Upgrading My DIY BLDC Winding Machine to Handle 4kg E-Scooter Stators

I’m currently pushing my DIY winding machine to its limits. The goal: scale up from small drone motors to the heavy-duty stators found in e-scooters.

Compared to standard drone motors, these stators are significantly thicker and heavier. My previous version handled 41mm to 83mm diameters with ease, but to move these beasts, we need some serious mechanical upgrades.

To handle the increased payload and torque, I’m swapping out the old T8 lead screw for a ball screw linear rail guide. This is the kind of hardware you usually see in industrial CNC machines, and it’s going to provide the rigidity and precision we need for these larger stators.

To get the precision I needed for this build, I sourced my custom components from JLCCNC, my go-to CNC fabrication service.

For the core of the rotation assembly. A 25mm diameter shaft will serve as the main stator axis. To ensure there is absolutely zero play, I installed it into these bearings and the timing gear using a freeze fit.

One of my biggest concerns was whether this BE4108 gimbal motor could actually handle the inertia of a 3.95 kg e-scooter stator.

To my surprise, it’s proving to be up to the task.

I’m currently running it under closed-loop angle control, with the driver-side voltage limited to just 3V.

There’s some visible wobbling—I’ll definitely need to spend some time optimizing the PID gains to smooth out the motion—but the core takeaway is that the BE4108 has more than enough torque to manage this heavy payload.

When you're winding a large stator with high torque, preventing horizontal vibration is critical for a smooth, consistent wind.

Industrial machines usually solve this by clamping the stator from the back or side using pneumatic actuators.

For a DIY build, however, pneumatics feel like overkill—they add too much cost and complexity. Instead, I want to tackle this using a purely mechanical approach with these gears.

To be honest, I haven’t finalized the entire design yet, and I'm not 100% sure if this specific gear-driven mechanism will work.

But my goal for the next iteration is to reach a minimum viable build and put this theory to the test.

If you have a project of your own that needs professional-grade parts, JLCCNC is an awesome choice. They offer high-quality fabrication, competitive pricing, and incredibly fast production times—all with the transparency you need when designing complex assemblies.

Thank you so much for reading!

For more about this project, go to the links below:

GitHub | Website | YouTube | Instagram | TikTok