Home

NEWFIRM User Manual

Contents

1. 5 Reboot NEWFIRM Instrument Controller aka NICC Guider must be powered off first and remain powered off until instrument controller has rebooted and reconnected to the network 1 Tum off power to the guider using the guider circuit breaker on the instrument interface panel facing the Cass cage door f there is not already an xterm window connected to nice nice kp connect to nice with ssh X nice nice kp lt er gt wi o 3 Log in as root and issue the shutdown command now The ssh connection will be terminated when shutdown is performed Go out to the Cass cage and tum off power to the instrument controller at its circuit breaker on the instrument interface panel The cold heads will stop 6 Wait 60 seconds 7 Tum on power to the instrument controller The cold heads will start back up 8 Wait 5 minutes for reboot of its self contained computer 9 Reconnect with ssh from mayall 2 following procedure A 3 1 10 start GUI menu with Juse local nice start menu lt er gt Isbin shutdown 100 101 11 Click on menu items to display GUIs for Temperatures Filters and Environmental Cover 12 At this point turn on power to the guider at the instrument interface panel 13 From the OA station in the control room restart the guider per separate Guider Setup Instructions Return to Error Recovery Procedures list 69 101 102 A8 Locations of rack mounted NEWFIRM computer power butto
2. a The images are in default directory ome data Move or copy the fits files and associated dat trig txt files to the desired directory Go to Supervisor Window Paths and Files and do mouse click lt er gt on the path to set it for future images Look for confirming message in the z super window A22 The IRAF el window in the DHS display has disappeared While the cl window appears in the DHS desktop is running on newfirm kp not on the nhs machines a It may have been backgrounded by a stray mouse click Check to see if it is hidden behind the DHS Supervisor window b If truly gone go down to the icons at the bottom of the DHS window and click on an xgterm icon to open an xgterm window Alternatively select New Xgterm window from the window manager menu w type cl lt cr gt to start IRAF in this window to go to the right task and ed to the current data directory If this problem occurs repeatedly halt the user interface in the command line window with and then ask the OA to reboot newfirm kp f Restart the user interface with 2 Check and reset if necessary image root name and data directory paths in DHS Supervisor Window Paths and Files 15 16 A23 The Ximtool in the DHS has disappeared or been accidentally closed 8 Go down to the icons at the bottom of the DHS window and click on an xgterm icon to open an xgterm window b Inthe xgterm window type ximtool a
3. Fig 2 Sereen shots of the lower left and lower right monitors atthe end of the startup sequence There are two different errors that may occur during startup Respond to them as follows Startup error condition 1 Ifthe windows do not appear and you get a sequence of error messages in the command Tine window about annot connect the host permissions need to be reset on mayall 2 Let the startup process run to completion returning the observer newfirm kp prompt in the command line window Type in the command line window eA lt er gt to stop all processes and close all NEWFIRM GUIs Alternatively click on Stop New in the startup GUL Type in the command line window Sit lt gt At the 4meter mayall 2 prompt type Reconnect to computer newfirm as before with In the command line window type Alternatively click on Start Newfirm in the startup GUL Now all the windows will appear and you can proceed Startup error condition 2 During the startup process watch the GUI called NEWFIRM Monsoon Supervisor Layer once it appears A string of about 14 blue RECV messages should appear If you do not see any RECV messages or if any red ERROR messages appear in the scrolling output in this GUE Let the startup process run to completion returning the observer newfirm kp prompt in the command line window Type in the command line window HRSIRESEESERS tha s en em ess ell Reset
4. d From this monsoon login on newfirm kp execute the command This clears DHS setup information from all NEWFIRM computers The next system startup will have a clean slate Exit the monsoon login and close the terminal window f Restart the user interface from the existing observer newfirm kp login with 2 Check and reset if necessary the root image name and the data directories in the DHS Supervisor Window Paths and Files h Take a short test frame or two see if all looks well Return to Error Recovery Procedures list 69 83 s A3 Instrument control errors These manifest as error messages in the Instrument Control System GUI and in the command line window A3 1 Instrument Controller GUIs not seen on Desktop 2 during instrument startup need to restart the connection to nice the instrument controller computer a Open an xterm window on Desktop 2 b Connect to nice with Start GUI menu with d Click on Temperature in the menu to open the temperature display GUI A3 2 No communication with Environmental Cover or filter wheels need to reboot the Instrument Controller A very rare circumstance Ask the OA to reboot the Instrument Controller This requires a halt and subsequent restart of the OA guider software The data system can be left running there is no need to stop or log out This may require moving the telescope to enable access to the cage This process takes 10 15 min
