AUTOFOC Test Results: 31 May 2001

Autofoc is the tool typically used to focus the Keck telescopes on slit jaws before spectroscopy. It operates by changing the secondary piston and taking guide images, calculating the image size and finally fitting a hyperbola to the resulting (focus, image size) pairs.

A series of tests on the ESI guider were performed on 31 May 2001 (UT) to investigate the robustness of the autofoc technique. Discussion follows.

Run #	Suggested secfocus	Notes
1	1.617 +/- 0.004	V=13.7 star, nominal parameters (10 arcsec box, star between echelle slits)
2	1.658 +/- 0.005
3	1.660 +/- 0.008
4	1.670 +/- 0.004	Edge of guide box purposely put on 0.75 arcsec slit
5	1.738 +/- 0.007	8 arcsec guide box
6	1.707 +/- 0.008	Using new echelle slits (with less vignetting)
7	1.721 +/- 0.007
8	1.718 +/- 0.012
9	1.829 +/- 0.015	On offset mirror (REF pointing origin)
10	1.825 +/- 0.009
11	1.815 +/- 0.006
12	1.840 +/- 0.013	On REF, brighter (V=12.6) star
13	1.833 +/- 0.015
14	1.834 +/- 0.012
15	1.848 +/- 0.006	Fainter star (V=14.8)
16	1.828 +/- 0.008
17	1.838 +/- 0.004

Analysis

The autofoc curves look fairly deep, and the values that come from the analysis look robust, to within the 0.010 mm range. If a slit encroaches on the edge of teh guide box it seems to make no difference. Using a smaller (8 arcsec) guide box did seem to make a difference, an effect that should be followed up.

There is apparently a 0.070 mm offset between the old and new echelle slits. The old slits had one of three measurements that seemed a bit off, so the error bar on this difference is larger than one would like.

The focus offset between the offset mirror and teh new echelle slits is 0.108 mm, corresponding to the offset mirror being 3.9 mm too high.

There appears to be little difference between stars of V=12.6-14.8 as far as accuracy of the focus.

Go To:
Instruments Home Page		Keck Home Page		Observer's Reference Shelf

ESI Master
31 May 2001