Torque ripple is the uneven torque an electric motor produces as its rotor spins. Instead of delivering a perfectly smooth, constant force, the motor’s output fluctuates slightly with each rotation, creating a periodic rise and fall in torque. These fluctuations stem from imperfections in the interaction between the motor’s electromagnetic fields, and they show up in every type of electric motor to some degree.
How Torque Ripple Works
An ideal electric motor would produce the exact same rotational force at every point during a revolution. In reality, the electromagnetic fields generated by the rotor (the spinning part) and stator (the stationary part) don’t interact with perfect uniformity. As the rotor turns, the magnetic flux shifts and introduces small harmonic distortions into the motor’s torque output. The result is a waveform that oscillates above and below the motor’s average torque, sometimes by a few percent, sometimes more.
One major contributor in motors with permanent magnets is something called armature reaction. As current flowing through the motor windings increases, it distorts the magnetic field in the motor’s core. This distortion changes the relationship between current and torque at different rotor positions, producing peaks and valleys in the output. Even slotless motors, which are designed to minimize magnetic interference, still exhibit torque ripple from this effect.
Torque Ripple vs. Cogging Torque
These two terms often get mixed up, but they describe different phenomena. Cogging torque happens when a motor is unpowered. It’s the “notchy” or jerking sensation you feel when you rotate the shaft of a brushless motor by hand. It comes from the attraction between the permanent magnets on the rotor and the steel teeth of the stator. Once you power the motor on, cogging torque as a standalone force effectively disappears into the broader torque ripple.
Torque ripple, by contrast, only exists when the motor is energized and running. It includes the effects of cogging but also adds in all the electromagnetic interactions that come with current flowing through the windings. So cogging torque is one ingredient in the torque ripple recipe, but far from the only one.
Why It Matters
At high motor speeds, torque ripple fluctuations happen so quickly that connected loads often smooth them out through inertia. The problem gets worse at low speeds, where the oscillations are slow enough to directly affect the driven load. A motor turning a precision stage at a few revolutions per minute, for example, will transmit those torque variations straight to the workpiece.
The consequences fall into a few categories:
- Noise and vibration. The high-frequency components of torque ripple excite structural resonances in the drivetrain, producing audible tonal noise and measurable vibrations. In electric vehicles, this is a significant concern because there’s no combustion engine noise to mask the drivetrain sounds. A noisy motor creates a perception of poor quality, and tonal whine is particularly unwanted.
- Gear and bearing wear. Repeated torque oscillations act like tiny cyclic loads on gears, bearings, and couplings. Over thousands of hours, this accelerates fatigue and shortens component life.
- Control accuracy. In robotics and CNC machining, torque ripple introduces position and velocity errors. Each ripple cycle nudges the motor slightly ahead of or behind its intended position. In one study of a robotic joint, gear-related torque ripple produced position-dependent force fluctuations of about ±0.2 Nm, enough to affect precision tasks.
How Motor Type Affects Ripple
Not all motors produce the same amount of torque ripple. Brushless DC motors, which use trapezoidal current waveforms, tend to have higher ripple because the sharp transitions between phases create more abrupt torque changes. Permanent magnet synchronous motors use smoother sinusoidal current waveforms and generally produce lower torque and current ripple as a result.
Switched reluctance motors, which work by sequentially energizing stator poles, are particularly prone to high torque ripple because torque production is inherently pulsed. On the other end of the spectrum, slotless brushless motors eliminate the stator teeth entirely, removing cogging as a ripple source, though armature reaction effects remain.
How Engineers Reduce It
Reducing torque ripple involves changes to both the motor hardware and the electronic control system.
Motor Design Changes
Skewing is one of the most common hardware approaches. Instead of aligning the stator slots (or rotor magnets) straight along the motor’s axis, engineers tilt them at a slight angle. This spreads each magnetic interaction across a wider range of rotor positions, averaging out the torque peaks and valleys. The tradeoff is a small reduction in overall torque output.
Choosing the right combination of stator slots and rotor magnet poles also makes a significant difference. Certain slot-pole ratios produce lower harmonic content in the torque waveform. Shaping the magnets themselves, using designs with tapered or bread-loaf cross-sections instead of uniform blocks, smooths out the magnetic field distribution and reduces the harmonics that drive ripple.
Control-Side Approaches
On the electronics side, advanced motor controllers can inject compensating currents that counteract known ripple patterns. If the controller knows the motor’s torque profile at each rotor position, it can adjust the current waveform in real time to flatten the output. This requires accurate position sensing and a controller fast enough to respond at the ripple frequency.
Field-oriented control, which is standard in most modern motor drives, inherently produces smoother torque than older six-step commutation methods. Adding harmonic injection on top of field-oriented control can push ripple even lower for demanding applications.
Torque Ripple in Electric Vehicles
Electric vehicle drivetrains put torque ripple under particular scrutiny. The motor connects to the wheels through a gear reduction, and ripple harmonics can excite resonances in the gearbox housing, driveshafts, and vehicle body. Research from the Institution of Engineering and Technology describes how the dominant low-frequency excitation in one EV drivetrain model came from the 36th harmonic of motor rotation, a combined effect of radial magnetic forces and torque ripple acting on the stator and rotor.
For a three-phase motor, the fundamental torque ripple frequency from phase commutation is six times the electrical frequency (twice the number of phases, times the electrical frequency). As motor speed increases, this ripple frequency climbs into the range where human ears are most sensitive, typically a few hundred to a few thousand hertz. EV engineers use a combination of motor design optimization and active noise cancellation to keep cabin noise acceptable.
How Torque Ripple Is Measured
Measuring torque ripple requires a sensor fast enough to capture the rapid oscillations, not just the average torque. Engineers typically mount a high-bandwidth torque transducer between the motor and its load, then record the real-time torque signal as the motor runs at a controlled speed. A large inertia disk is often placed between the motor and load to filter out speed fluctuations, isolating the torque variations from mechanical disturbances.
The ripple is usually expressed as a percentage of the motor’s rated or average torque. A common formula compares the peak-to-peak variation (highest torque minus lowest torque during one electrical cycle) to the average torque. A motor with 5% torque ripple, for instance, means the output swings by 5% of its average value above and below the mean. For precision applications like semiconductor manufacturing or surgical robotics, engineers push for ripple well below 1%. General-purpose industrial motors might tolerate 5% to 10% without issue.