5. A4 4 1 First recovery procedure e In the command line window type f A process will run in the Monsoon Supervisor Layer Window If it completes with no red ERROR messages if not go to second recovery procedure 2 Go to the DHS Supervisor Window Shared Memory Cache and click on Update Status h Watch the z super window to see if an image is processed to disk i Click on Update Status again to see if the memory has cleared 89 90 J If Update Status still shows DATA or META files in DHS memory go to A 2 5 procedures to clear the memory K When you have succeeded with this take a short single image to see if the system is now running normally A442 Second recovery procedure If the Monsoon Supervisor Layer Window continues to show red ERROR messages at step b above do a stop start cycle a Stop and restart the system with b In DHS Supervisor window click on Paths and Files tab and redefine the file root name and directory paths Take a short single image with a TEST script to see if the system is now running normally Debiasing and rebiasing the arrays may produce higher dark current or other transitory phenomena for tens of minutes Running a set of short integration time junk frames may help stabilize them It s not advisable to take low background data e g narrowband filters immediately after a debias rebias error recovery A443 Third recovery procedure If system is still hung with
6. Focus Steps is the number of steps Start Focus value should be several steps less than your initial estimate of best focus Step Size 50 100 and Focus Steps 7 10 are good choices Output images are analyzed with the focus analysis software tools described in Section 9 below Section 9 also describes observational procedures and rules of thumb for focusing eee a eee Pesaran aad este io Srp e scores connotano OO ee ee a oe Fig 14 FOCUS script This example takes one 10 second observation in the J filter at each focus step The sequence starts at telescope focus 10500 and takes 8 frames stepping by 50 units 10 end at 10850 The script takes nine images in all the first image is offset 30 arcsec in Dec and used for sky subtraction by the focus analysis routine Note that the wser assigned Script Name is not quite right Return to top of Sec 6 20 Return to Table of Contents 2 6 3 Simple observing scripts dither patterns The two scripts described here enable observing at a single telescope pointing including dithers small telescope motions and multiple integrations These are often used for targets that do not have extensive continuous nebulosity The science reduction pipeline 28 29 uses the image set to create a mean sky frame with point sources suppressed by dithers which is adequate for sky subtraction in the absence of extended sources The science redu
7. AO Kitchen sink error recovery If symptoms are unclear help is unavailable and or a return to science readiness without diagnosis is desired this sequence of steps will clear up many errors with data acquisition transfer and display a Halt the user interface with b Ask the OA to power cycle the Digital Head Electronics DHE also known as the Monsoon array controller This may require moving the telescope for access to the Cass cage where these electronics reside Open another terminal window on mayall 2 and connect to newfirm kp as d From this monsoon login on newfirm kp give the command This clears DHS setup information from all NEWFIRM computers The next system startup will have a clean slate Exit the monsoon login and kill the second terminal window GAZES from monsoon newrirm kp f Restart the user interface from the existing observer newyirm kp login with B terest compe 2 Check and reset if necessary the root image name and the data directories in the DHS Supervisor Window Paths and Files h Take a short test frame or two see if all looks well Return to Error Recovery Procedures list 69 B A 1 Global window management errors ALLL The noes start all process ran to completion returning observer newlirm kp prompt but no windows or GUIs appeared host permissions need to be reset on mayall 2 a Type to stop processes b Type to return to computer mayall At 4mete
