For a good while now, I’ve been wanting to try and build my own linear motors, and for an upcoming motion system build, I finally want to give it a go.

But before I baseline my design around them, I need to prove to myself that I can make consistently make decent-quality coils for those linear motors. Don’t get me wrong, I’m not fooling myself into thinking I’ll be making beauts like these commercial ones, but ensuring all of the coils in the drive are consistent will make the controls much easier and give better performance.

Soooo, of course my first instinct was “build a helper”.

This new helper had a handful of requirements:

• Rotate a mandrel or mandrels to wrap magnet wire into coil
• Provide bounding surfaces to control the outer geometry of resulting coil
  • Bounding surfaces must ‘release’ from standard CA adhesive
  • Thickness repeatability ≤ 250µm (give or take….this is a basement build)
  • Thickness parallelism ≤ 500µm (samesies)
• Maintain consistent wire tension throughout winding (~5N - 10N)
• Control winding count repeatably to better than half a turn
  • Can check this by comparing resistance (if my DMM has the resolution for this)

Ok…so did I have that written list when I jumped into Fusion and started pilin up pixels…no. But those all were generally the notions I had in mind as I worked on it. So I stand by it…as if it matters :)

After building my first iteration and testing it out a bit, I also added an additional requirement. The full winding assembly needed to be able to rotate 90 degrees. So the spinning axis can be either parallel to gravity or normal to gravity. That gives easier access for applying glue as I’m winding, and allows excess glue to drip out instead of settling on the mandrel surface.

Design

At the core of the system there are two main sub-systems:

• One for unwinding - Holds the spool of magnet wire and provides and axis of rotation for unspooling it
• One for winding - Provides the drive into the system and supports the mandrels that define the coil shape

In between those two sections sits a wire guide printed into the MainBody (the big green part in the above image). That guide is intended to accept wire from either end of the spool without introducing too sharp of a bending radius on the wire. The below cross-section shows an example wire path.