 |
|
|
Nanopositioning & Scanning Systems:
How to Select the Right Nanopositioning System
- Introduction to Piezo Flexure Nanopositioners and Scanners |
|
|
|
Piezo Nanopositioning and Scanning Systems vs. Traditional Micropositioning Stages
PI Nanopositioning and scanning systems with frictionless flexure guidance are decidedly superior to positioners with conventional guiding systems (crossed roller bearings, etc.) in terms of resolution, reproducibility, straightness and flatness. Due to their inherent friction and limited guiding precision, traditional positioners are best used in applications requiring repeatability on the order of 0.1 µm, even though encoder readout may indicate much higher resolution. In contrast, PI piezo-driven flexure nanopositioners can easily achieve repeatability and minimum incremental motion in the subnanometer realm.
Higher Speed
With piezo drives, capable of accelerations of up to 10,000 g, and their low moved mass, such piezo stages can provide significantly higher scanning speeds than motorized systems.
Why Flexures?
Flexure motion is based on the elastic deformation (flexing) of a solid material. Friction and stiction are entirely eliminated, and flexures exhibit high stiffness, load capacity and resistance to shock and vibration. Flexures are maintenance free and not subject to wear. They are vacuum compatible, operate over a wide temperature range and require neither lubricants nor compressed air for operation.
Excellent Guiding Accuracy
The multilink flexure guiding systems employed in most PI piezo nanopositioners (Fig. 2) eliminate cosine errors and provide bidirectional flatness and straightness in the nanometer or microradian range. This high precision means that even the most demanding positioning tasks can be run bidirectionally for higher throughput.
Lifetime / PICMA® Piezo Actuators
PI nanopositioning systems employ the award-winning PICMA® piezo actuators, the only actuators with cofired ceramic encapsulation. The PICMA® piezo technology was specifically developed by PI’s piezoceramic division to provide higher performance and lifetime in nanopositioning applications.
Multilayer piezo actuators are similar to ceramic capacitors and are not affected by wear and tear. PI nanopositioning systems are designed to be driven at lower voltages than most other piezo systems (100 V vs. 150 V). The research literature, PI’s own test data and 30+ years of experience all confirm that lower average electric fields, lead to longer lifetime.
|
|
P-733 low-profile XY and XYZ scanning stages.
Watch Piezo Stage Animation
Fig. 1. Long-travel (10 mm) multi-dimensional, Roberts-linkage flexure system prevents parallelogram errors in XYZ positioning applications.
|
|
|
|
Measuring Nanometers: Stage Metrology Selection |
|
|
|
|
|
Achieving nanometer and subnanometer precision requires more than a piezo stage capable of making moves on this precision scale. The stage internal metrology system must also be capable of measuring motion on the nanometer scale. The five primary characteristics to consider when selecting a stage metrology system are linearity, sensitivity (resolution), stability, bandwidth, and cost. Other factors include the ability to measure the moving platform directly and contact vs. noncontact measurement. Three types of sensors are typically used in piezo nanopositioning applications—capacitive, strain, and LVDT. Table 1 summarizes the characteristics of each sensor type.
PI capacitive sensors measure the gap between two plates based on electrical capacitance. These sensors can be designed to become an integral part of a nanopositioning system, with virtually no effect on size and mass (inertia). Capacitive sensors offer the highest resolution, stability, and bandwidth. They enable direct measurement of the moving platform and are noncontact. Capacitive sensors also offer the highest linearity (accuracy). PI's capacitive sensors / control electronics use a high-frequency AC excitation signal for enhanced bandwidth and drift-free measurement stability (subnanometer stability over several hours, see here). PI’s exclusive ILS linearization system further improves system linearity. If used with PI’s digital controllers, digital polynomial linearization of mechanics and electronics makes possible overall system linearity of better than 0.01%. Capacitive sensors are the metrology system of choice for the most demanding applications.
