ING Banner
Home > Astronomy > Telescope Operator Notes > Working with Autofib

AF2 & WYFFOS

AF2+WYFFOS is the multi-object, wide-field, fibre spectrograph at prime focus. It contains 150 science fibres each of 1.6" diameter, and 10 fiducial bundles for acquisition and guiding. At prime focus, the fibres are placed onto a field plate by the robot positioner AF2 at user-defined sky coordinates. Object light is transmitted along fibres 26m in length to the Wide Field Fibre Optical Spectrograph (WYFFOS). The path from Prime focus to the spectrograph consists of a prism, fibre button, 26 metres of fibre, finger, microlens and the facet block.

Startup: TCS  Autoguider  TV  
First night tasks: Rotator Centre  7-Star Calibrate  Gripper Centre  Aperture 0
Acquisition:  Configured Field   Standards
FibreGuider - for acquisition and guiding
Bad pointing
Arcs
Sky Exposures
AF2 Command Summary
Rotator Limits
AF2 GUI
AF2 MIMIC - needs adding
WYFFOS MIMIC - needs adding
Old Parameters
See also:

AF2 Main Page including recovery procedures
AF2 Commands complete list
FibreGuider Full Guide - get new link!

Important!

  • Remember to SLEEP the robot before taking science exposures: SYS> af2 sleep
  • When the dome temperature is cold (< 5 C), you need to exercise the gripper regularly: af2 wake, af2 sleep, af2 wake, af2 sleep, then it can be used as normal.

Startup

TCS:

  • USER> STATION PRIME
  • USER> INSTRUMENT AUTOFIB
  • USER> AGSELECT AF2_TV
  • USER> FOCUS 82.45-82.50 (Nov 2014) (82.5 Feb 2013)
  • ZEROSET (as usual)
  • CAL last/faint
  • ENTER AP 0 X Y
The optical configuration of the telescope is BROKEN CASS, to direct the light of the WYFFOS calibration unit via the nasmyth flat up to prime.

Focus (depends where you are looking!):

The best focus both on the WYFFOS detector and on the fiducial fibre TV is 82.25 mm. However, to compensate off-axis aberrations, the best focus in the outer parts (>=20 arc minutes off axis) of the field may be as high as 82.65 mm. Hence a compromise value of e.g. 82.45 mm is generally used. Fiducial stars right at the edge of the 29 arcmin field will be elongated by as much as 3-4 arcsec!

The focus can be checked by taking short AG exposures, and seeing if stars are round. You might want to make the focus best on the fiducials nearest the centre of the field if most of the science fibres are also near the centre of the field, i.e. weight the focus depending on where the science fibres are on the field.

On the mobile probe the best focus is at a position of about 83.20 mm. This may be important to know when acquiring faint sources with help of the TV, since the throughput is worse than on the Autoguider. However, for most targets, it won’t be necessary to change the focus (and it’s recommended not to change it so as not to forget to change it back before the science exposures).

Autoguider: AG5

See also FibreGuider Acquisition and Guiding Instructions

This camera looks at the output end of the 10 fiducial fibre bundles, also called "fiducial fibres", each with a FOV of about 8" on the sky. Each indiviual bundle has 10 smaller fibres which carry a proper image of the star to the camera. These fiducial stars will be included in the configuration file (done by the observer beforehand) and should be 14-15 mag. The astrometry must be VERY accurate in order that several fiducial stars coincide with fiducial fibre positions AT THE SAME TIME. The 10 bundles are arranged in a rough circle with numbers in this order: (clockwise from top) 144, 128, 96, 64, 16, 80, 160, 32, with 112 and 48 in the middle.

uDAS Autoguider commands and startup:

