This large section describes all the procedures and processes necessary to prepare a FLAIR plateholder for observation with FLAIR at the UKST including all the necessary preparation of the FLAIR mask (glass copy plate) of the field for observation prior to loading in either of the two FLAIR plateholders (PH's) 14/5 or 14/6.
This fully comprehensive section is intended as a training tool for visiting astronomers as well as an aide memoire for the UKST support staff. Individual copies of this section alone can be furnished on request and spare copies are kept at the telescope and in the fibre-ing laboratory.
The FLAIR copy plate of the target field is first marked-up with the target objects. This is done on a light table with the plate emulsion up! Chart output or other `overlays' can be placed under the plate to assist with object identification. A fine nibbed felt-tip pen is used to identify each object. The objects can be circled or tagged with a short line either side for unambiguous object identification. The object I.D. number is then appended beside the object. It is helpful to number the objects in increasing RA.
An image guide star is used to initially acquire the target field.
This should be a fairly bright star with diffraction spikes and
a dark inner halo but no outer halo. The star must be near the field
centre but preferably in the southern half of the field (if the old image
is being used as it is somewhat shorter than the new IG due to a breakage).
The chosen image guide will lead from this
position to the south edge of the plateholder where it is clamped. The
image guide is
~ 1.6mm in diameter (1.8arcmins on the sky)
and a clear
path should be left
from the image guide star to its anchor point at the south edge
of the plateholder. The guide star cannot be placed too close to the
southern edge because of the image guide's rigidity.
The chosen guide star can be identified with
a box and marked `IG'.
Fiducial stars are used to acquire the field in the telescope. These should be fairly bright stars of similar magnitude with diffraction spikes and a dark inner halo but not so bright that the outer halo forms. N.B. If old `bleached' 098 plates are used to make the FLAIR masks then it is as well to remember that stars will appear fainter than they really are - film copies of the survey fields should be examined to get a feel for the losses due to plate bleaching (performed because the old plates otherwise have excessively high fog levels). The chosen fiducial stars should be at least 2 inches in from the plate edges and well spread around the field (if the target objects are also uniformly spread) and centred generally on the image guide star. If there is a particular concentration of target objects in a given region then it is useful to select a fiducial in the vicinity. More fiducial stars than the number of available fiducial fibres can be chosen to allow flexibility during fibre-ing. PH 14/5 has five 33um fiducial fibres labelled A-E, whilst PH 14/6 has eight labelled A-H. For certain types of project it may be worth while trying to pre-select suitable bright fiducial stars according to the spectral type that most closely matches the target object colours for improved field acquisition and guiding (the current CCD FLAIR guiding camera is unfortunately an S-25 red response camera which is rather poor in the blue. Hopefully this will be replaced soon with a blue sensitive camera). Chosing such stars is simple if CHART has been used to produce the FLAIR finding charts as they are a by product of the process and a log of such stars is automatically produced. A cautionary note is that bright stars may have discernable proper motions after several years so the latest UKST exposures of each field should be used to produce the FLAIR masks wherever possible. Again the output from CHART should help to avoid the highest proper motion stars.
This section describes the procedures to prepare the selected PH at the start of a new FLAIR run, i.e. fibre-ing the field, tensioning the PH and securing the fibres prior to loading in the telescope.
The PH (14/5 or 14/6) is mounted on the X-Y table in the FLAIR lab to sit flush in the 3 locating points. The fibre feeds from the main fibre bundles and the fiducial star bundle must be `released' from their wound coils but the PH must also be pushed fully to the far left-hand side before securing the fibre feeds under the velcro strip on the right-hand edge of the X-Y table. Care must be taken to avoid hitting the gripper on the dark-slide handles. The chosen fibre bundle's output end (slit-vane assembly) can be fed to a `dummy' slit mounting on the second shelf of the table immediately right of the X-Y table. The fiducial bundle is simply attached to a clamp on the same shelf. These input ends are then illuminated with small desk lamps. (do not overilluminate - see later). The protective perspex covers can remain in place as sufficient light is still transmitted to illuminate the fibres. Spare cabling is kept coiled on the shelf.
