Most motion controllers determine a motion profile after the user enters certain points along the path. The real question is this: What path does the motor take between points? There are a number of ways to answer this question.

Connecting the dots

Point-to-point moves are accomplished by specifying speed, acceleration, and deceleration, and then following this familiar equation to calculate intermediate positions:

Although this is a common and useful approach for simple motion, it has drawbacks. For starters, motion system designers often are more concerned with time than velocity and acceleration. In other words, it may be more convenient to say, "I want the motor to be at point X at time T." Further, point-to-point moving makes coordinating multiple axes difficult, as the design engineer must also calculate total move time for each axis at each point. As handy as point-to-point motion can be, more advanced modes of motion are required for more complex systems.

Contour considerations

In contour mode, an engineer specifies both a travel distance and a delta time to achieve that distance. The motion controller then follows a linear path between two contour points. For example, if the motor needs to move 100 mm in 10 msec, then the commanded position will increment by 10 mm every millisecond. Add a second point to the path at 150 mm to be accomplished in another 10 msec, and the speed for this segment drops to 500 mm/sec.

It's important to know that speed will change instantly at the 100 mm mark from 1,000 to 500 mm/sec. These discontinuities in velocity can lead to an audible "ticking" during motion as the velocity instantly changes between contour points. Depending on inertia (and magnitude of velocity change), this can be detrimental. Contour mode makes coordinating axes easier by allowing each axis to have the same delta time, but is not viable in some systems due to the instantaneous change in velocity.

Pondering PVT mode

A third mode of motion, position-velocity-time (PVT) mode, allows the same ease of coordination as contour mode, but removes velocity discontinuities. In PVT mode, an engineer specifies a travel distance and delta time, plus a velocity to be achieved at the end of the delta time. With these three pieces of information, the controller then interpolates between PVT position points using a third-order polynomial. Using a polynomial for interpolation as opposed to using linear interpolation (as in contour mode) allows for a much smoother velocity profile. Why? In contour mode, acceleration between points is infinite while velocity is constant. In PVT mode, acceleration is constant and velocity follows a second-order polynomial.

Here, the velocity profile of a contour move and a PVT move that have the same position endpoints and delta time are compared. Though the velocity and position profiles between points are different, the ending positions for each segment are the same. In systems where endpoints are critical, but paths between points are not, PVT mode can easily replace contour mode for smoother motion.

Application example: Wafer positioning

Consider a two-axis stage for moving a silicon wafer along a specific path for cleaning. In this application, it is critical for the wafer to arrive at certain points at specific times to coordinate the cleaning procedure. However, the path between points is not critical. No jerky motion is allowed, as it may displace the wafer. This application is a good candidate for PVT motion.

The engineer must now select velocities for each PVT segment. Because the first two segments in the X axis are the same in this example, it is reasonable to select a non-zero value for the ending velocity of segment one. At the end of each of the subsequent segments, the X axis must stop, so the velocities are set to zero. Note that these are ending velocities; the velocity between points will be non-zero and follow a second-order polynomial path. The Y axis changes direction at the end of each segment; as a result, the ending velocities are set to zero.

By combining this data, a motion profile can be constructed. Plotting the full X versus Y profile reveals that each specified endpoint is achieved in the allotted time. Finally, the X-Y velocity profile (previous page) shows the X axis velocity (in blue) along with the Y axis velocity (in purple). Both velocity profiles are much smoother than a similar profile constructed in contour mode.

By taking advantage of the PVT mode of motion, system engineers are able to profile much smoother motion. The main drawback of contour mode — the discontinuity in velocities — is addressed by switching to PVT mode.

For more information, contact Galil Motion Control Inc. at galilmc.com or (800) 377-6329.

Beyond X-Y positioning

In certain applications, particularly those involving nanopositiong, simple X-Y positioning won't do. Alignment and testing of MEMS components, nanolithography, and microsurgery are all instances in which six-axis positioning is necessary for optimal results. As an example of this type of extremely precise positioning system, the F-206.S Six-Axis Hexapod Alignment System from Physik Instrumente, Auburn, Mass., provides 6-D motion with 33 nanometer resolution and advanced alignment software for optics and photonics components. The parallel-kinematics design with a virtual pivot point and vector-based 6-D motion-controller differentiates the F-206.S system from conventional stacked multiaxis positioners.

Due to the reduced moving mass (one common lightweight platform for all six actuators), the system can respond and settle much faster than traditional multi-axis stage stacks (serial kinematics). Its parallel-kinematics design and digital 6-D controller minimize unwanted crosstalk motion; the virtual-pivot-point capability built into the controller imparts flexibility. Because hexapod motion is not determined by fixed bearings, but realtime six-space control algorithms instead, this point can be chosen freely with a single software command. For more information, contact pi-usa.us.

Another new positioning tool is a wireless ballbar with volumetric testing capability from Renishaw Inc., Hoffman Estates, Ill. The QC20-W is a new design that features a linear sensor and Bluetooth wireless technology, allowing for testing in "closed door" manufacturing where openings for wiring can raise safety and procedural issues. The new design also allows testing in three orthogonal planes through a single reference point. Simple hardware setup quickens testing and the ability to produce a representative volumetric measurement of positioning accuracy. The new ballbar retains the operating principle of a simple CNC circular program, coupled with powerful software. The system can quickly diagnose and quantify machine positioning errors, including servo mismatch, stick-slip errors, backlash, repeatability, scale mismatch and machine geometry, and provide an overall circularity error value.