  • Check that the switch box for the autoguider has to be in position 4: (e.g. UDASDEV2 on whtdas18)
  • On the DAS machine see DEWARS, log in as 'whtobs'
  • Type 'obssys' then 'startobssys' in the yellow window
  • Then type 'startag AGX' where X is the number of the camera being used
  • Setup: Preserve during load (set all 3)
  • To make the star move in conventional cursor-arrow directions set in Zoom menu: Invert Y and 0 deg
  • Region -> load region -> choose the most recent region file in /home/whtobs/AF2/OldReg/ag5_date.reg The large circles are the FOV of the ficucial bundles (8”) and the smaller ones are the size of the science fibres (1.6”).
  • To shutdown the AG system use 'shutdownag AGX' where X is the same as above
To start guiding WITHOUT FibreGuider (old method):
  • Set integration time, probably a few seconds due to the low throughput
  • Set guide box size e.g. 75 pixels
  • Choose a bright, round fiducial near the centre of the configured field and note its fibre number
  • Select 'Manual select guide star' in SDSU AG window
  • Click 'Guide' and click on the fiducial you want to guide on
  • It may be necessary to enter 'AUTOGUIDE ON' (F9) on the TCS
Note that the switch-box for autoguider has to be set to the corresponding autoguider!

Acquisition TV: AG2

The camera looks at the output end of the mobile probe fibre bundle and can be moved in front of fibres to pick up stars. The FOV is approximately 20" x 30". The mobile probe is used for acquiring, especially standard stars on individual fibres and science fields if the telescope pointing is not good enough to see the fiducials in the fibres. With good pointing you may try to acquire science fields directly with the Autoguider. On the mobile probe the best focus is at a position of about 83.20mm.

uDAS TV commands and startup:

This is the same as for the uDAS Autoguider. The TV is called AG2 (see DEWARS for the DAS machine number).

  • On the DAS machine, log in as 'whtobs'
  • Type 'obssys' then 'startobssys' in the yellow window
  • Then type 'startag AGX' where X is the number of the camera being used, e.g. AG2
  • Setup: Preserve during load (set all 3)
  • Set up a window to view the mobile probe fibre, typically:
    SYS> window AG2 1 "[485:860,275:620]" (Oct 2015) (was "[155:515,25:490]" in Nov 2014 and "[460:900,65:460]" in May 2015) which can also be set using the Camera menu
  • To get the same directions as the conventional arrows, set in the Zoom menu: INVERT X  (ROTATE 90 is no longer needed)
  • Display the most recent region file in /home/whtobs/AF2/OldReg/af2_ag2_date.reg, marking the rotator centre (RC) and the gripper centre (GC)

Atmospheric Dispersion Corrector:

Check the PF atmospheric dispersion compensator is in 'ADC Only mode' (top right of prime mimic on ICS). If in doubt type:
  • SYS> PF_MODE ADC_ONLY
  • SYS> ADC_AUTO ENABLE

Other screens and consoles:

  • The AF2 TV screen (little monitor above VCR) is used to view the back illuminated fibres when a new field is set up. You see the white spots defocussing as the robot picks the fibre up.
  • Video Recorder. Used to record each new configuration setup in case of problems with fibres not being found etc.

First Night Tasks

Rotator Centre (RC) 

  • Make sure that CAL LAST, STATION and INSTRUMENT are set correctly
  • Mark the TV region, if not already done, which should cover the active (i.e. brighter) area of the mobile probe coherent fibre bundle. You can use the last known region and correct it if necessary. Make a note of the new region
  • Move to a 12th mag. star  (e.g. ft22+30 etc from the pointing grid)
  • USER>ROT SKY 0  (or 180, or any value you like)
  • Start the TV in continuous readout, viewing mobile probe
  • SYS> af2 wake
  • SYS> af2 viewsky 0 0   You could also start with using the viewsky X Y found last time (which may be 0 0 anyway). It’s possible that the RC will be too far from the centre of the TV and you will need to choose suitable X Y values so that the RC is closer to the centre of the mobile probe. This will ensure that when you rotate by 180deg you will see both positions of the star. You will have to see how it moves when you rotate.
  • When in position measure x and y pixel positions of the star
  • Rotate by 180 degrees, then measure x and y pixel positions again
  • If necessary, to keep the half circle on the mobile probe, move the star towards the rotator centre using the XY handset, to reduce the radius of the circle the star describes while rotating.
  • Calculate the new rotator centre (see HERE for the most recent values)
  • Mark the rotator centre and save the region file for future reference and write it in the logbook