The dark slide may now be removed from the PH. Care must be taken as there is not much clearance, especially when the fibres are bedded. In particular, the dark slide must always be raised or lowered parallel to the PH on removal or replacement. Take care to avoid hitting the gripper assembly on the dark slide's metal bar between the two slide handles as this can damage the gripper assembly. Initially all fibres will be housed in their `beds' along the north and south edges of the PH which should be in a de-tensioned state secured with 4 long Allen bolts and spacers. The Nitrogen supply spigot which squirts a supply of compressed nitrogen gas may be used to clean the mandrel of any accumulated dust etc with the nitrogen.
The whitened aluminium backing plate must now be inserted into the PH. Apart from providing extra rigidity during fibre-ing up the whitened face ensures sufficient light is reflected back through the plate to the CCD imaging camera for target acquisition. If the PH is still tensioned the 4 short Allen bolts are first removed and the two halves of the PH gradually levered apart til there is a sufficient gap to insert the 4 metal `spacers' which are placed over the 4 bolt holes. Two long Allen bolts are then screwed through the top half of the PH on the West side only, through the metal spacers and down into the lower half of the PH. The four metal spacers have a `lip' which must face inwards as these support the aluminium backing plate. With the curved `lip' along the south edge (i.e. the PH edge closest to you), the backing plate, white side up, is inserted from the left (East) side between the two halves of the PH. Each plateholder has its own collection of 4 long and short allen bolts and 4 metal spacers.
The marked up plate is now inserted emulsion up with the NE corner at the top left hand corner of the PH (standard orientation), keeping the left edge of the PH levered up slightly to allow smooth insertion. The aluminium backing plate and FLAIR mask can be cleaned quickly with Nitrogen gas prior to insertion. The two left side spacers (if removed) and the remaining two long Allen bolts are then replaced. The glass plate should be able to move slightly from all edges of the PH. N.B. If there is too much `play' with the plate movement in the PH then the plate can be moved by a fibre ferrule when sliding into position over a target object on the plate. True registration of object position can thus be lost. To prevent this cotton buds (cut in half) can be used to wedge in the top (and/or) bottom corners of the plateholder to prevent plate movement. The plate is now ready to be fibred-up.
Once the PH is ready, the wall socket power supplies to the computer, gripper assembly, LED field illumination, video recorder and mercury lamp can all be switched on. The gripper Nitrogen supply is also turned on by moving the lever labelled `gaseous nitrogen' to the right on the back wall. The video recorder is currently used to play back the CCD images on to the TV monitor as this prevents interference. The record button on the unit is simply activated before use.
A mercury lamp in the metal box/tube assembly to the right on the X-Y table provides a strong UV light beam to cure the optical cement used to glue the fibres. The lamp is operated from a power supply box on the bottom shelf of the table to the right of the X-Y table. Switch on the power supply (the current meter should register) and after a few seconds press the `start' button. The lamp should come on. The cooling fan inside the lamp assembly must now be switched on using the left hand switch on the double wall socket behind the Mercury lamp. At this time the `Ozone eater' which is attached to the mercury lamp assembly by suitable ducting is also switched on to prevent a build-up of harmful ozone gas in the fibering lab produced by the lamp. The computer must be off prior to switching on the Mercury lamp. The lamp is currently insufficiently shielded and a noise pulse crashes the computer if left switched on which could damage the PC's hard disk. UV safety goggles must be worn when the lamp is on and illuminating the plate. Check the focus of the UV beam. This is controlled by a bolt at the rear of the main beam tube which can rotate round the tube. Aim to get the beam in tight focus. The beam tube can also be rotated slightly in the X-Y plane to enable better ferrule illumination if required.