8. This takes about ten minutes See A 7 2 98 Reboot nfpan a kp and nfpan b kp d After both have been rebooted ask the Observing Assistant to tum on Monsoon power in the cage Log back in to newfirm and do a normal system start 95 96 followed by the normal setup activities in all monitors f Take a short single image with a TEST script to see if the system is now running normally it should be A 6 2 Raw image has negative counts and strange appearance the arrays aren t biased Counts may be several hundred ADU negative with light on the array and 1 2 thousand ADU negative with the dark slide in place Appearance is washed out with cracks and other defects bright not dark Bias the arrays with A63 Raw image shows a strong center to edge gradient or a vague shape covering the central part of the field condensation on the entrance window due to high local humidity This must be removed by turning up the dry N2 gas flow and waiting for the window to dry off First verify that it is condensation This will reduce the background sky level in the J band because it is blocking incoming photons but increase the background in K because it is warm and emissive a Take a 5 10 second test exposure on the sky in J Verify that the background level increases from the FOV center to FOV edge i e background is darker in the center b Repeat in Ks Verify that the background level decreases from FOV center to FOV e
9. ssh X monsoon nfpan a kp ii PN maeaea 6 Log in as root to perform a shutdown Lo sucer gt PW is 98 7 From root login give shutdown command sbin shutdown h t 0 now 8 Open another xterm window and do the same shutdown for PAN B 9 Wait a couple minutes for the ssh connections to drop out leaving you back at newfirm 10 Push front panel power button see note below on power buttons on each PAN computer to cycle power back on and reboot 11 Wait 5 minutes for reboot to finish 12 Tum on power to Monsoon at the circuit breaker on the instrument interface panel in the Cass cage 13 Inform the Support Team person that he she may continue with troubleshooting or recovery procedures A73 Reboot nfdca kp 1 Be sure the observer has halted the data system with user commands debias and noes stop all and done an exit from observer newfirm kp 2 Have the observer exit any other observer newfirm kp accounts he may have L 2 Alternatively use the monitor and keyboard provided in the computer room next to the NEWFIRM computer Tack selecting the desired machine with the rack mounted switcher box 4 Connect to nfdea kp i ssh X newfirm nfdca kp Gh Jir 5 Log in as root to perform a shutdown dii su lt cr gt iv PW is 6 From root login give shutdown command sbin shutdown h t 0 now 7 Wait a couple minutes for the ssh connection to drop out 8 Push front panel power butto
10. with lamp settings and integration times from Table 1 above Dome flats done from 4 5 pm with the observer present for required manual interactions A sequence of dark frames 10 each at 5 sec x 12 coadds 10 sec x 6 coadds 30 see x 2 coadds and 180 sec x 1 coadd using separate Dark scripts for each bundled into a superscript Sec 6 1 5 56 Observing times and coadds for dark frames are chosen to match typical use change to fit needs of your program The sequence above provides for short integrations on brighter sources 5 sec sky background limited integrations in H and Ks 10 sec and J 30 sec with 60 sec on source between telescope dithers and single background limited integrations in narrowband filters Darks run for 75 minutes while everyone is having supper Or at the end of the night left to run after telescope and dome are closed up and observer and operator go to bed Return to Table of Contents 2 56 57 9 How to check telescope pointing The instrument pointing center used to point the telescope must be at the science field center for the guider probe positioning calculations to work Unfortunately the science field center has no pixels in it due to the 35 arcsec gap between arrays in the 2x2 mosaic So we position the pointing check star off center then offset the telescope by an amount calculated to put the star at field center to initialize the pointing At start of
11. 50 7 Using the DHS windows while observing Two functions of this set of windows are to display and interact with raw images and to bbe sure that the image writing to disk is proceeding normally 7 1 Image display and interactions Ximtool image display features Image display shows the raw incoming pixel frame autoscaled to sky level Orientation is North up East left Click on the T and C boxes in the toolbar to open up two more toolbars Click on Pan WCS and ISM to get an image panner X Y coordinate display and pixel value display Mag will change the size of the panner box Hold down the right mouse key and pan around the image to change image brightness and contrast Click on Display to bring up a GUI with these and other capabilities he ae ae Pe Tia mae 2083 Fig 31 Ximtool image display with toolbars Pan WCS and ISM selected 50 Images on the disc can be displayed to the Ximtool with IRAF using msedisplay in the mscred package mscexamine or other tools can interact with the displayed image However if you are taking data the next raw frame to be read out will overwrite an image displayed from disc Imexam etc tools will not work on the autodisplayed raw frame since it was not retrieved from the disk It may be convenient to go to a separate desktop on the monitor open another connection to observer newfirm kp and then open a separate IRAF cl window and Ximtool for
