From beaulieu@ast.cam.ac.uk Fri Nov 3 21:05:09 2000 Date: Wed, 30 Aug 2000 16:22:17 +0100 (BST) From: beaulieu@ast.cam.ac.uk To: gil@ast.cam.ac.uk, jacco@ast.cam.ac.uk, rgm@ast.cam.ac.uk Subject: my CIRSI/WFIRC notes CIRSI COOKBOOK by Sylvie and Jacco Las Campanas Observatory for the 12-17 Aug 2000 run +++++++++++++++++++++++++++++++++++++++++++++++++++++++++++++++++++ People working on CIRSI and helping with problems at some point during the run: Mark Phillips (mmp@lco.cl) Eric Persson (epersson@ociw.edu) Christoph Birk (birk@ociw.edu) David Murphy Allan ? Frank ? Oscar : technician Herman : tel. operator Fernando : tel. operator +++++++++++++++++++++++++++++++++++++++++++++++++++++++++++++++++++ WARNING...this cookbook is a gathering of notes as of Aug 2000 by us. Eric, David, Mark and Christoph are currently writing the official cookbook. For this run, we only have access to J and H, i.e. that the cold box with the K-band is not part of the optics due to some misalignment problems most probably (is being ship back to Pasadena). =================================================================== All the passwords noted in Richard's CIRSI guide are now different. It is irrelevant to note them here because [1] they are now changed regularly and [2] it is safer not to write them down in any documents in order to prevent password from being obtained by hackers. Every workstation have their username and password written on the console. =================================================================== PRINTERS: lp file.ps lps file.txt enscript -2rG file.txt =================================================================== RELEVANT WEBSITES Richard's website: http://www.ast.cam.ac.uk/~rgm/cirsi/ociw/ioatime/ La Silla MeteoMonitor http://www.ls.eso.org/lasilla/dimm/meteo/meteomonitor.html#tab La Silla website: http://www.ls.eso.org/index.html =================================================================== Prepare finding charts of fields to observe, if not already done. =================================================================== LOG ON TO THE WFIRC COMPUTER AND TO THE DATA REDUCTION WORKSTATION: the two workstation on the righthand side of control room username and password writen on both workstation consoles =================================================================== On any other "public" workstation: Make a catalogue of the centre of fields that will be used by the WFIRC software: (PS: IF YOU WANT TO SAVE TIME, LET THE WFIRC SOFTWARE DO THE MOSAIC OFFSET INSTEAD OF PULLING THE NEW POSITION OFFSET FROM THE CATFILE i.e. ALL YOU NEED THEN IN YOUR CATFILE IS THE CENTRE OF YOUR FIELDS) New perl script written by Richard (cirsi_grid3.pl) modified by Jacco to fit Christoph's ascii format for reading the grid catalogue: ID | RA | DEC | MAG | COMMENTS with RA and DEC like hh:mm:ss.ss and dd:mm:ss.s . This script takes the fields centre and calculates the offset needed to build up the complete mosaic image. chmod +x cirsi_grid3.pl cirsi_grid3.pl --> show the default syntax i.e. cirsi_grid3.pl DUPONT cirsi-test.cat cirsi-test.grid where cirsi-test.cat is an ascii file with the centre of fields to be observed, e.g.: 12:00:00.00 +00:00:00.0 cirsi_field_1_ 12:00:00.00 +10:00:00.0 cirsi_field_2_ etc... and cirsi-test.grid is the resulting grid using Christoph's column format where we have set, in the script, the MAG as 0 and there is no COMMENTS: cirsi_field_1_D | 12:00:06.39 | +00:01:35.8 | 2000 | 0 | cirsi_field_1_A | 11:59:53.61 | +00:01:35.8 | 2000 | 0 | cirsi_field_1_C | 12:00:06.39 | -00:01:35.8 | 2000 | 0 | cirsi_field_1_B | 11:59:53.61 | -00:01:35.8 | 2000 | 0 | cirsi_field_1_X | 12:00:00.00 | +00:00:00.0 | 2000 | 0 | cirsi_field_2_D | 12:00:06.49 | +10:01:35.8 | 2000 | 0 | cirsi_field_2_A | 11:59:53.51 | +10:01:35.8 | 2000 | 0 | cirsi_field_2_C | 12:00:06.49 | +09:58:24.2 | 2000 | 0 | cirsi_field_2_B | 11:59:53.51 | +09:58:24.2 | 2000 | 0 | cirsi_field_2_X | 12:00:00.00 | +10:00:00.0 | 2000 | 0 | etc... BUT, since this output file is in ascii, we can later on add COMMENTS if we need to. Richard's script also create a list of files called: cirsi_field_1_.gaia, cirsi_field_2_.gaia etc... which are meant as finding charts informations. ====================================================================== How to transfer the