Start observing: 0. Turn on "MIRC Leach Power Supply" from "MIRC APC" tab on the Iceweasel web browser. If the MIRC APC tab prompts for user name and password, simply use the default values already filled in and hit "OK". Select the box next to "MIRC Leach Power Supply" and choose "On Immediate" from the pull-down action menu. Then hit the "Next >>" button, followed by "Apply" on the next screen. 1. open a terminal on wolverine, type observer, which allows you to log into mirkwood as oberve, then type astropci 2. open another terminal on wolverine, type spooler, which allows you to log into mirkwood as spooler, then type spooler again. 3. run gui_startup6T in another terminal, it will bringup 5 guis, the fiberexplorer gui is a little slow, be patient. 4(optional). type "shutter" in another terminal to bring up CHARA shutter gui, you need to enter the standard CHARA password. 5. check the beam order and reftel in pythonConfig gui, make sure they are right. Then click "send values to RT", and then click "get values from RT" to make sure the values are the same as you set. 6. Then check the little box to the right of "Delayline XX: beam" in gdtControl2 gui and make sure the correct beams are listed for each telescope. If it is not, close the gui, and bring it up again under ~/MIRC/mirc_python/IRCam6Tengineering2/gdt2 Then don't forget to select the right reference telescope by clicking the right circle under the same gui. 7. In IRCAM GUI, a. Press 'Update DAQ Parametes'. Wait until it is done. b. Press 'Config Camera'. Wait until it is done. c. Click 'Start exposure' d. Set the left-side display to 'cross section' mode, with 'Row # 3'. Click 'START'. e. Set the right-side display to 'Power spectrum' and click 'START' f. Now the two displays should be running. g. Close all shutters in the SHUTTERS GUI. Back in IRCAM, click 'Reset BG', then click 'Take BG' for a couple of times, e.g., 8. Note: wait for the camera to finish one cycle to click the next 'Take BG' The number of BG can be found on the astropci display (look carefully!) 8. On gdtControl2, click 'START LOOP'. 9. Fiber Explorer: a. In IRCam, set the 'Coherent Coadds' to 60 (or 100 if seeing is bad or star is faint). Click 'Update DAQ Parameters' b. In Fiber explorer GUI, click 'Detector Control' tab. In this tab, click 'Load Detecter Config' Update the pixel beam number locations. Click 'set xchan position'. c. If the target is bright, meaning photometric channels can see the light, then check all the stages, select the xy range you want to search, and click 'Perform Map(Xchan)'. It will automatically perform mapping for all fibers together. Remember to open all beams in SHUTTERS gui. Once the mapping is done, you can perform certain actions to certain stages by checking the box in front of them. For example, if you want add another image to beam 1 and beam 3, check beam1 and beam3, uncheck beam2 and beam4, click 'add to last'. If you want beam2 move to max pixel, check beam2 only and click 'Move to max'. If you want beam4 move to Gauss peak, check beam4 only and click 'Move to Gauss peak'. d. If the target is not bright enough for photometric channels to see, then check ONE stage at a time, set the xy search range and click 'Perform Map'. This will use MIRC quadrant to do search for the best position for each fiber. Remember to open the corresponding beam in SHUTERS gui. If there is a reasonable gauss fit, click 'Move Gauss peak'. If not and there is a bright spot which indicates where the light is, click 'Move to last max'. Do Perform Map and Add to Last Map, and Move to Gauss peak a couple of times if necessary. Don't obsessed to much on this. Check another Luminos Stage, and do the same thing above 10. Change coherent coadds to 14 for r0~10cm 16 for r0~15cm, or 12 for r0~5cm Change PS Coadds accordingly . 11. check the power spectrum on IRCam display, make sure there are no fringes on it (If there are some, one way to get rid of them is stop carts tracking). Then you click "CALCALL", it will take one minutes to calculate the threshold for each baseline, then you click "UPDATE THRESH" button, you will see the numbers in threshold are updated. 12. In gdtControl2: a. Type in the offsets and move all baselines that are being used. b. Make sure: We have light on all beams; We have delaylines in range for all carts, and metrology is good; GPS time and CHARA time are good; Offsets are right. c. Click LOOP for all baselines (except W2, the reference) to search for fringes. d. Change fringe detection shreshold if necessary e. Make sure the "lock fringes" button under gdtcontrol2 gui is on 13. When fringes are found, the corresponding button in the lower triangle of the matrix will turn yellow. You can then click the button in order to lock the fringe. Locked fringes are indicated with green color. If the fringe is lost while locked, then the button will turn red. Once a fringe is detected, the software will note the OPD difference between the involved delay lines in the upper triangle of the matrix. The software monitors the values in the upper triangle continously and uses them to make predictions for the delay line position of missing fringes. In order to make a prediction, the algorithm requires the OPD values from at at least one fringe involving the reference delay line and at least one cross-fringe. The predicted values appear in the row of buttons on the left side of the matrix. Clicking on the button will move the delay line to the corresponding position. In case of a false fringe detection, it is recommended to delete the wrong OPD value from the matrix, as spurious values will lead to wrong fringe position predictions. A spurious value can be detected from the upper triangle by simply clicking the button. 