Figure 1. All linear stages exhibit five off-axis error motion components: horizontal and vertical straightness, roll, pitch and yaw. (Image: Aerotech)

Advancements in silicon photonic and micro-optic technologies are driving the need to perform precision alignments down to sub-micrometer levels. As cutting-edge optical and photonic processes demand increasingly smaller, nanometer-scale tolerances, precision motion control is more important than ever. Achieving these alignments is crucial to quality – misalignments of even a few micrometers can result in significant power losses. As emerging chip-level functionalities and device miniaturization intensify the demand for nanopositioning technology, there is a correlated need for solutions that balance speed and precision – reducing process cycle times without compromising on quality.

Precision Motion Fundamentals

When considering a precision motion control solution for any high-tolerance photonics or optics application, it is helpful to think in terms of accuracy and repeatability. Accuracy describes how close a measured or achieved position is to the true or desired positional target. Repeatability characterizes the range of actual positions attained when the motion system is repeatedly commanded to a specific location. Because repeatability is inherent to a stage’s design and construction, a repeatable stage’s accuracy can be improved by means of error-mapping and calibration.

Several methods exist to define accuracy and repeatability, but not all supplier specifications are created equal. Metrology standards such as ISO 230-2, ASME B5.64 and others apply statistical treatments to a set of specifically defined positioning data. Beyond this, many motion suppliers have developed their own reporting methods. When in doubt, ask your motion supplier how they are defining this and request sample performance plots.

Sources of Error Motion

Accuracy and repeatability relate to errors in the direction of motion (X), but there are also five other degrees of freedom (DOF) to evaluate. For the linear stage, undesired linear motion perpendicular to the direction of travel is called straightness error, and it can occur in both the horizontal (Y) and vertical (Z) directions. Further, rotations about the X, Y and Z axes are called roll, pitch and yaw, respectively, as shown in Figure 1.

In the rotary stage, motion in the Z-direction along the axis of rotation is called axial error motion, while translation in the X and Y directions is known as radial error motion. Finally, rotation about the X and Y axes is referred to as tilt error motion, as shown in Figure 2. Note that error motions are different from total indicated runout (TIR) or often referred to simply as “runout.” Error motions describe the deviation of an axis from its ideal trajectory, whereas runout describes the measured deviation of a surface.

Figure 2. All rotary stages exhibit five off-axis error motion components: axial, radial (two directions) and tilt (two directions). (Image: Aerotech)

Understanding the causes of error motion is critical to achieving nanometer-level performance. Motor and amplifier heating result in errors related to thermal expansion and bimetallic effects. Bearing friction and cable drag forces appear as disturbances to the controller, while off-axis motor forces (e.g. cogging), sensor and electrical noise, and non-ideal bearing behavior can also generate errors. External factors including environmental temperature variation, insufficient vibration isolation and contamination further contribute. These are only a handful of possible error sources; that is why it is vital to work with a motion supplier who understands how to minimize them.

Error Budgeting

In a 6-DOF motion system, as illustrated in Figure 3, there are 36 total error motion contributions – one for each axis and degree of freedom.

Figure 3. In a 6-DOF serial-kinematic, multi-axis assembly, there are 36 total sources of error motion. (Image: Aerotech)

For multi-axis assemblies, any deviation in orthogonality and rotation-axis intersections further contributes to the total functional working point error. Estimating the total error from all of these sources is paramount to ensuring the chosen stages can achieve system-level requirements.

To begin estimating functional point error, gather fundamental information about the candidate positioning stages, including accuracy, repeatability and off-axis error motion, plus the distance between the functional point and each stage. The travel range of interest and axis alignment tolerances are also needed. Look for these details on supplier datasheets, and request clarification about how these items are defined and reported.

Next, calculate the contribution of each axis’ error motion in the X, Y and Z directions at the functional point, noting that each stage has a different distance to the functional point. Then, account for error contributions resulting from axis misalignment and intersection errors. Finally, sum the resultant errors of all axes in the X, Y and Z directions individually and combine using a root sum of squares technique. This technique provides a reasonable estimate of the functional-point error in Cartesian space. Slocum 1 provides a detailed exploration of this topic using homogeneous transformation matrices.

Compare the estimate to your process’ maximum allowable error. If you need to reduce the error estimate, first consider relocating the functional point closer to the stage assembly’s center to reduce off-axis error contributions. Additionally, operating the motion system over the smallest possible travel region can help. Further, single- and multi-axis error mapping together with precision assembly and metrology techniques should be considered. Lastly, you may need to reevaluate higher-performance stages and technologies. An experienced motion supplier can guide you through this analysis.

