We Couldn't Find the Right Bearings

So we made our own from 100 silicon carbide balls

Most of the Mikros One uses off-the-shelf components. Screws, seals, motors, plenty of suppliers make these well enough that custom versions would be engineering ego, not engineering necessity.

Bearings were different.

The Problem with Ready-Made Bearings

We needed bearings that could handle radial load (the grinding force pushing sideways on the burr carrier) while maintaining precise axial alignment (keeping the burrs perfectly parallel). Standard options fell into three categories:

Ball bearings: Too large. The slim profile we wanted wouldn't accommodate the outer race diameter.

Needle bearings: Wrong load direction. Great for radial loads, poor for maintaining axial rigidity.

Plain journal bearings: Close, but fundamentally flawed for our application.

That last one, plain journal bearings, was the most frustrating. A journal bearing is just a shaft rotating inside a sleeve with a small clearance gap. Simple. Cheap. Compact.

But that clearance gap is the problem.

Even a few thousandth of a millimetre of play allows the burr carrier to tilt slightly under load. When you're grinding light-roasted coffee, dense, hard beans that require significant force, that tilt translates to inconsistent burr spacing. One side of the burr grinds finer than the other. Your particle distribution widens. Extractions become less predictable.

We could have used journal bearings and accepted "good enough" alignment. Some hand grinders do. But "good enough" wasn't the brief.

The Solution: Custom Silicon Carbide Ball Bearings

Instead of accepting inadequate off-the-shelf options, we designed a custom dual-bearing system using 100 silicon carbide balls, 49 in the top bearing, 51 in the bottom, each 2.5mm in diameter.

Why Silicon Carbide?

Silicon carbide is one of the hardest materials available short of diamond (9.5 on the Mohs scale, compared to hardened steel at ~7-8). It's what manufacturers use for grinding wheels and cutting tools. As bearing balls, it offers:

• Extreme hardness: Near-zero wear, even under high contact stress
• High stiffness: Minimal deformation under load maintains precise geometry
• Low friction: Smooth, polished surface and inherent material properties
• Corrosion immunity: Won't rust or degrade from moisture or coffee oils
• Temperature stability: Properties don't change with grinding heat

Why Different Ball Counts? (49 vs. 51)

This wasn't arbitrary. When two bearings with the same number of balls rotate, they create resonant frequencies, audible as a hum or whine. By using 49 balls in the top bearing and 51 in the bottom, we ensure their rotational frequencies never align. The result is significantly quieter operation.

It's the same principle used in high-end machine tool spindles and precision instruments where noise control matters.

Why Different Contact Angles?

The tubular shaft experiences different types of loading at different points:

Top bearing: Primarily handles axial loads (downward grinding pressure pushing through the burr carrier). We designed this bearing with a steeper contact angle to resist axial forces efficiently.

Bottom bearing: Primarily handles radial loads (sideways forces from uneven grinding resistance and crank input). This bearing uses a shallower contact angle optimized for radial stiffness.

By optimizing each bearing for its specific load profile rather than using identical bearings throughout, we achieved better performance with less friction and wear.

The Engineering Process

We didn't arrive at silicon carbide balls immediately. Early prototypes used POM (polyoxymethylene) polymer balls. They worked, quiet, low friction, adequate for moderate use. But long-term testing revealed gradual wear under the sustained contact stress of dense light-roast grinding.

Silicon carbide solved this. Harder than steel, harder than ceramic, nearly as hard as the sapphire bearings in luxury watches. Overkill for a coffee grinder? Perhaps. But overkill in durability means the bearings outlast everything else in the assembly by decades.

Why Not Just Use Steel Balls?

Standard bearing-grade steel (52100 or similar) would work, if we accepted noise, required lubrication, potential corrosion from moisture exposure, and eventual wear. Silicon carbide eliminates all of these while delivering rigidity that steel can't match at this scale.

The trade-off? Cost and sourcing complexity. Silicon carbide balls at 2.5mm diameter aren't catalogue items. They require specialized manufacturing and careful quality control for sphericity and surface finish. But the performance was non-negotiable.

Engineering Decisions, Not Marketing Ones

We could have used journal bearings and marketed the grinder as "precision-engineered" anyway. Most companies do. Most customers would never notice the difference, until they compared grind consistency side-by-side with a grinder that actually maintains alignment.

We're not interested in "most customers won't notice." We're interested in building the best version of the thing, even when it costs more and takes longer.

100 silicon carbide balls in custom races with optimized contact angles and harmonic spacing. Because off-the-shelf wasn't good enough.