ripple must be filtered appropriately during acquisition.

Brushless DC and Permanent Magnet Synchronous Motors

For BLDC and PMSM samples, the step-response test is performed under closed-loop commutation with a servo drive configured in a transparent, unity-gain voltage mode. The drive must not introduce its own current or speed loop dynamics that would mask the natural mechanical response of the motor.

Servo Motors and Direct-Drive Torque Motors

Integrated servo motors with encoders allow direct use of the built-in feedback device for speed measurement. Frameless torque motors with large inertia and low winding resistance typically exhibit longer mechanical time constants, requiring longer acquisition windows and highly stable voltage sources.

Specialty Actuators

Galvanometer scanner motors, voice coil actuators, gimbal motors for unmanned aerial vehicles, and limited-angle torque motors are also tested for mechanical time constant, because their positioning bandwidth is directly governed by this parameter. For limited-angle devices, the equivalent time constant is extracted from the angular velocity transient rather than continuous rotation.

Testing Items

A complete mechanical time constant characterization typically includes the following measured and derived quantities:

• Mechanical time constant tau-m, determined by direct step response or by indirect calculation
• Winding terminal resistance R, measured with the four-wire method and corrected to the reference temperature
• No-load speed n0 and no-load current at rated voltage
• Back-EMF constant Ke, from the driven-generator method at controlled speed
• Torque constant Kt, from locked-rotor torque measurement or from Ke in SI units
• Rotor moment of inertia J, from the added-inertia method or torsional oscillation test
• Electrical time constant tau-e, from winding inductance and resistance
• Friction and windage torque, from the coast-down retardation test
• Ambient and winding temperature at the time of measurement

Testing Methods

Method 1: Direct Step-Response Measurement

A step of rated voltage is applied to the motor at standstill through a low-impedance solid-state switch, and the instantaneous shaft speed is recorded with a high-resolution optical encoder and a counter-timer or data acquisition system. The time at which the speed first reaches 63.2 percent of the steady-state no-load speed is identified directly from the trace. This is the preferred method because it captures the real, assembled behavior of the motor including friction, cogging, and damping effects.

Method 2: Indirect Parameter-Based Calculation

The mechanical time constant is computed from independently measured parameters using the relation tau-m equals J times R divided by (Ke times Kt). This method is useful when the speed transducer bandwidth is insufficient for very fast motors, or when the motor cannot be step-excited safely. Its accuracy depends on the combined uncertainty of the inertia, resistance, and constant measurements.

Method 3: Coast-Down Retardation Method

The motor is accelerated to steady no-load speed, the supply is disconnected, and the free deceleration of the shaft is recorded. The retardation curve yields the friction and windage torque and provides a cross-check of the inertia value used in the indirect calculation. This method also validates the damping assumption underlying the first-order model.

Step-by-Step Experimental Procedure

Step 1: Sample Preparation and Conditioning

• Perform a visual and dimensional inspection of the motor, verify the nameplate data, and record the serial number and test conditions.
• Mount the motor rigidly on a vibration-free test fixture with the shaft axis horizontal and the encoder or tachometer aligned to within the manufacturer's specified concentricity tolerance.
• Condition the sample in the test environment at 23 plus or minus 5 degrees Celsius per the ambient requirements of IEC 60034-1, and perform a run-in at rated voltage for the duration recommended by the motor specification to stabilize brush contact and bearing friction.
• Allow the motor to cool to ambient temperature before the transient test so that the winding resistance corresponds to a known reference temperature.

Step 2: Equipment Setup and Calibration

• Use a calibrated low-noise DC power supply with output impedance low enough that terminal voltage droop during the inrush current is less than 1 percent of the step value.
• Employ an optical incremental encoder with at least 1000 counts per revolution, or a calibrated DC tachometer generator, connected to a counter-timer or data acquisition device with a sampling rate at least ten times the reciprocal of the expected time constant.
• Calibrate the speed measurement chain against a reference tachometer and record the calibration status of all instruments, including the four-wire micro-ohmmeter, the digital storage oscilloscope, and the temperature sensors.
• Verify that the solid-state switch used to apply the step voltage has a rise time shorter than one tenth of the electrical time constant of the motor.

Step 3: Determining the Reference No-Load Speed

Apply the rated voltage to the motor in the continuous mode and record the steady-state no-load speed n0 and the no-load current after thermal stabilization. Repeat the measurement in both directions of rotation where applicable, and confirm that the ripple in the steady-state speed trace is within acceptable limits. The value of n0 obtained in this step defines the 63.2 percent threshold used in the transient analysis.

Step 4: Capturing the Speed Step Response

• Confirm that the rotor is completely at rest and that the winding temperature is at the reference ambient value.
• Trigger the solid-state switch to apply the full rated voltage in a single clean step while simultaneously starting the acquisition of the encoder speed signal.
• Record the complete acceleration trace from standstill until the speed has settled within 2 percent of n0, ensuring that the terminal voltage is monitored in parallel to confirm the absence of supply droop.
• Repeat the step test a minimum of five times, allowing the motor to cool and return to standstill between runs, and randomize the starting rotor position to average out cogging and detent effects.

