Positioning Precision Takes Center Stage

Paula M. Powell

These basic laws of physics can frustrate any system designer’s best efforts to move a component from point A to point B rapidly and to stop it on the desired target. In most cases, the positioning system must repeat this action over and over with some precision.

Bill Hennessy, president of Alio Industries in Loveland, Colo., said that the positioning challenges are increasingly compounded by the demands for motion in the submicron and nanometer ranges. “This trend is prevalent in numerous positioning-sensitive industries, from fiber optics, biomedical and micromachining to semiconductor and optical component processing, as well as research and development. Microelectromechanical systems and semiconductor devices, in particular, are now requiring line-density line spacing below 100 nm.”

Regardless of the application, most positioning systems still rely on three basic drive systems: stepper or servomotors with ball-screw drives, linear motors and piezoelectric-driven stages. All have a place in helping photonics engineers design processes, but they come with varying levels of accuracy, repeatability and cost.

In nanometer-scale applications, for instance, piezoelectric scanning technology is common. One example is a near-atomic-scale topography system that is used to measure 20-nm-deep indents on fused silica (PI’s MTS Nano Indenter XP). The system relies on a subnanometer-precision parallel-metrology piezoelectric scanning stage with capacitive positioning sensors.

In addition, conventional multiaxis positioning systems have offered both linear and rotary motion via a system of stacked stages. This process, according to engineers at Melles Griot in Ely, UK, can cause a buildup in motion errors — especially in a six-axis stage.

Error in the stack typically comes from the misalignment between the measurement and the motion axis. One problem is cosine error related to the fact that the X-axis of the first stage in the stack is often not collinear or orthogonal with that of the second stage. One must add to that any Abbe errors, which are related to the fact that the height of the second stage will amplify any pitch error in the first stage.

The Melles Griot engineers report that measurement error continues to build with any orthogonal motion in the multistage system, as well as with any structural deflections within the sensor. Even flexure stages that are used for precise, small-scale positioning applications are prone to error buildup in a stacked arrangement — something newer systems are increasingly designed to avoid.

Flexure-based stages, which have been around for some 30 years, are based on the ability of certain solid materials to deform elastically under force and to return to their prior state when the force is removed. This design is frictionless and stictionless, with no sliding or rolling, which, according to engineers at PI USA in Auburn, Mass., enables them to offer better resolution, straightness and flatness compared with conventional roller- or ball-bearing designs. For example, the straightness and flatness of a precision X-Y flexure scanning stage is often on the order of a few nanometers, with tilt of only a few arc seconds.

So what’s the catch? It is that flexure designs are limited in their travel length and, according to some reports, may not be as durable as other technologies. One solution might be to rely on flexure-type designs for precision positioning and on other stage technology for “coarse” travel. Engineers at Aerotech Inc. in Pittsburgh say that a successful positioning strategy ultimately will depend on how well system designers understand the unique problems of a particular application, such as laser seam welding. More in-depth discussion on such issues can be found in the following articles.