E-Bike Hub Motors Don’t Have to Be Loud—We Proved It

Walk through any city bike lane in Amsterdam, Berlin, or San Francisco, and you’ll hear it—that persistent, high-pitched whine from e-bike hub motors. It’s not loud enough to violate noise regulations, but it’s there. Always there.

At TO7 Motor in Suzhou, we started hearing the same feedback from OEM partners and riders: “Can you make it quieter?”

The challenge is this: Direct-drive hub motors are already among the quietest propulsion systems in urban mobility. We eliminated gears years ago, which removed the primary source of mechanical noise. Yet that residual acoustic signature—what acoustic engineers call the “electromagnetic harmonic signature”—persists.

Here’s what we discovered early in our simulations: A hub motor casing doesn’t just sit there passively.

It behaves like a cylindrical shell—essentially, a very precisely machined bell.

When electromagnetic forces from the rotating magnetic field interact with the stator (the stationary part of the motor), they create what’s called Maxwell stress—tiny pressure waves that ripple through the metal structure.

Even though the rotation itself is smooth, these forces contain “harmonics”—repeating patterns that occur at specific frequencies.

Think of it like plucking a guitar string.

You hear the fundamental note, but there are also overtones—higher frequencies that color the sound. In a hub motor, these electromagnetic harmonics can excite specific vibration patterns in the motor housing.

What does that mean in plain English?

This explained why simply adding more damping material (foam, rubber mounts) wasn’t enough. We weren’t just dealing with vibration amplitude—we were dealing with modal radiation efficiency.

Some vibration patterns radiate sound far more effectively than others, even when the physical movement is tiny.

Using COMSOL Multiphysics and boundary-element acoustic modeling, we ran hundreds of simulations to determine which design parameters actually matter for noise reduction.

We tested:

But here’s the catch: These are simulation results under ideal conditions. The predicted baseline in a perfectly isolated, anechoic environment was 38 dB at 300 RPM. In real-world applications—with mounting systems, frame coupling, and manufacturing tolerances—we expect to add 8-15 dB to that figure.

That means realistic operational levels would be in the 46-53 dB range, which is still competitive with the best direct-drive motors on the market but not the dramatic breakthrough the raw simulation suggests.

One of the most useful findings came from our harmonic force decomposition analysis. Not all electromagnetic harmonics contribute equally to noise.

We found that the 6th and 12th temporal harmonics were the dominant contributors, aligning with structural resonance frequencies in the 600-1200 Hz range.

Meanwhile, lower harmonics (2nd and 4th) produced negligible acoustic radiation because their corresponding mode shapes had low radiation efficiency.

Here’s what the testing protocol looks like:

Test Motor:

TO7 Motor ZM05-175DL —our 3000W-rated, 4000W peak direct-drive hub motor with 120 Nm torque and 82% efficiency.

Structural Measurements:

Tri-axial accelerometers mounted at multiple locations on the motor housing. We’ll use impact-hammer testing to measure frequency response functions (FRFs) and compare the measured natural frequencies with our simulated values.

Acoustic Measurements:

Free-field microphone array in a semi-anechoic chamber, measuring sound pressure levels at ten points around the motor following ISO 3744 protocols. Background noise will be subtracted to ensure ±1 dB accuracy.

Signal Processing:

Short-time Fourier transform (STFT) to extract order-tracked spectra—basically, separating motor-generated noise from environmental noise.

In shared urban spaces, a quieter motor means:

For OEM partners, lower acoustic emissions mean:

What’s Next: Our 12-Month Development Roadmap