Документ взят из кэша поисковой машины. Адрес оригинального документа : http://www.apo.nmsu.edu/35m_operations/35m_manual/Instruments/Echelle/PDF/echelle_quickstart.pdf Дата изменения: Mon May 23 02:58:34 2005 Дата индексирования: Sun Apr 10 09:30:12 2016 Кодировка: Поисковые слова: star trail

A Quick Start Guide To The ARC Echelle Sp ectrograph
John C. Barentine Apache Point Observatory

1.

Introduction

The ARC Echelle Spectrograph (ARCES) is a medium-high resolution, cross-dispersed visible light spectrograph that operates in the wavelength range 3500-9800° in 107 stacked A orders. It has a resolving power of R=37,500 corresponding to a velo city resolution of 8 km s-1 . The sensitivity of the instrument is 1 count s-1 pixel-1 °-1 for a star of mV = 16. (note: A this needs to be verified for current detector). With the recent (late 2002) upgrade of the slit viewing/autoguiding camera detector, users can consistently autoguide on point sources down to mV = 15, contingent on seeing. Depending on science requirements, a selection of entrance apertures ("slits") is available. These are as follows: 1.6 x4.7 , 1.6 x3.2 , 1.6 x1.6 and 0.9 x0.9 , where the units refer to the size of the slit in pro jection on the sky.

1.1.

Instrument Setup Requests

When contacting APO in advance of your observing run, make sure to include whether you would like any slits other than the "default" (1.5 x1 ) installed in the instrument. Also specify whether you intent to log in early to take cals. Under certain circumstances, it is possible to take cals with the Echelle during the daytime or while other observers are using the telescope and other instruments. Check with your Observing Specialist as to when you may begin using the Echelle for taking cals.

2.

Operating The Instrument via Remark

The Echelle is operated in a manner very similar to the other Remark-controlled instruments DIS and GRIM; users experienced with these instruments will find much of the following discussion familiar. Because each instrument has its own operational peculiarities, new Echelle users should read and review all the instructions in this do cument. In the Instruments menu within Remark, cho ose Echelle. A window will pop up headed "Camera - echelle" (see Figure 1); if you have used other instruments under Remark, you will note that this window is identical in content to the "Camera" window of the other instruments. The Camera window gives information about the current settings of the Echelle,


­2­

Fig. 1.-- The top-level ("Camera") Echelle control window. including the filename convention, sequence number, FITS comment card contents and elapsed time of the current exposure. This window is also where the exposure time is set for your spectra. Buttons for starting, pausing, stopping (abort with readout) and abort (with no readout) are included along with a pull-down tab with presets for biases, darks and an "Expose" setting. The latter is the default for general-purpose images of any given exposure time. To start an exposure, click the button to the right of center in the figure labelled "Expo 1". Note, however, that depending on the preset selected and number of images to be taken specified by the user, this button may contain other text such as, e.g., "Bias 5". The user-operated features of the Echelle are the calibration mirror, calibration lamps, and a filter wheel. Control of these features is done in the Echelle Configuration window (Figure 2); to open this window, click the "config" button in the Camera window.

Fig. 2.-- The Echelle Configuration window.

2.1.

Setting Up Exposures

Before starting to take exposures, be sure to select and set a destination directory for your data on tycho. The default directory is /export/images but, for convenience, you may


­3­ make a subdirectory off this path and store your data there. If your subdirectory do esn't already exist, use the UNIX mkdir command to create the directory AND the chmod 777 * command to make the directory writeable by users other than visitor1. Go o d choices for subdirectory names are, among others, your program abbreviation, your initials/name, or the date. Example: mkdir /export/images/jan01 chmod 777 /export/images/jan01 Once your directory has been created and the permissions changed, it must be set as the "image directory" in order for your data to appear there. This is done via an MC no de, which can be accessed from the visitor1 account on tycho by typing mcnode at the prompt. After starting the no de, the command imdir (path) will change the destination directory. Issuing imdir without an argument in the MC will return the current setting of the destination directory. Also, be sure if you are the scheduled first-half observer on a given night to reset imdir /export/images before you log out as a courtesy to the observer who follows you. Setting the user-defined image characteristics of filename convention, sequence number, and FITS comment is done in the Configure Echelle Exposures window (Figure 3). If

Fig. 3.-- The exposure configuration window. taking certain cals such as biases, where a sequence of images is desired, change the value in the "Number of Exposures" box at the top of the window. Enter filename and sequence information and an optional FITS comment; leave the binning pull-down tab set to 1x1 and


­4­ do not click any of the buttons in the "Reduction Choices" box. The "File Type" pull-down tab should be left set to "FITS".

2.2.