A strain gauge sensor is a resistive film bonded to a piezo stack or—for enhanced precision—to the guiding system of a flexure stage. It offers high resolution and bandwidth and is typically chosen for cost-sensitive applications. As a contact type sensor, it measures indirectly, in that the position of the moving platform is inferred from a measurement at the lever, flexure or stack. PI employs full-bridge implementations with multiple strain gauges per axis for enhanced thermal stability. PI's PICMA® drive technology also enables higher performance of actuator-applied strain gauge sensors.
LVDT sensors measure magnetic energy in a coil. A magnetic core attached to the moving platform moves within a coil attached to the frame producing a change in the inductance equivalent to the position change. LVDT sensors provide noncontact, direct measurements of position. They are cost-effective and offer high stability and repeatability. |
|
Fig. 2. Wire-EDM cutting process provides highest-accuracy flexure guiding systems in compact nanopositioning stages.
Fig. 3. Response of a PI Nanopositioning stage to a square wave control signal clearly shows the true sub-nm positional stability, incremental motion and bidirectional repeatability.
Measured with external capacitive gauge, 20 pm resolution. |
| |
Table 1
| Sensor Type |
Sensitivity* (Resolution) |
Linearity* |
Stability* / Repeatability |
Bandwidth* |
Metrology Type |
Excitation Signal |
| Capacitive |
Best |
Best |
Best |
Best |
Direct / Noncontact |
AC |
| Strain |
Better |
Good |
Good |
Better |
Inferred ** (Infect) / Contact |
DC |
| LVDT |
Good |
Good |
Better |
Good |
Direct / Noncontact |
AC |
|
| |
* Note. The ratings describe the influence of the sensor on the performance of the whole nanopositioning system. Resolution, linearity, repeatability, etc. specifications in the PI product data sheets indicate the performance of the complete system and include the controller, mechanics and sensor. They are verified using external nanometrology equipment (Zygo Interferometers). It is important not to confuse these figures with the theoretical performance of the sensor alone.
** Strain type sensors (metal foil, semiconductor, or piezoresistive) infer position information from strain. |
| |
 to the top |
|
|
|
|
| Parallel and Serial Designs, Controller Choice |
There are two ways to achieve multi-axis motion: parallel and serial kinematics. Serial kinematics (nested or stacked systems) are simpler and less costly to implement, but they have some limitations compared to parallel kinematics systems (see here for more information).
In a multi-axis serial kinematics system, each actuator (and usually each sensor) is assigned to exactly one degree of freedom. In a parallel kinematics multi-axis system, all actuators act directly on the same moving platform (relative to ground), enabling reduced size and inertia, and the elimination of microfriction caused by moving cables (Fig. 4). This way, the same resonant frequency and dynamic behavior can be obtained for both the X and Y axes. The advantages are higher dynamics and scanning rates, better trajectory guidance as well as better reproducibility and stability.
Direct Parallel Metrology:
Multi-Axis Measurements Relative to a Fixed Reference
Parallel kinematics facilitates implementation of Direct Parallel Metrology—measurement of all controlled degrees of freedom relative to ground. This is a more difficult design to build but it leads to clear performance advantages.
A parallel metrology sensor sees all motion in its measurement direction, not just that of one actuator. This means that all motion is inside the servo-loop, no matter which actuator may have caused it, resulting in superior multi-axis precision, repeatability and flatness, as shown in Fig. 5. Direct parallel metrology also allows stiffer servo settings for faster response. Off-axis disturbances—external or internal, such as induced vibration caused by a fast step of one axis—can be damped by the servo.
Analog and Digital Controllers
PI manufactures a large variety of analog and digital nanopositioning controllers (see here).
State-of-the-art PI digital control systems offer several advantages over analog control systems: coordinate transformation, real-time linearity compensation and elimination of some types of drift. Digital controllers also allow virtually instant changes of servo parameters for different load conditions, etc. However, not all digital controllers are created equal. Poor implementations can add noise and lack certain capabilities of a well-designed analog implementation, such as fast settling time, compatibility with advanced feed-forward techniques, stability and robust operation.