7-Star Calibrate

  • SYS> af2 viewsky 0 0 (or X Y, using the same values as before to view the rotator centre)
  • USER> ROT MOUNT 'angle' (to stop the rotator moving and avoid problems with the rotator limits)
  • USER> CAL FAINT
  • TV in continuous readout, viewing mobile probe
  • Centre up all stars on the rotator centre and hit HANDSET after each one is centred

Gripper Centre (GC)

This should always be done on the first night, but it is usually also checked on subsequent nights, maybe once a night, or more often if you have acquisition problems (especially for standards). The gripper centre is needed to acquire a standard star on an AF2 science fibre, correcting for a possible offset between GRIPPER centre and MOBILE PROBE centre and also for a possible small displacement of the mobile-probe fibre at the AG CCD end. It is also used to find fiducial stars if the pointing is not good enough to see them directly.
  • The rotator centre should have been found already and the 7-star calibrate done too.
  • You must have a fiducial fibre somewhere in the field to determine the gripper centre.
  • If you have a configured field setup (AFSETUP already done) then this will be the case.
  • If there is no field already set up, you can still find the gripper centre by placing just one fiducial fibre at the centre of the field:, e.g. to move fibre 80:   SYS> AF2 movefibre 80 0 0   moves fibre 80 to position 0,0 and you can find the gripper centre here, using Sky PA 0 (or any Sky PA).
To determine the gripper centre:
  • USER> gocat 'standard star' e.g. one from the pointing grid will be fine
  • rot sky XXX (if it's a configured field you need to use the specified Sky PA)
  • Move the star to place it under a fiducial fibre - choose one near the centre of the configured field, or the one that you have placed at the centre:
    • SYS> af2 wake
    • SYS> af2 offsettel SKYPA STARNAME FIDUCIALFIBRENR
    • Answer the questions (first 'n' then 'y') - this offsets the t/s to put the star down the fiducial fibre and moves the mobile probe to see this fibre
  • Centre the star on the TV where you think the gripper centre is. Try using the most recent value, or one from the previous run, or if you have no idea then use the rotator centre
  • SYS> af2 park (otherwise you won't see the star on the autoguider)
  • Centre the star well on the fiducial fibre on the autoguider
  • SYS> af2 viewobject FIDUCIALFIBRENR (same fiducial fibre as above)
    • See where the star ends up on the TV
    • Mark this position as the new gripper centre and save the region if required
    • See HERE for the most recent values and some old ones

Aperture Offset Setup

Using a multi-object instrument the aperture 0 offset should be defined , so the field rotates around the instrument centre. In very old notes, the instrument centre was defined as the centre of the mobile probe, but this seems at least arbitrary and probably incorrect. The real centre of the instrument is the GC, this is the point where you want to acquire stars onto. So by moving a star from the RC to the GC you will measure the required aperture offset and everything will be centred on the GC rather than the RC. This AP 0 can be left in the TCS for all the run.

