Holding Stability in Beamline Positioning with Minimum Incremental Motion

Table of Contents

Study Objective
Test Setup
Results
     Metrology Loop (Structural) Noise Floor
     Vibration Noise Floor
     Minimum Incremental Motion Testing
Minimum Incremental Motion Testing Conclusions
     Short-Term Stability (In-Position Jitter) Testing
Short-Term Stability Testing Conclusions
Long-Term Stability Testing
Long-Term Stability Testing Conclusions
Conclusion

Study Objective

This study evaluates the performance differences of motor and feedback technologies on minimum incremental motion (achievable mechanical step size), long-term stability, and short-term stability.

Short-term stability, or in-position jitter, is the amount of motion measured at the point of interest over a short period of time, which is four seconds for this study, while long-term stability is measured over a larger period of time, which is 60 minutes for this study.

Long-term stability tests are performed when the stages are at equilibrium and directly after a move-sequence in order to replicate typical motion and move sequences that exist when samples, optics, or other equipment are aligned.

In order to evaluate and isolate these aspects of performance, Aerotech has designed a mechanical positioning stage based on its ANT180-L platform that could be fitted with different drive technologies, including a linear servomotor, a 12 mm diameter x 1 mm pitch ball screw driven by a rotary servomotor, and a 12 mm diameter x 1 mm pitch ball screw driven by a stepper motor.

Crossed-roller bearings with an anti-cage-creep mechanism were used in the ANT180-L stage for all tests. For position feedback and commutation on the servomotors, optical encoders are used. The encoders include direct-metrology linear encoders (glass scales, 20 µm signal period, 1 Vpp output) in line with the direction of motion and rotary encoders (1000 lpr, 1 Vpp output) mounted to the motors.

A common mechanical platform with modular drive and feedback technologies effectively controlled variation in results caused by different bearing types, variations in assembly or manufacturing, and preload and friction.

Table 1. shows a summary of the stage motor and feedback configurations tested.

Table 1. Stage motor and feedback configurations tested

Stage Type/Drive Description of Motor/Feedback
Linear Motor Motor: 3-phase linear motor (BLM-142-A)
Linear encoder: 20 µm signal pitch encoder on glass; 1 Vpp output (0.3 nm interpolated electrical resolution)
Ball Screw and Stepper Motor Ball screw: 12 mm diameter x 1 mm pitch
Motor: 0.9°/step bipolar stepper (Lin P/N: 5709G-01P)
Full step resolution: 2.5 µm
Microstepping resolution: 0.5 nm (x5000)
Ball Screw and Servomotor (No Linear Encoder) Ball screw: 12 mm diameter x 1 mm pitch
Motor: 3-phase slotless AC servomotor (BMS60)
Rotary encoder: 1000 cnts/rev; 1 Vpp output (0.5 nm interpolated electrical resolution)
Ball Screw and Servomotor (With Linear Encoder) Ball screw: 12 mm diameter x 1 mm pitch
Motor: 3-phase slotless AC servomotor (BMS60)
Rotary encoder: 1000 cnts/rev; 1 Vpp output (0.5 nm interpolated electrical resolution)
Linear encoder: 20 µm signal pitch encoder on glass; 1 Vpp output (0.3 nm interpolated electrical resolution)

Figure 1 shows the solid models of the various stage configurations used for testing.

Figure 1. (A) ANT180-L stage with ball screw and servomotor, (B) ANT180-L stage with ball screw and stepper motor, and (C) ANT180-L stage with linear motor.

Test Setup

An air-isolated granite table in Aerotech’s Engineering Lab was used to conduct the tests. Aluminum vacuum rails on the granite were mounted with the ANT180-L stage. The work point was located at about 25 mm above the surface of the linear stage for the tests.

This is a best-case scenario in most beamline experiments, as several stage stacks and offsets can be hundreds of millimeters in length. However, the test setup used removes stage stack arrangements that are application-dependent and allows a true comparison of the drive and feedback technology.

To minimize the effects of temperature fluctuations on the metrology loop, the metrolofy fixturing was manufactuered out of Invar. To ensure that in-position stability measurements are not influenced, the test structure was designed in such a way that the structural loop stiffness was adequately high. The first resonant frequency of the metrology loop fixturing is about 615 Hz (bending in the Y- direction).

