Reducing pileup with subframing mode

As discussed above, pileup depends (non-linearly) on the source flux (Table 6.22): for a given exposure and frametime, decreasing the source flux leads to a smaller pileup fraction, essentially because the number of events per frame is lower. In a realistic astrophysical situation, the observer does not have the option to vary the source flux; however, if a method can be found to reduce the number of detected events per frame, the observed pileup will be lower. This method is called ``subframing'' and consists of reading out only a fraction of the CCD rows centered on the aimpoint. Since the CCD is sitting on the source for less time, the number of photons detected during each frame will be lower.

A few tests during the Phase F of the XRCF calibration operation were indeed performed in subframe mode, giving us a chance to test this method to control pileup. Table 6.24 reports the Phase F tests that were used here. They corresponds to data acquired with the BI chip (w97c1) for a fixed source flux of 0.0133 c/s/cm² and exposure 2183.41 s at the Al-K$\alpha$ energy. (Similar tests with the FI chip could not be used because the energy switched to another value after the first two tests.) The number of rows, and corresponding frametimes, were varied as reported in Table 6.24.

We accumulated spectra and analyzed them in the way described above. Table 6.24 lists the measured pileup fraction for two reference grade selections, G02346 (the most commonly used for the in-flight data) and the unbinned ACIS grades All-255, for comparison. The pileup fraction defined in eq. 6.14 was used. It is apparent that subframing mode can be used to control pileup. For the tests in Table 6.24, the pileup is reduced by 55% in the case of no grade selection, and by 44% for G02346.

We conclude that subframe mode can be used to effectively control pileup in most astrophysical situations when a bright point-like source of a given flux is observed. Of course the penalty paid for sub-framing is a reduced field of view because only a portion of the CCD is being processed for X-ray events. Thus there will be a smaller sky coverage when subframe mode is used.
 

Table 6.24: XRCF Phase F Pileup Data 

Test Name	Frametime	# of Rows		Pileup Fraction
				All-255	G02346
F-I2C-CR-2.001a	0.113	18		0.260	0.105
F-I2C-CR-2.002	0.227	38		0.410	0.172
F-I2C-CR-2.003	0.318	54		0.487	0.180
F-I2C-CR-2.004	0.660	114		0.580	0.186