In applications such as robotic joints, precision rotary tables, and high-performance motion systems, the precision characteristics of a harmonic drive reducer can directly determine the performance ceiling of the entire system.

When selecting a harmonic drive reducer, engineers often encounter terms such as backlash, lost motion, transmission error, positioning accuracy, and repeatability. Although these parameters are closely related, they describe different aspects of motion accuracy and should not be treated as interchangeable.

Through years of experience in AGV drive wheels, servo drives, and motion-control systems, Plutools has found that understanding the physical meaning of these key component specifications is essential for achieving the right match between the reducer, motor, controller, and overall motion system.

Below, we take a practical approach to explain these five key precision parameters.

1. Five Key Precision Parameters Explained

1.1 Backlash - How Much Mechanical Clearance Exists?

Backlash refers to the amount of angular movement at the output side when the input side is locked and the output is moved alternately in the forward and reverse directions.

In a harmonic drive reducer, backlash is primarily associated with mechanical clearance between the teeth of the flexspline and circular spline.

Thanks to the elastic deformation of the flexspline and the simultaneous engagement of multiple teeth, harmonic drive reducers can achieve extremely low mechanical backlash.

Typical values for mainstream harmonic drive reducers are around:

≤1 arcmin

High-precision products may achieve:

≤0.5 arcmin

Some products may even be specified as having "zero backlash" under their stated measurement conditions. However, this does not necessarily mean that the physical backlash is absolutely zero. In most cases, it means that the measured value is below the resolution or detection capability of the specified test equipment.

It is also important to distinguish backlash from load-dependent elastic deformation. Backlash itself is primarily a mechanical clearance characteristic, while the total angular displacement observed under load can be significantly affected by the stiffness and elastic deformation of the reducer.

In simple terms: Backlash tells you how much mechanical free play exists when the direction of motion changes.

1.2 Lost Motion - How Much Angular Position Is Lost During Reversal?

Lost motion provides a more realistic description of angular displacement during direction reversal.

Under specified test conditions, the output shaft is subjected to a defined torque, and the angular difference between the two stabilized positions when the torque direction is reversed is measured.

Unlike pure mechanical backlash, lost motion can include multiple sources of deformation and clearance, such as:

Gear tooth clearance

Flexspline elastic deformation

Bearing clearance and deformation

Housing deformation

Shaft and structural compliance

Therefore, lost motion is generally greater than or equal to the measured mechanical backlash under comparable test conditions.

As the applied torque increases, elastic deformation becomes more significant, and the measured lost motion may increase accordingly.

Typical values are often in the range of:

1–3 arcmin

However, the actual specification depends heavily on the manufacturer, reducer size, test method, applied torque, and measurement conditions.

For system selection, engineers should therefore pay particular attention to lost motion under a defined load, rather than relying solely on an unloaded backlash specification.

In simple terms: Backlash describes mechanical free play; lost motion describes the total angular displacement that appears during loaded direction reversal.

1.3 Transmission Error - How Accurately Does the Reducer Transfer Motion?

Transmission Error (TE) is one of the most important indicators for evaluating the motion transmission accuracy of a harmonic drive reducer.

The theoretical output angle can be calculated from the input angle and the reduction ratio. The difference between the actual output angle and the theoretical output angle is the transmission error.

A typical measurement system uses high-precision encoders on both the input and output sides. The reducer is then operated through a specified angular range while the actual transmission deviation is recorded.

Because the engagement conditions of the flexspline and circular spline change periodically during operation, the transmission error often exhibits a characteristic periodic waveform.

Manufacturers may specify transmission error using different definitions, such as peak-to-peak error or maximum deviation, so the test method and specification definition should always be checked before making a direct comparison.

Typical high-precision requirements may be around:

≤1 arcmin

Transmission error is particularly relevant to applications involving smooth, accurate, and continuous motion.

In simple terms: Transmission error tells you how accurately the reducer converts input rotation into the expected output rotation.

1.4 Positioning Accuracy - Can the System Actually Reach the Target Position?

Positioning accuracy is no longer purely a reducer-level specification. It is a comprehensive performance indicator of the entire motion system.

Starting from a reference position, the commanded motion is executed toward a target position. The difference between the actual final position and the commanded target represents the positioning error.

System-level positioning accuracy can be affected by many factors, including:

Harmonic drive transmission error

Motor encoder accuracy

Servo drive control error

Mechanical backlash and lost motion

Shaft and coupling alignment

Structural stiffness

Installation accuracy

Thermal expansion and deformation

Load characteristics

Control-system compensation

Therefore, the positioning accuracy of a complete system cannot be directly inferred from the transmission error of the reducer alone.

In practical motion-control applications, installation quality and structural rigidity are sometimes underestimated. Even a high-precision reducer may fail to deliver its expected system performance if the mechanical structure is insufficiently rigid or the installation accuracy is poor.

In simple terms: Positioning accuracy is about how close the complete system can actually get to the commanded position.

1.5 Repeatability - Can the System Stop at the Same Position Every Time?

Repeatability, or repeat positioning accuracy, focuses on the consistency of repeated movements.

