[Last Change: 01 Mar 2011 (rev 2)]

Results of Crude Optical Axis Test

---

Maxim Markevitch (for Terry Gaetz and myself), 8/17/99

(A preliminary report, more details are coming later today)

We have obtained 9 pointings to PKS 0637-752 at different offsets, to look for the position of the mirror optical axis. The purpose of the test is to define the position at which the accurate focus measurement should be taken. To determine the axis position on the detector, we have searched for a minimum of the PSF width. SIM has been set -0.5 mm from its current best focus to avoid pileup and other complications.

The table below summarizes the measured average positions of the source in the chip coords (relative to the current aimpoint at 256, 512) as well as 50% and 90% encircled energy radii measured from the sky (aspect-corrected) images. The fraction of the flux is relative to the total within r=13 pixels which excludes the jet. Background has been subtracted for both the coordinate and the PSF width determination. Errors for the PSF radii are only quick estimates -- they were obtained by dividing several representative observations into 4 time intervals, calculating their rms scatter and dividing by 2. Errors that appeared too small were then multiplied by 2 to account for systematics (e.g., shifts) and because we should not be able to measure the radii to better than 0.1 pixel anyway. One of the exposures, 1057, was 4 times shorter than the others which resulted in a comparably big error.

The 50% and 90% radii as a function of two coordinates were fitted by a parabolic function. A fit to the 90% radii resulted in the best-fit position of the axis of (-0.24', 0.28'), or a 22" offset, from the aimpoint in the S3 chip. F-test gives only an 87% confidence that this position differs from the present aimpoint (of course, our errors are rather arbitrary). A fit to the 50% radii resulted in the best-fit offset of (-0.06',0.05'), or 5". The figure below shows how the PSF widths change along the two diagonals that were sampled on the detector (errors are 1-sigma):

Continuous lines are parabolic fits to the data with the coordinate offset fixed at zero. One can see that the nonzero offset obtained with the 90% radii (the offset is approximately along the diagonal shown in the right panel) is mostly driven by one point.

Thus the result is consistent with no offset from the current ACIS-S aimpoint at (256,512) in the S3 chip coordinates, to the accuracy required for the focus test (and the focus test is now proceeding at the zero offset). A more accurate axis position will be obtained later with HRC-I (whose pixels are smaller), but it is already clear that the axis is well within 0.5' of its assumed position.

---