A typical CCD spectral response function is illustrated in Figure 3. This pulse-height spectrum was obtained from a flight-candidate device using a radioactive Fe source, which produces characteristic K and L X-rays of Mn at 5.9, 6.4 and 0.65 keV; only events in which the charge is confined to precisely one or two pixels (so-called one- and two-pixel events) are included in the histogram. The various parameters to be constrained by the spectral response function measurements are indicated in the figure. These include the location and asymmetric shape of the primary photopeak, the magnitude of the Si K-escape and fluorescence features, and the amplitude and shape of the extended low-energy tail. For the spectrum shown, about 2.5% of the events occur outside of the primary photopeaks. The magnitude of the various features of the spectral response function depend on a variety of parameters, including X-ray energy, detector operating temperature and bias levels, and event selection criteria.
The objective of the spectral response function measurements is to constrain the various response function components as functions of these parameters. The specific accuracy objectives of the measurements are as follows:
Some details of the spectral response function measurements made for nominal operating conditions are listed in Table 1. Reflection-grating spectrometers are used to sample the response function nearly continuously with incident energy below 2 keV. Discrete characteristic lines are used at higher energies. The apparatus used to make these measurements are described in other contributions to these proceedings [5,6].
Note that each ACIS CCD contains four output amplifiers, and each of these amplifiers serves one (rectangular) quadrant of the device. Table 1 lists the number of counts obtained for each CCD quadrant.
Table 1: Spectral Response Function Measurements