Lion capacitive sensors (Driver P/N: CPL290, Probe P/N: C9.5-5.6-2.0) and an Agilent 35670A Dynamic Signal Analyzer were used to capture position data for both the in-position stability and minimum incremental motion tests. Unless noted, data were low-pass filtered at 500 Hz (below the first metrology loop structural resonance). During long-term stability measurements, thermocouples were used to capture the temperature of the granite, air, carriage, and motor.

Linear amplifiers (Aerotech’s Ensemble HLe) were used to perform the tests. The Ensemble control platform was used as it is the Aerotech controller platform typically found in synchrotron applications. Linear amplifiers were selected in order to prevent noise from amplifiers, such as pulse width modulated (PWM) switching noise, inducing into the stage or measurement.

Figure 2 shows the test setup. It must be noted that the coordinate system shows the positive direction of the capacitive sensors used in the testing. The direction of travel is oriented in the X-direction for all of the tests.

Figure 2. Test setup.

Results

Metrology Loop (Structural) Noise Floor

The structural metrology-loop noise, including the cap sensor noise, was measured on all X, Y, and Z probes before starting the minimum incremental motion and stability tests. The test setup where the stage was replaced with a series of fixture blocks, is shown in Figure 3.

Figure 3. Test setup for measuring capacitance and metrology loop noise.

The data, captured for four seconds, were low-pass filtered to a 500 Hz measurement bandwidth. The results obtained from this measurement are shown in Table 2.

Table 2. Measured metrology loop (structural) noise

Direction RMS pk-pk
X 0.14 nm 1.11 nm
Y 0.16 nm 1.10 nm
Z 0.18 nm 1.45 nm

Vibration Noise Floor

An air-isolated granite table was used to perform the tests. The vibration noise floor, in addition to measuring the metrology-loop noise floor, was also measured before measuring the vibration levels at the granite surface. The acceleration time data measured on the granite in the X, Y, and Z directions is shown in Figure 4. The resulting acceleration spectral density for each direction is shown in Figure 5.

Figure 4. Acceleration measurements taken on the air-isolated granite.

Figure 5. Resulting acceleration spectral density (ASD) of measurements taken on the air-isolated granite.

Minimum Incremental Motion Testing

Table 3 shows the stage configurations and minimum incremental motion test results. The tests were initiated at the fundamental resolution of the stage and the commanded step size was increased until the stage was able to perform 10 forward and 10 reverse steps reliably and with reasonable accuracy.

Table 3. Stage configurations and achieved minimum incremental motion results

Stage Type and Drive Linear Encoder Resolution Rotary Encoder Resolution Measured Minimum Incremental Motion Figure Number
Linear Motor 0.3 nm N/A 1.5 nm 6
Ball Screw and Servomotor (No Linear Encoder) N/A 0.5 nm 50 nm 7
Ball Screw and Servomotor (With Linear Encoder) 0.3 nm 0.5 nm 1.5 nm 8
Ball Screw and Stepper Motor N/A 0.5 nm1 250 nm 9

1. Equivalent resolution achieved when microstepping.

Minimum Incremental Motion Testing Conclusions

  • A 1.5 nm minimum incremental motion was achieved in linear motor stage
  • In the case of the ball screw and servomotor without a linear encoder, as shown in Figure 7, a step size of 50 nm was approximately the limit where the inaccuracy of each step did not make up a large percentage of the step size. In this test case, about 25 nm of backlash existed in the positioning stage
  • Addition of the linear encoder to the ball screw and servomotor stage allowed for a 1.5 nm step size, which is the same as that of the linear motor stage
  • The stepper motor stage can obtain approximately 250 nm step sizes with no major step size inaccuracy. Smaller steps were discernible, but significant step inaccuracy was present. The vibration at the start of each step was due to the stepper motor resonance. By applying electronic damping or by using a mechanical damper attached to the stage shaft, the vibration can be minimized, but neither of the two strategies was used in this study

Figure 6. Minimum Incremental Motion step plot for the ANT180 linear motor stage.

Figure 7. Minimum Incremental Motion step plot for the ANT180 ball screw and servomotor stage without linear encoder.

Figure 8. Minimum Incremental Motion step plot for the ANT180 ball screw and servomotor stage with linear encoder.

Figure 9. Minimum Incremental Motion step plot for the ANT180 ball screw and stepper motor stage.

