Observing With MIRC ---------------------------------- JDM 2006Jun21 JDM 2006Sep02 update MZ 2007May15 update Before observing ------------------------ 0. Pump on dewar. Try to get vaccuum to <5e-4 mbar. Use UM gauges as close to dewar as feasible (don't trust CHARA gauges). You can use the nice new turbo pump (chara's). 1. Fill MIRC external vessel w/ LN2 using metal tube every day (dewar can last ~48 hours). Top off inner vessel with funnel every few days. [at beginning of run, fill internal vessel using the provided metal tube] 2. Align CHARA lab to MIRC table in normal way using retros. See MIRC alignment document for details. 3. Fibers and Pickoff delays should be at nominal locations. Turn on White light source to the black line (this marks the maximum voltage you are allowed to supply). 4. Use fiber explorer to align fiber with white light source a. make sure server is running on mirkwood. Login lothlorien as observer and start astropci b. Start IRCam, set up parameters for IRCam (e.g., for H band, Nreads=5, Coadds=100, ps adds=15 ...), config and start exposure, take background with white light on. c. Run fiber explorer (under mircdev@lothlorien or under observer account). 'Load detector config', and explore fibers. 5. run tiptilt server (take makebias if necessary). Check whether light from white light is centered on CCD startracker, if not, use pico2 to center it. 6. repeat for other two mirc beams (using visbeamsgtk ) [ at the beginning of run, it proably would be wise to also re-find the lab fringes on beams 1-2,2-3,3-4 to make sure visibilities are high and ealing mounts are in the right places.] 7. open domes early. 8. get vmes running for ople. Check that oples move ok. Beginning of night --------------------------- 1. do m10 alignment when dark enough. 2. acquire stars with tiptilt control. 3. fiber explorer on each beam. start with 36mu (steps 6mu). Once you've found the right part of psf then use 24 mu (steps 4 mu). Sometimes you have to go bigger or smaller. If good image quality then you generally only need 20 mu(steps 4 mu) to tweak up from source to source. Using the 'standard' detector config then you want 2000 counts or so to see fringes. using the 'load config' option, we want like 100-200 to easily see fringes. 4. start shut_gtk ( >source go_zoot, then >shut_gtk) 5. run astropci. Note: if it says "Failed to get delayline offsets", then probably need to restart opel, and even mircwood. 6. run spooler 7. run the IRCam. Current version: observer> mirc_python/IRCam/IRCam_Main soon to be another new version. 8. run ./delaylines (same directory, or delaylines_S1E1W1W2) and ./gdt (or gdt_S1E1W1W2) 9. run shuter_matrix and mirc_Fiberexplorer under fiber_explorer/ 10. run Parameter_setup (something like this, need to confirm the name of the GUI), set up parameters for the array config, threshold(e.g., 50), fringe peak positions, wavelength, etc. Note: a. follow the instruction in /documents for the path of the above GUIs. Or, use startGui (modify to the latest GUI versions) to start them all. b. New thing (not tested) python/IRCam/v10/gdt2.py (replaces ./ delaylines) still buggy -- 11. setup camera. Current setup: 16 rows 160 cols (for K band: 28 rows 160 cols) row off: 5 col off: 38 (for K band: 19, 38) nread 8 --> this gives a 5.45 ms t_int per frame coherent coadds = 140 ps coadds = 10 frames per reset = 1000 to start up camera do the following clicks: update DAQ params Config Camera hit start exposure button. now check image -- hit start on the left display (2d image). you should see data -- if its mostly black/zero then stop exposure and re-do config camera. you can use cross section to tell this. and start right displayon 2-d PS to see fringes. 12. once data directory is created on mirkwood /data, then you can go into it and run % lstail to see the most recent file for monitoring data. During night ----------------- Running chara ---- a few important things. 1. to get offsets at the beginning of run,we used NIRO as a fringe finder. a) first, find fringes on MIRC using white light source and retros on neighboring beams like s1s2 or w1 w2. b) then move beam switchyard and send the beams to NIRO. Find fringes on niro (IRimage). now the combiners are 'phased'. c) find fringes on niro. record offsets. these should directly translate to offsets on mirc even if beam order is different since astromod takes care of the details. NOTE: the offsets of NIRO is in CM, but MIRC uses MM!!! 2. other details to get ople working a) run astromod by hand. e.g., astromod -o CHARANUM -f REFTEL -b beams [this should be obsolete by next run] b) run primary ople server (next run might be secondary..) c) w2 and s1 share delay line cpu so one has to be reference. cant' run both at same time. use REF: xx to move reference cart to desire location. but typically the best thing to do is not to 'track' at this location but to turn that cart off and then to use REF: AUTO for astromod (this will auto reference to the reference cart actual position). d) also make sure all the 'MAN' buttons are clicked in on the ople/control gui when doing fringe tracking searching. e) after invoking astromod, then 'start astromod' on ople gui. Check HD number in ople gui. Running MIRC. ----- 1. after tiptilt closed, then check fluxes, usually with 'quick' fiber explorer (10-15 min total). Baseline changes ------------------------- curfrent procedure is to: 1. change reftel and beam_order arrays in RT code. 2. you must also change in ./delaylines (childWindows) the names of the labels and the beam order ... a few spots.. search for 'S1'. or use the new gdt2 (IRCam/v10) which is nearly working. RT versions and GUIs: -------------------------- follow the instructions in /document directory.