Temperature Rating Selection for Magnetic Encoder Disks Under High/Low-Temperature Conditions: Injection-Molded Ferrite Vs. Rubber Magnet—Don’t Just Look at The Word “Ferrite”

After a collaborative robot runs continuously for several hours, its position feedback begins to drift slowly. When the joint is disassembled, the magnetic encoder disk looks intact, and its magnetization shows no obvious abnormality, but when it is retested at room temperature, the angular error does not return to its initial value. The problem often lies not in the encoder chip, but in the fact that the high-temperature behavior of the disk material has been underestimated.

In the material selection for magnetic encoder disks, injection-molded ferrite and rubber magnet are the two most commonly compared options. A common misjudgment is: both are called “ferrite magnets,” so their temperature resistance should be similar. The reality is exactly the opposite—what determines their temperature rating is mainly not the ferrite magnetic powder, but the “binder” that holds the magnetic powder together. The Curie temperature of ferrite magnetic powder itself can be as high as 450–470°C, but the maximum temperature at which an encoder disk can work long-term depends almost entirely on how high the matrix materials—nylon, PPS, nitrile rubber, or hydrogenated nitrile rubber—can withstand.

1. The “Temperature Truth” of the Two Materials: Magnetic Powder Sets the Floor, Binder Sets the Ceiling

Injection-molded ferrite is made by mixing ferrite magnetic powder with thermoplastic resin (nylon, PPS, etc.) and injection molding it. Its temperature ceiling is determined by the resin, not the ferrite:

• Nylon 12 (PA12) matrix: maximum working temperature approximately 120°C;

• Nylon 6 (PA6) matrix: approximately 150°C;

• Polyphenylene sulfide (PPS) matrix: can reach approximately 180°C.

Rubber magnet is a flexible magnet made by filling nitrile rubber (NBR), hydrogenated nitrile rubber (HNBR), or other elastomers with ferrite magnetic powder. Its temperature resistance is likewise dominated by the rubber formulation, and the spread is larger than many people imagine:

• Conventional NBR formulations on encoder disks are usually rated at a long-term working temperature upper limit of around 80°C. Short-term exposure at 80°C can keep magnetic decay at a relatively low level, but further heating significantly increases the risk of irreversible demagnetization;

• NBR formulations for industrial encoders can reach 140–150°C, and HNBR formulations can further reach 160–180°C;

• Through high-temperature matrices such as silicone rubber or special magnetic powder modification, the short-term tolerance ceiling of rubber magnets can be even higher, but such formulations pay a price in dimensional stability, modulus, and cost.

• In other words, both statements—“rubber magnets can only reach 80°C” and “rubber magnets can also reach above 150°C”—can be true. The difference lies in the formulation, not in the broad category name “rubber magnet.” If selection only says “rubber magnet” without locking down the rubber grade and the supplier’s actual long-term temperature curve, the risks under high-temperature conditions cannot be stopped by incoming inspection.

2. The Core Dividing Line in High/Low-Temperature Selection: Not “Whether It Can Be Used,” but “Whether Demagnetization Is Reversible”

The effects of temperature on magnetic encoder disks fall into two categories. Conflating them is the most common source of misjudgment in high-temperature validation.

Reversible demagnetization means that as temperature rises, remanence and coercivity temporarily decrease according to the temperature coefficient, and when temperature recovers, magnetic performance basically recovers as well. The remanence temperature coefficient of injection-molded ferrite is on the order of −0.18%/°C to −0.20%/°C. This means that from 25°C to 100°C, the remanence drop caused solely by the temperature coefficient is more than ten percentage points. This drop itself does not necessarily mean the disk is “bad,” but at high temperature the effective signal amplitude seen by the sensor becomes smaller, and the same noise, eccentricity, and external magnetic interference are amplified more in the angle calculation, manifesting as angle jitter, zero-position drift, or speed glitches under high-temperature conditions.

Irreversible demagnetization means that after the temperature exceeds a certain threshold, the magnetic domain structure undergoes permanent change, and magnetic performance still does not recover after returning to room temperature. For injection-molded ferrite, coercivity is relatively low, and the risk of crossing the operating-point knee at high temperature is higher; for rubber magnets, softening or aging of the binder at high temperature changes the dispersion state of the magnetic powder particles, which can likewise solidify a temporary decline into permanent loss.

Therefore, the true criterion for high/low-temperature selection is not “what nominal temperature rating the material has,” but “whether, at the maximum working temperature, the angular error and magnetic field residual after returning to room temperature remain within the allowable range.” Judging “demagnetization” solely by how much the magnetic field amplitude drops in a high-temperature chamber, or selecting a material solely by one maximum temperature number on the supplier’s TDS, is insufficient to cover the actual risks of the encoder system.

3. Selection by Operating Temperature Band: An Actionable Decision Framework

Rather than asking “which has better temperature resistance, injection-molded ferrite or rubber magnet,” it is better to use the maximum operating temperature as the first cut and validation capability as the second cut.

