Count Rate

It is important to accurately measure the count rate of the largest emission features of the ExtCalSrc, mainly the Al, Ti, and Mn K$\alpha$ lines for two reasons. First, the accuracy of the gain calculations needed for continual ACIS calibration depends on how well the line centers are measured, which in turn depends on the number of counts in the peak. With an accurate count rate in hand, we can calculate the time needed to collect the counts for the desired level of gain accuracy. Second, knowing the count rates will allow adjustment of the spatial windows to avoid saturating telemetry and to maximize the amount of calibraiton data. The ExtCalSrc was carefully measured on 23 October 1997 during the first ISIM thermal vaccuum testing (ISIM-TV1). Because both spatial and spectal information were desired and full-frame data on six chips would have saturated telemetry, we acquired a mininum of $\sim$1500 HST frames for each chip. Gaussian lines were then fit to the K$\alpha$ peaks, and the count rates simply equal the gaussian area/integration time. Table 4.82 present the count rates and errors, where the errors are the 1 $\sigma$ Poisson noise.
 
 

Table 4.82: Observed Counting Rate from the External Cal Source on 23 October 1997

 

	Mn K$\alpha$	Ti K$\alpha$	Al K$\alpha$
Chip	(cnt s-1 chip-1)	(cnt s-1 chip-1)	(cnt s-1 chip-1)
I0	89.69 $\pm$ 0.13	25.65 $\pm$ 0.07	25.86 $\pm$ 0.07
I1	89.44 $\pm$ 0.13	26.02 $\pm$ 0.07	25.89 $\pm$ 0.07
I2	90.57 $\pm$ 0.19	26.93 $\pm$ 0.10	26.48 $\pm$ 0.10
I3	89.58 $\pm$ 0.13	27.29 $\pm$ 0.07	26.38 $\pm$ 0.07
S0	87.34 $\pm$ 0.13	24.54 $\pm$ 0.07	25.55 $\pm$ 0.07
S2	92.08 $\pm$ 0.13	28.10 $\pm$ 0.07	27.14 $\pm$ 0.07
S4	89.48 $\pm$ 0.13	28.77 $\pm$ 0.07	26.36 $\pm$ 0.07
S5	87.61 $\pm$ 0.13	28.40 $\pm$ 0.07	24.98 $\pm$ 0.07
S1	59.12 $\pm$ 0.11	22.14 $\pm$ 0.06	29.46 $\pm$ 0.07
S3	72.98 $\pm$ 0.08	27.45 $\pm$ 0.05	31.04 $\pm$ 0.05

Two general facts are immediate noticable. First, the ratio of Mn K$\alpha$:Ti K$\alpha$:Al K$\alpha$ is $\sim$4:1:1 for the FI chips and $\sim$2:1:1 for the BI chips. The large difference in the ratio and the differences in the actual count rates are due to the general variations in quantum efficiences between BI and FI chips. Second, the fluctuation in count rate between FI chips is much larger than the stastical errors. In some cases this is at least partially attributable to slightly different QEs, but the largest contributor to this effect is the physical distribution of incident photons (see the section below on spatial uniformities).

The observed flux from the ExtCalSrc will exponentially decrease with time as the source ⁵⁵Fe radioactively decays. Using the published half-life value T_1/2 = 2.73 yr, we have calculated what the count rates for the three primary lines will be during the early days of AXAF's orbital operation. Table 4.83 lists the count rates for a typical FI chip and both BI chips.
 
 

Table 4.83: Predicted Count Rate for 01 April 1999

 

	Mn K$\alpha$	Ti K$\alpha$	Al K$\alpha$
Chip	(cnt s-1 chip-1)	(cnt s-1 chip-1)	(cnt s-1 chip-1)
<FI>	62.1	18.7	18.1
S1	41.0	15.4	20.5
S3	50.7	19.1	21.6