Skip to content. | Skip to navigation

Personal tools
You are here: Home draft The WFCCD in Multi-Grism Mode


Ray Weymann, Bill Kunkel, Andy McWilliam, Mark Phillips, September 1999

Revised Miguel Roth, Mark Phillips, Nidia Morrell, December 2003

                                                               Last modified by Nidia Morrell, June 2010

Table of Contents


1. Introduction

This manual describes the use of the Wide-Field CCD (WFCCD) camera for obtaining multislit grism spectra direct images and long-slit spectra on the 100-inch du Pont telescope. The WFCCD reimages a 25 arcmin diameter field onto the WF4K CCD camera, with a scale of about 0.484 arcsec/pixel. It produces good images over the full field with a modest radial degradation, and has good transmission from 3750-9000 A. Field distortion is minimal so that the camera can be used for drift scanning in strips close to the equator. It also produces a collimated beam with a 70 mm pupil in which filters, grisms, or a Fabry Perot etalon can be placed.


This manual provides a description of the advanced preparations, setup, and operation of the WFCCD for multislit and long-slit spectroscopy.


Remember that you must send in your set-up form prior to your run.


2. Slit Masks


There are several slit mask holders, together with the screws and holddown bars kept in a wooden box labeled "WFCCD SLIT MASK HOLDERS" which is in the WFCCD cabinet in the small room behind the mezzanine. Please keep all the hardware in this box and return the box to the cabinet. Please do not leave bits and pieces lying around in the control room! Several long slits permanently mounted in their holders are available. In particular there are 2 useful long slits. These are:

      1. A 1.65 arc-sec slit (150 microns), and

      2. A 8.65 arc-sec slit (750 microns) used for spectrophotometric standards.


These two long slits are off centered in such a way as to get a full wavelength coverage on the detector.


Preparation of the metal masks is covered in a separate document on this same web page . Make sure to send in your  mask generation files as far in advance as possible. Masks submitted later than 6 weeks prior to the run can not be guaranteed. Avoid painful experiences...


The metal masks are fabricated on Las Campanas, and will be handed to you upon arrival.


To mount a mask, orient the mask frame holder so that the curved piece is away from you and the small screws holding the hold-down bars are up. There is then a single pin on the right border of the frame and two pins on the lower border. (These pins define the position of the metal masks precisely). Using a 3/32" Allen wrench, remove the hold-down bars (only two of the hold-down bars are currently used for the metal masks). The metal slit mask should be oriented so the single notch bears against the pin on the right side of the holder, and the two lower notches bear against the two pins along the bottom side of the holder. Place the hold-down bars over the mask and screw them in. If the holes in the slit mask interfere with the screws, you may need to slightly enlarge them with a small rattail file.



3. Mounting the Mask in the Aperture Wheel

Up to 3 mask frames can be mounted into the aperture wheel; the 4th position should be left open. The wheel is accessed by loosening the two red thumb screws and sliding off the cover. You will be holding the mask with the curved slot facing down, and the observer looking up toward the dome during this operation. Push the frame gently into the wheel and the wheel will "grab" the frame. Someone on the technical crew will assist you with this and other set-up procedures.


4. Grisms


There are presently two low-resolution grisms, both of which give a FWHM resolution of about 375 km/sec. The undeviated wavelength of the blue grism from a central slit is about 4700 A and for the red grism is about 6000 A. The blue grism gives about  2  A/pix, and the red about  3 A/pix. Slits .6 arcmin from the center of the field will enable coverage of roughly 3800-7600 A for both grisms.


A higher-resolution grism sometimes referred to as the "H & K grism" is also available. This grism has a dispersion of  0.8 A/pix, corresponding to a FWHM of about 130 km/s. The undeviated wavelength is near 3700 A. Spectral coverage depends on the location of the slit; however, wavelengths longer than ~6300 A are inaccessible.


An echellette is also available.


Let the LCO staff know well in advance of your run (via the WFCCD instrument setup form) what items you wish mounted in the filter/grism wheel. The grisms have been carefully mounted in their cells; alignment of the grisms in their cells and the cells in the filter wheel is described in a separate technical document.


Normally you will mount one grism and 3 filters in the grism/filter wheel.



5. Taking Data with the WFCCD


The WF4K CCD should be mounted and powered up. Your account on the data acquisition computer (clarity) will be initialized.

The data acquisition program is started by typing 'ccd' on any terminal from clarity. This will launch up the startup window:

                                                         wfccd startup


In this window you have to update the observer's name, select 'WF4K-1' in the 'CCD' pull-down menu, and WFCCD for 'Instrument'.

The option "WF2K-2" for 'CCD' would start the 2 amplifier readout mode, which is not fully operational yet, thus only  the one amplifier readout mode is recommended for the moment being.  The number of overscan lines and columns can be modified, although the default value of 128 is recommended.

