Observations of bright stars for the Las Campanas / CfA stellar library project with the FIRE spectrograph at Magellan Baade. A brief step-by-step guide Date: 2013/07/14 Author: Igor Chilingarian Introduction. ^^^^^^^^^^^^^ Our stellar library contains stars having J magnitudes between 0 and 12. We use a special approach in order to remove the effects of the atmosphere seeing from the data and also to achieve similar count levels for stars of very different apparent magnitudes. When we observe a star, it is scanned across the slit at a constant rate first in A (1st exposure), then in B (2nd exposure) dithering positions using non-siderial tracking of the telescope. The effective exposure time is defined as a ratio between the slit width and the scan rate and ranges from 1/20 sec (hardware limit) for 0th-1st mag stars to minutes for fainter targets. Because of the specific nature of our observations, we always observe with the slit oriented along the East-West direction which corresponds to the "89.8 EQU" setting in the Magellan target catalogs. Therefore, we use a modified list of telluric standard stars "FIRE_tellurics_EQU89.8.cat" which is identical to the standard one with only the rotator setting changed Before you start, please copy the attached script "firescan_ab.sh" into the home directory of your obsX account and make it executable by launching the following command in the terminal window: chmod 755 firescan_ab.sh Please write the log -- either on paper or in a text file on a computer. This is essential. Science Observations. ^^^^^^^^^^^^^^^^^^^^^ (1) We observe stars only through the 0.45-arcsec slit in the Echelle mode of FIRE, therefore the first step before taking any exposures is to switch the instrument into this mode and align the slit as described in the FIRE observing manual. At the end, in the slit mode of the FIRE Guider interface click on the *slit center* (NOT in the "A" or "B" position) - THIS IS IMPORTANT! We always use the "full well" mode of the detector corresponding to the GAIN of 3.8e- and Fowler-1 readout - THIS IS ALSO IMPORTANT. (2) Target catalogs. At present, we make the NIR follow-up of a sample of stars having high-quality optical spectra from the INDO-US stellar library. There are two target catalog files: "IR_lib_south.cat" and "IR_lib_20d.cat" containing stars in the Southern hemisphere and those between 0 and +20 deg of declination correspondingly. We will update the files in future marking the stars which have been observed. (3) Target selection. It is recommended to take one of the target files mentioned above and start from any object with the R.A. value around the current siderial time (+/- an hour or so) and then continue along the list avoiding the large telescope slews. The lists are R.A. ordered and the density of targets correspond more or less to the rate of the data acquisition, therefore the telescope will point to the similar position on the sky during the entire observing session. At this stage I would recommend to avoid any stars fainter than J=8 mag (values are given in the "comment" field in the catalog files). (4) Observational procedure. These steps should be repeated for every scientific target being observed as well as for telluric standards. (a) Target acquisition: ask the TO to slew to the desired target and start guiding. Wavefront sensor corrections should be done once every 20-30 minutes, i.e. once every 4-5 stars. They DO NOT have to be done for every target because it will add another minutes to the overheads. (b) Switch the FIRE Guider Interface into the TCS mode and set the minimal possible exposure time (which is never shorter than 10 sec because of the read-out cycle of the guider. Even with the shortest exposure, the faintest stars in the sample will look like huge balls with the saturated black centers. It is recommended to turn ON the "Loop" exposure mode by pressing the corresponding button. (c) Click in the center of the saturated star on the guider image. A window should appear to confirm the TCS offsets. Confirm the offsets by clicking the button in the window. Wait for the next guider acquisition and check that the star is located in the slit center. Repeat the procedure if necessary -- what is important for us it to have a star aligned in the East-West direction (i.e. along the slit) with the precision of 0.5 arcsec or better. The North-South direction is less important because we will anyway scan across the slit. IMPORTANT: sometimes TCS offsets are computed incorrectly because of the bug in the Guider software that's why it is important to check where the star is after the alignment. (d) Prepare the FIRE Engine window (target name, observation type: science, correct file number, etc.) as described in the FIRE Observing Manual. Please put the target name corresponding to the star name in the Magellan target catalog. (e) Check with the TO that the WFS corrections have been made if they were enabled in step (a), then start the telescope control script "firescan_ab.sh" from the terminal window as follows: ./firescan_ab.sh 0.45 for example ./firescan_ab.sh 0.45 10.0 to get the effective exposure of 5 seconds on the slit (appropriate for 6-th magnitude stars under good seeing conditions). Below we provide a table with suggested exposure times for stars of different magnitudes and spectral types. See the exposure table at the end of this document. (d) The script will stop guiding/WFS and then prompt you: ------------- A position offset telescope 1.5E 1.35N Taking exposure of 10 sec across 0.45 arcsec slit in the A position (+1.5arcsec) please start the exposure of 60 sec and press any key when ready ------------- At this stage you should put the exposure time indicated in the second line (60 sec in this case) in the FIRE Engine window and start exposure there. Once the exposure has been started, IMMEDIATELY go back to the terminal window and press any key to continue (once). (e) The script will tell you: ------------- Sleeping for 30sec waiting for the FireEngine to finish reset+readout Press ENTER to continue of wait for 30sec ------------- If you switched from the FIRE Engine Window to the terminal window promptly (within 3-5 sec), DON'T DO anything, just wait for the next messages which will show up with the delay between them corresponding to the exposure time: ------------- start track at rate 0.045 ttime= 60 stop track sleeping for 30 sec before moving to the B position press Enter to continue immediately ------------- Within 30 sec from the last message, the readout should have finished in the FIRE Engine window, then it means that the 1st exposure had been successfully taken. (f) Then you will see again in the terminal window: -------------- ====================================================================== B position offset telescope 1.5W 0.0N offset telescope 1.5W 0.0N Taking exposure of 10 sec across 0.45 arcsec slit in the B position (-1.5arcsec) please start the exposure of 60 sec and press any key when ready -------------- Please repeat steps (d) and (e), i.e. start the second exposure and quickly come back to the terminal and press any key. You will see within the next seconds: -------------- Sleeping for 30sec waiting for the FireEngine to finish reset+readout Press ENTER to continue of wait for 30sec start track at rate -0.045 ttime= 60 stop track press any key to finish the sequence and move the telescope back -------------- Now press any key again. If all the sequence worked well, the star will end up exactly in the same position where it was sitting at the time when you started the sequence. Congratulations! The spectra have been obtained. Please check the frames in DS9 and make sure that the counts level in the maximum is between 6000 and 24000. The optimal values are around 15000-18000 Calibrations. ^^^^^^^^^^^^^ FIRE has known problems with the repeatability of the slit position when the slits are changed, however once it has been set, it remains pretty stable. Therefore, we DO NOT need ThAr/Flat exposures after each star. We do need them either immediately AFTER the 0.45 arcsec slit alignment procedure or immediately BEFORE you switch to another slit. If our stellar library program runs until the end of the night, it will be OK to take arcs/flats during the day time with the only condition: SLIT MUST NOT BE CHANGED. Darks can be taken only once during the run. We need the following calibrations: Flat-QH 45sec -- 10-20 frames ThAr 20sec -- 4-5 frames ThAr 120sec -- 3-4 frames Dark 60sec -- 5 frames Dark 120sec -- 5 frames Dark 180sec -- 5 frames Dark 300sec -- 5 frames If possible, one can take twilight flats -- either early in the evening, or right before the sunrise, however it is not really critical. Telluric standards. ^^^^^^^^^^^^^^^^^^^ Please follow the telluric standard selection procedure described in the FIRE Observing manual with the following small modifications: (a) By default the "find_tellurics" script sets the minimal magnitude for a star to be 7 or 8 which is a bit faint for our purposes. In order to select telluric stars between mags 3 and 6, use the "--tmin 3 --tmax 6" keywords and choose something not too far away in coordinates in order to avoid overheads with the wavefront sensor acquisitions. (b) Use the "FIRE_tellurics_EQU89.8.cat" modified catalog file and not the original one -- please tell this to the TO. (c) Follow the "Science Observations" procedure described above for the telluric star and do not forget to change the target name in the FIREEngine window. Exposure time guide for bright stars ^^^^^^^^^^^^^^^^^^^^^^^^^^^^^^^^^^^^ Use this table as a guide for the exposure time values which have to be specified for the "firescan_ab.sh" script. Please adjust the values correspondingly for worse seeing conditions and/or bad transparency. Seeing: 0.6-0.8 arcsec (guider seeing) OBAF - early-type stars (tellurics will fall into this category) GKM - late type stars J-mag OBAF GKM 0.0 0.05 0.05 0.5 0.05 0.05 1.0 0.07 0.05 1.5 0.15 0.1 2.0 0.2 0.15 2.5 0.3 0.2 3.0 0.5 0.3 3.5 0.8 0.5 4.0 1.2 0.8 4.5 2 1.3 5.0 3 2.5 5.5 6 4 6.0 15 10 6.5 25 15 7.0 30 25 7.5 45 30 8.0 60 45 8.5 90 60 9.0 120 90 9.5 180 120 10.0 300 180