Motor Back-EMF Testing and Waveform Analysis: Constant Measurement, Harmonic Decomposition, and Commutation Optimisation

Back-EMF: The Motor’s Electrical Fingerprint

When a motor shaft rotates, the moving magnetic field induces a voltage in the stator windings even with no current flowing. This is the back-electromotive force (back-EMF), and it is the most information-dense single measurement available from a rotating electrical machine. The amplitude reveals the flux-linkage per pole pair and therefore the motor’s ability to produce torque per amp. The waveform shape determines whether the machine is best driven by sinusoidal or trapezoidal currents. The harmonic content predicts torque ripple, acoustic noise, and commutation error. And any asymmetry between phases—a deviation in amplitude or phase shift from the ideal 120° spacing—flags manufacturing defects in winding turns, magnet placement, or magnetization quality before the motor ever enters service.

Back-EMF measurement is therefore performed at three points in a motor’s lifecycle: during initial design validation (characterise the design), during pre-production samples (gate production), and on the end-of-line production test bench (screen every unit). Each context has different measurement speed, accuracy, and interpretive requirements.

What Back-EMF Testing Reveals

The back-EMF test bench drives the motor under test (no current in its windings) at a controlled speed using a prime-mover dynamometer, then captures all three line-to-neutral voltages simultaneously at high sample rate. From this single capture, several properties can be extracted:

• Back-EMF constant (kE): RMS or peak-to-neutral voltage divided by shaft angular velocity (V/(rad/s) or V/krpm). Directly proportional to the torque constant kT = kE for SI units; deviation from design target indicates wrong magnet grade, turn count errors, or incorrect airgap.
• Waveform shape: Sinusoidal (PMSM with distributed winding), trapezoidal (BLDC with concentrated winding), or intermediate. Shape determines the optimal drive current profile.
• Total harmonic distortion (THD): Dominated by 5th and 7th harmonics for surface-mount PMSM; 3rd harmonic for some fractional-slot concentrated-winding machines; 6th harmonic (referred to the fundamental) for machines with skewed slots.
• Phase balance: Amplitude ratio and phase angle between UA, UB, UC; should be 1.0 and 120° for a balanced three-phase machine.
• Zero-crossing position: Critical for BLDC commutation timing; zero-crossing of the phase back-EMF occurs 30° before the commutation instant in a 6-step trapezoidal drive.

Core Test Items

1. Back-EMF Constant Measurement (kE)

The motor is spun at a defined test speed (typically 1,000 or 3,000 rpm) and the RMS or peak line-to-neutral voltage is measured with a precision power analyser or high-resolution oscilloscope (12-bit or higher). kE is computed as Urms / ω (where ω is the electrical angular velocity in rad/s). Multiple speeds are tested to verify linearity (kE should be constant with speed for a linear machine; deviation indicates eddy-current or saturation effects at high speed). The result is compared to the design specification and to the batch mean from previous units.

Key metrics: kE within ±2% of specification for production acceptance; variation coefficient (σ/μ) < 0.5% within a production batch.

2. Harmonic Decomposition and THD

A fast Fourier transform (FFT) of the back-EMF waveform decomposes it into fundamental and harmonics. For a PMSM driven by sinusoidal current, the torque ripple amplitude is approximately proportional to the product of harmonic order and harmonic amplitude, so a 5th harmonic at 3% of fundamental contributes roughly 15% torque ripple at 6× the fundamental frequency (6th harmonic of the torque). The test bench characterises each harmonic up to at least the 20th order and documents which harmonics fall above the acceptance threshold.

Key metrics (typical PMSM): 5th harmonic < 3%; 7th harmonic < 2%; THD < 5%.

3. Phase Balance and Asymmetry Detection

Three channels of the data acquisition system simultaneously capture UA, UB, UC at 1 MHz or higher. Amplitude asymmetry (one phase higher or lower than the other two) flags missing turns, shorted turns, or wrong magnet polarity. Phase-angle asymmetry (not exactly 120° apart) indicates mechanical issues—rotor eccentricity, incorrect magnetization sector alignment, or wrong winding pitch. The test bench reports both amplitude ratio (UA/UB, UB/UC, UC/UA—all should be 1.0 ± 1%) and phase angle deviations (should be 120° ± 1°).

4. Speed-Dependent Back-EMF Linearity

The bench sweeps speed from 10% to 100% of rated speed in steps and plots kE versus speed. An ideal machine has flat kE across speed; eddy-current losses in the magnet or rotor iron cause kE to appear to decrease at high speed (the flux effectively “leaks” into eddy currents). Machines intended for field-weakening operation are characterised at speeds above the base speed, where back-EMF should not exceed the supply voltage limit at which the inverter loses current control.

5. BLDC Commutation Timing Optimisation

For trapezoidal BLDC motors, the zero-crossing of the unenergised phase back-EMF is the reference signal for commutation. The bench characterises zero-crossing position versus load and speed to determine the optimal advance angle (typically 0–30° electrical). A commutation advance that is too early causes the incoming phase to fight against the back-EMF; too late causes torque dips and efficiency loss. The bench runs the motor at multiple load points, sweeps advance angle, and measures efficiency and torque ripple to find the optimum.

Key metrics: efficiency improvement from advance angle optimisation: 2–8% depending on speed; torque ripple reduction: 30–50% at optimised advance vs. zero-crossings commutation.

6. Temperature Coefficient of kE

For permanent magnet machines, kE decreases with increasing magnet temperature due to the negative temperature coefficient of Br (typically −0.11%/°C for NdFeB). The bench measures kE at 20°C, 60°C, 80°C, 100°C, and 120°C (using a thermal chamber or winding-resistance-based temperature estimation). The measured coefficient is compared to the magnet data sheet; a coefficient significantly more negative than expected indicates a lower-grade magnet than specified. This test is also used to calibrate thermal models used in the motor controller for flux observer accuracy.

Production End-of-Line Back-EMF Testing

On the production line, back-EMF measurement must complete in under 5 seconds per unit. The end-of-line bench spins the motor to a fixed test speed (e.g., 1,000 rpm) in 1–2 seconds, captures a 50–200 ms window of back-EMF data, and computes kE and phase balance. Units outside the tolerance band are flagged for rework or scrap. The measurement uncertainty of the production bench (including speed uncertainty, voltage measurement accuracy, and data processing latency) must be small relative to the acceptance tolerance—for ±2% tolerance on kE, the bench measurement uncertainty should be < 0.3%.

What This Means for Test Bench Selection

Back-EMF measurement equipment ranges from a simple oscilloscope and a tabletop dynamometer for early R&D work to a fully integrated end-of-line tester with automated pass/fail reporting and statistical process control (SPC) data export. The critical specifications are measurement bandwidth (must capture harmonics up to at least the 20th order), simultaneous three-channel capture (to accurately compute phase angle), speed measurement accuracy (uncertainty in ω directly multiplies into uncertainty in kE), and—for production—cycle time and data traceability.

Our motor test systems include back-EMF characterisation as a standard test item in both R&D and production configurations. For motor designs where back-EMF quality is a primary design gate, we can configure dedicated characterisation benches with analysis software tailored to your motor topology. To discuss your requirements, talk to our engineering team.