The camera accepts two sizes of plate holders containing either 10 x 10 or 5 x 7 inch plates.
The holders are designed to take 10 x 10 x 0.06 inch plates, the flat focal plane being defined by the emulsion surface. Sheet films 8 x 10 inch also can be exposed by affixing them to the surface of 10 x 10 inch plates. The plate holders are completely light-tight (to 5 minutes in bright sunlight) and are fitted with a gas inlet valve so that they can be purged with nitrogen after loading. Each holder (except number 7) is individually numbered by a series of spots which can be read in the dark. All the plate holders are parfocal.
The dark slide can only be opened if a release button is pressed and it can only be removed completely from the plate holder by pressing the release button again. At very large ZD's and if the plate holder orientation is appropriate, the weight of the dark slide may cause it to slide back into the closed position. This is particularly unfortunate if it happens during a long exposure made without an observer in the PF cage. When remote operation is intended it is prudent to fix a strip of tape across the open dark slide to prevent accidental closure.
Three grooves in each plate holder base radiate from the optical axis at angles of 120° and locate on three studs on the camera body. The plate holders are held against this kinematic mount by spring loaded clamps but positive confirmation of location of the plateholder on to the studs by firmly bedding the holder down is strongly recommended since mislocation is all too easy.
A focusing knife-edge is available which fits on to the camera pedestal like a 10 inch plate holder. The knife-edge is normally left in position to protect the corrector lenses when the camera is not in use and at the end of a session. It should always be in place during the top end change.
The 5 x 7 inch holders are now rarely used. They are of lighter construction than the 10 x 10 inch holders and are not so light tight. When loaded, they should be kept from bright direct illumination. They have no provision for gas purging. The optical axis is 25 mm off centre on the 5 x 7 inch plates, the idea being to leave space (by not fully opening the dark slide) for pre- or post-calibration of an unexposed part of the plate. However, plates in these holders cannot be calibrated by the built-in PF sensitometer during the telescope exposure.
The 5 x 7 inch holders (and knife edge) are magnetically latched into an adapter plate which locates like a 10 x 10 inch plate holder.
Knife edge focusing is provided for both the 10 x 10 plateholders and the 5 x 7 plateholders, although experience shows both that they are confocal and that the 10 x 10 knife edge is easier to use. A circular depression is machined in the 10 x knife edge plate to accommodate a small telescope to more easily access the Maxwellian view of the primary mirror. However, it is this writer's opinion that this is more trouble than it is worth and direct viewing of the knife edge during focusing, while less comfortable, is quicker and easier.
Two methods of knife-edging are commonly used; both give identical and consistent results. In one method the knife edge is set on an out-of-focus image of a star so that it obscures half the Maxwellian field. The focus is shifted in and out by the observer until the knife edge appears to dissolve and the field is judged to be uniformly illuminated. The telescope is then in focus. In the second method the focus is moved in steps and at each step the knife edge is moved in and out of the field. The focal setting where the field brightens and darkens uniformly is taken to be the optimum focus. Focus readout, to within 0.01mm, is only available at the console, and the observer is asked to focus several times to obtain a satisfactory mean value. In very good seeing the focus is reproducible to ± 0.01mm; in ~ 2 arc sec seeing values of ± 0.03mm are normal. Once determined manually, the focus is thereafter monitored by the telescope control computer. This system automatically compensates for changes in telescope attitude and also for temperature changes by sensing the position of the end of an invar wire between the primary mirror and the telescope top end. Under stable conditions the telescope need only be focused at the beginning of the night. If however there is a large temperature shift, or if the seeing improves during the night it is prudent to refocus.
If the focus was determined without a filter, an adjustment is needed when one is inserted to allow for its optical thickness, which is the measured thickness of the filter divided by its refractive index. To a first approximation, the focus setting must be increased by the effective optical thickness of the filter, bearing in mind that the focus readout value increases as the top end is raised. However, because the corrector and plateholder are moved by the focus drive, and the correctors have some power as lenses, the adjustment for true focus differs slightly, but significantly, from that implied by the optical thickness. For the triplet, the motion required is 0.90 x , for the doublet 1.04 x and the aspheric plate (just) 1.01 x, that implied by the calculated optical thickness. The optical thickness of all the large filters available for prime focus is given in Table 1.7.
Such problems are largely avoided by setting focus through one of the visually clear filters, such as GG495, and adopting that value for the other filters having the same thickness (most of the coloured glass filters). However, for the triplet and aspheric plate, there is some advantage in re-focusing for different wavelengths. (see the sections on the Triplet and Aspheric plate for more details).