Under the same operating conditions, the system repeatedly moves toward the same target position. Repeatability describes the amount of variation among the actual stopping positions.

Unlike absolute positioning accuracy, repeatability does not primarily ask whether the system reaches the correct absolute position. Instead, it asks whether the system can return to essentially the same position each time.

Repeatability may be expressed using statistical methods such as ±3σ, depending on the applicable standard and test methodology.

A typical high-precision requirement may be around:

≤±0.5 arcmin

For applications such as robotic assembly, welding, precision handling, and automated docking, repeatability can be particularly important because consistent motion is essential for maintaining process stability.

A system can have excellent repeatability while still having a systematic positioning offset. In other words, it may repeatedly stop at nearly the same position, but that position may not be exactly the commanded target.

In simple terms: Repeatability tells you how consistently the system returns to the same position.

2. Understanding the Relationship Between the Five Parameters

Putting these five parameters into one framework makes their different roles much easier to understand.

The table can be understood as a progression from component-level mechanical characteristics to complete system performance.

Mechanical clearance is only the starting point. Elastic deformation, transmission characteristics, motor and encoder errors, mechanical installation, structural stiffness, thermal effects, and control algorithms can all influence the final motion performance.

Four Important Relationships to Remember

1. Lost Motion is Generally Greater Than or Equal to Backlash

Backlash mainly describes mechanical clearance, while lost motion also includes elastic deformation and other mechanical compliance.

As the applied torque increases, elastic deformation can become more significant, making the difference between the two parameters more obvious.

2. System Positioning Accuracy Depends on More Than Transmission Error

Transmission error represents the motion-transfer performance of the reducer itself.

Positioning accuracy, however, is determined by the complete motion chain. Motor, encoder, servo control, mechanical structure, installation accuracy, thermal effects, and load conditions can all contribute to the final error.

3. Positioning Accuracy and Repeatability Describe Different Things

Positioning accuracy focuses on the deviation from the commanded target, while repeatability focuses on the consistency of repeated movements.

A system can therefore have excellent repeatability but relatively poor absolute positioning accuracy if a systematic offset exists.

4. Transmission Error and Lost Motion Have Different Purposes

Transmission error primarily evaluates how accurately the reducer transfers motion during continuous operation.

Lost motion is particularly relevant when the direction of motion changes and the system must reverse under a defined load.

They are related, but they should not be treated as interchangeable specifications.

3. Three Common Misunderstandings

Misunderstanding 1: "Zero Backlash Is All We Need"

A low backlash value does not automatically mean that the reducer has extremely low lost motion.

In high-load robotic joints and precision motion systems, elastic deformation of the flexspline, bearings, shafts, and housing can contribute significantly to the total angular displacement during reversal.

If selection is based only on the unloaded backlash specification while ignoring lost motion under load, the system may experience unexpected reversal errors, vibration, or overshoot.

For this reason, engineers should evaluate both backlash and load-dependent lost motion whenever possible.

Misunderstanding 2: "Low Transmission Error Guarantees High System Positioning Accuracy"

Transmission error is only one part of the complete motion chain.

A high-precision reducer still needs to work together with an accurate motor encoder, properly tuned servo drive, rigid mechanical structure, and precise installation.

In real-world applications, positioning problems may ultimately be traced to factors such as shaft misalignment, insufficient structural rigidity, thermal deformation, or inappropriate load-inertia matching rather than the reducer itself.

Converting reducer-level precision into actual system-level precision requires coordinated mechanical design and motion-control optimization.

Misunderstanding 3: "High Repeatability Means High Absolute Accuracy"

High repeatability only means that the system can repeatedly reach nearly the same position.

It does not necessarily mean that this position is the correct target position.

For example, a system may repeatedly stop at a position that is consistently offset from the commanded target. Its repeatability can still be excellent even though its absolute positioning accuracy is poor.

In precision servo systems, systematic errors can often be compensated through calibration and control algorithms. Once these systematic deviations are properly addressed, high repeatability becomes an important foundation for consistent high-precision operation.

Conclusion

From backlash and lost motion to transmission error, positioning accuracy, and repeatability, these five parameters form a progressive framework for understanding motion precision-from individual mechanical components to the complete motion-control system.

For engineers selecting harmonic drive reducers, there is no single "best" precision parameter.

The right approach is to evaluate the specifications according to the actual application requirements:

Backlash - How much mechanical free play exists?

Lost Motion - How much angular displacement occurs during loaded reversal?

Transmission Error - How accurately does the reducer transfer continuous motion?

Positioning Accuracy - How close can the complete system get to the commanded target?

Repeatability - How consistently can the system return to the same position?

With extensive experience in AGV drive wheels, servo motors, servo drives, and motion-control solutions, Plutools understands that component-level specifications ultimately need to be translated into real system-level performance.

There is no universal "best" specification-only the right combination of clearance, stiffness, transmission accuracy, positioning accuracy, and repeatability for a specific application.

Understanding what each parameter actually represents is the first step toward selecting the right components and building a high-performance motion-control system.