calibration results by item detector hrc build2 AUTHORS : Marco Sirianni, Gabriel Perdue, William Jon McCann
DETECTOR : HRC - Flight Build 2
PURPOSE :
To calculate the CCD gain (uV/e-) of the HRC flight build 2 device (7282LND01-02).
DATA :
The images used were taken in the February of 1998 at the Ball Aerospace PI bench as part of the screening test procedures. The CCD was operated at -193 K and the clock timing was : HRCFLTB 2X parallel NON MPP. All the images used in this test were taken at 800nm with a filter ND=3.3 and an electronic gain of 1.39 uV/DN.
METHOD :
The gain is calculated using the photon transfer method. The data required for this test consist of sequence of pairs of flat field frames so that the exposure times of both images of a pair are the same. The sequence of the images have an illumination level which spreads evenly over the entire dynamic range from just above the bias to the saturation. A script, pt.pro, written by Marco Sirianni, was used for the data reduction,. The program takes a series of paired images and perform the bias subtraction . To eliminate every flat field variation, in each pair of images, both with the same exposure time, one frame is subtracted from the other. The result of the subtraction is used to compute the variance and the mean frame of the bias subtracted images of each pair to evaluate the mean signal. The transfer curve is plotted as the variance along the y-axis versus the the mean signal accross the x-axis. The slope of the curve, below saturation, is the reciprocal of the conversion factor of the CCD system in electrons/DN.
In order to eliminate every local dependence which could lead to a wrong value of the system gain, several 40x40 pixels subarrays have been selected across the chip and the conversion factor has been calculated for each different region. Before preforming the photon transfer calculation, in each subarray, deviant pixel values were substituted with the median value of the entire subarray, and the "cleaning" loop was repeated until all the pixel values were within 4-5 sigma from the median value. The CCD gain in units of uV/e- is calculated from the total system gain and the electronic gain :
CCD Gain (uV/e-) = Electronic Gain (uV/DN) / System Gain (e-/DN).
In addition, a similar program, acsgain.pro, written by Jon McCann, was used to verify some of the results.
RESULTS :
Table 1 lists the files used and the results obtained for each amplifier. Two plots are provided for each aplifier : the signal vs. variance plot with the fitted slope and the distribution of the gain found in each subsection of the chip.
Few flat fields were taken on Nov-Dec 1998 at GSFC with this CCD in the camera. For each gain setting only two signal levels are available (see Table 2). These results have hence to be taken into account only as approximate values for the total system gain (e-/DN) of the camera. A more accurate determination of the gain setting will be available after the next thermal vacuum campaign.
TABLES :