The camera head carries all the controls necessary for setting up on a field, including the autoguider control panel, guide probe and calibrator (Fig. 3.1). The camera head can be rotated around the optical axis to any of the four cardinal points where it is retained by a spring-loaded latch. The rotation is limited to a single 270 degree range because of cable wrap constraints.
Fig. 3.1 The controls on the prime focus camera head. The small carriage with the autoguider eyepiece and optical adjustments is attached to the probe and moves in the X direction as the X position is shifted.
The camera head is moved by swinging the clamp lever to UNCLAMP and then pressing the INDEX LATCH RELEASE (marked PRESS) while rotating the camera head. The latch drops every 90 degrees. Intermediate positions are indicated by a red warning UNCLAMP light on the camera - the autoguider will not work with the camera unclamped. A four-quadrant indicator shows the real ( i.e. sky) orientation of the guide probe with respect to the optical axis. For example, the camera head might be rotated so that the probe and its eypiece are near the parked chair, which parks in the north. The quadrant indicator will show S (south) because it refers to the image of the sky reflected by the primary mirror. Plates taken in this configuration will have the shadow of the guide probe along the southern edge of the image.
The ideal observing position is one with the autoguider eyepiece in front of you and the dark slide pulled out to the right. It is also perfectly possible to work with the eyepiece to the right, in which case the dark slide pulls out opposite the observer. If the eyepiece is across the plate holder from the observer it can be difficult to reach especially at large zenith distances, though a specially extended (and seldom used) eyepiece tube is available for this contingency. This position is further complicated by the calibrator lamp housing which sticks out of the camera at crotch level. Having the eyepiece on the left means that the dark slide must be withdrawn towards the observer. It is possible manoeuver oneself to remove the dark slide completely and thus work in this position but great care must be taken to avoid stray light entering the dark slide slot. In the end, the camera position selected is usually dictated either by the position in the field of suitable guide stars or the intended orientation of the object or position of the calibration spots on the plate.
The PF observer has available three ways of preventing light reaching the plate.
A probe ranges over an area approximately 60 x 120mm (Y x X) (16 x 32 arc min) of the focal plane to one side of the field. Two interchangeable probes (``short'' and ``long'') double the area of field available, in the Y direction. The short probe is designed for use with the triplet and the long with the doublet. At the cost of some vignetting of the guide star image the short probe can be used with the doublet with no intrusion of the probe shadow into the unvignetted field and in practice the short probe is used for both doublet and triplet. If no suitable guide stars are found in the area available to the probe then the camera may be rotated, possibly at the cost of some inconvenience in reaching the eyepiece.
The range of focus in the guiding optics is not sufficient to cover all possible applications of the prime focus camera. For direct photography this is not a problem since the range covers all the focus settings likely to be found in practice with telescope filters of optical thickness up to 1.85mm (equivalent to about 6mm of glass).
For a variety of mechanical reasons some image tubes or CCD's cannot be mounted with their light sensitive surface in the nominal focal plane, so the telescope focus must be shifted substantially to accommodate these detectors. Such a shift may take the focal plane outside the range of the guide probe focusing mechanism. Each probe is therefore fitted with an adapter ring to accept small supplementary lenses to bring the probe focus back into range. These lenses are kept in the prime focus optics box which is normally at PF access.
There are three basic ways of finding stars which are suitable for autoguiding i.e. brighter than about 14th mag.
The simplest method and the one usually employed for direct photography is to remove the plate holder (and if necessary an optically dense filter) and look directly at the field of stars in the focal plane of the telescope. The X and Y probe drives are operated so that the probe, whose shadow can be seen against the sky, is moved over a suitable star. Generally if the star is bright enough to see clearly it is bright enough to use for guiding. This method is quick and very easy once mastered, but some novices find it troublesome at first. Until this familiarity has been achieved, guide star positions must be precalculated by one of the following methods. These alternative techniques are essential if it is impractical to remove the plate holder or if an image tube or CCD is fixed to the camera.
Plastic transparent overlays are available in all our chart rooms which indicate the area of sky covered by the various PF correctors but reduced to the SERC/ESO/POSS plate scale. The overlays also indicate the area covered by the short guide probe and can be used to locate the approximate position (in terms of X and Y on the guide probe controls) of suitable stars from the sky surveys. Two types of transparent overlay have been produced to match the scales of doublet and triplet. One of these is shown in Fig. 3.2. Transparent film copies are available from the Observatory.