PI digital controllers can download device-specific parameters and calibration information from ID-chip-equipped nanopositioning stages, facilitating interchangeability of nanomechanisms and controllers.
All PI nanopositioning controllers (analog and digital) are equipped with one or more user-tunable notch filters. A controller with notch filter can be tuned to provide higher bandwidth because side-effects of system resonances can be suppressed before they affect system stability. For the most demanding step-and-settle applications, PI’s exclusive Mach™ InputShaping® implementation is available as an option. |
|
Fig. 4. Principle of a PI XY-Theta-Z, minimum-inertial-mass, monolithic, parallel kinematics nanopositioning system. Accuracy, responsiveness and straightness/flatness are much better than in stacked multi-axis (serial kinematics) systems. See here for more details.
Fig. 4b. Parallel Kinematics XY-Theta-Z Nanopositioning Stage Animation.
Fig. 5. Flatness of an active-trajectory-controlled nanopositioning stage over 100 x 100 µm scanning range is about 1 nm. |
|
|
|
| Controllers / Interfacing |
Digital Dynamic Linearization (DDL)
The E-710 digital controller (p. 6-14 ff.) features an internal algorithm that can eliminate tracking errors and nonlinearity in repetitive waveforms to increase linearity and effective bandwidth in scanning applications by up to 1000-fold (Fig. 6a, 6b).
Choice of Interface:
Digital or Analog?
Analog interfacing provides high bandwidth and remains the most common way of commanding piezoelectric motion systems. It is usually the choice when the control signal in the application is provided in analog form. A key advantage of analog interfacing is its intrinsic deterministic (real-time) behavior, contrasted to the difficulty of accurately timing high-bandwidth communications on present-day multitasking PCs.
However, when analog control signals are not available, or when a significant distance between the control signal source and the nanopositioning controller would affect signal quality, digital interfacing, which must not be confused with digital control, is the preferred choice.
Digital signals can be transferred through copper wires, or for complete EMI immunity, through optical fibers.
Five types of digital interfaces are typically used in piezo-nanopositioning applications: parallel-port, RS-232, IEEE 488, USB, Fiber Link and, with some digital controllers, direct DSP links. For dynamic, high-precision applications, the exact timing of an interface is more important than the data transfer rate.
Interface Bandwidth vs. Timing
Piezo-driven stages can respond to a motion command on a millisecond or microsecond time scale.
That is why synchronization of motion commands and data acquisition have a high impact on the quality of many applications, like imaging or micromachining. The USB, for example, was designed to transfer huge blocks of data at high speeds, but exact timing was a much lesser concern. While insignificant in less responsive positioning systems, this kind of non-deterministic behavior may not be tolerable in high-speed tracking or scanning applications. Each motion command—comprising just a few bytes—must be transferred instantaneously and without latency. A lower-bandwidth bus with higher timing accuracy may perform better in many applications.
There are several factors that affect the response of a digital interface: the timing accuracy of the operating system on the controlling computer; the bus timing protocol; the bandwidth of the bus; and, the time it takes the digital interface (in the piezo controller) to process each command. Parallel-port interfaces do not require command parsing and offer the best combination of throughput and timing accuracy.
In addition, to the interface properties, the bandwidth of the nanopositioning system (mechanics and servo) matters. A slow system (e.g. 100 Hz resonant frequency) will not benefit from a responsive interface as much as a high-speed mechanism.
CE Compliance
All standard PI nanopositioning systems are fully CE compliant. |
|
Fig. 6a. Six-axis digital piezo controller with Super Invar 6D- nanopositioning stage. All PI nanopositioning systems and controllers are fully CE compliant.
Fig. 6b. Rapid scanning motion of a P-621.1CD (commanded rise time 5 ms) with the E-710 controller and DDL option. Digital Dynamic Linearization virtually eliminates the tracking error (<20 nm) during the scan. The improvement over a classical PID controller is up to 3 orders of magnitude, and increases with the scanning frequency. |
| to the top |
|