Clicking 'OK' will start the WFCCD command window:

                                                            WFCCD  gui

                                                                                                    The WFCCD camera GUI

The first thing to do within this window  is defining the directory where your data will be stored. This is done by clicking the 'Options' pull-down menu and choosing 'Data_Path'. The working directory will usually be a subdirectory of DATA on the observers account on clarity. If this directory does not exist, it will be automatically created.

 Before taking images, you need to edit the 'Aperture' and 'Filter' names to reflect the masks and filters that are effectively mounted on the instrument. To do that just click on the 'Edit' buttons near the corresponding windows in the GUI. Then, apertures and filter/grism are changed by clicking on the small button within the 'Aperture' and 'Filter' windows.


CAUTION: There may still be occasional problems with either the aperture or filter/grism wheels not properly seating in their detents. Whenever the position of either of these wheels is changed, you must check the screen of the PC wheel controller program. A fault will be indicated by the particular button in question remaining "red". If this should occur, go to another position of the wheel until you get all blue. Then send the wheel to the desired position again.  Nevertheless, this problem will very seldom, if at all, appear.

"ExpTime" is the exposure time in seconds for the next observation. (For this parameter and the ones for Loops and File# the entry window will be highlighted red after the number is typed. The value is not recorded by the program until the user hits Enter). The exposure time can be changed any time during the exposure by entering a new number and hitting Enter.

Loops: Setting up a loop (= number of integrations wanted)  can be a convenient way to take flat field series or make synoptic observations. 

ExpType: The exposure type (saved in the fits header) can be set to "Object", "Bias", "Dark", "Flat"  or "Focus" . Selecting "Bias" will automaticaly set the exposure time to 0 seconds. Selecting "Dark" will disable the shutter for the exposure. Selecting "Focus" will start a focus sequence. The 'loops' parameter will automatically be set to 7, although you can modify this if desired.

Binning: The CCD can be binned by factors of 1, 2, 3, or 4 in rows and columns.

Subraster: Some readout time can be saved by setting a subraster. This is especially useful when observing wiht a long slit. The default "Full" frame mode readout is indicated by the "Full" label in the right side of the CCD camera GUI. Clicking on the button next to the "Full" label will start the subraster window. The subraster section can be entered as XcYcWH (X_center, Y_center, Width and Height) or X1X2Y1Y2 (starting and ending column and row numbers) , selected by clicking in the 'CoordMode' button within the subraster  window. These coordinates are in binned pixels if the CCD is read in a binned mode. The full requested number of overscan columns will be  added. After typing the subraster values, click the "APPLY" button, and then "DONE" to activate the subraster mode. It is also possible to use the option "Save" to save the selected subraster to a file that can be later invoked with "LOAD". When using a subraster, we have to set the 'save mode' to "Full" which will fill in whith zeroes the part of the image outside the subraster., thus keeping the original image dimensions.  This is needed in order to the alignment tasks to work.


The readout speed can be set to "Slow", "Fast" or "Turbo", having the following characteristics (for 1 x1 binning):

Speed     Gain   Noise (e-)   Readtime full (s)  Readtime subraster (s)      Saturation                                               

Slow       0.85        3.4                 254                      87                          60,000 DN

Fast         0.75       3.6                 144                      51                           60,000 DN

Turbo      1.40       4.8                 100                       38                           45,000 DN

"Readtime subraster" refers to the subraster normally used for long slit observing; i.e. [1:4064, 1400:2650].

The dark current for the WF4K CCD is about 3.8 electrons/hour/pixel.

Linearity and shutter time tests have yet to be carried out for the WF4K CCD.

 From the "Modules"  pull-down menu in the CCD Camera GUI, a quick look tool can be invoked, where the succesive exposures will be automatically displayed.
                  wfccd quick look
                                                                                      WFCCD  Quick-Look Tool

More details about the Quick Look Tool can be found at:


Also open from the 'Modules' pull-down menu, the Dewar_Status window will provide information on the detector temperature along the observations.  


5.1 Starting IRAF


IRAF is used to display the data and to measure alignment errors for the aperture masks. IRAF and DS9 are started by means of an strong-box icon on the lower bar on the screen of clarity.

In order to load the IRAF tasks that will be used at night for aligning the field and slitlet masks, type:


 In order for the existing WFCCD alignment software to work, all images that are used in the alignment must be flipped in the x-pixel direction after stripping off the bias columns. This is accomplished with the IRAF task wflip in the wfccd package. To flip an image, just type:

wf> wflip ccd0175 <cr>

this will produce the image ccd0175f.fits which you must invoke in the alignment procedures.