If the RA and Dec coordinates of suitable stars are known, a simple program available at Epping and Siding Spring (and all Starlink nodes) can be used to calculate X and Y for both long and short probes for a given plate centre. The routine is called AATGS and further details can be found in Starlink User Note 6.2. Similar routines are also available on the CCS.
The design principles of the AAO autoguider are described by Kobler and Wallace (1976) though there have been some changes to the system since that publication. While short exposures are possible without using the autoguider, it has been found that small periodic errors in the telescope gearing are almost impossible to correct by guiding manually. This is especially noticeable in good seeing and it is recommended that the autoguider be used for all imaging applications at the prime focus where exposures exceed a minute or so.
Once a selected guide star is visible in the autoguider eyepiece it can be centred by the X and Y guide probe knobs. The star image, which might at this stage be out of focus, is seen together with an illuminated graticule consisting of two pairs of crossed lines. The central square that these form is 2 arc sec on a side when the probe is used (as is normal for photography) without an auxiliary lens. This square can be used to estimate the seeing.
To set up the optical system, first select one of three available eyepieces and focus it on to the graticule by rotating the upper smaller knurled ring. Then focus the guide star on to the graticule by rotating the LENS FOCUS knob on the guide panel. The autoguider optics are confocal with the graticule, which is illuminated by a green light whose brightness is adjustable. Do not move the LENS FOCUS knob once guiding has begun - there may be some shift of position.
A labelled three-position lever-switch operates a prism assembly to direct the guide star light to the guide EYEPIECE, to the autoguider PM TUBE or to BOTH with a beamsplitter which splits the light roughly 90% to the eyepiece, 10% to photomultiplier tube. The BOTH position is preferred if the guide star is bright enough so that the observer can monitor changes in the seeing. The 3 prism positions do not exactly coincide in positioning a star image on the graticule or autoguiding sensor, therefore the prism position must not be changed during an exposure.
NOTE: The following discussion of the autoguider system is somewhat out-of-date. The guide head now carries a commercial ST4 CCD camera, which is used to derive guiding corrections for the telescope. Acquisition of guide stars is carried out as before, however now the night assistant controls guiding from a PC. Searching for guide stars is only necessary if the guide star falls outside the ??' x ??' field of view of the ST4 camera, in which case the observer will have to move the guide star into the cross hairs manually. There is no equivalent to the `STAR-TO-PROBE' mode, with the ST4. A more detailed description of the use of the ST4 control software is in preperation.
PMT specific notes : The autoguider, when so commanded by the night assistant, searches an area 12 arc sec square in a raster scan. It senses the position of the brightest star and proceeds to guide on this star by scanning in a cross pattern 6 arc secs by 6 arc secs over the star image. The autoguider gives a display to the console operator of the seeing profile of the guide star.
The astronomer must be aware of the two ways in which the guide star can be acquired by the autoguider's image dissector. In the PROBE-TO-STAR mode the telescope position at the time guiding began is taken to be correct, and the guider then maintains the guide star at its initial, arbitrary position. Once guiding has occurred, as the guide star drifts with respect to the telescope, the telescope will be moved so the guide star will be brought back to the initial position until OFF is selected by the operator, at which time the arbitrary position is lost. Any subsequent guiding will then, of course, begin by defining a new position corresponding to the telescope orientation during the first few seconds after re-acquisition. Temporary loss of the guide star due to cloud will not cause the guiding position to be lost unless OFF is selected.
In some cases - where the telescope has been accurately set to some particular position ( e.g. with the program star centred in an aperture) - the PROBE-TO-STAR mode must be used to maintain the telescope position as initially set.
The second method of acquiring a guide star, the STAR-TO-PROBE mode will cause the autoguider to shift the telescope a few arc secs as it acquires the guide star and brings it to a standard position on the photocathode. This nominally corresponds to the centre of the guiding eyepiece graticule but may in practice not do so precisely. Where a choice is possible, e.g. during direct photography, the STAR-TO-PROBE mode is preferred as there is no risk of double exposure if the guiding position is lost due to malfunction or misoperation. If the autoguider should fail during an exposure it is possible to continue by guiding manually without risk of losing position if the operation is started in the STAR-TO-PROBE mode.