Turn on the PC and start up AutoFred
The gripper assembly directly above the X-Y table allows semi-automatic fibre positioning under PC control. The control software is called `AutoFred'. Turn on the PC using the red switch on the right hand side of the main PC box. Once booted an `AUTOFRED>' prompt appears. N.B. At this stage copy the appropriate ferrule-number to fibre-number look-up-table (LUT) for the chosen fibre bundle and/or PH being used to the file FIBRES.DAT. There are 3 such files of LUT's named FIB55.DAT and FIB100.DAT for the two fibre bundles for PH 14/6 and FIB92.DAT for the 100um bundle for PH 14/5 (unfortunately the fibre numbers on the slit do not correspond to the consecutive ferrule numbers on the plateholders - hence the LUT's).
Type `AF' to start the AutoFred programme.
Hit enter twice. AutoFred then asks if you wish to overwrite the AF.LOG
file (a running log of the fibre-ferrule-object set up). Enter `Yes" at
the start of a new FLAIR run or enter `No' if continuing fibre-ing
the same field after a break or computer crash. The AutoFred (`AF')
program is fairly self-explanatory and moderately user friendly operating
as a series of menu screens with multiple options. Most commands
are single-key operations which are executed immediately that key is struck
(rather than waiting for a carriage return).
At the start of a new run it may be necessary
to set up the gripper height(s) and to centre the gripper-fibre-ferrule
correctly in the field of view of the CCD camera (consult
the FLAIR instrument scientist or FLAIR support astronomer for full
details of the AF program). Entering `CTRL/C' interrupts the
AF program at any point and returns to the `AUTOFRED>' prompt. Once the
AutoFred software has started, the CCD camera images an
7 x 5 arcminute
field of view of the plate to be fibred-up. N.B. Each PH requires
slightly different focus setting to properly image the emulsion surface of
the mounted FLAIR mask. The AutoFred CCD camera focus is controlled by a
knurled knob adjacent to the main camera assembly atop the
X-Y table on the right hand side and should be adjusted till the emulsion
surface is in sharp focus.
This section describes how to set-up the gripper, how to affix fibre-ferrules on the appropriate images on the plate and how to affix the image guide socket.
The gripper has three heights; Load `L', where
ferrules are loaded between the gripper jaws and where glue is applied to
the ferrule base; Cruise height `C',
above the plate; and Plate height `P' where the gripper descends to the
plate for gluing on the ferrules. The gripper's plate height
and cruise height may need to be set for a new run. This `set-up' option
offered whenever AF is restarted. The AF program asks
``Is the Gripper properly centred?''
If you answer `Yes' then AF asks ``What height is this (L,C,P)?'' The normal
response would be to type `L' since the gripper is always left at load height.
The next AF menu screen then
states ``All I need is the Gripper height''. If `Enter' is now pressed then
the menu screen ``Acquire next object'' is displayed and the fibre-ing
process can begin. If you entered `No' to the question ``Is the Gripper
properly centred?'' this takes you to the image scale set-up
option which allows a ferrule to be properly centred in the field of
view. With this option load a ferrule into the gripper (via the `G' manipulate
gripper menu) and then press `Enter' each time the gripper finds the fibre
and moves around the field (the software should automatically find the fibre
core and position the cursor over it. The ferrule ends up centred in the
Finally answer `Yes' to question `Do you want to fix the scale?'
The `U' and `D' keys can be used to move the gripper up or down in the Z-axis. The default Z-motion is 10 x that of the X-Y gripper motion. Tapping the space bar stops the motion and `H' can be used to define the new gripper height.
The 3 gripper heights can be checked by simply typing `C', `P'
and `L'. At cruise height the Gripper jaws should be the same height above
the plate as the thickness of an image guide socket, i.e.
(see section 4.4.4.).
When at plate height the gripper should descend not just to
plate level but a mm or so beyond. The gripper mechanism
will then push back on the plate against a spring. This ensures that the
ferrule assembly makes good contact with the plate when descending to
plate height. The load height is the least crucial but should be high enough
off the plate to easily insert the ferrule and apply glue.