≤ 80°C ambient/low-temperature range: Both are usable. Injection-molded ferrite has advantages in magnetic performance, dimensional accuracy, and multipole magnetization consistency; the flexibility of rubber magnets helps with impact resistance and adaptation to irregular mounting, and the cost is relatively friendly. At this point, selection should focus more on resolution, pole pitch, and structural integration requirements, rather than temperature resistance. It should be noted that rubber magnets may harden at very low temperatures (e.g., −40°C). If the application involves low-temperature startup or thermal shock, the formulation’s retention of flexibility at low temperature should be confirmed, not just the “low-temperature resistant” label.

80–120°C medium-high temperature range: This is the interval where the two materially diverge. Conventional rubber magnet formulations have already entered a zone requiring careful evaluation in this range; the PA12 matrix of injection-molded ferrite has a nominal upper limit of about 120°C, but when used as an encoder disk, it is advisable not to treat the “material upper limit” directly as the “design operating point.” If the operating peak is close to 120°C, PA6-based injection-molded ferrite (approximately 150°C) will provide more ample margin, and irreversible loss should be confirmed through the post-temperature-cycle residual after returning to room temperature.

120–150°C: PA6-based injection-molded ferrite is the natural candidate for this temperature band, while rubber magnets require explicitly selecting a high-temperature NBR or HNBR formulation and requesting the supplier’s long-term aging data at the target temperature, rather than relying only on short-term instantaneous tolerance values. This band is also where “formulation makes all the difference” is most evident: for the same “rubber magnet,” conventional NBR formulations and high-temperature NBR/HNBR formulations for encoders may behave completely differently in long-term behavior near 150°C.

> 150°C: PPS-based injection-molded ferrite (approximately 180°C) is the more realistic option among the two material types. Rubber magnets are not absolutely unusable, but they require a high-temperature rubber matrix or specially modified formulation, and the trade-offs among flexibility, dimensional stability, and cost become more prominent. If the operating condition exceeds 180°C long-term, both types of organic-binder magnets begin to approach the boundaries of the material system, and the selection discussion should shift to other magnet systems or system-level thermal management, rather than continuing to fine-tune between “injection-molded ferrite or rubber magnet.”

The low-temperature end likewise cannot be judged only by “minimum temperature.” The coercivity temperature coefficient of ferrite differs in direction from the remanence temperature coefficient, and some ferrite grades have a mechanism of irreversible low-temperature demagnetization at low temperatures; the low-temperature risk of rubber magnets is more concentrated in hardening and embrittlement of the rubber matrix. Under alternating high/low-temperature conditions, angular zero-position stability after temperature cycling reflects the usability of the disk in actual operating conditions better than magnetic performance at a single temperature point.

4. The Step Most Easily Missed in Validation

High-temperature validation of magnetic encoders has a “seemingly correct” trap: the magnetic field amplitude drops after heating, does not fully recover after cooling, and is then judged as irreversible demagnetization. But amplitude residual may come from the magnet, or from assembly stress release, air-gap changes caused by structural thermal expansion, or sensor front-end drift. Conversely, a temporary amplitude drop at high temperature that recovers after cooling does not mean the system is necessarily safe during the high-temperature period—after the effective signal becomes smaller, the angular error may already have exceeded the limit during the high-temperature process, and it merely “looks good” after cooling.

A more robust approach is: before and after temperature cycling, use angular error and zero-position drift as the primary criteria, and magnetic field amplitude as an auxiliary diagnostic quantity. If the angular error and zero position after returning to room temperature do not return to baseline, then break down whether it is irreversible magnet loss, binder aging, or a change in assembly state. For bonded magnets such as injection-molded ferrite and rubber magnet, binder state changes after high-temperature exposure often do not manifest as “the magnet suddenly disappears,” but as slow degradation of the angle signal, which is easily missed by routine “magnetization OK, appearance OK” checks.

Conclusion

Under high/low-temperature conditions, the selection of magnetic encoder disks is essentially not a binary choice between “injection-molded ferrite vs. rubber magnet,” but two sequential questions: the maximum operating temperature and cycling mode determine the binder system of the candidate materials; the validation data—especially the angular residual after returning to room temperature—determine whether the candidate material can truly enter the design. The temperature ceiling of injection-molded ferrite is written in the resin grade, the temperature ceiling of rubber magnet is written in the rubber formulation, and the reliability of the encoder is written in the angular error after temperature cycling. Only by keeping these three layers separate will selection not stop at the overly rough impression that “ferrite is high-temperature resistant.”

Quick Selection Checklist

• First lock down the maximum operating temperature and number of cycles; do not use average temperature in place of peak temperature.

• For injection-molded ferrite, ask clearly about the resin matrix: PA12 ≈ 120°C, PA6 ≈ 150°C, PPS ≈ 180°C; use the supplier’s measured long-term temperature curve as the authority.

• For rubber magnet, ask clearly about the rubber grade: the long-term upper limits of conventional NBR formulations and high-temperature NBR/HNBR formulations can differ by tens of degrees Celsius; do not just write “rubber magnet.”

• Require the supplier to provide post-aging angular error/magnetic field residual after returning to room temperature at the target temperature, not just instantaneous magnetic flux at high temperature.

• For low-temperature conditions, also confirm the low-temperature embrittlement risk of the rubber matrix and the low-temperature irreversible demagnetization tendency of ferrite.