catalogue, cirsi-test.grid to the WFIRC Computer: The catalogue is transfered to the WFIRC Computer using the window program ftp at the WFIRC Computer - we cannot transfer the catalogue from the Unix environment. click on the "START" icon, bottom left box of Windows on bottom bar click to "run"--> this open a window type "ftp" then click "ok" ftp>lcd wfirc ftp>lcd objects -->these two lines put you in the dir where you will put your catalogue file ftp>open canopus --> or the computer where your file is ask username and password where your file is on canopus ftp>cd mydir --> your directory where the file is ftp>get cirsi-test.grid cirsi-test.objects (extension to be .objects) ftp>quit Then, you should be able to see your file renamed cirsi-test.objects in the wfirc/objects area. Or maybe name it .objects right from the beginning when you create it. ====================================================================== CREATE AN ENVIRONMENT TO SEND THE DATA TO BE SAVED ON DISK on the data reduction workstation, create a directory on /data1/wfirc/ where your data will be sent after being read. It is a good habit to create subdir within this area for each night: /data1/wfirc/august14, /data1/wfirc/august15 etc... --> remember this path because you need to setup the WFIRC software to indicate where to send the data (see below "RUNNING AN OBSERVATION WITH WFIRC" for more details). ====================================================================== The observations are to be conducted as: We have only J and H available. In order to minimize the dust movements in the detector, we will do one filter per night. Galactic bulge fields: a mosaic has 4 positions - each position will be done with a rot-9 dithering of 11", with 5 exposures each (cosmics) at 20sec exptime per exposure. So, this come up with: 9dith x 5exp x 20sec x 4 pos= 60min total exptime with an additional overheads time (moving + reading ccds). See at end of document for statistics on efficiency. STD fields: USE THE MACRO DICE-5 (NOT ROT-5/9 FOR STD) done at beginning of night and a few times during the night done with a larger dice-5 dithering offset of 25", only one exposure of 5sec on chip 1 5dith x 1exp x 5sec=25sec plus overheads XGal. fields: same setup as the Gal. bulge's fields ====================================================================== CHANGING FILTERS AND GOING TO DARK POSITION: Turn the filter knob clockwise in order to move the filter wheel. Look at the lights of the grey filter box. The initial position (1) is the DARK position and both red and green should be ON. Then, from that position, start counting each time the green light turns ON, which means that you are at a filter position. Then, just reach your desired filter position. The filter positions are written on a box next to CIRSI. IMPORTANT: you need three knob turns for each filter position mmoves AND, for the first move (i.e. from the DARK position to position 2) there is a "fake" green bleep, to be ignored, the next green bleep is the correct position 2. This only happends for the move from position 1 to position 2. The positions are: (1) Dk Dark (2) Jc Carnegie --> we use in this run (3) Jg J (Gemini consortium spec) --> we make test dome flats (4) Hs short H (5) Hx H (Gemini consortium spec) --> we use in this run (6) Hl long H (7) Jn narrow OH (8) Ks short K If the filter position is not right i.e. in the socket, we will see it when taking an image. There will be angled shadows. ====================================================================== RUNNING AN OBSERVATION WITH WFIRC: Get your logsheet ready and remember to accuratelly mark each image taken with its mosaic position, i.e. from the centre position, offsetting to N, then W, then S etc... Add information like RN and Gain, bias and dark level, detector's scale, mosaic offset value for both object fields and STD, and some information about weather from time to time. For the weather info, use La Silla's weather monitor for a close approximation of conditions at Las Campanas. There is also a small white meteo box outside on the catwalk, situated righthand side as you get out. There is two humidity indicators, just take the mean of the two values. Useful if La Silla's MeteoMonitor is offline, as it happends from time to time. Click on Camera icon