Short-Term Stability (In-Position Jitter) Testing

For short-term holding stability (in-position jitter), four different stage configurations were tested. Before taking the measurement, the stage configurations were allowed to equilibrate in the environment for a minimum of 12 hours, with the motor enabled. The stepper motor configuration was tested with 0% and 20% holding currents.

The RMS and pk-pk jitter values are reported at a 500 Hz measurement bandwidth, after the data was captured for four seconds. The stage configurations and results are shown in Table 4. Figures 10-14 illustrate the in-position jitter time captures and cumulative RMS jitter results for each test.

Table 4. Stage configurations and achieved short-term in-position jitter results.

Stage Type and Drive X-Direction Y-Direction Z-Direction Figure Number
RMS pk-pk RMS pk-pk RMS pk-pk
Linear Motor 0.52 nm 3.82 nm 0.16 nm 2.09 nm 0.22 nm 2.87 nm 10
Ball Screw and Servomotor, (No Linear Encoder) 0.17 nm 1.47 nm 0.22 nm 1.54 nm 0.20 nm 1.64 nm 11
Ball Screw and Servomotor, (With Linear Encoder) 0.21 nm 1.68 nm 0.18 nm 1.53 nm 0.25 nm 1.78 nm 12
Ball Screw and Stepper Motor, (0% Holding Current) 0.16 nm 1.24 nm 0.15 nm 1.12 nm 0.30 nm 2.39 nm 13
Ball Screw and Stepper Motor, (20% Holding Current) 0.16 nm 1.25 nm 0.18 nm 1.31 nm 0.25 nm 2.18 nm 14

Short-Term Stability Testing Conclusions

  • The jitter measurements for all stage configurations were below 4 nm pk-pk in the X, Y, and Z directions.
  • The screw-driven stages performed slightly better than the direct-drive stage in this test.
    • It should be noted that Aerotech also builds linear-motor stages and drives that allow the direct-drive motor jitter to rival that of the screw-based systems.

Figure 10. In-position jitter measurements for the ANT180 linear motor stage.

Figure 11. In-position jitter measurements for the ANT180 ball screw and servomotor stage without linear encoder.

Figure 12. In-position jitter measurements for the ANT180 ball screw and servomotor stage with linear encoder.

Figure 13. In-position jitter measurements for the ANT180 ball screw and stepper motor stage with 0% holding current.

Figure 14. In-position jitter measurements for the ANT180 ball screw and stepper motor stage with 20% holding current.

Long-Term Stability Testing

Long-term holding stability was tested for the four different stage configurations for over 60 minutes. First, a baseline test was conducted, where the stage was allowed to equilibrate to the environment for a full day, with the motor enabled. The temperature of various parts of the measurement setup and stage were captured and the in-position stability was measured in the X, Y, and Z directions.

Once the long-term baseline test was complete, the stages were commanded to perform a relatively benign move profile back-and-forth over 50 mm for 45 seconds at 6 mm/s and an acceleration/deceleration rate of 200 mm/s2.

Then, the stage was brought into the capacitance probe range to capture the in-position stability for 60 minutes. This move-profile test replicated positioning of a sample or movement of the beamline motion platform before the experiments were performed.

Table 5 shows a summary of the long-term stability tests. Figures 15-24 show the measured drift and temperature data for the tests.

Table 5. Stage configurations and achieved long-term stability results.

Stage Type and Drive Move Routine Prior to Test pk-pk X Stability (nm) pk-pk Y Stability (nm) pk-pk Z Stability (nm) Figure Number
Linear Motor No 10.9 nm 9.3 nm 12.7 nm 15
Ball Screw and Servomotor (No Linear Encoder) No 16.5 nm 9.9 nm 18.8 nm 16
Ball Screw and Servomotor (With Linear Encoder) No 8.2 nm 9.5 nm 16.5 nm 17
Ball Screw and Stepper Motor (0% Holding Current) No 7.0 nm 6.7 nm 15.8 nm 18
Ball Screw and Stepper Motor (20% Holding Current) No 10.1 nm 5.5 nm 19.0 nm 19
Linear Motor Yes 11.7 nm 6.4 nm 21.5 nm 20
Ball Screw and Servomotor (No Linear Encoder) Yes 115 nm 7.2 nm 28.1 nm 21
Ball Screw and Servomotor (With Linear Encoder) Yes 38.2 nm 4.9 nm 15.8 nm 22
Ball Screw and Stepper Motor (0% Holding Current) Yes 283 nm 14.3 nm 136 nm 23
Ball Screw and Stepper Motor (20% Holding Current) Yes 288 nm 15.0 nm 134 nm 24

