Previous: System checks
Up: Low level operations and calibration checks
Previous Page: System checks
Next Page: FOS Operation
Experience shows that FOS is very stable, having been removed from the cass cluster and replaced several times with no noticable change in focus and alignment (within the measurement errors). Focussing of FOS is done by moving the CCD detector in 3-D using the 3 motor micrometers (labelled s r and b). These are not adjustable in normal operation, but their status can be obtained by reading the position sensors (LVDTs) next to each micrometer. MSTAT can be used to get these values, as well as focus and tilt angles (HA,VA) of the CCD. These should be within 10 units of the specification. e.g.
focus motor positions are:
MIKES=378 MIKER=481 MIKEB=291
hence FPOSIT=383 HANGLE=95 VANGLE=2
If these values are outside the tolerance, inform the support astronomer. As a further check, expose a copper-argon comparison lamp and, if available, a mercury comparison lamp, as follows:
Using the PROFILE option in PHOTOM you can find the position and width of five lines at the following wavelengths:-
The line at 5461Å is from the mercury comparison lamp. All the other lines are from the copper-argon comparison lamp.
Use SLICE (from the FOS reduction package) to get accurate vertical and horizontal profiles with rectangular boxes of dimensions [2,8] and [8,2] respectively. Alternatively, use STARFIT; this routine is really designed to measure stellar images, but works at least as well on arc lines taken through a narrow dekker. Values for the chip installed in May 1988 are given in table 2.1 The centroids can change absolutely by about 0.5 pixels, and by more than 1 pixel if the folding flat has been moved. Drift with telescope attitude is about 0.03 pixels per hour, and relative stability is of the same order.
Table 2.1 focus and alignment goals (1st order)
NB. V is in the dispersion direction
H is in the cross dispersion direction
The sigma/fwhm estimates are a good indication of focus if you are concerned about it, but remember that FOS just undersamples, so fwhm and 2.32 sigma may not quite agree - two different methods are used to determine these quantities, and the results will also depend quite strongly on the position of the line within the pixel.