Where Does The Phase Deviation of A High-Speed Rotating Magnetic Ring Come From? – Troubleshooting Guide for Robot Magnetic Encoder Sensor

In high-performance motion control applications such as servo motors, robot joints, and electric spindles, magnetic encoders are gaining popularity due to their small size, oil resistance, vibration tolerance, and low cost. However, many engineers encounter a thorny issue during high-speed rotation testing: the angle reported by the encoder deviates from the rotor’s actual position -a "phase deviation" that worsens with increasing speed. At low speeds it may cause only slight speed ripple and increased noise, but at high speeds it can lead to control instability.

I. How Does a Robot Magnetic Encoder Sensor "See" the Angle?

Let’s briefly review the basic structure of a magnetic encoder.

A typical magnetic encoder consists of three parts:

1. Magnetic ring: mounted on the rotor end or shaft extension, with multiple pole pairs (N/S alternating) distributed around its circumference.

2. Magnetic sensor: commonly Hall, AMR, GMR, TMR, etc., placed close to the ring surface to sense magnetic field variations.

3. Signal processing circuit: amplifies, conditions, and interpolates the weak sensor signals, ultimately outputting angle or quadrature pulses.

As the magnetic ring rotates, the sensor detects periodic magnetic field changes and outputs quasi-sinusoidal/cosinusoidal signals. By resolving the phase of these two signals, the rotor angle is obtained.

Ideally, the magnetic field distribution is perfectly sinusoidal, and the sensor output strictly corresponds to the mechanical angle. However, under high-speed rotation, this "ideal correspondence" is disrupted by multiple factors, and phase deviation emerges.

II. What Is Phase Deviation and Why Does It Become More Pronounced at High Speed?

Simply put, phase deviation is the difference between "the angle perceived by the encoder" and "the true rotor angle".

For example, if the rotor is actually at 30° but the encoder outputs 32°, that 2° is phase deviation. At low speed, this deviation might be only 0.1° and barely noticeable; but as speed increases, the deviation can amplify to several degrees or even over ten degrees, significantly degrading control performance.

Why does deviation amplify at high speed? The main reasons are as follows.

1. Mechanical installation errors are "magnified"

Perfect coaxial alignment between the magnetic ring and the shaft is impossible. Suppose the ring has an eccentricity of 0.05 mm -at low speed, the sensor can still tolerate the resulting field distortion; at high speed, centrifugal force and vibration increase the eccentricity, distorting the magnetic waveform and causing periodic phase deviation.

Similarly, runout of the ring end face, axial play, and poor dynamic balance introduce additional high-frequency disturbances at high speed.

2. Sensor response speed is insufficient

As the ring speed increases, the frequency of the magnetic field variation rises accordingly. For example, with a 32-pole-pair ring at 10,000 rpm, the field frequency is about 5.3 kHz; at 30,000 rpm, it exceeds 16 kHz.

Common Hall sensors may have bandwidths of only tens of kHz or even lower -at high speed, output amplitude attenuates and phase lags, resulting in angle errors. Even if the sensor itself is fast, excessive delay in subsequent filter circuits can also cause phase lag.

3. Degradation of the magnetic ring’s field quality

At high speed, eddy current losses inside the ring cause localized heating. As temperature rises, remanence and coercivity change, reducing field strength and distorting the waveform.

In addition, non-uniform magnetization, excessively wide transition zones between poles, and improper selection of pole-pair numbers all increase harmonic content in the magnetic field at high speed, enlarging phase-calculation errors.

4. Electromagnetic interference and signal integrity

At high speed, high-frequency PWM currents in motor windings, shaft currents, and nearby high-power lines generate interference. The sensor’s weak output signals, once contaminated, cause jitter and error in zero-crossing detection or phase calculation.

III. Troubleshooting Strategy: Mechanical, Electrical, and Algorithmic Approaches

To address the above causes, corrective actions can be taken at three levels: mechanical installation, sensor and signal chain, and algorithm compensation.

1. Mechanical level -lay a solid foundation first

Many phase deviation problems originate not from electronics but from mechanical installation.

(1) Strictly control coaxiality and end-face runout

After mounting the magnetic ring, coaxiality should be kept within 0.03–0.05 mm, and end-face runout within 0.02 mm. Use shrink fitting, adhesive bonding, or dedicated fixtures for positioning, and avoid hammering that could tilt the ring.

(2) Perform dynamic balancing

Although the ring is lightweight, centrifugal force at high speed is not negligible. After installation, conduct overall dynamic balancing to reduce vibration-induced signal errors.

