Some of my recent results seem to indicate why we are seeing a landau like distribution even though we are saturating.
 

This is the result of "RUN1175" using sensor D35-S943 (no rotation), dcol=6.
The horizontal scale is 0..59 (Pixel row number) and vertical scale is 1000..2500 ADC.
Note that the log of the intensity is plotted.

This plot shows on the vertical scale ADC counts and on the horizontal scale the pixel number. In the first plot (left), the raw readout data is plotted. The thick read line is the pedestal (at approx. 2000ADC) and the green line below (approx. 1500 ADC) is the signal. It can be seen that there is a variation from pixel to pixel in its pedestal. This variation seems to depend on the position of the pixel in the column. The double column consists of 2 columns of pixel (0-29 and 30-59). It can be clearly seen that the pedestal level decreases as we go to the end of the column (pixel 29/30)and then keeps rising again, showing a symmetric distribution. (I am not sure why we are seeing such a correlation). If a shift of the pedestal of a pixel calculated with respect to the average pestestal level, then one can correct for this variation in pedestal from pixel to pixel and bring all pixel pedestals to the same level, this was done in the plot in the middle. (Note that pixel 58 is a noisy pixel).
At further correctin for the common noise gives the plot on the right.
It can be seen that eventhough a correcton was done for the shift of the pedestal, each pixel still has a different gain. If one looks close in a region where the gain of the pixels seems to be comparabale with each other like the region from pixel (10-19)

and then lools at the signal, one gets the following result (right).

now if one compares this signal height distribution of 10 pixels with the "same" gain to the signal height of the whole double column:

the signal height distribution looks more "Landau like" here because of the different gains of the pixels.

This is the result of "RUN1173" using sensor D35-S943 at a 20deg rotation angle, dcol=6.

This is the result of "RUN1200" using sensor D19-S91 (no rotation), dcol=3.
The horizontal scale is 0..60 (Pixel row number) and vertical scale is 0..2500 ADC.

the root program 'readout3.c' was used
 

Back to the Beamtest 2000 page.
Last up-date 12-14-00 by Carsten Rott:carott@physics.purdue.edu