14. Once all fringes are locked, click DATA on ShutterMatrix GUI to take data for a given number of files. 15. Once DATA is finished and the green 'DATA' light is off, click SHUTTER to take shutter sequence data. When switching to the next target, you should click the button "Clear Matrix" in order to clear the fringe detection OPD values in the matrix and to open the fringe locking states for all baselines. -------------------------------------------------------------------------------- Trouble shooting 1. sometimes the spooler will crash, and the server gives "wrong length message". What you need to do is set the save number to 0 (which is usually true) in IRcam gui, and click "Update DAQ Parameters" button. This should stop taking data. Then you restart spooler by just typing "spooler" and hit enter, and try to save data again. If it is still not working, then repeat from the beginning. IT should work within 2 tries. 2. If the gdtcontrol2 crashes, just open it from terminal, and hit "GET OFFSETS" button, it will bring back the previous offsets. And then hit "Update threshold", it will bring back the previous threshold. 3. Check the xchan positions (Photometric channels position) under "Detector Control" tab on the fiber explorer gui and make sure they are right. If not, you have to manually input the numbers and click "set xchan pos" button. The way to find the xchan position is to either use white light source in the lab during day time or use sky target during night, and do fiber explorer with the old method ("Perform Map" button under Fiber Mapping tab) with only one CHARA shutter open to find best fiber position for that beam. Then you should be able to see photometric channel light in IRcam display, zoom in and find out the column number of the light, and update the xchan position of the corresponding beam under the "Detector Control tab". Write down how many columns it has moved, and it should be the same number for the rest of the beams. 4. FLAT FIELDING for low vis objects, you must use flatfield. setup: 1. get light from all 4 telescopes after fiber exploring. 2. Set coherent coadds high (like 500) 3. enter 10000 for save number. Hit 'update daq parameters' [this will take a set of data for flat fielding] 4. wait 10 seconds or so. 5. repeat #3-4 a few times. 6. you should see the powspec image look noisy and the fringe weights in the gdt control should be smaller than before. 7. reset save number to 0 you can reset the flat field by entering 10001 in save number (followed by 'update daq parameters') you can toggle between two flat field methods using 10002 in save number. -------------------------------------------------------------------------------- Running the mirc_coadd script from wolvervine 1. There should be a directory named /home/observe/mirkwood. If that directory is empty then create a mount point for mirkwood through this directory: cd /home/observe sshfs observe@192.168.3.131:/data/ mirkwood 2. Change directories into mirkwood - you should now have access to the mirkwood data disk: cd /home/observe/mirkwood/2011Sep 3. In another terminal on wolverine, move to the dayafter Reduction directory: cd /home/observe/Reduction/Users/dayafter 4. Start IDL from the dayafter reduction directory: cd /home/observe/Reduction/Users/dayafter idl ../../idl_startup_mirc6.pro 5. Start mirc process 1: IDL> .r mirc_process1.script Select proper directories from a gui prompt: DATA DIRECTORY: /home/observe/mirkwood/2011Sep/2011Sep02 COADD DIRECTORY: /home/observe/DATA_TEMP/MIRC_COADD/ 6. Create the *.mirclog file: cd /home/observe/Reduction/Users/dayafter idl ../../idl_startup_mirc6.pro IDL> mirclog, date='2012Sep03', coadd_path='/home/observe/mirkwood/2012Sep/', /verbose 7. If the above directory selections are followed: The *.mirclog file will be saved in /home/observe/Reduction/Users/dayafter. The coadded files will be saved in /home/observe/DATA_TEMP/MIRC_COADD. Note: svn directories accessed by idl_startup_mirc6.pro are in: /home/observe/Reduction/MIRC/ /home/observe/Reduction/Libraries/ /home/observe/Reduction/Users/ -------------------------------------------------------------------------------- Mounted data disks on mirkwood - as of 2012 Dec 8: /dev/sdb1 367G 295G 55G 85% /data /dev/sdc1 1.8T 866G 876G 50% /staging1 /dev/sdd1 1.8T 102G 1.7T 6% /staging2 --------------------------------------------------------------------------------