N.B. Always check the 3 set gripper heights (L,C,P) at the start of each
fibre-ing session or whenever the AF program has been restarted.
Figure 13: Side view of the gripper mechanism showing a needle shank of a ferrule being held in the gripper jaws.
Once the gripper set-up procedure is finished ferrules can be glued onto the plate.
As well as the automatic AutoFred ferrule number to object number logging there are is a target object logging sheet (enabling extra comments etc) and fiducial fibres and sky fibre MAP sheet where a cartoon of the approximate locations of the fiducials and sky-fibres can be drawn. Both the fiducial letters and ferrule/fibre numbers for the sky fibres should be indicated along with their approximate field locations. There is also an additional type of target logging sheet which permits straightforward fibering of alternate fibres on the slit by identifying the associated ferrule numbers to use.
To aid in optimum object-ferrule selection there are also fibre numbers alongside the ferrule numbers on each FLAIR PH. This makes it easier for the fibre-er to choose where on the chip they want particular objects or sky fibres to lie. Hence the brightest objects can be selected for the poorest transmitting fibres etc. They can also be sequenced together on the chip. Sky fibres can also be arranged in sequences or placed in equal intervals along the chip etc. With PH 14/6 for example the top and bottom 7 100um fibres imaged on to the FLAIR spectrograph's CCD have double the spacing of the remaining fibres. This is because as only 73 x 100um needed to be imaged on to the CCD c.f. 92 x 100um fibres with PH 14/5 there is the opportunity for greater fibre spacing at the edges. N.B. Because of the problem of scattered light between the closely spaced fibres the brightest objects can thus be chosen to occupy these doubly spaced fibres and minimise the scattered light directly.
The following steps are repeated during fibre-ing for each object.
N.B. It is important to carefully read the instructions on the AutoFred menu screens until fully familiar with the program.
~1mm of bare metal shank at the front end between the gripper and the plastic ferrule. Close the gripper and then use a syringe loaded with Norland Optical adhesive (code 81) to dispense a small blob of cement on the teflon pad on the ferrule's base. The azimuth of `withdrawal' of the gripper assembly can be controlled by entering `A' and then using the cursor to indicate the desired withdrawal direction.
N.B. If the gripper jaws become sluggish or stop working completely check that the nitrogen supply line has not become `kinked', possible if large gripper head azimuth changes are made and the supply line wraps around the gripper head.
~8um on the plate). If the residuals remain too large then it is necessary to go through the process once more.
Sometimes, even after manual re-positioning of the gripper ferrule, the rms residuals may remain too high (i.e. >1pixels). The AF program then asks ``Carry on Anyway?''. If `No' is entered AF immediately goes into manual mode and a flashing cursor appears directly over the position of the object. The image of the fibre can then be brought co-incident with the flashing cursor using the arrow keys (see section 4.4.4 item 11 below). N.B. The UV lamp slide must now be brought out into position to illuminate the fibre since hitting `Enter' will cause the gripper jaws to open and ascend to load height leaving the unglued ferrule `loose' on the plate. This is due to a small bug in the current version of AutoFred.
N.B. It is important not to over-illuminate the fibre slit otherwise the images of the fibre cores may become too extended and saturated. When the software then tries to find the fibre core (by assessing an intensity histogram of a grabbed CCD frame of the area) it may pick up on a saturated region which does not correspond to the physical true centroid of the fibre core. This will lead to positional errors. Hence reduce the core image intensity until the core achieves a minimum diameter and yet is still quite bright and distinct (can be done by pointing the desk lamp obliquely at the slit output end).
When all the ferrules are affixed, exit the AF program and obtain a print-out of the object numbers allocated to each ferrule (offered as an option when quitting AF). Check these against the list on the note-pad and by eye on the plate. Any errors can be corrected using the EDT editor to edit the current AF.LOG file. The corrected AF.LOG file can be renamed to a new file, e.g. ``FIELDnnn.LOG'' which can, if desired, be copied via floppy disk to another computer.
