This section will describe how TDI images2 can be used to measure conversion factor. To acquire two TDI images, the test bench is set at the setting used to record flats. There is one particular condition; the test bench must be able to open the shutter during the reading of the CCD. The monochromators are set to a wavelength of 630 nm and 10 nm bandwidth. In this condition, the photon response nonuniformities of 44-82 e2v CCDs are negligible (~1%). Most of the noise is photon noise. Equally important, the illumination of the CCD has to be homogeneous all across the surface (the ESO test bench achieved this criteria. it produces a flat illumination with fluctuations which are less than 1%). Under these conditions, it is not necessary to flat field out the TDI images. Otherwise, the method described in this paper is still usable, but it would involve more computation process stages.
The principle and the mathematical tools are the same as in the standard photon transfer curve technique . The variance in Analog Digital Units is related to the signal (ADU) by the following linear equation,
2 A TDI acquisition consists in collecting light and continuously reading out the detector one row of pixels at a time from the bottom (where the output register is located) of a detector chip. Under constant illumination, this reading out mode will create an image (see Fig. 1) with the required response.
Digital Units is related to the signal (ADU) by the following linear equation,
Sa : Average pixel value (ADU)
The variance of the signal will be plotted versus the mean signal. To extract these parameters the following manipulations must be carried out.
To measure the mean signal, the two TDI images are averaged. In the sensitive area, the mean of each row, Sa, is measured and the bias subtracted.
To measure the variance one TDI image is subtracted from the other (see Fig. 2). For each row of the sensitive area, the variance, (72, is measured (( = ( / 2. The variance is divided by two to obtain the variance of the line in a single frame). .
The plot of the (Sa , () pairs will provide the photon transfer curve and the linear fit of this plot, the slope which is the inverse of the conversion factor.
Was this article helpful?