Performance and limitations

Performance

The measuring device can measure areas with a maximum size of 300 mm by 300 mm with repeatability in Z direction of 1 mm RMS on the whole area. The position accuracy in X and Y is limited by the 130 mm diameter of the laser spot of the triangulation sensor. However, X and Y accuracy is not as critical because CCD surfaces are very flat over distances much larger than the spot size. The instrument can measure 100 points equally distributed over the area of a standard 2k x 4k CCD,15 mm (30 mm x 60 mm) in less than 1 minute.The tip and tilt of the CCD can then be determined to an accuracy of 0.1 mrad.

The process of a measurement is as follows: After the dewar is put underneaththe device and aligned with respect to the X and Y axis, the triangulation sensor is moved to a position above the CCD to be measured in a manual position mode. Then the triangulation sensor is moved in Z direction so that the CCD surface is in the middle of its measurable range (5 mm). The laser spot is moved to the 4 corners of the CCD and these positions are recorded. The user selects the distance between the points to be measured and the CCD is then scanned with the defined resolution. This process is repeated for each CCD in a mosaic. It is also possible to scan circular and annular areas, as it is necessary to measure a dewar flange or window as a reference surface.

The best-fit plane is calculated from each set of measurements of the referencesurface and the CCDs. The tip, tilt and offset parameters of a CCD with respect toa reference surface indicate the alignment and corrections to be applied. Thetilt and offset of the CCDs in a mosaic can be displayed in surface plots,as shown below:

Limitations

The measuring device is not perfect, since the window in front of the CCD influences the measurement in two ways:
  1. The deviations of the optical thickness of the window between sensor and CCDdirectly compromise the accuracy. The window needs to be optically flat within thedesired accuracy (2 mm to10 mm).

  2. There is a systematic deviation if there is a relative tilt between the CCDand the window. Analysis of this effect shows that it influences only the absolutedistance measurement and not the tilt of the CCD.
If required, one way to avoid all influences of the window is to take measurements,with and without the window, between the CCD and the triangulation sensor andcompare the results.

 

Tilts of the CCDs inside WFI at roomtemperature

Tilts of the CCDs inside WFI at room temperature
The best fit planes for each of the 9 CCDs inside of the 8k x 8kWide Field Imager are plotted. This measurement was done right after the integrationof the CCDs. The CCDs are assembled in two rows of 4 CCDs each. The single CCD on theleft is used for tracking the telescope.
[High resolution view,255 kB]

 

Back to   "CCD Surface Measuring Device"