to open it up: This will first open a window which tell the online/offline status of... -in Camera-Control: =select the DATAPATH within "options", this is the path for your data from telescope to disk. MAKE SURE THAT IT IS UPDATED EVERYNIGHT TO THE SUBDIR FOR THAT NIGHT =also in "options", select the "macro-progress" window =in "modules", select the Guider, TeleGui and Display. this will open there window. =change the display window size to 512x512 and drag the display to the right hand side monitor. also set the display to "auto" to automatically load the images as they come in and also, set chip-1 as the prefered chip to load first, this avoids the display trying to load a chip that has not been read yet from the detector (chip 1 is the first one to be downloaded). =then, fill in the white boxes according to your chosen setup. readmode should be "double" and chip should be "all". click on "mos", "std" or "foc" to select a macro and each need to be setup as well. For the STD macro, use the dice-5 macro and NOT the rot-5/9. -in Telescope: =get the skymap, airmass and objectlists from within "files". this will open three new windows. =setup "equinox" and let the telescope operator know what you will be using. =setup the values of dx and dy =from "object list", load our .objects file from the file menu, click on an entry to select an object, then click the "sel pos" and then the "move to" switchs. this will update the RA and DEC in the window and also send the coords to the tel. operator. it also show the skymap. PS: THE NO. OF CHARACTERS FOR THE OBJECT NAMES IS 12, SO WE ARE MISSING THE LAST BIT OF INFO OF OUR OBJECT NAME. WE ASKED CHRISTOPH TO INCREASE FROM 12 TO 20 FOR HIS NEXT VERSION. =choose a prefered catalogue for the skymap (the GSC is the default and probably the best one to use) and also adjust the zooming of the skymap so that you can see the 4-chips pattern. The guider's window is loaded from the menu of skymap, "parameters" and "show guider" (IT IS AT THE MOMENT UNCLEAR WHAT IS THE ORIENTATION OF THE GUIDER'S WINDOW. THIS SHOULD BE CHECK PROPERLY WHEN SETTING UP THE GUIDER) Do a telescope pointing calibration using a bright star from the Almanac. (A new pointing calibration check may be needed in the middle of the night.) Reset the Guider to zero, this is the blue box with a black screen on the left hand side of the control room. The resetting is done using the two switch under "OFFSET GUIDER", [1] press "reset counter" and then [2] press "home". The two x,y positions on that monitor should then be zero each. Then setup the guider from the WFIRC system as follow: --> the next section on EGuider currently does not work so take the alternative method explained right after "EGuider" -EGuider: the guider needs to be setup at the begining of the run: click on a star (preferably near the centre of the guider's window using the left mouse button. this give you the star's name and its mag. Then, click on the star again but with the middle mouse button, this give you the x,y guider's coord. system. If you click again with the middle mouse button but twice, it send this information to the guider's window. you need to verify if the guider did receive the x,y position on the monitor of the tel. operator. then, if all is ok, click "zero" on the guider's window. this is the last step to setup the guider. -Guiding: alternatively, we can guide as follows: The WFIRC guider's window should be on "hold" setup the x,y to 0 in the guider's window and also on the guider monitor system at the begining of the night. After pointing to our field, choose a star on the skymap guider's box. click on the middle mouse button to read of the x and y position then go to the guider monitor system and, using the handset for the guider, set the x and y on the black monitor. then put the star into the white box of the guiding monitor and then, press F3 on the guiding keyboard. then, go back to the WFIRC guider's window and press "hold" to bring it to "run". then, change the x,y entries in that window to those of the black monitor. IF YOU FORGET TO DO THIS LAST UPDATE OF X AND Y, YOU MAY GET A POPUP WINDOW WHICH WILL WARN YOU THAT THE GUIDER IS NEAR THE LIMIT. Focus: to focus the instrument, follow these steps, there is a macro in Camera to do this: =first, make sure