The following notes on image orientation hold for flipped images only:

The orientation of the CCD image (and the TV guider) will be a function of the Cass ring rotation angle as illustrated in Figure 1. However, with DS9 it is simple to switch polarity of the display. With the Cass ring angle set to 270 degrees, east will be towards the bottom of the screen, and north to the left. With the Cass ring angle set to 180 degrees, then north is to the top, and east to the left.



Figure 1. Orientations of CCD and TV Guider for the WFCCD for flipped images



6. Daytime Mask Exposures


Take a short un-dispersed image using the dome flat lamp or the light leaked into the dome and check the slits. You may want to check that a dispersed image of a flat field or residual dome light produces spectra which don't overlap. If they do, then the grism will need to be re-seated-- This should only be performed by knowledgeable members of the LCO staff.


For both the blue and red grisms, wavelengths increase with increasing column numbers. The wavelength solution is not well approximated by a linear fit. Use both available arc lamps, a HeAr and and a Neon lamp to get a good solution. If you want to take flat fields of for your slits, ask the person assisting you to set the telescope for flat-fields.


7. Pointing and Focusing the Telescope



7.1 Initial Pointing of the Telescope


The mask design program prints out the RA and Dec of the position at the nominal mask center. The pixel on which this position of the mask lands does not coincide with the instrument rotator axis and therefore will depend upon the Cass Ring angle. Currently the center of rotation lies at X=1869, Y= 2020 on the flipped image. Have the operator point the telescope to a bright setup star. Take a short (~1 sec) frame, flip it with the wflip IRAF task and display it in the DS9. Now compute the appropriate telescope offset to place the star at the zero point coordinates (the image pixel scale which is 0.484 arcsecs/pixel). Alternatively, you can use the IRAF task "cobject" in the WFCCD package to calculate the correct offset.  Once the star has been centered, ask the telescope operator to update the zero point of the telescope pointing.

 If you do a CSET at a position which is different from the center of rotation of the instrument, this will work only if the Cassegrain Ring angle is not substantially changed during the night.


7.2 Focusing the Telescope

The telescope should be focused in the normal way by choosing "Focus" in the ExpType  button in the ccd command window. Set the exposure time (usually 7 - 10 s) and then use the START button to begin the focus sequence.


8. Alignment of a multislit mask


  • Tell the operator the desired angle for the Cassegrain Rotator (270 degrees is the value needed when slits run in the E-W direction).
  • Select a multislit mask from the aperture wheel, and a filter (V) from the filter wheel. If you wish to save time, you can do the acquisition in Turbo readout mode,  but then, make sure that you don't forget to switch back to Slow or Fast before taking your spectra!
  • Take a short image of the mask and measure the size (in pixels) of the guide star boxes (this will be needed for alignment; and can be done in the afternoon).
  • Tell the operator to slew to your target coordinates.
    • Select the "open" position of the aperture wheel, and a filter (V) from the filter whee, and take a short exposure of the field (10 - 30s, depending on objects brightness).  Again, you can do this in Turbo readout mode.
    • Put the multislit mask in and take another short exposure (do not use an image of the mask taken in the afternoon).
    • Within the IRAF window, set the align  task parameters to something like this:                                                                                                

                             align parameter set


    • mask and field parameters refer to images taken with the multislit in and out respectively; box_file is the name of the file where the box coordinates will be stored; ring indicates the angle of the Cassegrain rotator; xtvold and ytvold are the current coordinates of the TV guider (xy5 in the guider camera window),  xsz and ysz indicate the size of the guide star boxes measured before. The remaining parameters should not vary.
  • Run the  align task and follow the instructions, as follows:
    • The task will display the image of the mask, where you should mark the positions of the boxes.
    • The field image is then displayed overlaying the positions of the boxes at the current telescope pointing. In it, you must first identify the alignment star corresponding to the first box marked in the previous setp.
    • The task shifts the boxes' positions accordingly and displays them now with circles. All your alignment stars should fall into these circles. You must now mark  all stars following the boxes order starting again from the first star (a smaller circle will now be drawn around each of them). If all goes well, the ds9 display will look like:


      •     On the IRAF screen the task will display the result of the alignment fit, issuing the  offsets in RA, DEC and CassRotator position to be applied. The offsets in RA and DEC can be preformed either by moving the telescope or the guiding box.
      • Ask the telescope operator to rotate the Cass ring accordingly. As the console display of the absolute value of the rotation angle lacks precision, the differencial encoder is a better choice. Be aware that guiding must be turned off during this procedure.
      • Ask the telescope operator to offset the telescope as indicated. For large offsets it is better to slew from the telescope console and for small offsets (~1 arcsec or less) the TV guider box positions is a better option, normally used for the second iteration.
      • Take another "open" and "masked" image. As the previous offsets should have brought the stars into the boxes you can now run falign on the masked image. The output file including box positions from align is now one of the input parameters. On the IRAF graph term you will see the fit to the alignment stars in their boxes, one at a time. Each should resemble:



  • The output of  falign is similar to that of align, and these new rotation angle and telescope offsets should have placed stars in the center of their respective boxes.  You can confirm this by taking a new  image and checking that all stars are in their slitlets (recommended). If you are not happy with it, it is advisable to iterate with falign.
  • Select the grating from the filter/grism wheel
  • Set the readout mode to Fast or Slow, as desired.
  • Take your spectra.


