Telescope Focus

Telescope Focus

Focus Maintenance System
Calibrating autofocus “fine focus offsets” for the instruments
NB: the latter includes a standard protocol for measuring the offsets.

Focus Maintenance System

One of the most common factors degrading the images of an astronomical telescope is defocus. In the case of UKIRT the goal of the telescope Upgrades Programme was to reduce such degradation by all identifiable factors to less than 0.”03 rms; in the case of defocus this implies that the secondary should be positioned in Z to within ±0.004 mm (4 microns) of the correct location.

All structures change shape and dimensions with temperature and attitude shifts, and this is true of UKIRT. Finite Element Analyses carried out at the UKATC (then ROE) showed that elastic length changes of up to 1.0 mm occur between the zenith and an altitude of 30°; a length change of 0.5 mm occurs when the telescope steelwork changes temperature by 5°C, as can readily occur in the first half of a night. Such shifts are hundreds of times larger than the tolerances allowed by the performance goals.

UKIRT is accordingly equipped with a simple model which corrects for these effects using the attitude information from the Telescope Control System and temperature information from sensors attached to the truss steelwork. Position corrections are sent to the hexapod secondary mirror positioner whenever the expected length change exceeds 1 micron for more than 10 seconds.


The autofocus system complements the focus maintenance system and allows instrument focus settings to be checked during observing far more rapidly and accurately that is possible with a standard “through-focus” run with an imager or spectrometer.

How autofocus works

The Fast Guider re-images the telescope focal plane on the CCD using one of two lens systems mounted in a wheel. A single lens is employed for Normal Guide, Acquisition and Focus modes. For auto-focus a second, similar lens is equipped with a 2×2 array of 4 f/100 Shack-Hartmann lenslets mounted immediately behind it. Instead of a single image these produce 4 images of the star on the instead of one, in a plane somewhat nearer the lens wheel than the is single image used for normal guiding.

The longitudinal position of the image formed by the front lens on its own depends on the overall focus of the telescope. Consequently the diameter of the converging pencil in any plane between the lens and its image is a measure of the telescope focus. The four sub-images are formed from sub-pupils of the converging cone from the single lens, so the radial separation of the four images is also a measure of the overall telescope focus setting. In auto-focus mode the four images are sensed by a 24×24 array of pixels, which are again binned up 3×3, now into 8×8 superpixels forming four (rather than one) 4×4 guiding arrays.

The larger readout area slows the read process somewhat: while in normal guide mode the read rate can be 100 Hz or more, in autofocus mode the image positions are sensed at 60 Hz. The position references are the four centrepoints of the central quartets of pixels, and for each readout of the array the displacement of the centroids of the sub-images from the reference positions is computed. The average radial component measures the current defocus. The X-Y component averaged over the four sub-images measures image movement, just as in normal fast guiding, so that tip-tilt correction is still available, with a closed-loop bandwidth of ~6 Hz.

Averaging periods in autofocus

In principle the focus corrections could also be applied to the secondary mirror as fast as the CCD is read out, giving Adaptive focus correction as well as the usual tip-tilt image stabilisation. In practice the piezo-electric actuators do not have enough throw to correct for the focus excursions actually seen, so the focus corrections are applied via the hexapod, which is a lot slower (< 1 Hz) than the tip-tilt system (~10s of Hz).

The 60 Hz measures are averaged over a specified interval, during which the RMS of the focus fluctuations (“Zrms“) is also determined. Because the 4×4 sensor is non-linear at large excursions, such as might accompany a change of instrument (see below), when autofocus mode is initiated the 60-Hz focus corrections are averaged, and then applied, over consecutive periods of 2, 4 8, 16 and 32 seconds, first to facilitate convergence and then to allow seeing-induced fluctuations to be averaged out. (The 32s averaging is then repeated indefinitely.)

Calibrating autofocus for the instruments

  1. Autoguider fine focus: Each scientific instrument has a slightly different optimum telescope focus setting. As we have seen the guider in autofocus mode can only correct the telescope focus to bring the images into coincidence with the reference points on the CCD. The single and quadruple lenses are both carried in the same lens wheel, which can be moved towards and away from the CCD. This adjustment is called “Fine Focus” on the Bottom-End Control screens and is a measure of the lens wheel position (in mm) relative to an arbitrary zero.The lens wheel can be translated. This adjustment is called “Fine Focus”. A movement of the lens wheel towards or away from the guider CCD will change the radial spacing of the images on the CCD. Thus, if the “Fine Focus” setting is changed while the Fast Guider is in autofocus mode, the result will be a change in the Z position of the secondary, so as to bring the four images back to their reference positions. The change in secondary position then causes a change in the focus at the instrument.A series of measurements with a science instrument at different autofocus Fine Focus settings can then be used to determine the optimum setting of the Shack Hartmann position (Fine Focus), i.e.that which gives the best working focus for the instrument. This measurement is made for all the facility instruments, with and without relevant accessories (FPs, etc.) according to a standard protocol. Current default estimates for the optimum fine-focus settings are given here. NB: The IRPOL polariser module, being above the dichroic, is in both the instrument and autguider beams and its focus shift will be the same in both beams (modulo the difference in the refractive index of its material in the visible and in the IR). Differences between the optimum autofocus fine focus settings with and without IRPOL should therefore be small. This is not the case for, e.g., the Fabry-Perot etalons, however, as these effect only the focus of the IR beam, so that a new focus determination is needed.
  2. Different Fine Focus for autofocus and normal guiding modes: The lens combinations used in “normal guide” and “autofocus” guider modes have different effective focal lengths: the “autofocus” lenses have their combined focal plane nearer to the lens wheel than that of the “normal guide” lens. Consequently different Fine Focus settings are used for autoguiding and for normal fast guiding. The settings for “normal guide” have no effect on the telescope focus and are selected simply to give the best Fast Guider performance.

Back to Acquisition & Guiding, or to “Other guider modes”.