Pileup experiments and determination of $\hat{\epsilon}$ and $\epsilon$

The experimental technique to measure $\epsilon$ and $\hat{\epsilon}$used the High Energy X-ray Source (HEXS) at MIT's CCD Calibration Facility. HEXS is the same source used for the ACIS quantum efficiency calibration, so careful measurements of pileup using that source are particularly important for the AXAF program. The HEXS source uses fluorescence from 12 different targets ranging from Al to Ge as shown in Table  4.20. The fluorescence spectrum is generated by the bremsstrahlung spectrum from a commercial electron impact x-ray tube using a Mo target. The tube current and voltage are independently adjustable and are temporally stable to within one percent over relevant measurement periods. The maximum tube power is 9 W, with a maximum voltage of 30 kV.
 
 

Table 4.20: HEXS targets, energies, and x-ray penetration depths.

 

Target	Energy	X-ray mfp in Si
	(eV)	($\mu$m)
Al	1487	8.0
Si	1740	12.4
P	2015	1.6
Cl	2622	3.1
Ti	4508	13.6
V	4949	17.7
Fe	6399	36.9
Co	6925	46.4
Ni	7471	57.8
Cu	8040	71.6
Zn	8630	88.0
Ge	9874	130.7

All detectors used for this analysis are flight-like ACIS CCDID-17s, produced by MIT Lincoln Laboratories, which have a 1024 column by 1026 row array of 24 $\mu$m square pixels. As mentioned above, each CCD is divided into 4 readout quadrants of 256x1026 pixels. Since the gain from each quadrant can be different, most analysis is conducted on a quadrant basis rather than on the entire CCD. Each CCD was flight qualified for ACIS although only w140c4r was selected for the flight focal plane.

A series of 11 pileup measurements were conducted using various CCDs, HEXS configurations, electronics and exposure times. The different configurations are listed in Table  4.21 . For each configuration, the CCD was exposed to x-rays from most of the 12 available targets. For each target, a sequence of exposures was taken with 4 or 5 different x-ray fluxes generated by using different tube currents (the tube voltage was held constant at 15 kV). The range of fluxes covered an approximately even spread up to twice the nominal flux used during the ACIS calibration.
 
 

Table 4.21: Pileup measurement configurations.

 

CCD	Date	Electronics	Exposure Time	Comments
			(secs)
w103c4	03jun96	Lbox	7.22	Early HEXS configuration
w103c4	16jan97	Lbox	7.22
w103c4	12feb97	Lbox	7.22
w140c4r	22jan97	DEA	3.28	Back sided CCD
w163c3	27sep96	DEA	3.28
w190c3	26nov96	DEA	3.28
w203c2	07may97	DEA	3.28
w203c2	08may97	DEA	7.15	Lbox exposure time
w203c4r	16jan97	DEA	3.28
w210c3r	12may97	DEA	3.28
w210c3r	12may97	DEA	7.15	Lbox exposure time

An important assumption for the process is that the x-ray flux from HEXS varies linearly with the x-ray tube current. This was checked two different ways. First, the actual tube current was measurered using three different current meters and compared to the front panel display. All agreed within error. Secondly, the total electron charge detected by the CCD was found to be linear with tube current. The total charge is a reliable quantity since is does not depend on any event selection criteria. The data are presented in the next subsection.