Slitviewer And Autoguider

The Echelle features an autoguiding camera to automatically maintain the position of your ob jects on the slit. Control of the slitviewing camera and autoguider is accessed through the "guide" button in the Camera window (Figure 4). Basic controls in this window

Fig. 4.-- The exposure configuration window. include starting/stopping slitviewer exposures, subframing, automatically taking a series of exposures and trailing and autogiding features. The camera is equipped with its own filter wheel containing, among others, a series of neutral density (ND) filters useful for observing bright stars without saturating the detector. In the top-center of this window is a representation of the slitviewer frame, with the small rectangular slit in the center. The filter wheel controls are lo cated just below this representation. A pull-down tab allows the user to change which filter is in the slitviewer beam, a status line shows the current filter wheel position, and buttons command or cancel filter wheel moves. Be sure to click Move after requesting a new filter wheel position. To take a slitviewer image, set the exposure time in the Seconds textbox and click Expose. A new window will open displaying the image; for details on how to change the image scaling, see Section () below.

2.2.1. Centering Targets On The Slit With "Shift-Click" Aligning targets with the slitviewer has been greatly simplified with the Shift-Click algorithm. With one mouse click, the offset between target and slit is calculated and issued


­5­ to the telescope without any measuring required on the part of the observer. To use this feature, select the slitviewer image display by clicking once in the window; your cursor will turn into a small "bullseye" target, indicating that Shift-Click is active. Place the bullseye over your ob ject, hold down the Shift key and click the mouse once. Wait for the beep from Remark signifying that the commanded offset has been executed, then take another image. Your ob ject should now be reasonably well centered on the slit. Occasionally, a second Shift-Click iteration is needed to perfect centering.

2.2.2. Adjusting Display Parameters In Remark If a target is particularly bright or faint, the client built into Remark for displaying slitviewer images may inappropriate autoscale the resulting image. First check that you have not set the exposure time to o long or short; also check the state of the slitviewer filter wheel (Figure 5) if your ob ject appears to be `missing' ­ you may have inadvertantly left a neutral density filter in the light path. Failing these explanations, the slitviewer image scaling can be changed to better match the dynamic range in the image. This is done with an interactive to ol for changing the displayed limits of the image pixel histogram (Figure 5), accessed by pressing (Open Apple)+H. The figure shows a section of the histogram for

Fig. 5.-- The "histogram" to ol. an actual slitviewer image: the blue line represents the data, and the display limits are designated with the vertical dotted lines. Use your cursor to grab and drag these lines to change the display limits; even in the absence of an ob ject in the slitviewer field of view, by properly adjusting the scaling one should see the slit silhouetted against the sky. The displayed range of the image can also be coarsely adjusted with the Least Significant Bit (LSB) pull-down tab. Large LSB values correspond to larger display ranges and vice versa. The other pull-down tabs and buttons in this window are not particularly useful, although


­6­ the "log/lin", used to change the scaling function between logarithmic and linear, can help in situations such as stars embedded in nebulosity.

2.2.3. Subframing, Repeating Exposures and Autoguiding Once you have lo cated your target with the slitviewer camera, adjusted the display properties for monitoring purposes and centered the ob ject on the slit, you're ready to set up and start the autoguider. There are three steps to this process: selecting and positioning a subframe, setting the exposures to repeat, and beginning guiding. Subframes of the full Echelle slitviewer field read out much more quickly than full frames, allowing for faster guiding corrections. The subframe checkbox, to the right of the slitviewer frame representation, turns software's subframe capability on or off; clicking the box will make a small dashed square appear on the frame representation roughly centered on the slit. The size and position of this selected area can be changed by the user by clicking and dragging the corners and interior of the box, respectively. Typically, the box size will be a small fraction of the total frame area and centered on the slit. Next, click on the AutoExp checkbox and hit "Expose". The Expose button will then change to "Stop" and the software will automatically take successive images until that button is pressed again. Autoguiding requires this continued sequence of frames and will end when "Stop" is pressed; to begin guiding, click the AutoG checkbox in the lower right corner of the Echelle Slit View window. Confirmation that guiding is in progress comes in the way of the beep Remark makes every several seconds when an offset command is issued. The Echelle autoguiding algorithm takes a running average of suggested offsets before sending an actual offset command to the telescope. This approach has the benefit of higher accuracy in variable seeing conditions as it is able to discriminate actual tracking drift from smearing due to rapid seeing variations. By default, three successive exposures are averaged before an offset is sent to the telescope; depending on seeing, this number can be changed in "Guiding Frame Average" textbox in the slitviewer/autoguider window.

