Repository navigation
Calibration Methods
#Primary Author: Laura Bergsten (lbergsten@brandeis.edu) and Hannah Herde (hherde@brandeis.edu)
The stage auto-calibrates, as long as you run the "homing" cycle the each time you turn on the stage controller and camera. Aerotech has many resources on the homing cycles and its consequences here. Each axis tracks out to its mechanical stop and then back to its centre. Axes must be enabled in order to move. While it is possible to disable the axis after each move completes, this is not recommended. Disabling an axis halts the current flow to that axis, allowing it to float freely and thus casting doubt on the stage's true position. It is important to note that we can still read out accurately the stage's position while the axis is disabled; that position will just change in time based on environmental factors, as shown in the plot.
Single target position measurements that were done overnight presented temperature dependence. In particular, the y position is dependent on the temperature of the invar and the granite table. The following documentation presents a feasible correction for this temperature dependence. Another alternative solution is to automatically control the fans in the surrounding room of the clean room all the time.
Camera calibration consists of three stages:
- Checking optical axis alignment
- Optimizing image capture height to ensure sharp images
- Determining the pixel-to-micron conversion to interpret camera coordinates in the real world.
The current camera calibration is Version 2, consistent with the V2+ of the code. Version 1 documentation is provided as legacy.
The camera's vignette allows us to assess the optical axis' alignment quality, since we use a coaxial light as our primary illumination. The coaxial light travels the same optical path as it passes out of the optics and reflects off of a piece of test silicon, lying flat on the granite table surface. An aligned axis, perpendicular to the test silicon surface, shows a symmetric, centred vignette. The optical alignment is relatively stable, and should be checked on the order of once a week or once every few weeks. There was no guiding software for this check in V1 of the code.
- Optical alignment check facilitated by determining the brightest part of the image and then assessing its eccentricity. The user can either run this VI independently, or run it through the visual inspection tool in Utils/Standalone, ContinousStageAndCameraV2.
- Camera height calibration hinged on the standard deviation of the intensity in the image.
- Pixel-to-micron conversion leverages the regularity the strips to extract the strip pitch in pixels, which can then be compared to the known pitch in microns.
- Users could test the optical alignment quality via ContinuousStageAndCamera by examining the vignetting directly. There were no assistive procedures in the first version of the code.
- Camera height calibration depended on the measured edge length of the dark-light transition in the circular fiducial marks on the bus tape.
- Pixel-to-micron conversion used a geometric match to a template of strips of a known conversion. The geometric match could scale and rotate, and by applying the same scale and rotation to the template's conversion, we could extract the pixel-to-micron conversion.
Calibrating the stave-stage basis transformation requires that we determine the stave's origin in XYZ stage coordinates and the angle of the stave with respect to the stage's x axis. The approach consists of locating several characteristics of the stave along a line (either locking points or bus tape fiducial circles), and running a fit to those positions to determine the origin and angle.
