How To Interpret Resolver Angle Accuracy And Electrical Error?

In motor control, new energy vehicle electric drives, servo system selection, or testing, engineers most frequently ask: "What is the accuracy of this resolver?" However, datasheets often list parameters like "Electrical error ≤ ±10′", "Angle accuracy ±15′", "Resolution 16 bits", etc., which can easily confuse people.

Are electrical error and angle accuracy the same thing? How large is 10′ exactly? How should you read the datasheet without being fooled by the numbers?

I. First, distinguish two concepts: Angle accuracy ≠ Electrical error

A resolver itself outputs analog sine/cosine signals, not a digital angle directly. It requires a subsequent RDC (resolver-to-digital converter) or software demodulation to obtain the angle value.

Therefore, what we commonly call "angle accuracy" is actually a system-level specification that includes at least:

• The resolver’s own electrical error

• Decoder (RDC) error

• Mounting eccentricity and assembly tolerances

• Errors from cables, shielding, and signal conditioning

"Electrical error" is the inherent angular deviation of the resolver body under ideal decoding conditions. It can be understood as the resolver’s "raw accuracy."

An analogy:

• Resolver electrical error is like the graduation error of a ruler itself;

• Angle accuracy is like the final measurement error when you use that ruler to measure a workpiece.

• Whether the ruler is accurate depends on electrical error; whether the measurement is accurate also depends on how steady your hand is and how finely you read it.

So, electrical error is the core intrinsic specification of the resolver, but it is not the whole story for system accuracy.

II. What is electrical error and where does it come from?

Under ideal excitation and ideal decoding conditions, the deviation between the angle calculated from the resolver output signals and the true mechanical angle is the electrical error.

The common unit is arc-minute, denoted by ′.

Conversion:

•  1° = 60′

•  10′ = 10/60° ≈ 0.167°

•  1′ ≈ 0.0167°

For example, if the datasheet says "Electrical error ≤ ±10′", it means that at any position, the deviation between the decoded angle and the true angle does not exceed 10 arc-minutes, approximately 0.167°.

Major sources of electrical error include:

These factors combine to form the electrical error curve of the resolver.

III. How to read "Electrical error" on a datasheet?

When you see "Electrical error ≤ ±10′" on a datasheet, don’t just look at the number. Ask three questions first:

1. Is the unit mechanical angle or electrical angle?

This is especially important for multi-pole resolvers.

For a single-pole resolver, mechanical angle equals electrical angle, so no distinction is needed. But for multi-pole resolvers:

Electrical angle = Mechanical angle × Number of pole pairs

If a 4-pole-pair resolver specifies "Electrical error ≤ ±10′" and it refers to electrical angle error, then the equivalent mechanical angle error is only:

10′ ÷ 4 = 2.5′

The mechanical accuracy is actually higher.

But if it specifies mechanical angle error of ±10′, then the equivalent electrical angle error becomes:

10′ × 4 = 40′ ≈ 0.667° electrical

This could be approaching the upper limit for FOC (field-oriented control) applications.

Therefore, when you see electrical error, first confirm whether it is mechanical angle error or electrical angle error. Many resolver manufacturers default to electrical angle domain error, but practices vary. If necessary, ask the original manufacturer for clarification.

2. What are the test conditions?

Electrical error is not fixed; it is affected by temperature, speed, excitation voltage, and frequency.

Common datasheet statements include:

· Room-temperature electrical error ≤ ±10′

· Full-temperature electrical error ≤ ±15′

· 0~10000 rpm dynamic error ≤ ±20′

Looking only at the room-temperature specification is insufficient. For new energy vehicle electric drives, high-temperature servo applications, etc., you must pay attention to the electrical error over the full temperature and full speed ranges.

3. Is it maximum error, peak-to-peak, or RMS?

Electrical error can be expressed in several ways:

•  Maximum absolute error: the maximum absolute deviation among all measured points

•  Peak-to-peak error: the difference between the maximum positive deviation and the maximum negative deviation

•  RMS error: root-mean-square value, reflecting the overall error level

Datasheets usually specify "maximum electrical error" or "peak-to-peak electrical error." If it simply says "Accuracy ±10′", it is generally understood as maximum absolute error.

IV. How to interpret the error curve?

If you can obtain the resolver test report, the best way is to directly inspect the error curve.

A typical electrical error curve has mechanical angle on the horizontal axis and error on the vertical axis. The shape of the curve helps identify error sources.

Common error characteristics:

General rule:

The smoother and more regular the error curve, the more stable the resolver manufacturing process, and the error is predictable and compensable.
If the curve has many glitches or large jumps, first check the test environment, shielding, and excitation signal quality, rather than rushing to judge the resolver’s accuracy.

In servo drives or electric drive systems, some solutions implement error compensation tables. However, compensation requires that the error be repeatable. If the error varies each test, compensation is meaningless.

V. Brief practical method for measuring electrical error

If you want to measure resolver electrical error yourself, the basic approach is:

1. Use a high-precision encoder, optical grating, or indexing head as the angle reference.

2. Provide standard excitation signals to the resolver.

3. Simultaneously acquire the resolver’s sine/cosine outputs.

4. Decode the angle via an RDC or algorithm.

5. Compare with the reference angle and record the error over 360°.

6. Calculate the maximum error, peak-to-peak, RMS, and plot the error curve.

Precautions during testing:

• Keep speed low to avoid mixing in dynamic errors.

• Use short cables, differential acquisition, and shielding.

• Set excitation frequency and voltage to match actual operating conditions.

• For multi-pole resolvers, clearly record whether the result is electrical or mechanical angle.

VI. How to use these specifications in selection?

Example:

Suppose a motor has 4 pole pairs, and the FOC control requires electrical angle error to be controlled within 1° electrical.

If the resolver’s electrical error is ±10′ electrical, i.e., about ±0.167° electrical, the body error is small. After adding decoder and mounting errors, the overall system electrical angle error can still be kept within 1°, so the solution is feasible.

However, if another resolver specifies mechanical angle error of ±10′, then for a 4-pole-pair motor, the electrical angle error will be amplified to about ±0.667°. After adding decoding and assembly tolerances, the overall electrical angle error may approach or even exceed 1°, requiring a re-evaluation.

Therefore, when selecting a resolver, do not simply compare numbers; convert all to electrical angle uniformly, and then add system error budget.

Conclusion

Electrical error is the core indicator of resolver body accuracy, but it is not the whole story for system angle accuracy. To properly understand resolver accuracy, three key points are essential:

1. Distinguish between electrical error and system-level angle accuracy;

2. Confirm whether the unit is mechanical or electrical angle;

3. Consider test conditions, error curves, and actual operating conditions.

Whether the number looks good on paper is one thing; whether it is stable and accurate when integrated into the system is the ultimate criterion.