2.2.4. Trailing Exposures Under ordinary seeing conditions, a point source barely fits within the 1 -wide slit at half maximum. Better seeing will critically undersample the ob ject on the slit, allowing more sky photons in. For sufficiently bright sources (usually mV < 10), the sampling of the star on the slit can be improved by trailing the star across the slit during the exposure. When trailing is turned on, by clicking the Trail checkbox in the Echelle Slit Viewer window, alternating guiding offsets are issued to deliberately "smear" the centroid of the ob ject around the


­7­ software boundary box of the slit. The illumination pattern of the resulting Echelle orders is much more uniform than would otherwise be the case, ultimately preserving more of the signal which seeing variations would otherwise deplete. If you cho ose to trail your Echelle exposures, the motion of the ob ject relative to the slit will be evident in your slitviewer images and should not be cause for alarm.

3.

Calibrations

The Echelle has a mo dern, SITe TK2048E back illuminated 2048x2048 CCD with 24 µm pixels. The bias level is stable over a perio d of days to weeks, and dark frames are not strictly required with this chip due to the very low noise (Idark = 0.002 e- sec-1 pix-1 ). Some users taking exceptionally long exposures, of order one hour, may wish to take a few darks in order to sample cosmic ray events statistically. While the instrument is highly stable from night-to-night, a full set of calibration frames may be required nightly depending on demands of your science such as extreme velo city sensitivity. The cals most sensitive to night-to-night variations are flat fields and lamp comparison spectra.

3.1.

Filters And Lamps

The Echelle slit head is enclosed in a light-tight box also containing the calibration lamps, a filter wheel and flat pickoff mirror for changing between "Source" and "Calibration" mo de in the Echelle Configuration window. The light sources are the already mentioned quartz lamp and a Th-Ar arc for wavelength comparisons; a number of go o d Th-Ar atlases exist for identifying lines in the wavecals. Flat fields taken with the Echelle require a little more effort than with DIS. This is a result of the response of the detector versus wavelength and the inherently red color of the quartz lamp used to illuminate the spectrograph for flats. In order to reach sufficient S/N in the bluest orders without saturating the bright red orders, it is necessary to take two sets of flats each of which reach sufficient S/N for blue and red orders without saturating. To achieve sufficient signal in the blue, a CuSO4 filter is placed in the beam via a pull-down tab in the Echelle Configuration window. Exposures on the "red flats" are kept short to avoid ghosting of the bright red orders showing up in the blue orders. Similarly, Th-Ar lamp spectra must be kept necessarily short in order to avoid ghosting of the bright near-infrared lines. Sample exposure times are given below.


­8­ 3.2. Suggested Exposure Times

The following figure represents the S/N per second achieved in ARCES spectra as a function of the mV of point sources. This curve was determined empirically from 191 ob-

Fig. 6.-- S/N per second of integration as a function of mV for point sources observed with ARCES. servations of stars with the spectrograph with a variety of slits and observing conditions. Signal-to-noise for ARCES is determined from the following formula: N S/N = = N = 8(AD U ) (1 ) N where N is the number of electrons per pixel, equal to 8 times the number of ADU. The factor of 8 is derived from the following: 2 pixels per Nyquist-sampled resolution element в 4 e- ADU-1 detector gain.

3.3.

How Many Of Which Kind?

The following table summarizes the types of calibration data taken with the Echelle including advice on the number and exposure times for each. The recommendations are


­9­ based on practical experience with the instrument. Type N Exp Bias 3- 5 Th-Ar Arcs 3-5 "Red Flats" 3 "Blue Flats" 3 osure(s) 0 30 7 240

"Red flats" are defined as those taken without filtration, whereas the "Blue flats" are taken with the CuSO4 "blue" filter. You may, of course, wish to adjust the times as needed. Thirty seconds is suggested for arcs, but because a number of bright lines in the near-IR saturate in only a few seconds, users may wish to take arcs with a range of exposure times in order to avoid saturating those lines while recovering enough flux in the blue orders for adequate global dispersion solutions.

4.

After Your Run Ends

At the end of each night, don't forget to turn off any calibration lamps you may have used!

4.1.

Data Storage And Retrieval

Your data will remain on tycho.apo.nmsu.edu:/export/images/your subdirectory for 7 days before being automatically deleted; observers with special circumstances requiring more than this length of time for data storage should contact Craig Lo omis (clo omis@apo.nmsu.edu) at APO to make arrangements.

4.2.

Data Reduction

Reducing Echelle data is a tricky pro cess as the format of the images is very different from conventional slit spectrographs. Prior experience reducing 1-D spectral data is very helpful, with the understanding that working with Echelle data is much like reducing many 1-D spectra simultaneously. Fortunately, Julie Thorburn (UChicago) has written a go o d do cument for getting started with ARC Echelle data reduction, available as a PDF file at: http://www.apo.nmsu.edu/Instruments/echelle/ARCES data guide.pdf.


­ 10 ­ This information was used to generate an IRAF proto col for reducing Echelle data, available from the Echelle instrument page linked off the APO home page (http://www.apo.nmsu.edu). For any further information regarding the Echelle, please contact Russet McMillan (mcmillan@apo.nmsu.edu).