There is also a `hold' facility called FREEZE/THAW. The night assistant selects FREEZE on the control panel which stops the autoguider from inserting corrections but does not exit from the guiding loop. Selecting THAW will resume guiding instantly, without the need to acquire the star again. This is a very useful mode when there is cloud around, or for speeding up re-acquisition when doing remote exposures. (See section 3.7). The autoguider is described in more detail in the AAO Technical Manual (TM 7) covering the telescope control system. XXXX
To use autoguider:
|Position||Colour Filter (Knob C)||ND Filter (Knob D)|
|1||Clear 350-750nm (2mm WG280)||Clear|
|2||Blue ~400nm (1mm BG25 + 1mm BG39)||ND=0.7|
|3||Green ~ 550nm (1mm GG495 + 1mm BG39)||ND=1.4|
|4||Red ~ 630nm (2mm RG630)||ND=2.0|
It is easy for the filters (either ND or colour) to get stuck `half-way' between their positions. If you do not plan on using the filters, ensure that both are set to position `1'.
The observer can confirm that the autoguider works by offsetting the telescope about 5 arc sec with the guiding handset and with light splitting prism set to BOTH. Watch the autoguider return the star to its original position.
There are occasions when it is advantageous to start and stop an exposure remotely, without an observer in the prime focus cage. The most obvious use for this is when a very long ( e.g. > 2 hr) exposure is planned and the few minutes lost to setting up and the two or three extra slews can be justified.
To set up a remote exposure, find a guide star as described in the previous section and ensure that it is bright enough to guide on. Check that the autoguider is in STAR-TO-PROBE mode and ask the night assistant to `mark' the telescope position. This command enables the telescope to return to the selected position to within one arc second, even after a lengthy slew (see UM-1, section 3.7.7.). If the autoguider accepts the guide star, clamp the probe X and Y controls, and ask the night assistant to FREEZE the autoguider. The slew to prime focus access can now be started and the desired filter and plateholder attached in the normal way.
Ask for the PF dust cover to be closed but LEAVE THE CAMERA SHUTTER OPEN. At access, under dim light, open the dark slide and tape open if necessary to prevent accidental closure during slew or exposure. Turn out all lights and step out of the cage. Ask the night assistant to return to the object. When the telescope has settled at its marked position open the PF dust cover. As soon as the guide star appears on the monitor, THAW the autoguider. Guiding will then start. If the STAR-TO-PROBE mode has been selected, the telescope will then shift slightly to to bring the star image to its standard reference point, but the double image will not be visible on the processed plate.
The exposure is terminated by closing the PF dust cover and returning the telescope to PF access. Before using strong lights ensure the plate holder dark slide is closed and the autoguider is switched to MANUAL.
SOME IMPORTANT POINTS
The rotating camera head is fitted with a projection-type sensitometer constructed as described by Schoening (1976). The calibrator projects 16 square spots, each about 6mm on a side in a pattern 28mm square. The pattern is projected into a corner of the plate in a region protected from sky illumination. Projection is through the telescope filter in use and only reaches the plate when the roller-blind shutter is open. An area roughly 45 x 60mm measured from the edge of the plate is vignetted by the calibrator, which can be removed in a few minutes if necessary.
A stabilised power supply for the sensitometer is built into the PF cage and its control unit is fixed to the camera head near the lamp house. The control unit has digital readout of lamp voltage and current and a battery-powered audible alarm which indicates lamp or power failure or an open circuit. The calibrator is normallly left on throughout the night. The light source is a miniature tungsten-halogen lamp, Atlas type M29 (or similar) 6V 10W, run at 5V to give an effective colour temperature in the integrating sphere of 2975 K and an anticipated life of 930 hours (manufacturer's data). The optical system contains a filter (presently Tiffen 840 B) to transform the colour temperature of the output to ~ 6000 K. However the filter is only effective over the range 400 - 750nm. Outside this range it absorbs much of the near UV light (as do the glass optics) and transmits more in the near IR than would a perfect conversion filter.
The sensitometer spots cover an intensity range of about 1000:1 (3 log E units) and have been calibrated photoelectrically. Calibration results are obtainable from the Observatory. A rotatable aperture wheel gives an additional stepped brightness range of 150:1; however, well calibrated plates are obtained without altering the standard 2.4mm aperture setting for most plate filter combinations, but long exposure I band plates require a 0.8 mm setting and U band plates require the 4.8 or 9.6mm aperture.
The sensitometer which produces these spots has been in use since 11 March 1978 (plate 1590) without modification and is known as `Sensitometer 1'. If planned improvements are implemented, Sensitometer 1 will be replaced Sensiometer 2. New intensity values, hopefully without a colour equation, will be published when the new sensitometer is installed.
A few early plates were calibrated before or after exposure with the Mt. Stromlo tube sensiometer. Calibration values for this device can be supplied on request.