(3) Avoid axial forces on the magnetic ring

If the ring is close to the motor end cover or bearings, thermal expansion at high speed may cause axial displacement. Reserve appropriate axial clearance in the design, or adopt an elastic preload structure.

2. Electrical and signal chain -keep signals clean and timely

(1) Choose the right sensor type

For high-speed applications, prefer TMR or GMR sensors, as their bandwidth typically reaches hundreds of kHz or even MHz -far higher than ordinary Hall sensors. Also ensure the sensor’s magnetic sensitivity and saturation field match the ring’s characteristics.

(2) Use differential signal output

Differential output suppresses common-mode interference and improves signal integrity. If the sensor only provides single-ended output, add a differential amplifier stage at the front end.

(3) Reduce filter delay

RC filters and digital filters introduce phase delay. In high-speed applications, set the filter cutoff frequency sufficiently high, or use low-delay, linear-phase filters. For high-frequency noise that truly needs removal, consider smoothing after angle calculation rather than over-filtering the raw signals.

(4) Optimize the air gap between ring and sensor

Excessive air gap reduces signal amplitude; too small a gap risks rubbing due to vibration at high speed. Typically choose an air gap of 0.3–1.0 mm based on the ring specifications and sensor sensitivity, and determine the optimal value experimentally.

(5) Implement proper shielding and grounding

Use shielded twisted-pair cables for sensor wiring, ground the shield at a single point, and keep the cable away from motor power lines and high-current traces to reduce electromagnetic coupling.

3. Algorithm level -use "software" to compensate for "hardware" deficiencies

After hardware improvements, residual deviations can be further corrected by algorithms.

(1) Offline calibration and lookup-table compensation

Under low-speed, steady-state conditions, use a high-precision reference encoder (e.g., optical encoder) to perform full-circle calibration of the magnetic encoder, record the deviation at each position, and build a compensation table. During actual operation, look up the correction based on the current position.

This method effectively eliminates periodic phase deviations, but has limited effect on additional dynamic deviations that appear at high speed.

(2) Dynamic phase compensation

If deviations mainly arise from a fixed delay in the signal chain, calculate the angle delay by:

where w is angular velocity and t_d is the total system delay. Feed this compensation forward to the output angle, which can significantly improve high-speed accuracy within a certain speed range.

(3) Adaptive filtering and observers

For random jitter at high speed, use Kalman filters, sliding-mode observers, or similar methods to fuse the encoder angle with a motor model, suppressing high-frequency noise while preserving the true angle signal.

(4) Temperature compensation

Temperature-induced changes in remanence affect signal amplitude. Place a temperature sensor near the encoder to establish a temperature-amplitude/phase compensation model and apply real-time correction.

IV. Brief Troubleshooting Case Study

A servo motor exhibited a periodic phase deviation of about 3.5° from its magnetic encoder at 15,000 rpm, causing current-loop oscillation. The team followed these steps:

1. Mechanical inspection: The ring’s coaxiality was found to be 0.08 mm; after re-installation it was reduced to 0.03 mm, and the deviation dropped to 2.1°.

2. Sensor replacement: The standard Hall sensor was replaced with a higher-bandwidth TMR differential sensor, and the filter cutoff frequency was raised from 20 kHz to 100 kHz -deviation further reduced to 0.9°.

3. Algorithm compensation: Fixed-delay feedforward compensation was added, followed by full-circle offline calibration on the test bench. Final high-speed phase deviation was controlled to within 0.3°, meeting control requirements.

This case illustrates that a single measure often solves only part of the problem -a combination of mechanical, electrical, and algorithmic improvements yields the best results.

Conclusion

Phase deviation of the magnetic ring in a robot magnetic encoder sensor at high speed is inherently a multifactorial problem. Mechanical installation errors, insufficient sensor bandwidth, signal delays, degraded magnetic field quality, and electromagnetic interference are all amplified at high speed.

When troubleshooting, follow the sequence: mechanics first, then electrical, then algorithms:

· First, ensure coaxiality, dynamic balancing, and air gap are solidly established;

· Then optimize sensor selection and the signal chain to guarantee clean, low-delay signals;

· Finally, use calibration and compensation algorithms to "fine-tune" residual errors.

For most high-speed applications, after such systematic remediation, the phase deviation of the magnetic encoder can be kept within an acceptable range, allowing the servo system to remain stable and accurate even at high speeds.