During fibre-ing fiducial fibres must be glued to images of the selected fiducial stars. To ensure access to these stars, which may not be possible if too many surrounding object fibres have been affixed, it may be useful to glue these fibres in position early on. The five 33um fiducial fibres are labelled A-E on FLAIR PH 14/5, whilst the eight fiducials are labelled A-H on PH 14/6. The fiducials can also be identified by their blue fibre jacketing colour. For PH 14/5 three of the fiducials are seated on the northern edge and two on the southern edge of the PH. All 5 fibres can extend right across the plate. For PH 14/6 the eight fiducials can only extend half way across the field. There are 4 each in the Northern and Southern beds. The fiducial fibres are particularly fragile due to their small diameters and single protective outer jacket. They should be handled with extreme care as they are vital for field acquisition. When AF enquires for the object number after gluing the fiducial a `0' is entered to prevent the fiducial stars being logged by AF as they are not part of the fibre object bundle.
This process would normally be performed by an experienced member of the UKST staff, usually the FLAIR support astronomer for the run in question. The procedure however is quite straightforward.
There are 6 image guide sockets (IGS) available. An IGS is used to
hold the image guide in place over a bright
image guide star. The field of view is
These are perspex blocks coated with
teflon pads on their bases with a double prism assembly at one end on a
thin perspex extension. This
faces onto a drilled tube within the main perspex block through which the
image guide is eventually inserted and abutted. The silvered back surface
first prism has a fiducial mark which must be aligned with the
chosen image guide star. It is crucial that this is done accurately or else
may be difficult or impossible to acquire the field in the telescope.
The following procedure must be followed:
~15minutes to enable the UV to penetrate to the glue through the perspex. After a few minutes the cement should be sufficiently cured to release the gripper and bring it to load height where the dummy ferrule is removed (go into the gripper menu by typing `G' and toggle the space bar to open the gripper jaws. Press the `escape' key to exit the gripper menu and enter `L' to send the gripper to load height). Remember to refocus the UV beam.
During fibre-ing the following points should be bourne in mind. They offer tips and guidelines which lead to safe, reliable and more accurate operation and can save considerable time during fibre-ing.
~10). This is important to ensure good sky subtraction from object spectra. These should be placed uniformly over the field and in areas of high target object concentration. Sky fibres can be used to get out of a `tight spot' where there are too many short fibres and not enough objects in the vicinity. During logging sky fibres are usually given the number `999' to distinguish them from ordinary object fibres. Use the sky and fiducial map sheet to roughly sketch the locations of the sky-fibres in the field.
Once again this procedure will usually be performed by the duty FLAIR support astronomer.
Once all the fibres and the image guide socket are firmly attached the PH can be tensioned. The following procedure should be followed:
Loose strands of fibre must now be secured to the plate using pieces of black electrical tape found under the X-Y table on an aluminium palette or on the spool. These are cut to various lengths for convenience. The aim is to prevent fibres from moving to vignette object prisms or from catching on the dark slide. The tape can be moved with forceps, but the fibres should only be manipulated by hand or with wooden toothpicks. The longer tape lengths are used to anchor down fibres near their exit slots as the fibres are not self tensioning and could easily come out further than needed and catch on the dark-slide. Once the fibres are secured the dark-slide can be replaced.
The PH is now ready for use. Detach the fibre bundle output ends from their clamps, coil-up the bundles and place them on top of the PH. Lift the PH onto a trolley to take it to the dumb waiter and thence to the dome. The FLAIR support astronomer will supervise or more usually carry out this task.
During a normal FLAIR run (i.e. where only one field is observed per night), after the night's observations the next PH must be prepared for the subsequent nights work. This is usually done during the day shift. With two PH's available two fields can now be fibred-up well in advance. Furthermore, when observing just one field per night the second PH can be prepared for the following night during the observations. This relieves pressure on the UKST day staff. When attempting two fields a night this procedure is essential.
