Nanomag GmbH

Technology

Two decisions determine what a pole ring can do: how it is magnetised and what it is made of. We made both differently.

The process

Five steps — one of them is the decisive one.

Machine the carrier

A stainless-steel carrier ring is machined, the bore to tolerance class H7. This establishes the mechanical reference before anything is magnetised.

Apply the SmCo layer

A thin high-performance samarium-cobalt layer is applied to the carrier. No binder, no polymer, no adhesive — the bond is fully metallic.

Measure the runout

The part's actual geometry is measured against its intended axis of rotation. Every eccentricity is then known rather than estimated.

Magnetise with runout compensation

The pole pitch is written to the axis of rotation, not to the part. The measured error is compensated during magnetisation — this is the step that does not exist elsewhere.

Verify and document

Runout deviation, field strength and pole pitch are verified and documented. You get measured values, not assurances.

The decisive step

Runout compensation

Mechanical eccentricity can never be fully machined away. Every fixture, bearing and fit carries a residual error. Conventional processes write the pole pitch to the part — and thereby carry that error into the magnetic field.

Without compensationpole pitch follows the part geometrypitch error Δφsensor signal: amplitude and phase errorWith runout compensationpole pitch written to the axis of rotationpitch constantsensor signal: clean, constant amplitudeDashed = axis of rotation · outer ticks = nominal pitch grid · schematic, deviation exaggerated

What the sensor sees

A sensor does not measure geometry — it measures field transitions over time. An uneven pole pitch therefore produces exactly what disturbs a control loop most: amplitude variation and phase error once per revolution.

  • Runout deviation ≤ 25 µm — guaranteed, not merely typical
  • This makes runout a known quantity in your error budget, not an assumption
  • At a ring radius of 25 mm that corresponds to 206 arcseconds of angular error
  • Of 23 ring products reviewed, none states a guaranteed runout
  • Higher system accuracy at lower manufacturing cost on your side
0010832016530248403305041360495µmrunout deviation — and the resulting angular error on a ring of 25 mm radiusNanomag SmCo≤ 25guaranteedSIKO: 50 µmprecondition for ±0.1°sintered / elastomer-basednot specifiedEccentricity error, independent of the sensing principle: error [″] = (e / R) × 206,265 — e = runout, R = ring radius.Of 23 ring products reviewed, none states a guaranteed runout. SIKO quotes 50 µm —but as a precondition for its own accuracy figure, not as a guarantee on the ring (as of 09/2026).
Additional angular error per 10 µm of extra runout, from error [″] = (e / R) × 206,265. R is the ring radius, half the diameter.
Ring ⌀Penalty [″]Penalty [m°]Penalty [°]
20 mm20657.30.057
25 mm16545.80.046
44 mm9426.00.026
50 mm8322.90.023
57 mm7220.10.020
100 mm4111.50.011
120 mm349.50.010

The penalty is constant in absolute terms: at ⌀ 50 mm, 10 µm more always costs 83″ — whether from 25 to 35 µm or from 100 to 110 µm. In relative terms, starting from our ≤ 25 µm it is +40 % at any diameter.

The material

Why samarium-cobalt — and why without a binder

The choice of material determines the limits your application runs into. SmCo moves them in both directions.

bore H7⌀ 44 mm · wall 3 mmalready manufactured · scale 1 : 1SmCo 0.1 mmlayer structure (detail, to scale)1 mm = 100 px0.51.01.50mmSmCo magnetic layer0.1 mm · radially magnetisedstainless-steel carrier2 – 3 mm · machined to H7break line — carrier continues to full wall thicknessmanufactured: ⌀ 44 mm with a 3 mm wall and 3 mm ring heightwall thickness from 2 mm — smaller diameters on request, with no loss of robustnessfully metallic — no binder, no polymerwall thickness and ring height freely adaptable to the application
Cross-section: thin, radially magnetised SmCo layer on a stainless-steel carrier machined to H7.

No spin reorientation

Sintered NdFeB undergoes a spin-reorientation transition below roughly −138 °C and loses coercivity there. SmCo has no such transition — magnetisation stays stable down to cryogenic temperatures.

No binder that can fail

Elastomer- and polymer-bonded magnets live and die by their binder: it embrittles in the cold, degrades in the heat and is attacked by solvents. A fully metallic design has no such weak point.

Resistant to aggressive media

Resistant to all common oils and greases, organic solvents, saltwater, salt solutions and aqueous alkaline solutions. After years of exposure more than 94 % of field strength remains and runout is unchanged.

Inert to radiation and EMP

Radioactive radiation and electromagnetic pulses leave the material unaffected. Demagnetisation is practically impossible — relevant wherever a failure cannot be repaired.

Speed and temperature together

No restriction on speed, none on temperature — and, unlike most alternatives, none on both at once.

Efficient use of material

The high energy density of SmCo allows very thin layers. That saves magnet material, lowers mass and moment of inertia and makes extreme lightweight design possible.

Operating range

From −273 °C to +300 °C — verified

Every magnet class has two limits: one low, one high. For most of them those limits sit uncomfortably close together.

-273-200-1000+100+200+300temperature in °CNanomag SmCo pole ring−273 °C+300 °Cverified operating rangesintered NdFeB-138+180spin reorientation below ≈ −138 °Cferrite-40+250coercivity drops in the coldelastomer-bonded-40+150binder limits both endsliquid heliumliquid nitrogenambient
Comparison of operating ranges. The figures for sintered NdFeB, ferrite and elastomer-bonded magnets describe typical properties of the respective material class.
>94%of field strength retainedAfter years of exposure tobrake fluidoilscoolantssaltwateralkaline solutionsorganic solventsgreasessalt solutionsrunout: unchangedfully metallic — no binder that can be attacked

What this means

What you get from this in your system

  • Clean sensor signal with no per-revolution amplitude or phase error
  • Less effort for calibration and compensation in software
  • Lower demands on the surrounding mechanics — the axis is relieved
  • No corrosion protection, coating or encapsulation required
  • Smaller moment of inertia and therefore higher achievable dynamics
  • Less installation space for the same bore — or a larger bore in the same space
  • One material for cold and heat instead of two variants in your portfolio
  • A fraction of the magnet material per ring — less exposure to rare-earth raw material markets

Let's talk about your application.

Send us your requirements — diameter, pole count, temperature range, media exposure. You will get a substantive technical answer, not a brochure.