Magnetic Pole Pairs: Too Many Is Wasteful, Too Few Is Insufficient—Where Are The Selection Limits of Injection-Molded Ferrite Disks?

When selecting magnetic encoder disks, a common intuition is: the higher the number of pole pairs, the higher the resolution and the better the performance. At the sensor level, this judgment is broadly valid. But once the magnet material is fixed as injection-molded ferrite, the situation becomes more nuanced. The magnetic performance ceiling of injection-molded ferrite, the process limits of multi-pole magnetization, and the easily overlooked fact that "resolution ≠ accuracy" together draw a selection curve that is not simply "the higher, the better."

First, clarify a premise: resolution is not equal to accuracy

TI application literature gives a concise relationship: when using a single Hall latch, the pulses per revolution equal the number of pole pairs of the magnet; when using two Hall latches for quadrature decoding, the counts per revolution are twice the number of poles. From the perspective of "graduations," pole pairs do directly determine the smallest angular step the system can resolve.

But a reminder from an open-source magnetic encoder project deserves attention from every selector: "Resolution != Accuracy"—you can obtain a theoretical resolution of 0.017°/step at a 2.4 mm pole pitch, but these steps are not uniformly mapped around the circumference. In other words, pole pairs determine "how fine the graduations are," not "how accurately the graduations are drawn." Angular error comes from pole pitch consistency, magnetization boundary clarity, sensor installation eccentricity, and the material uniformity of the magnet itself. Pole pairs themselves do not eliminate these errors; instead, they may amplify them as the number of pole pairs increases.

The "magnetic performance ceiling" of injection-molded ferrite

The positioning of injection-molded ferrite is clear: it is injection-molded from ferrite powder (usually strontium ferrite) combined with a polymer matrix such as nylon PA12/PA6. Its advantages lie in integrated molding of complex shapes, good dimensional consistency, corrosion resistance, and low cost, but its magnetic performance is at the low end among commonly used encoder magnet materials.

In terms of specific data, the residual induction Br of injection-molded ferrite is roughly 1.7–3.0 kGs, the intrinsic coercivity Hci is about 2.3–4.0 kOe, and the maximum energy product BHmax is about 1–2.2 MGOe. For reference, the Br of bonded NdFeB is typically 4.5–7.5 kGs, and BHmax can reach 5–8 MGOe or even higher. This means that, under the same sensor air gap and detection threshold, the magnetic field "signal strength" that injection-molded ferrite can provide is clearly weaker than that of NdFeB systems.

Surface magnetic field data after multi-pole magnetization further illustrates the constraint: the surface magnetic field of thin-wall inner-diameter oriented injection-molded ferrite multi-pole magnetic rings is about 600–1100 Gs, while anisotropic sintered ferrite can achieve higher values. As the number of pole pairs increases and the arc width of each individual pole narrows, the "area" available for establishing the magnetic field shrinks, edge effects and inter-pole magnetic field superposition become more significant, and the signal-to-noise ratio deteriorates further. This is not a problem that can be solved by adjusting the current of the magnetizer; rather, it is a soft upper limit imposed by the material's energy product on the number of pole pairs.

Geometric constraints among pole pitch, thickness, and pole pairs

An empirical relationship given in a European patent is quite instructive: the thickness of the magnetizable element should be about 1.25 times the pole pitch (±25%), corresponding to a thickness range of about 0.3–1.5 mm; the pole pitch itself is on the order of 0.25–1.25 mm. This geometric constraint ties the number of pole pairs to the disk diameter: for a given diameter, the higher the number of pole pairs, the smaller the pole pitch, and the thinner the required magnet thickness.

For injection-molded ferrite, thinning the wall further compresses the effective magnetic flux. At the same time, once the pole pitch shrinks to the sub-millimeter level, the tolerance window for magnetizing fixture pole-head precision and magnetic circuit design rapidly narrows. Industry materials clearly point out that the higher the pole count, the smaller the single-pole arc width, and the higher the requirements for magnetizing fixture pole-head precision and magnetic circuit design. Problems such as under-saturation at pole tips, magnetic field distortion, and inter-pole zero-position shift will then appear.

This explains why common specifications of injection-molded ferrite encoder magnetic rings on the market are concentrated at positions such as 62/64 poles and 126/128 poles, rather than being stacked upward indefinitely. These positions are balance points that have already undergone engineering convergence among material performance, magnetization process, and cost.

Selection logic: work backward from "how fine is needed" to "how high can be made"

A practical selection path should be reversed: first determine the resolution the system truly needs, then evaluate whether injection-molded ferrite can provide usable signal quality at that pole-pair count.

Low pole-pair range (4–12 poles): The arc length of a single pole is large, and the magnetic flux is sufficient, so the weak-magnetic-field shortcoming of injection-molded ferrite is compensated by geometric dimensions. This range is suitable for scenarios that do not require stringent resolution but require low cost and good consistency, such as fan motors, smart water meters, and appliance BLDCs. In these applications, the role of pole pairs is more to provide a commutation reference than a precise angular scale.

Medium pole-pair range (16–32 poles): This is the threshold for entering encoder-level applications. Injection-molded ferrite can still maintain a usable signal-to-noise ratio in this range, making it suitable for medium-resolution angle feedback, such as humanoid robot joint encoders and some servo auxiliary feedback. During selection, the key verification item is pole pitch consistency rather than simply pursuing a higher pole-pair count—a 32-pole magnetic ring with uniform pole pitch distribution may achieve better angular accuracy than a 64-pole magnetic ring with blurred magnetization boundaries.

Higher pole-pair range (48–64 poles and above): Injection-molded ferrite begins to approach the upper limits of process and material. This range is better suited to dual-track design: the main track is responsible for high-resolution incremental or absolute position, while the secondary track provides revolution count or coarse position reference with coprime pole-pair counts. The 62/64-pole and 126/128-pole dual-track magnetic rings already available on the market are products of this approach. However, it should be noted that when the number of pole pairs increases further, surface field attenuation and inter-pole interference of injection-molded ferrite will degrade signal quality. At this point, bonded NdFeB is usually the more reasonable choice, despite its higher cost.

Two key reminders for engineering implementation

First, include magnetization capability in the selection review. The pole-pair count on a magnetic ring drawing is a design value. Whether the actual magnetic pole distribution is uniform and whether the inter-pole zero positions are clear depend on the magnetizing fixture and process. Before determining the number of pole pairs, the magnet supplier should be asked to provide the measured surface field distribution curve and pole pitch consistency data at that pole-pair count, not merely a schematic of the magnetic pole arrangement.

Second, dual-track is not a shortcut for "stacking pole counts." The value of dual-track lies in using a combination of coprime pole-pair counts to simultaneously obtain high resolution and an absolute position reference. However, it places higher demands on phase alignment between the two magnetic tracks and on magnetization crosstalk control. If the pole-pair counts of the main and secondary tracks are improperly selected (for example, too close to each other), the position-solving advantage brought by coprimality will be offset by signal crosstalk. The pole-pair combination of the dual tracks needs to be designed together with the decoding algorithm, rather than being decided unilaterally by the magnet manufacturer.

The essence of selection is to find the intersection among resolution requirements, material capability, process window, and cost. Injection-molded ferrite is not a "universal magnetic ring material." It provides a highly cost-effective solution in the low-to-medium pole-pair range; but when the pole-pair count advances beyond 64, selectors need to honestly face its magnetic performance ceiling—either accept a loss in signal quality or turn to a material system with a higher energy product. Pole pairs are a ruler, but whether the graduations on the ruler are drawn accurately depends on whether the material and process can support them.