Results for the flight devices.

The technique described above was applied to all frontside illuminated flight devices. Depletion depths were extracted from the Fe⁵⁵data sets taken at the HIREFS spectrometer chamber. These data have lower flux rate than other chambers, and, hence, very low level of pileup. The computation of the depletion depth included correction for the presence of the $K_{\beta}$ line in the source spectrum as well as correction for the undetected photons according to equation (4.40). No pileup correction was made because the level of pileup in these data is low. Results are summarized in Table 4.30.
 

Table 4.30: Depletion Depth Estimated from 5.9 keV branching ratio

 

Device	Location	Depletion Depth, $\mu m$								Remarks
		$\Phi$ PI = +2,+10V (standard)				$\Phi$ PI = -5,+5V (reduced dark current)
		c0	c1	c2	c3	c0	c1	c2	c3
w203c4r	I0	66.4	66.9	66.8	66.4	47.6	48.1	48.0	48.7
w193c2	I1	67.4	66.5	67.8	66.6	48.3	47.4	47.5	47.9
w158c4r	I2	66.8	67.2	66.5	66.3	47.7	47.9	47.8	47.4
w215c2r	I3	66.1	67.3	66.5	66.2	48.7	47.8	48.5	47.9
w168c4r	S0	64.8	65.5	65.5	65.3	44.9	45.8	45.5	45.4
w182c4r	S2	79.5	79.1	80.6	79.9	57.7	58.2	58.1	58.3
w457c4	S4	73.3	74.2	73.9	73.6	53.4	53.5	53.0	53.4
w201c3r	S5	73.0	74.4	74.7	74.4	52.1	53.4	52.9	53.2
w198c1	ref.	73.7	73.9	74.7	72.8	52.3	51.6	52.2	51.3
w461c4	ref.	81.2	81.3	80.2	80.6	59.2	58.9	58.2	58.4
w203c2	ref.	68.1	68.1	67.0	67.2	48.0	49.0	48.9	47.8

For standard image section voltages (+2,+10 Volts) every measurement was made at three different temperatures, namely, -130, -120, and -110 degrees C. Since depletion depths should not depend on temperature (at least in the narrow temperature range), for every quadrant of every chip the results were averaged over three measurements. Typical standard deviation of the result is 0.35 microns. As can be seen in Table 4.30, some of the chips show quadrant-to-quadrant variations of the depletion depth on the order of 1.5 microns. Such variations exceed statistical error and, most likely, are real and caused by nonuniformity of the substrate resistivity. They translate into approximately 2% variations of QE at 8.4 keV and provide an explanation why spatial variation of QE at this energy is much larger than at lower energies.