that the GSET on the tel. console has been done. check with night assistant. This will reset three indicators of the focus on the tel. console [green and black numbers on screen and red indicator on the middle console (SELECTED DATA). =use chip 1 and do a sequence of 8-10 frames with focus steps of 5, then use IRAF/imexamin "r" and note the FWHM. Choose the focus which converge. =Once the focus has been set, then type FSET on the tel. console (again, this will be done most probably by the tel. operator) Each time we move the telescope we have to check that the green reading of the focus is moved to the black value using the handset. This will be done by the tel. operator. A new focus check may be needed in the middle of the night The focus value will be in the range 4450 to 4500 Darks and Bias: Eric reckon that they are irrelevant but if you really want them, do these in the afternoon. The steps are: =put the filter system to DARK position (both red and green lights will be on when reaching position (1) = Dome flats: also do these in the afternoon. The steps are: =ask tel. operator to point on the dome screen WE HAVE TO TAKE TWO KIND OF DOME FLATS [1] WITH LIGHT ON AND [2] WITH LIGHT OFF --> BUT REMAINING ON THE SCREEN =[1] set the light level at 0249 on the blue rack lower monitor and do a 15-loop at 3sec exptime. this should give you a flat level at around 8000 counts in Hgem =[2] turn off light and do the same 15-loop at 3sec exptime. This gives a flat with a level of around 15 counts. Jcar = lamp level 300 with 3sec --> ~8000 counts Jgem = lamp level 250 with 3sec --> ~8000 counts Hgem = lamp level 250 with 3sec --> ~8000 counts Sky flats: matter of choices, some say they are not useful, dome flats are better. Richard reckon they should be done at beginning and end of nights. The steps are: Start an observation: =setup the Camera settings to: exptime=20 (NOTE THAT THE MIN. EXPTIME SET IN THE SOFTWARE IS 3SEC) readmode=double objtype=astro (or std or flat ...) filter=Hx loops=5 chip=all filename=001 (at the beginning of the first night and increment automatically) VERY IMPORTANT NEVER TO CHANGE THIS ENTRY FOR THE REST OF THE RUN!!! It is preferable to save unwanted images on disk and later discard them during data reduction than risk overwriting good data by mistake during the run! macro=mosaic_1_9 which is our own macro created using the "mos" setup with: step=194" but ignored w/ none-1 mos. pattern=none-1 (no mosaic) dithering steps=11" dithering pattern=rot-9 guider=auto =the guide star system should be on "hold" =from objectlist, get the object we want to observe and click on "setpos" and "move to", this send the info to the tel. operator and set the RA and Dec in the Telescope RA and Dec white area =setup the Telescope settings to: equinox=2000 --> and tell the tel. opreator that you use this equinox RA and Dec should be set to your chosen object wait for tel. operator to say "OK" i.e. telescope on position press "get" then, we set dx and dy to 97 (which is half of 194) for the first part of the mosaic press "get" again =reset the guiding system like: take the x,y value of black screen and enter them in the WFIRC guiding window (under "position"), then choose a star on skymap and get its x,y (middle mouse button) and move the guider so that the x,y on the black is now at those values. now, there should be your star plus perhaps a few others. choose a good, not too bright star to put in the white box then press F3 and press "hold" so that it change it to "run" (see above for details of guiding) =if all seems ok, click on "execute" or "go" in the Camera --> "go" if you are not using a macro --> "execute" run a given macro =at the end of the sequence, you should be able to see the 4-chips pattern with all its dithering position in green on the skymap and the images will start loading on the display window (with chip-1 as its default, if set that way). =switch guider to "hold" and press F1 on the guiding monitor keyboard to switch it OFF =then, move the 4-chips pattern 194" to N, so you need to change the dx,dy values to 194 and click on "N". You then should see the 4-chips pattern move to the upper position on the skymap for your first sequence and fill the missing northern gap (an the graphic will be in red until its been