Usual method:
  • go to a pointing grid star e.g. ft20+22
  • SYS> AF2 viewsky 0 0  (assuming you are able to view the rotator centre like this)
  • centre the star on the rotator centre with the HANDSET in X_Y mode
  • Using HANDSET in APOFF mode, move star to the GC, then exit HANDSET mode
  • USER> store ap 0
Alternative method, if you need to use a VIEWSKY X Y (not 0 0) to view the rotator centre:

  • go to a pointing grid star
  • SYS> AF2 viewsky X Y
  • centre the star on the rotator centre
  • move the mobile probe in steps of e.g. 300 to af2 viewsky 0 0 while maintaining the star on the rotator centre using HANDSET in APOFF mode, then exit HANDSET mode
  • USER> store ap 0
  • Test by going to another star, which should return close to the previously marked rotator centre (within normal telescope pointing error) and when rotating should describe a circle around this

If a known good fiducial fibre can be placed on the instrument centre (which is NOT possible if a field is setup), you can use this method, which also finds the gripper centre:

  • go to a pointing grid star
  • SYS> af2 viewsky X Y (same as before)
  • centre the star on the rotator centre
  • Move the mobile probe in steps of e.g. 300 to af2 viewsky 0 0 while maintaining the star on the rotator centre using HANDSET in APOFF mode
  • park the robot and centre the star on the fiducial fibre 
  • USER> store ap 0
  • SYS> af2 view sky 0 0. 
  • mark the star position as the gripper centre 

Acquiring a configured field

It is recommended to GOCAT to the field position BEFORE doing the fibre setup. For reconfiguring or for parking fibres, it’s OK for the t/s and rotator to be tracking, but not slewing. Also it’s recommended to do it at el >45 (in case the fibres fall down). The AF2 TV screen is used to view the back illuminated fibres when a new field is set up. You see the white spots defocussing as the robot picks the fibre up. The astrometry used by the observer to generate the configuration, must be accurate in order that several fiducial stars coincide with fiducial fibre positions at the same time.

  • GOCAT to the field and check for rotator limits
  • Set the Sky PA and AP 0
  • The observer must then run AF2 SETUP to configure the field. This takes about 15 minutes
  • Centres of the Fiducial Fibres: The fibre positions and/or shapes on the AG camera change with time, perhaps they depend slightly on EL. Before acquiring a new field check and if necessary move the marks of all fibres, illuminating the fibres using:
  • SYS> wcallamp He (W, Ne or Hg are other options) or  SYS> wcallamp 4  (4=He)
  • The regions should then be saved to the file /home/whtobs/AF2/OldReg/af2_date.reg, using Region -> Save
Quick acquisition when the pointing is very good:
  • The telescope pointing must be reasonably accurate (Calibrate Sky RMS <6" approx.) to put the fiducial stars directly into the fibres without any offsets. 
  • If it's not good enough, do a short spiral search with a step size of 10" with X_Y handset mode
  • You can test the pointing on a bright pointing grid star by putting the "mobile probe" in the center of the field using AF2 VIEWSKY 0 0. Moving this star onto the GC using handset will tell you the pointing correction. If the pointing star is close to the target field, you can apply this correction at the field position and the fiducial stars should fall into the fields of view of the fiducial fibre.
  • A rotator offset must be applied. If the pointing is accurate (or the above correction is applied) then fiducial stars close to the centre of the field will show up in the corresponding fibres. If there are no fiducial stars close to the centre then you may need to move the rotator about -0.2 degrees using the HANDSET. Wait for 3 readouts of the acquisition camera to see the updated position. 
  • use FibreGuider to complete the acquisition and start guiding
  • af2 sleep
Longer acquisition when the pointing is not good enough to put the fiducial stars directly into the fibres using the above method. This uses the larger field of the moving probe to first get a fiducial star close to the correct position, i.e. on GC (and it helps to find the rotator offset from scratch, but we don't use this method because the offset is always close enough to -0.2deg).  

