Phase H data

We first analyzed data collected during Phase H (i.e., ACIS + the High Resolution Mirror Assembly). We considered only data with high signal-to-noise ratio beam normalization spectra, and analyzed only the ACIS telemetry mode observations; the High Speed Tap data lead to similar results (e.g., Kastner, Allen, & Glotfelty 1998).

Table 6.21 lists the Phase H data used in this study, in order of increasing source flux. Data from both chip I3 of ACIS-I, corresponding to a Front-Illuminated (FI) CCD, and from chip S3 of ACIS-S (Back-Illuminated, BI, CCD), were considered. Data from FI and BI chips are listed in Table 6.21a and 6.21b, respectively. The TRW IDs of the tests are listed in column 1, while columns 2 and 3 list the energy of the source (Al-K$\alpha$, 1.486 keV; O-K$\alpha$, 0.525 keV) and the exposures. The latter correspond to the net integration times after screening of the data, and after removing intervals of data dropouts. All the data in Table 6.21 were accumulated in staggered mode with only 38 rows centered on the aimpoint being read out. All data were taken with a well collimated, in focus beam.
 
 
 

Table 6.21: XRCF Phase H Pileup Data

Test Name	Energy (keV)	Exposure (s)	BND Flux (counts s-1 cm-2)	Predicted ACIS counts/frame a
				All-255	G02346	G023	G0
a) Front-illuminated CCD
H-IAI-CR-1.001	1.486	7555	0.0045	0.321	0.319	0.283	0.231
H-IAI-CR-1.003	1.486	2624	0.0166	1.194	1.187	1.150	0.859
H-IAI-CR-1.005	1.486	818	0.0530	3.807	3.784	3.363	2.739
H-IAI-CR-1.007	1.486	595	0.1017	7.435	7.370	6.566	5.348
H-IAI-CR-1.009	0.525	7145	0.0342	0.489	0.489	0.452	0.369
H-IAI-CR-1.011	0.525	2248	0.0987	1.417	1.415	1.306	1.152
b) Back-illuminated CCD
H-IAS-CR-1.002	1.486	7459	0.0037	0.323	0.312	0.202	0.115
H-IAS-CR-1.004	1.486	2419	0.0129	1.129	1.089	0.702	0.402
H-IAS-CR-1.006	1.486	932	0.0432	3.834	3.700	2.236	1.366
H-IAS-CR-1.008	1.486	533	0.0838	7.344	7.084	4.570	2.616
H-IAS-CR-1.010	0.525	2910	$\cdots$b	$\cdots$	$\cdots$	$\cdots$	$\cdots$
H-IAS-CR-1.012	0.525	7215	0.0065	0.344	0.335	0.239	0.124

a=Predicted unpiled ACIS count rate per frame as a function of grade (see text);
b=Poor quality data.
 

We define the frametime t_f as the time during which the CCD is integrating data during one frame:

where T is the net exposure (in seconds), f is the number of frames during the exposure, t is the frame transfer time (41 ms), and n is the fraction of rows that are read out (n=0.0371 in 38 row staggered mode). For the data in Table 6.21, we find t_f=0.1153 s for all cases except test I-HAS-CR-1.006, where t_f=0.1167 s.

Columns 4 of Table 6.21 reports the source flux at the HRMA entrance, determined from the beam normalization detectors (BND). Columns 5-8 list the predicted incident (unpiled) ACIS counts/frame for different grade selections: unbinned ACIS grades excluding grade 255 whose events are most likely due to cosmic rays, and the ASCA-like grades G02346, G0234, and G0. The counts/frame were derived by multiplying the BND fluxes for the timeframe t_f and a factor including the effective area of the mirror, the detector quantum efficiency, and the filter transmission. The detector quantum efficiency is grade-dependent (Chartas et al. 1998), so we list the counts/frame corresponding to each grade selection; for reference, for G02346 and at Al-K$\alpha$ the HRMA/ACIS effective area is 619.45 cm² for FI chips and 733.18 cm² for BI chips, and at O-K$\alpha$ it is 124.3 cm² for FI and 446.8 cm² for BI. The effective areas were obtained from the Telescope Scientist Team web page: http://hea-www.harvard.edu/MST/mirror/www/home.html (Note that the effective areas provided at this site are the result of scaling the XRCF calibration data to the raytrace effective area curve. Calibration data differ from raytrace as much as 15%.) The ACIS I3 and S3 CCD quantum efficiencies were derived from the MIT Web pages:http://acis.mit.edu/cal/w215c2r_eff_897.qdp for the FI chip, andhttp://acis.mit.edu/cal/w134c4r_eff_pre_997.qdp for the BI chip. The transmission data for the ACIS OBF's are taken from the PSU anonymous FTP site ftp.astro.psu.edu in /pub/gc/filters/acis_i.data for the Imager, and in /pub/gc/filters/acis_s.data for the Spectrometer. The quantum efficiencies for the BND detectors were taken from the SAO Web pagehttp://hea-www.harvard.edu/MST/simul/xrcf/HXDS/index.html. Relatively large systematic errors of up to 10% remain between the model fits and the measured BND quantum efficiencies. The newly released BND quantum efficiencies differ considerably from the values used in earlier papers (e.g., Kastner et al. 1997).

Tests H-IAI-CR-1.001, 1.003, and 1.005 were previously analyzed by Kastner et al. (1998). Among the tests done with ACIS-S in Table 1b, test H-IAS-CR-1.010 provided a spectrum of poor quality, and will not be considered any further.