4-m MOSAIC II CCD IMAGER

COOKBOOK

K. Olsen, 7/28/03

DON'T PANIC !!

You are sitting in front of ctioa8, the Mosaic II observing computer. An observer support person (Alberto Alvarez or Claudio Aguilera) has probably already set up the instrument for your observing program. Here's how to verify the setup:

Three windows that look like the images below should be on your screen. If they are not, then you will need to start up these windows. See the documentation on Using VNC with Mosaic. Note that these are virtual desktops, one for taking observations, one for the automatic transfer of images to ctioa8, and one for the real time display. Do not use the observing desktop for anything other than taking observations! (Why? The observing desktop runs on ctioa1, which is a very old machine. Your data are transferred to ctioa8 automatically by the DCA. If you type e.g. "display obj001.fits 1" in the Arcon window, you are asking ctioa1's slooow CPU to display an image that it accesses across the network. Bad idea!).
The following images show the VNC windows:

Now it's time to check the relevant parameters in the Arcon parameter files. Since Arcon uses an IRAF interface, these are IRAF parameter files. There are five parameter files you will need to check out: obspars, instrpars, wheel1, detpars, and telpars. Typing "lpar obspars" will list the contents of obspars, for example, while simply typing "obspars" or "epar obspars" will allow you to edit the values. The exception to this last statement is detpars: to edit detpars, DO NOT type "detpars" or "epar detpars"--you must instead type "setdet force+", so as to ensure that your changes are updated in the instrument.

obspars:

      picture = 1001            Picture number of first exposure
    (template = "")             Image name template
  (autopicnum = yes)            Generate picture number automatically ?
   (detupdate = yes)            Update detector parameters automatically ?\n\n#
     (postpic = "")             Post processing command for single picture
     (postseq = "")             Post processing command for sequences
     (restart = yes)            Restart the server on every command ?\n\n# ADDI
   (observers = "Galvez, Bonati") Observers
      (propid = "007")          Observing proposal ID
    (comments = " ")            Comments
     (comfile = "")             Observer header comments file\n
        (mode = "ql")

What might you want to change in obspars? The fields "observers", "propid", and "comments" are all added to the header and to the automatic observing logs, and should correspond to your program. If you leave "template" blank, then your images will have our standard prefixes ("obj" for objects, "dflat" for dome flats, "zero" for biases, "sflat" for sky flats, and "focus" for focus frames) plus the picture number given by "picture" (which is automatically incremented for you as long as "autopicnum" is set to "yes"). Some observers like to use one template for all types of exposures, which they enter in the "template" field. It's also popular to change the "picture" number at the beginning of each night (say 1001 for the first night, 2001 for the second, etc.) so as to easily distinguish images taken on different nights.

instrpars:

      filter1 = "VR"            Filter in wheel one
      (wheel1 = "")             Filter info. pset for wheel one\n\n# SELECTING 
  (setfilters = "yes")          Query and set filters?\n\n# SETTING FOCUS FOR E
    (setfocus = "auto")         Query and set focus?
 (temperature = 5.7)            Telescope truss temperature (C)
   (basefocus = 17600.)         Telescope focus base value
     (reftemp = 5.2)            Telescope temperature for base focus value (C)
    (tfrcoefs = "1 -135.0")     Coeficients of Temperature-Focus relationship\n
    (tv1focus = -6.203)         Focus for camera 1
    (tv2focus = -2.499)         Focus for camera 2\n\n# SHUTTER READY POSITION
(shutter_read = "auto")         Position of shutter when NOT exposing\n
   (instrname = "mosaic")       Instrument name
        (mode = "ql")

In instrpars, the parameters of greatest concern to you will be "setfocus", "temperature", "basefocus", and "reftemp". By setting "setfocus" to "auto" (RECOMMENDED), you allow the telescope to update its focus for changes in temperature (using the coefficient in "tfrcoefs", where "1 -135.0" means a change of -135 microns of focus for a 1 degree change in temperature). This procedure is described in detail below, under the focus procedures. You should also normally leave "setfilters" to "yes", so that the Arcon will ask you for the filter whenever you take an exposure. The ONLY reason you would want to set it to "no" (in which case Arcon uses whatever filter is in "filter1"), is if your program uses only one filter.

wheel1:

         (id1 = "U")            Short identifier for filter in position 1
       (name1 = "U c6001")      Full name of filter in position 1
      (focus1 = -200.)          Focus value for filter 1 \n
         (id2 = "B")            Short identifier for filter in position 2
       (name2 = "B Harris c6002") Full name of filter in position 2
      (focus2 = 10.)            Focus value for filter 2 \n
         (id3 = "V")            Short identifier for filter in position 3
       (name3 = "V Harris c6026") Full name of filter in position 3
      (focus3 = 30.)            Focus value for filter 3 \n
         (id4 = "R")            Short identifier for filter in position 4
       (name4 = "R Harris c6004") Full name of filter in position 4
      (focus4 = 0.)             Focus value for filter 4 \n
         (id5 = "I")            Short identifier for filter in position 5
       (name5 = "I c6028")      Full name of filter in position 5
      (focus5 = 0.)             Focus value for filter 5 \n
         (id6 = "z")            Short identifier for filter in position 6
       (name6 = "z SDSS c6020") Full name of filter in position 6
      (focus6 = -15.)           Focus value for filter 6 \n
         (id7 = "M")            Short identifier for filter in position 7
       (name7 = "M Washington c6007") Full name of filter in position 7
      (focus7 = 65.)            Focus value for filter 7 \n
         (id8 = "D51")          Short identifier for filter in position 8
       (name8 = "DDO 51 c6008") Full name of filter in position 8
      (focus8 = -35.)           Focus value for filter 8 \n
         (id9 = "VR")           Short identifier for filter in position 9
       (name9 = "VR SuperMacho c6027") Full name of filter in position 9
      (focus9 = 30.)            Focus value for filter 9 \n
        (id10 = "o3")           Short identifier for filter in position 10
      (name10 = "o3 OIII c6014") Full name of filter in position 10
     (focus10 = 130.)           Focus value for filter 10 \n
        (id11 = "g")            Short identifier for filter in position 11
      (name11 = "g SDSS c6017") Full name of filter in position 11
     (focus11 = 30.)            Focus value for filter 11 \n
        (id12 = "r")            Short identifier for filter in position 12
      (name12 = "r SDSS c6018") Full name of filter in position 12
     (focus12 = 120.)           Focus value for filter 12 \n
        (id13 = "i")            Short identifier for filter in position 13
      (name13 = "i SDSS c6019") Full name of filter in position 13
     (focus13 = -20.)           Focus value for filter 13 \n
        (id14 = "u")            Short identifier for filter in position 14
      (name14 = "u SDSS c6021") Full name of filter in position 14
     (focus14 = 230.)           Focus value for filter 14 \n
   (first_pos = 1.)             First Filter positions in wheel (read only)
    (last_pos = 14.)            Last Filter positions in wheel (read only)
        (mode = "ql")

The wheel1 parameter file contains all of the information about the filters: their short identifiers (the name you type when requesting that Arcon use a certain filter, and the name that appears on the GUI), their full names, and their focus offsets with respect to the R filter. The focus offset is applied automatically when taking an exposure; Arcon assumes that the base focus value that appears in instrpars (see above and the section on the focus procedure) corresponds to the R filter. Normally, you should not need to change anything in the wheel1 parameter file; we recommend, however, that you check to make sure that the focus offsets are set to the measured values listed on the filter page.

detpars:

        (xsum = 1)              pixels summed in X direction 
        (ysum = 1)              pixels summed in Y direction 
       (ysize = 8192)           Height of centered ROI (full width) 
     (pixsize = 15.)            Pixel size in microns
    (nxpixels = 2048)           Detector size in X 
    (nypixels = 4096)           Detector size in Y 
     (detname = "Mosaic2")      Detector identification
        (mode = "ql")

The detpars file is where you can specify binning, if you want it, and set regions of interest (ROIs). Note that setting ROIs does not lower the readout time as much as you would expect; talk to you staff contact before using ROIs. Also, as stated above, do NOT change the parameters in detpars by typing "detpars" or "epar detpars"; you MUST use "setdet force+".

telpars:

     telfocus = 158211          Telescope focus
       nfexpo = 7               Number of focus exposures
   freference = 17500           Focus value
       fdelta = 100             Focus increment
   (ra_offset = 0.)             RA Offset (arcsec)
  (dec_offset = 0.)             Declination Offset (arcsec)\n\n# PARAMETERS FOR
      (fnrows = 30)             Focus number of rows to reverse shift
       (refis = "middle")       Reference is first, middle or last ex\nposure?
     (focmode = "auto")         Focus mode\n\n# FOR INFORMATION ONLY
     (telname = "ct4m")         Telescope name
     (station = "pf")           Focal station
      (fratio = "f/2.9")        Focal ratio
  (platescale = 18.)            Plate scale (arseconds/mm)
        (mode = "ql")

The current telescope focus is stored in telpars (do not change "telfocus" by hand, as it is updated automatically as long as "setfocus" in instrpars is set to "yes" or "auto"), as are the parameters that control how focus sequences are taken. Depending on the seeing, you may want to change the focus step ("fdelta") between successive exposures in a focus sequence, or the number of focus exposures in a sequence ("nfexpo"). The values listed above are good all-purpose values.

Now that you have set up the Arcon environment, take a look at the MCCD Configuration window. Are there green check marks by the words "TCS" and "MSE"? If either of these are highlighted in red, you'll need to contact observer support (x420). Likewise, if any of the temperature indicators (Dewar, Fill neck, or CCD) are red, there may be a problem--contact observer support. There should be a thick red line at the top of the window, indicating that the shutter is closed. There should also be two green arrows between the shutter graphic and the blue line indicating the filter, which means that light is getting to the guide cameras.