  • SYS >AF2 WAKE
  • To acquire, move the robot to a bright fiducial object close to the instrument centre:
  • SYS> AF2 VIEWOBJECT#
  • you should see the star on the TV
  • centre the star with the XY HANDSET on the GC.
  • (optional step, not needed) Now move the mobile probe to another fiducial star at the edge of the field; this time move the rotator to centre the star as well as possible, then park the robot
  • SYS > AF2 PARK, now you should see most of the fiducials in the fibres on the autoguider
  • use FibreGuider to complete the acquisition and start guiding
  • SYS > af2 sleep

Acquisition and Guiding with FibreGuider

A new acquisition tool for AF2, FibreGuider, was written by Richard Bassom in semester 2013A. It was written to enable more accurate and repeatable acquisition with AF2 and WYFFOS. It calculates the required XY translational shift and rotation to best centre all stars in the fiducial fibres. A brief outline of how to use the tool is given below:
  • first use the above acquisition procedures to ensure that at least 3 (or maybe even just 2) fiducial stars are seen in their fibres
  • start the acquisition tool:
    DAS> fibreguider &
    at the fiducial AG DAS prompt
  • stop the TV
  • in the Acquisition Tab click 'Calculate' to take an image and automatically determine the required shift in XY and rotation
  • then with the three checkboxes (apply XY, rotation and use TCS guide) ticked, click 'Apply'. This will send the rotation and XY offsets to the TCS as a TWEAK and then it takes another image
  • repeat as required so that the XY and rotation offsets are <0.01 degrees for rotation or <0.5" for X and/or Y)
  • then in the Autoguider Tab, start the autoguider guiding on the arbitrary position 250 250 with GUIDE. You should see these commands arrive at the TCS. 
  • when the guiding RMS stops decreasing, you are ready for the science exposures
  • if you need to stop guiding (e.g. for arcs), click STOP in the Autoguider Tab
  • to begin guiding again click GUIDE, which will gradually bring the fiducial stars back into position - check that the guiding RMS is as low as it can get

If you don't see the fiducials just after the gocat, you can press the Spiral Search button in FibreGuider and it will stop every time the software finds a star inside the fiducial regions (or hot pixels). You can then continue with the search or stop it. However this is not efficient yet. It is a lot quicker to try and move a bit with handset (only a few tries) and then wake the robot and see the mobile probe on the acquisition TV.

If you happen to need a spiral search using the acquisition TV, this button is useful since you can select the offset you want and it's easier to perform a spiral search without human mistakes, although you always have to continue/stop the search since this is not implemented in the software and can either stop with hot pixels or not stop even if you have found the star close to the fiducial fibre.

Acquiring standard star onto science fibre

A configured field will already have been set up before doing standards. There is no guiding for standards, as there won’t be any guide stars located in the fiducial fibres.

In the following steps, the detour via a fiducial fibre is to check or re-determine the GC, as the fiducial fibres have the same optical centre as the science fibres; this way you cannot miss the standard star.

  • USER> gocat star
  • USER> rot sky xxx
  • SYS> af2 wake
  • SYS> af2 offsettel skyPA standardname fiducialfibre#
    Answer the questions (first 'n' then 'y', since the first 'y' would do another GOCAT)  this moves the star from the centre of the field onto one of the fiducial fibres
    (preferably close to the centre of the FoV)
  • centre star on the GC using the handset
  • SYS> af2 park
  • move the star to the centre of the fiducial fibre using the HANDSET
  • SYS> af2 viewobject#   to view this fibre on the TV
  • mark this position - it’s the new GC
  • SYS> af2 offsettel skyPA standardname sciencefibre#    this moves the star to put it down science fibre
  • Answer the questions first ‘n’ then ‘y’ (NB: if you say ‘y’ to the first question, it will do a GOCAT).
  • recentre it on the gripper centre
  • SYS> af2 sleep
Repeat just these last 4 steps to put the same standard star down different science fibres as required

Then to go to a different standard star:

  • SYS> af2 offsettel SKYPA STARNAME SCIENCEFIBRENR
  • Answer the first question with 'y', which does a GOCAT
  • Centre the star on the TV on the gripper centre
  • Then answer 'y' to the second question. The telescope will offset to put the star in the fibre.
  • Centre the star again on the TV on the gripper centre. This second recentering will be more pronounced for fibres further from the field centre.