8. 3. Hints for acquisition of spectra and calibrations

  • The alignment needs to be checked every so often. Depending on telescope inclination, it might need some realignment after 30-40 minutes.
  • He-Ne-Ar comparison lamps, and Flatfield images  should be obtained  immediately before/after  the science images to avoid flexure  displacements.


9. Alignment for long-slit operation

For longslit observations, it is recommended to use a subraster to save readout time. A suitable subraster is:

X= 1 - 4064

Y = 1400 - 2650

While setting the subraster mode, keep the 'SaveMode' in "Full", so the resulting frames will still have dimensions [4064:4064]. Otherwise the alignment software will not work.

Make sure that the wfccd package is loaded in your IRAF session in clarity.

Remember that in order for the existing alignment routines to work, all the images used in alignment nedd to be flipped by using the IRAF task wflip.

As mentioned before, the more commonly used long slits are off-centered from the center of the chip (called center of rotation above). You can either use the  pre-determined coordinates of this center, which are X = 1869 and Y = 2020 on the flipped image,  or determine them from scrath by  taking images at different Cass ring positions and determining the center. You should also determine the position of the slit and the point on the slit where you want your spectrum.

In the afternoon, take an image of the slit without the disperser in (through any of the filters). Flip the image, display it and choose the slit position where the object will be placed.

At night, set the pointing and focus the telescope as explained in section 7. Then ask the telescope operator to point to your target.

Since you will probably be using parallactic angle, alignment of an object is not so trivial. Here is a short cook-book type recipe:

 Use the IRAF routine called pangle. Enter the requested values for hour angle (estimate the hour angle you will be at when starting the exposure), declination and expected exposure time. The output will be the parallactic angle, the Cass position angle and the "pa" on the TV guider. The operator will slew the telescope and set the Cass angle and other values.

Since your object is going to be close to the center of rotation and you will be at an arbitrary Cass angle determined by pangle, use the task 'toslit'  in order to get the object close to the center of the slit. You will need to enter the coordinates of the center of rotation, those of the center of the slit and the Cass angle. Give the output generated by toslit to the telescope operator and ask him to offset the telescope by those values.

 After offsetting by the values predicted by toslit, take a short image of the field.
It is recommended to take also a short image of the slit, for every different object observed at night, and measure there the X and Y slit position where the object should be placed.  Those may show slight variations from different telescope positions which may result in a poor object centering.


Next flip the object image, and  use the task cobject to get the object into the slit. cobject is interactive, works in every way like align and is pretty self explanatory. Iterate if needed. The parameter set for cobject will look like this:


                    cobject parameters


As the final alignment (usually after the second run of cobject) is usually done with the TV guider box, make sure to give cobject the current values of it (XY5 on the guider screen).

Finish the alignment procedure by taking an undispersed image through the slit and check that your object is located where you want it to be. You will have to use ad-hoc z1 and z2 parameters for the display command.




Table 1. Relative Efficiencies of Grisms


red low-res / blue low-res

blue hi-res / blue low-res

4000 .



4500 .



5000 .



5500 .



6000 .



6500 .



7000 .



7500 .



8000 .



8500 .



9000 .



Appendix B: Observing Hints




  • Try to avoid different raster or sub-raster schemes during one night.

  • Watch out for wrong GAIN setting. Notice that the default gain is 1. If you are well aware of what's going on, you can risk doing alignments with Gain 3 and therefore cutting down your cutting overhead. DONT FORGET TO CHANGE BACK TO GAIN 1 WHEN EXPOSING A SPECTRUM. Taking a long exposure and reading out with the wrong gain can be infuriating.

  • It has happened to many of us: After a superb alignment, don't forget to put in the grism if you want to see spectra...

  • When aligning, keep in mind that the positions of the masks may vary slightly with time and zenith angle. Never use slit mask images taken in the afternoon. Take you slit mask image and field images in consecutive order.

  • Check the alignement periodically if you are taking several spectra of the same field.

  • Comparison spectra (and dome flats) are taken in situ. Ask the operator to close the dome when the CCD starts reading out the last exposure. Closin the dome takes about the same time as reading out (although the light that indicates that the dome is closed takes significantly longer to turn green). You can start the arc exposure as soon as the CCD has read out.



Document Actions