This following section describes the general PH `decommissioning' procedures. The `used' plateholder should already be returned from the dome and placed on the X-Y table in the FLAIR lab as in section.2.1 (normal procedure at the end of the night).
Carefully remove the dark-slide. Remove the tape used to secure loose fibres on the plate using tweezers and place on the palette for storage. The fibres are fragile and must not be touched with the tweezers. The blue 33um fiducial star fibres are especially delicate. Replace the dark slide to protect the fibres during the next stage.
This activity will normally be done by the UKST support staff or FLAIR support astronomer.
Use the Allen key to unscrew the Allen bolts each by half a turn, corner
corner until the tension in the PH has eased. The 4 short Allen bolts can
then be completely unscrewed and removed.
The top half of the PH is now lifted
~ 2cm to further
de-tension the PH and allow access.
This must be kept parallel otherwise it will jam on the
locating dowels on the right hand side. Wooden clothes pegs can
be used as wedges to assist at this stage. When there is sufficient room,
the 4 spacers can be inserted in line with the bolt holes. The 4
longer Allen bolts are now inserted and tightened to secure the assembly.
The dark slide may now be removed again.
An efficient and safe technique for removal of the ferrules is to use the forceps to lever-off each ferrule directly. The method is simply to insert the narrow tips of the forceps under the metal shank of the ferrule and then to gradually push the forceps under the shank. This levers off the ferrules both rapidly and with much greater control and is the recommended technique for ferrule removal. Check the teflon pad on the ferrule base for any remaining cement, often left as un-cured liquid, removing any residue with a cotton bud. The fibre is then gently fed back into its exit slot, taking care not to touch the prism surfaces with your fingers.
Once all the ferrules are removed, the image guide socket can be removed. The IG can be gripped by hand and twisted off the plate. If stubborn the `dummy' image guide metal tube can be inserted into the socket. The IGS can then be carefully levered and twisted off the plate. Ask the day crew for assistance if unsure. Clean off any un-cured cement on the base of the IGS and then store in the plastic cannister provided.
Remove the two longer Allen bolts from the East edge and slide the spacers apart slightly. Slide the old plate out to the left. Place in its envelope and store with the other old FLAIR plates in the yellow Kodak boxes. Clean the mandrel of dust using the Nitrogen gas blast. The aluminium plate (curved lip to the south edge) is now inserted as in section 4.2.3. The marked-up copy plate of the next field is now be inserted as in section 4.2.4. again after cleaning with Nitrogen.
N.B. At this stage the gripper head should be checked for build up of UV curing cement. Any build up should be cleaned away or else ferrules might not seat properly on the plate during the subsequent fibre-ing run. The easiest way to clean the gripper jaws of glue is to soak the jaws in a bath of a suitable solvent (such as dichloromethane) for about twenty minutes. A cut-down 35mm film canister is usually employed with the gripper sent to cruise height above the plate and the gripper jaws set open. If time is important a razor blade can be employed to scrape the glue carefully off the base of the gripper. It is important to ensure that the blade does not cut in to the soft brass of the gripper so this technique should only be used if absolutely necessary.
After the last night of FLAIR-II observations the process in sections 4.7.1 to 4.7.4 is followed. Once each PH has been de-tensioned, all fibres removed and `bedded' and the IGS removed and stored, the FLAIR mask is removed as in section 4.7.5. The two halves of the PH are then gently squeezed together and the 4 short Allen bolts re-inserted just finger tight. The mandrel's surface can be quickly cleaned with a blast of Nitrogen gas. The dark slide is replaced and the PH can be left sitting on the X-Y table ready for the start of the next run.
N.B. Once familiar with the fibre-ing process a briefer summary document is available to remind you of the most important points to remember when fibre-ing. This summary is given as an appendix to this user guide.