I Engineered a 2mm Micro-Bearing D20 Ring That Free-Spin for 20 Seconds

Over the past few months I’ve been experimenting with how small a functional bearing-based mechanism can get while still feeling smooth, durable, and useful. This project started from a simple question: How thin can a real, free-spinning bearing be while still handling continuous rotation?

Most “spinner rings” you see online cheat by simulating rotation — there’s no real bearing, just loose tolerance metal sliding on metal. True micro-bearing rotation needs precision, tight tolerances, and high surface finish, which is difficult when everything needs to be wearable on a finger.

So I tried to push it in the opposite direction and ended up making this:

A 2mm-thick stainless steel ring with an internal micro-bearing track and 20 steel balls that free-spin for 20+ seconds with a single flick.

Mechanical details: • CNC machined inner race with ~0.01mm tolerance • 20 micro steel balls loaded through a lateral channel • Outer ring pressed onto the bearing shell • No plastic, no bushings, no lubricant • Built to withstand everyday wearing forces (compression, torsion, micro-impacts) • PVD variant for color durability • Outer surface can be marked 1–20, turning it into a tiny randomizer

Originally the goal was purely mechanical — to see if a bearing this thin could be made. But it ended up being surprisingly functional for solo tabletop RPG use: when you need a quick random result but don’t have table space, or when dice are too loud (playing in bed, on a commute, etc.). The ring spins silently and lands pointing at a single number.

This wasn’t meant to replace dice; it just became a neat side effect of the engineering challenge.

Why I’m posting here

HN tends to appreciate: • micro-manufacturing • tolerances • machining challenges • precision mechanical design • unusual “why does this work?” projects

I’d love to hear feedback on: • improving durability • minimizing friction losses • alternative ball materials • raceway finishing • any tricks for increasing spin time without adding thickness

If anyone has experience with miniature bearings or wearable mechanical assemblies, I’d appreciate insights. Happy to answer questions about the build process, the tolerances, or the failures along the way.