12. Settle in seconds is an open loop wait time after a telescope offset and before the start of the integration This allows telescope vibrations to damp out 63 2 DitherStare This script has the same configurations and parameters as Standard but the dats are not flagged as standard star in the archive Use it when observing science targets with a single telescope pointing plus a simple dither pattern Fig 15 STANDARD script GUI Settle time set to 0 reduces overhead at the expense of some image blurring from telescope shake following a move See Fig 17 below for more information on the SPX dither configuration Fig 16 DITHERSTARE script GUL Seutle time set to 15 sec is very conservative See Fig 20 below for more information on the Random dither configuration 6 3 3 Dither patterns The dither geometry choices in Dither Configuration define the pattern of motion Dither pattern size is set by RA and Dec parameters which have variable meanings All pattems are centered on the initial telescope pointing and return the telescope there at the end SPX is an X pattern with integrations at each comer and in the center RA and Dec offsets are the distances from center to corner not corner to comer RA x Dec is a rectangular pattem An m x n grid pattem is stepped through in successive pointings beginning in the southeast comer and offsetting north and west 30 aL RAand Dec offsets are the
13. Telescope Configuration in which most of the parameters have been disabled Relative offsets are enabled for some scripts Absolute pointing to RA Dec is disabled in all scripts Scripts are stored in home observerlexec 2 wigntnat quiceep Syria eehtapninshy Moamapsoarse Dit Fig 9 Seript editor NGUI with buttons for script selection 6 1 1 To create name and save a script To create a script i e to set parameters in an existing script GUD click on the appropriate script selection button at the bottom of NGUL This will open the script specific GUL In every script GUI the Script Configuration allows you to enter an Object Name and a Script Name The Object Name is the title keyword in the image FITS header It should e short descriptive and meaningful to you and for that observation You will execute the script by typing the Script Name in the command line window Name your script something specific and logical to you and not too long To save a script click OK at the bottom of the script GUL If a script with that name already exists you will be asked whether or not you wish to overwrite the previous version What s This opens a short description of the purpose and operation of the script IF you decide not to create a script after all click Cancel All scripts are saved in the exec directory home observer exec The Instrument Scientist may delete scripts at this le
14. Tool to retrieve a tarfile The tarfile contents include a README file with instructions for installation on your system While this downloadable copy of NGUT is useful for practice in making scripts the resulting scripts may not work at the telescope NGUI uses the filter position number not the filter name to create scripts The list of filters in this demonstration version of NGUI may not match the list actually installed in NEWFIRM at the time of your observations Return to top of Sec 6 20 Return to Table of Contents 2 6 2 Calibration and telescope setup scripts Secs 6 2 6 3 and 6 4 describe all the NGUI scripts from simplest to most complex New users are strongly urged to read through each of these descriptions and examine the associated script GUL Each successive script description assumes knowledge gained by reading its predecessors Here we describe scripts used to take dome closed calibration data and to get the telescope pointed and focused on sky 25 NOTE the Quick Reduce Pipeline requires appropriate Dark and Dome Flat Sequence data taken with these scripts Be sure to do this the afternoon of your first night See Section 8 62 1 Dark Dark is the simplest script with the smallest number of user adjustable parameters To start click the Dark button in the script selection NGUL This script has three Configuration sections Script Observation and Monsoon Configuration Consult with your start
