Boller & Chivens CCD Page 2


General Information and Page Index

UV Flooding

Last UV flooding has been made in date February 23th, 1998 by the ODT on La Silla.
The last short test on the CCD (June 13th, 1998) show no particular change in the CCD sensitivity since the last UV flooding. One of the (undesired) effects of UV flooding is the broadening of the spectral lines and, therefore, of the instrumental line width. The amplitude of this effect decreases with wavelength. The following plot gives an idea of the relationship between the real slit width and the instrumental FWHM. The measurements were made with grating #26 (grating angle 19o) and using 8 He-Ar lines form 492.1nm to 573.9nm. Dots represents the FHWM of the 8 lines, as measured by mean of the ALICE MIDAS context, as a function of slit width in pixels. The line represents an empirical fit to the data. The empirical relation is shown at the top of the plot.

[Click here
for the PS file]

Bias Frames

The bias frame shows a small gradient in both column (i.e. along the dispersion) and row (i.e. along the slit) directions. The structure is due mainly to the warming up of the amplifiers when reading the CCD and depends on the windowing applied. In a bias frame the difference between the regions with highest and lowest counts is at most 3 ADU (note the R.O.N is about 4.5 ADU). Because of the structure, the bias frame changes slightly when changing the windowing of the CCD. The results presented in this page have been obtained with a window defined [801,1:1150,2060] (MIDAS notation) but they are qualitatively valid for any window. The mean bias level is stable at a level of 0.15% (0.3 ADU) on short time scales (1-2 hours) and at a level of 0.3% (0.6 ADU) on time scale of a week. The mean bias level is about 177.5 ADU in the centre.
Plots and Images
Raw bias frame (fits file or gif file).
Average bias frame (180 frames) (fits file or gif file). Note that on this frame the rms around the bias value is of the order of 0.32 ADU and therefore the vertical structures that is possible to see here are much below the R.O.N of raw frames.
Bias profile along the column direction. By averaging all columns of the average bias frame (see previous blue point) we obtain a high S/N column profile. A gradient is present toward the upper part of the frame (i.e. toward the red) of about 0.0006 ADU/pixel. Also a faint periodic signal is present. The power spectrum of the Fourier transform indicates that the spatial period of this signal is of 29.41 pixels.
Bias profile along the row direction. By averaging all rows of the average bias frame we obtain a high S/N column profile. In the first 20 columns a steep gradient is present. Looking the profile in more detail, we can also notice the gradient of about 0.0011 ADU/pixels and the narrows peaks giving rise to the vertical structure visible in the average bias frame
Bias level as function of time on raw frames and average frames. The mean value as been computed in a region of 95x70 pixels in the center of the CCD. Each average frames has been obtained combining 10 raw bias frames (a - clipping algorithm has been used to avoid cosmic rays).
Read Out Noise as function of time. The same area 95x70 pixels has been used to compute this value on raw frames.

ESO 1.5-m
Home Page
2p2 Team
CCD Page
2p2 Team
Home Page
Last update: Oct.98.
Comments to 2p2team@eso.org