Bad pointing (guide stars not found)

If the pointing is too bad even to see the star using the long acquisition method, first check the obvious (petals open, check t/s focus, check encoders) then put the "mobile probe" in the center of the field with AF2 VIEWSKY 0 0.  GOCAT to a pointing grid star which is very close to your field. Then move this star onto the GC and  the offset will be the approximate pointing correction. Since the star is close to the target field, you can apply this correction at the field position and the fiducial stars should fall into the fields of view of the fiducial fibre. 

Alternatively, if you can't see the fiducial stars, perform a spiral search. In this case it is recommended to re-define the gripper centre on the mobile probe.


Arcs

When you take arcs, either offset the telescope away from the object or close the petals, as otherwise you get the two signals superimposed.

Sky Exposures

  • stop guiding by pressing STOP on FibreGuider
  • offset t/s as required
  • take sky exposures
  • offset back to the original position
  • press GUIDE on FibreGuider, to pull the fiducials back into position

 

AF2 Commands

For a full list of AF2 commands, see here. These are a few of the most useful commands:

SYS> AF2 WAKE  (turn on the robot so it can be used for acquisition)
SYS> AF2 PARK  (move the robot to its parking position at the edge)
SYS> AF2 SLEEP (move the robot to its parking position at the edge AND turn its power off)
SYS> AF2 VIEWOBJECT 112 (or any of the fiducial fibres)
SYS> AF2 VIEWSKY 0 0   or VIEWSKY X Y  (view a particular area with the TV)
SYS> AF2 MOVEFIBRE 16 0 0  (to place fibre 16 at the centre, 0 0 position)
SYS> AF2 OFFSETTEL (for doing acquisition)

Rotator Limits

The Mount PA has a range of -163.37 to 193.63, so there is a blind spot of 3 degrees. It is useful to be able to determine the time at which a rotator limit would be reached for a new field, without having to move the telescope and read the information off the TCS. This can help to plan the observations during the night.

To calculate when a rotator limit will be hit
  • Use Staralt to plot the Sky PA during the night for the proposed field
  • If the parallactic angle is increasing throughout the observations, the rotator will be moving towards a negative limit
  • If the parallactic angle is decreasing throughout the observations, the rotator will be moving towards a positive limit
  • Use the formula:

    Parallactic Angle = Sky PA - Mount PA

  • Given the Sky PA of the observations, and the Mount PA of either the positive or negative limit, obtain the parallactic angle for the time at which the rotator limit will be reached
  • Use the staralt plot to read off the time for this parallactic angle

To determine if a rotator limit will be hit for fixed time observations

For a planned observation with fixed Sky PA and for a definite start and end time, we can determine in advance if a rotator limit will be reached during the observations:
  • Use Staralt to plot the Sky PA during the night
  • Read off the parallactic angle for the start and end of observations
  • Use the formula:

    Mount PA = Sky PA - Parallactic Angle

  • Make 2 calculations, one using the parallactic angle at the start of observations, and one at the end
  • Obtain the Mount PA for both these times: if it falls within the permitted rotator range then no limit will be reached

AF2 GUI

Troubleshooter

If a fibre gets "lost" you have to move the t/s to access park: Move first to AZ 300, EL 15; Switch to ENG; Press AP2 on platform; Open gate; Press AP3 on platform 

Then follow the instructions on the AF2 main page


Old Parameters

Gripper centre
(viewsky 0,0)

172,213

181,221

151,219

82,239

21/11/05

17/08/05

20/04/06

21/02/13

Rot. Center Af2
viewsky -600,500

viewsky 0,0

148,116

202,208

192,174

59,247

21/11/05

17/08/05

20/04/06

21/02/13

Telescope aperture '0'
0.6,8.9
21/11/05



Top | Back

Contact:  (Observing Support Assistant)
Last modified: 17 November 2015