Stage Technologies and Selection

When considering high-performance stages for photonics applications, expect to navigate tradeoffs in bearing technologies, drive mechanisms and the kinematic architecture. Bearings constrain the motion to the desired degree of freedom and minimize unwanted motion in the other directions. Rolling-element bearings are often used in stages for photonics applications. Crossed-roller bearings deliver smoother motion and tighter geometric performance, whereas recirculating ball bearings offer longer allowable travel ranges, higher stiffness and load capacity, and tend to have a lower price point. Air bearings yield the smoothest and most precise motion performance, but they are more costly, require a clean, dry air supply and are more sensitive to debris and contamination. Flexure bearings also offer excellent geometric performance but are limited to short travel ranges, typically 1 mm or less.

Drive mechanisms play a key role as well. Indirect drives including ball screws, lead screws and belt drives are cost-effective and capable of generating considerable force or torque. However, they are susceptible to backlash and drive-screw pitch errors, require regular maintenance and lubrication, and can exhibit wear over time. Plus, they are connected to a drive shaft via a coupling, resulting in windup-related errors. Direct-drive mechanisms not only eliminate such errors but also require virtually zero maintenance and achieve higher speeds and smoother motion. Direct-drive stages are advantageous in photonics applications because the high speed and smooth motion contribute to increased throughput and quality, resulting in a lower ownership cost over time. Ultimately, multi-axis systems can use both direct- and indirect-drive stages, with the former allocated to the most critical-performance axes and the latter reserved for less critical supporting or adjacent motion.

Serial- and parallel-kinematic architectures are a key consideration in selecting positioning mechanics. In serial-kinematic arrangements stages are stacked, with each stage’s orientation corresponding to a direction of motion. Parallel-kinematic architectures, such as hexapods or Stewart platforms (shown in Figure 4), use multiple actuators in parallel to position a single platform. Hexapods require six linear inputs – one for each strut – and kinematic transformations transpose these inputs into three linear and three rotational outputs of the moving platform.

Figure 4. Hexapods are parallel-kinematic devices that use six actuators working in parallel to position a common platform. (Image: Aerotech)

Both architectures facilitate three-dimensional programming and virtual pivot-point rotations. Serial-kinematic arrangements offer intuitive visualization and straightforward programming, and they can be more accurate and repeatable at the functional point. They also offer superior design modularity and greater efficiency in 6-DOF use cases. In contrast, parallel-kinematic architectures can provide higher stiffness, smaller form-factors and clearer payload access. Because most hexapods are screw-driven, they tend to be slower than direct-drive, serial-kinematic architectures. Choosing the optimal architecture depends on the application’s priorities. An experienced motion supplier can help guide this decision.

Controller Technologies and Selection

The controller choice and setup are equally as important as stage selection. Deciding between pulse-width modulation (PWM) and linear power-stage amplifiers is key. Linear amplifiers offer extremely low noise and sensitivity to electromagnetic interference, making them ideal for nanometer-level stability and minimum incremental motion. However, they are larger and costlier than PWM drives. While PWM drives are smaller and more economical, they can exhibit switching noise. Opt for a controller architecture in which linear and PWM drives can coexist within a system.

Trajectory and servo feedback rates are another important consideration. A high trajectory generation rate, on the order of 20 kHz or faster, helps to facilitate high speed and precision. Slower trajectory rates result in having too few points to fully define the desired path, resulting in dynamic position errors. Remedies include operating at a slower velocity or increasing the trajectory rate. Some controllers even allow for a spline interpolation between trajectory points to further minimize the following error.

Choosing a controller that offers automated alignment algorithms is an enormous benefit to throughput. These algorithms are especially useful for identifying first light and then efficiently searching for peak power transmission. Many different alignment techniques exist, such as spiral, raster, hill-climb scans and more. Leading motion suppliers can assist with choosing and optimizing a search routine for your unique process. Still, alignment algorithms can only be as precise as the stages they’re controlling.

Working with Motion Suppliers

Defining a motion system for optical and photonic alignment applications requires working closely with a supplier who understands your technical and commercial challenges. A proficient supplier asks detailed questions to evaluate your priorities and offers multiple solutions with clear tradeoffs, helping you make informed decisions. A supplier who recognizes your needs and is invested in your success will likely present a framework of multiple solutions, especially if the tradeoffs are nuanced. Do not hesitate to ask how a supplier has addressed similar motion challenges in the past, and be sure to request relevant test data and performance plots to reduce technical risk. Assess a supplier’s ability and experience to help you transition from the lab to the fab. Scaling up in a technically and economically sound manner is critical but often overlooked. Furthermore, consider the supplier’s global footprint to ensure they can service localized R&D efforts and larger, global production facilities.

Precision motion control is a fundamental aspect of many cutting-edge photonic processes, and it deserves ample consideration. Whether you are manufacturing, inspecting, aligning or bonding, it is essential to have a fundamental understanding of precision motion principles to effectively engage with suppliers and choose one who will maximize your effectiveness. An ideal motion supplier will navigate your precision motion control journey alongside you as a partner who is invested in your longterm success.

This article was written by Brian M. Fink, Product Manager, Aerotech, Inc., (Pittsburgh, PA). For more information, visit here .


  1. Slocum, A. (1992). Precision Machine Design. Society of Manufacturing Engineers.