Long-Term Stability Testing Conclusions

  • All stages performed in a similar manner in the baseline long-term stability tests (no move routine). The differences in measured drift between tests (5 - 10 nm) were within the expected amounts due to temperature variations in the room over the measurement time of 60 minutes
  • When a move routine was carried out, the measured drift in the motion direction of the stepper-motor stage was 23 times higher than the linear-motor stage, and about 7 times higher than the servomotor with linear encoder
  • For the test case with the ball screw/servomotor and move routine, approximately 115 nm of drift was observed in the X-direction with minimal temperature increase at the measured points on the stage (motor experienced about 0.1 °C). The cause of this drift is probably due to heating and hysteresis in the ball screw, since the measured displacement did not return to the starting position after being allowed to equilibrate for 2 - 3 hours
  • As expected, the linear encoder (direct-feedback) helps to substantially reduce drift in the direction of motion

Figure 15. In-position stability (long-term stability) measurements for the ANT180 linear motor stage (baseline)

Figure 16. In-position stability (long-term stability) for the ANT180 ball screw and servomotor stage without linear encoder (baseline)

Figure 17. In-position stability (long-term stability) for the ANT180 ball screw and servomotor stage with linear encoder (baseline).

Figure 18. In-position stability (long-term stability) for the ANT180 ball screw and stepper motor stage, 0% holding current (baseline).

Figure 19. In-position stability (long-term stability) for the ANT180 ball screw and stepper motor stage, 20% holding current (baseline).

Figure 20. In-position stability (long-term stability) measurements for the ANT180 linear motor stage with move routine prior to testing.

Figure 21. In-position stability (long-term stability) for the ANT180 ball screw and servomotor stage without linear encoder with move routine prior to testing.

Figure 22. In-position stability (long-term stability) for the ANT180 ball screw and servomotor stage with linear encoder with move routine prior to testing.

Figure 23. In-position stability (long-term stability) for the ANT180 ball screw and stepper motor stage with move routine prior to testing, 0% holding current.

Figure 24. In-position stability (long-term stability) for the ANT180 ball screw and stepper motor stage with move routine prior to testing, 20% holding current.

Conclusion

A ball-screw-driven stage with a linear amplifier performs best in applications that require the highest levels of in-position jitter (shorter-term stability). Direct-drive linear-motor stages rival ball-screw-driven stages with not much change in-position jitter (~4 nm compared to ~2 nm pk-pk). Aerotech​ has standard linear-motor stage designs with in-position jitter values in the 1-2 nm range.

Direct-drive linear-motor stages substantially outperform screw-driven stages in their ability to make small mechanical movements (minimum incremental motion) and maintain better holding stability over longer time periods. Both can be improved by adding a linear encoder to a ball-screw servomotor stage; however, speed and smoothness of motion typically suffer.

For instance, several new X-ray imaging applications need brighter X-rays (higher brilliance). However, the samples can get damaged if not imaged quickly with these brighter X-rays. Therefore, many applications are opting for scanning and avoiding the step-measure-step-measure approach. Direct- drive stages are better for such scanning applications because of a noncontact drive mechanism.

Stepper-motor stages suffer from a major drawback - heat - although they provide a cost-effective and simple way to perform positioning. When sitting at rest, it is quite common to reduce the current to lower levels (sometimes 0% holding current).

However, as this testing has shown, even 45 seconds of stepper-motor stage movement will cause the temperature to rise by a few degrees causing hundreds of nanometers of drift and taking much more than one hour to come to equilibrate.

Vacuum-based applications usually exacerbate this drift as thermal management is more difficult. Servomotor stages are inherently advantageous as they only use current when needed, effectively minimizing heating effect and thermal drift effect and allowing for a more stable positioning system.

As the effort to image samples at higher resolution progresses, the positioning system will play an increasing role in realizing high-quality images. In applications where submicron or even nanometer precision in sample and optics positioning is critical, the capability for extremely fine positioning and stability makes rotary and linear servomotors an ideal choice.

This information has been sourced, reviewed and adapted from materials provided by Aerotech, Inc.

For more information on this source, please visit Aerotech, Inc.

Tell Us What You Think

Do you have a review, update or anything you would like to add to this article?

Leave your feedback
Submit