observed) =again, acquire a guiding star like above... =repeat this for W and S =then, the mosaic should be completed With our current setup for a field and exptime, it should take about 1506sec to do each sequence of the mosaic. So, we get 4x1506sec=1h40m to acquire a mosaic for one field and that includes almost all the overheads i.e. I have not counted the moving time from one position the next for the creation of the mosaic. Must not be a lot...This setup runs at 72% efficiency. ====================================================================== BACKUP OF DATA AFTER EACH NIGHT Perform a backup of the data after each night. Make 2 copies, one to stay here (and leave in the Cambridge Cabinet) and one to bring back to Cambridge. We have DLTs which can hold close to 50Gb. After a backup, run a verification of the tape to see if the data has backuped properly. Procedure for backups on DLTs: cd /data1/wfirc/ ls --> to see your subdir tar cvf /dev/rmt/0cn august14 ---> this make the backup tar tv >&! august14.log ---> this run a table-list check of the tape Run 14.August.2000 tape 303671 destroyed 303691 the only DLT tape with the 14.aug data DLT Tapes unused, left at LCO: 303693-303915-303918-303932-111992 Since the DLT drive is broken, we were only able to do a DLT backup (one copy) of 14.aug. This copy is left at LCO, in the Cambridge cabinet (cabinet no. 5, top shelf). Thus, we had to do the backups on DAT and only one copy, which is being brought back to Cambridge. **************** FTP / Backup on DAT: ftp to canopus from rigel using "obs100" username cd /export/data1/obs100/ir_data/august17 bin prompt mput irx* quit can only transfer part of the night's data, canopus disk too small so did it by chunk of 99 files tar cvf /dev/rmt/0ln * DAT tapes details: 120 G107 3E2C CIRSI 14.AUGUST.2000 00001-->00129 * 120 G117 3E3L CIRSI 14.AUGUST.2000 00130-->00210 * 120 G117 3E2P CIRSI 15.AUGUST.2000 00001-->00092 * 120 G117 3E4J CIRSI 16.AUGUST.2000 00001-->00099 * 120 G107 3F6Y CIRSI 16.AUGUST.2000 00100-->00199 * 120 G117 3E1B CIRSI 16.AUGUST.2000 00200-->00225 * 120 G117 3E5K CIRSI 17.AUGUST.2000 00001-->00071 * write error at end... with 00072 120 G117 3C4L CIRSI 17.AUGUST.2000 00072-->00086 * write error at end... with 00087 120 G117 3C5Q CIRSI 17.AUGUST.2000 00087-->00102 --> it died on file 00103 Miguel cleaned the tapedrive in the morning, it has cleared the problem 120 G117 3D4K CIRSI 17.AUGUST.2000 00103-->00197 120 G117 3C3G CIRSI 17.AUGUST.2000 00198--> should be to 00263 but must leave before it finish so will check in Cambridge ====================================================================== The system gain ie DN to electrons for this run is, according to Eric, around 2e- The detector scale is 0.2"/pix ====================================================================== DETECTORS alignment are: N --------- | 4 | 1 | E --------- | 3 | 2 | --------- and each chip is formed of 4 quadrants In IRAF the NE alignment for each chip is...? ====================================================================== EXPOSURE TIMES: We started with an exptime of 40sec but got a very high background of ~19 000 counts and lots of saturation stars. We decided to lower the exptime to 20sec but increase the number of exposures from 3 to 5. Now, the background is down at 7 000 to 10 000 counts (caused by a gradient from unresolved stars). There is still some saturated stars but much less. The efficiency has dropped from ~72% to 61% but we reckon it is still the best thing to do (see next section). ====================================================================== OBSERVATION EFFICIENCY: For a 9dith x 3exp x 40sec x 4pos setup, we get 72min total exposure. With overheads included, we endup with 1506sec for a one position. Roughtly speaking, we end up with 1506sec x 4pos = 1h40m, plus a bit of overheads involved in the move to create the mosaic --> 72% is thus calculated for the 4pos without that small extra overhead For a 9dith x 5exp x 20sec x 4pos setup, we get 1hr total exposure. With overheads included, we endup with 1473sec x 4pos = 1h38m, plus again a bit of overheads involved in the move to create the mosaic --> 61% is thus calculated for the 4pos without that small extra overhead ======================================================================