Taking exposures

First you'll need a directory to store your data. In the Arcon Acquisition window, go to the mosaic directory on one of the free disks on ctioa8 (e.g. /ua83/mosaic, /ua84/mosaic,/ua85/mosaic--NOT /ua80/mosaic), make yourself a subdirectory, and cd to it.
Start the autologger. Click on the autologs button and enter the relevant information in the window that pops up. Once you click ok, logs will pop up containing information for each image that you take.
Take a test exposure by typing "test", specify that you want to take a "zero" frame. As the bias frame reads out, watch the Arcon status window, the DCA windows, and the real time display (RTD). Are the "Buffers read" in the Arcon status window ticking over? Are the "Pix Pkts" changing in the DCA control window in time with the "Buffers read"? Is the image appearing on the RTD? Once the zero frame is read out, does it look more or less like this? If there are nasty-looking patterns in your zero frame, take a few more to see if they go away. If not, then contact observer support.
If your first zero image looks ok, then you are set to go ahead with calibrations. Take more zeros by typing "zero" or "observe" and then "zero" (make sure that the dome is dark for this), then ask the night assistant or observer support to point the telescope at the white spot and turn on the flat field lamps. Take ~6 dome flats through each filter, using the exposure times listed on the board. The command "doobs" will make doing this easier, once you've taken one exposure through each filter to verify that the exposure times are right (using the "observe" command or simply the "dflat" command). You don't want more than 10000 counts in your flats, since the saturation levels for some of the CCDs are ~20000. If the standard dome flat exposure times produce fewer than ~10000 counts, then the dome lamps may be misaligned or one or more may have burned out. Please ask observer support to check the dome lamps if this is the case.
Observer support will open the dome 1 hour before sunset to clear the air out of the dome. Right after sunset, prepare to take twilight sky flats. Ask the night assistant to point the telescope 1-2 hours to the east. Take 1-second "test" exposures, starting with the filter that you expect will require the brightest illumination, until the counts drop to ~10000. Then use "sflat" to take successive sky flat exposures, asking the night assistant to move the telescope E by 30 arcseconds between exposures. With the 100-second readout time, you will need to increase your exposure time by a factor of ~1.5 between exposures to maintain the same count level.
Once it is too dark for twilight sky flats, ask the night assistant to move to a bright star to zero point the telescope. Take a 1-second "test" object exposure, and calculate the offset needed to bring the star to the center (the pixel scale is 0.27 arcsec/pix). Alternatively, there is a task in CTIOs iraf that calculates the RA and Dec offset for you. Just type "zpt" and then click the mouse on the center of the star. You can have the night assistant do the offset, or do it yourself using the "teloffset RA_offset_arcsec DEC_offset_arcsec" command, bearing in mind that DEC increases with X and RA with Y. Take another 1-second test to verify that you moved the telescope in the right direction, and by the right amount.
Telescope focus procedure. Start the focus procedure by typing "settemp"--this command reads the temperature of the upper truss and records it in the "temperature" parameter of instrpars, described above. It also takes a guess at the best focus value--use this value as the focus value of the middle of your sequence. Now type "focus", specify an exposure time of 10-15 sec, and use the R filter (if you use another filter to do the focus, bear in mind that Arcon will add the focus offset listed in wheel1 to the focus when taking an exposure). Using the parameter values specified in telpars, Arcon will now take a sequence of N exposures, stepping the focus and shifting the charge by a number of rows between each one. After the exposure is done reading out, you will see an image containing columns of stars on the RTD. It will look something like this and like this if you zoom in.

In an IRAF window on ctioa8, type "mscfocus". You will now select stars to use in calculating the best focus. Place the cursor on the top star in a cleanly separated, unsaturated sequence and type the "g" key (always use the top star, regardless of whether or not there is a double step between it and its neighbor). A window will appear displaying the FWHM as a function of focus for that star. Type "q", place the cursor on the top star of another column, and type "g" again. Repeat this sequence until you have measured the focus profiles of 1-2 stars on each CCD. With the cursor in the ximtool, finally type "q". You will now see a window overlaying the focus sequences for all of the stars you just measured. If there are strongly outlying points, place the cursor on them and type "d", which deletes the information for the whole star associated with that point, Once you are satisfied, type "q", and the program calculates the best focus for you. Enter this value in the "basefocus" parameter in instrpars, and change "reftemp" to equal the value in "temperature". If "setfocus" is set to "auto", Arcon will calculate the difference between "temperature" and "reftemp", multiply this by the temperature coefficient, and add it to "basefocus" before taking an image. Until you type "settemp" again, this focus adjustment will be 0. Later during the night, if you find that the temperature has changed significantly, type "settemp" before taking an exposure, and "temperature" will be updated.
If you don't like using "mscfocus", you can always use "mscexam" instead. "mscexam" works just like "imexam"--you can examine stars with the "r" or "e" key to pick which of the stars in the focus sequence is in best focus.

Observing tips

For taking dither sequences to remove the gap signature, "mosdither" is a useful command. Type "epar modither" to look at its parameters. The files characterizing various dither sequences are stored in /ua80/mosaic/dithers/. You will need a 5-point dither to cover all areas with at least 3 images.
If your program involves a large number of short exposures, then "mosocs" could save you a lot of time. See the MOSOCS documentation for help.
If Arcon or the DCA crash, check to make sure that /tmp on ctioa0 isn't filling up with *xpim* files. If it fills up, the only recourse is to reboot ctioa0.