Step 5: Data Extraction and Calculation

• Filter the raw speed trace with a low-pass filter whose cutoff is at least ten times the expected dynamic bandwidth, taking care not to introduce phase lag into the 63.2 percent crossing region.
• Identify the time t63 at which the speed first equals 0.632 multiplied by n0; this value is the directly measured mechanical time constant for that run.
• As a verification, linearize the response by plotting the natural logarithm of (1 minus n divided by n0) against time; the slope of the fitted line should equal minus 1 divided by tau-m, and the fit residual confirms the validity of the first-order model.
• Compute the arithmetic mean and the standard deviation of the valid runs and report the coefficient of variation as part of the measurement uncertainty budget.
• If the electrical time constant is not negligible compared with tau-m, fit a second-order transfer function to the trace instead and report the dominant pole as the effective mechanical time constant.

Step 6: Indirect Determination Cross-Check

Measure the terminal resistance with the four-wire method at the reference temperature and correct the reading using the copper temperature coefficient of 0.00393 per degree Celsius referenced to 20 degrees Celsius. Determine the back-EMF constant by driving the motor at a controlled constant speed with an external prime mover and measuring the open-circuit terminal voltage, then compute Kt from Ke in SI units. Determine the rotor inertia by the added-inertia method, in which the acceleration response is measured with and without a calibrated flywheel of known inertia, or by the torsional oscillation method. Calculate tau-m from the formula tau-m equals J times R divided by (Ke times Kt) and compare with the direct measurement; agreement within 10 percent is generally considered satisfactory.

Step 7: Reporting

The test report shall state the motor identification, the applicable standard references, the instrument list with calibration traceability, ambient conditions, the raw speed-time traces, the value of n0, the individual and averaged values of tau-m, the indirect calculation with all constituent parameters, the uncertainty evaluation, and any deviations from the standard procedure together with their technical justification.

Safety and Precautions

• Guard all rotating parts and ensure that couplings and flywheels are mechanically secured before energizing the motor.
• Limit the number of repeated step tests to avoid overheating the winding, and monitor winding temperature throughout the sequence.
• Discharge any capacitive energy in the supply and verify zero energy state before changing the test configuration.
• Ensure that braking or coast-down tests are performed in a controlled area, since high-inertia rotors store significant kinetic energy.

Uncertainty and Error Sources

Principal contributors to measurement uncertainty include encoder quantization at very short time constants, temperature drift of the winding resistance between runs, voltage source droop during inrush, misalignment of the speed transducer, and cogging-related variation with rotor starting position. Best practice requires controlling the ambient temperature, using four-wire resistance measurement, averaging over multiple starting positions, and cross-validating the direct result with the indirect parameter calculation. Typical expanded uncertainties of 3 to 8 percent are achievable in well-controlled laboratory conditions.

Applications

• Servo drive selection and control loop tuning in CNC machine tools and industrial automation, where the mechanical time constant sets the achievable velocity bandwidth
• Galvanometer optical scanners for laser marking, welding, and medical imaging, where scan speed is limited by rotor dynamics
• Robotics joint actuators and humanoid drive units requiring predictable acceleration response
• Aerospace control surface actuators, thrust vector mechanisms, and antenna positioning systems governed by stringent dynamic specifications
• Unmanned aerial vehicle gimbal stabilization motors requiring millisecond-level response
• Medical devices such as surgical handpieces, infusion pumps, and laboratory automation
• Motion control component qualification, incoming inspection, and failure analysis in accredited third-party laboratories and manufacturer test centers

FAQ

Why is the mechanical time constant defined at 63.2 percent of final speed?

Because the acceleration of an ideal DC motor follows a first-order exponential response, the speed reaches exactly (1 minus 1/e), or 63.2 percent, of its final value after one time constant. This single-point criterion provides a physically meaningful and reproducible measure of the system dynamics.

How is the mechanical time constant different from the electrical time constant?

The electrical time constant, equal to winding inductance divided by resistance, describes how fast the current rises in the motor, while the mechanical time constant describes how fast the shaft speed builds up against the rotor inertia. The mechanical time constant is typically one to two orders of magnitude longer than the electrical time constant in well-designed servo motors.

What is the relationship between the mechanical time constant and servo bandwidth?

As a first-order approximation, the velocity loop crossover frequency cannot substantially exceed 1 divided by (2 pi times tau-m). A motor with a mechanical time constant of 5 milliseconds therefore supports a fundamental dynamic ceiling in the range of roughly 30 Hz, before additional margins for phase lag and disturbances are applied.

Can the mechanical time constant be measured on a motor already installed in a machine?

Yes, provided the coupled load inertia is known and included in the analysis. The measured value then represents the combined drive train time constant rather than that of the bare motor, and the report must clearly state the loaded condition.

Which standards govern this test?

General test conditions follow IEC 60034-1 and its national adoptions such as GB/T 755, with micro-motor specific requirements addressed by GB/T 7345. Where contractual or industry-specific procedures apply, they should be referenced in the test plan agreed between the manufacturer, the customer, and the accredited third-party laboratory performing the measurement.

What sample size and repetition are recommended?

A minimum of five valid step-response runs per motor, with randomized starting rotor positions, is recommended. For production qualification programs, additional samples across manufacturing lots should be tested to characterize the statistical distribution of the mechanical time constant.