New CCD Clock Patterns
New CCD clock patterns installed on 2004-Apr-23. These new clock
patterns correct a problem that caused the images from the FIERA CCD
controller to have 48 pre-overscan pixels and 52 post-overscan pixels,
instead of the normal 50 and 50. This is not a major problem, really
only affecting only reduction software, but easy to account for (as is
done automatically in the official FEROS-DRS
package). This problem affects ALL images acquired between the dates of
2003-October-01 and 2004-Apr-23.
The following images show that the new patterns indeed result in the
correct image format.
Figure 1: plot of row 2000, columns -50 -- 2. The images are CCDTest flatfields. The blue line is an image from immediately before the installation of the new clock patterns, the red line an image from immediately after. Clearly we see that the new clock patterns give SCIENCE data starting in column 0 instead of column -2.
Figure 2: plot of row 2000, columns 2045 - 2098. The images are CCDTest flatfields. The blue line is an image from immediately before the installation of the new clock patterns, the red line an image from immediately after. Clearly we see that the new clock patterns give SCIENCE data untill in column 2048 with the overscan pixels starting in column 2049 instead of column 2047.
Figure 3: plot of row 2000, columns 1250 - 1350. The images are CCDTest flatfields. The blue line is an image from immediately before the installation of the new clock patterns, the red line an image from immediately after. This image shows the bad column which is nominally in the 1300th column of the SCIENCE data.
Figure 4: plot of row 2000, columns 1250 - 1350 (here the x-axis is in units of mm as set by the FEROS-DRS prered.prg script). The images are CCDTest flatfields. The blue line is an image from immediately before the installation of the new clock patterns, the red line an image from immediately after. This image shows that the DRS software properly 'corrects' the bad column which is nominally in the 1300th column of the SCIENCE data. The correction of this (and other) bad column is made by averaging neighbooring pixels.
The results of the CCDTests carried out immediately before and after the installation can also be compared. CCD tests were done with the RED LEDs in both 225kHz,1,low and 625kHz,1,med readout modes. Comapre the two sets of results obtained on 2004-04-23.