15. coverage is still continuous Fig 26 Sky coverage of the RA x Dee pattern of Figs 24 and 25 The sky is sampled with twelve overlapping telescope pointings 4l Fig 27 Observation with an RA x Dee pattern time on source The single pointing FOV is shown in red An area of sky within this blue is sampled at every pointing for maximum time on source except for the eross shaped region where the array gaps produce stripes of lower Integration time and SNR The total sky coverage is outlined in green Integration time and SNR taper stepwise to lower values in the frame defined by the blue and green outlines In general any dither pattern that is bounded by a box will produce an area of maximum sensitivity that is the single pointing FOV less a frame of half the box width on all sides The regular pattern of RA Dec offsets used for Figs 24 27 emphasizes the stripes of lower integration time across field center by precisely overlaying the row and column array gaps multiple times This reinforcement can be reduced by using a Random pattern or by creating a file of offset positions chosen so that the array gaps do not overlap on sky between any telescope pointings This will produce spatially smoother SNR in the final registered and stacked composite image B Return to top of Sec 6 20 Return to Table of Contents 2 6 4 Scripts with complex telescope motions Previous scripts have enabled a pattern o
16. data inspection 7 2 Monitoring data logging This is done through the DHS Supervisor Window and the z super window In the DHS Supervisor Window click on the Shared Memory Cache tab This displays the data presently in memory A complete data set for a single image is five lines of text all with the same exposure ID number at the end ExpID column Lines of text will appear and disappear as data are read into memory and then processed onto the disc Fig 32 DHS Supervisor Window Shared Memory Cache display 51 Processing steps are displayed in the z super window The main thing to look for is the highlighted box with message Postproc DONE updating at regular intervals This indicates that a raw image has been written to disk as a multiextension FITS file Fig 33 zsuper window with Postproc DONE message Tf lines of text begin to build up in the DHS Supervisor Window without disappearing 10 lines or more try clicking the Update Status button at the bottom to flush data to disk It does no harm to click this more than once to flush multiple frames You should see processing occur in the z super window as text lines disappear from the DHS Supervisor Window If processing stops and text lines continue to build up or if the Shared Memory Cache page turns red there is a problem Use the procedures in Appendix A 5 93 to abort any script that s running and then recover from the error condition with the
17. data directory with h Doa directory listing then delete old data with the command clean lt er gt Note the leading underscore i This will remove any FITS image older than seven days J Take a short test image to see if the data are processed and stored to disk properly Ifthe raw image shows all pixels having exactly the same value go to Procedure A61 k Put an entry into Service or have the OA make an entry describing briefly what you ve done The Instrument Scientist may want to do further data deletion after checking with the Archive Scientist A2 6 2 Second recovery procedure if data disks aren t full but data aren t being written to disk as fits files a Halt the user interface and exit in the usual way This will return the mayall 2 prompt b Exit from any other observer newfirm kp logins that may be open on mayall 2 or mayall 3 Ask the OA to shut down and reboot the Data Capture Agent computer nfdea See A 7 3 98 Reboot nfdea kp Then log back in as observer newfirm kp and restart 9 80 A2 7 nffocus IRAF routine hangs leaving the IRAF cl window unresponsive to mouse or keyboard This occasionally happens during the catalog creation process Kill the process and or the IRAF window and start a new CL session A 2 7 1 First recovery procedure a Type repeatedly in the CL window 5 or more times b I it returns a prompt type fpr lt er gt a couple of times Recovery is complete instead th
18. for data directories is fnfidata observer Directories have numerical names in the form year month day based on the local MST date at startup e g 20120231 You ll see directories from previous nights there as well as the one for the present MST date You may create and use different directory names and or subdirectories for raw and processed data using the IRAF command mkdir in the IRAF cl window Irecommend creating a subdirectory for processed data within your raw data directory This will receive output from the Quick Reduction Pipeline Once created set the path to your data directory in the IRAF el window with the usual IRAF command 4 1 3 Rearrange the window layout to your liking 10 The Figure shows one arrangement Raw images are automatically displayed from memory in the image display The z super window shows the progress of data from memory to disk with a highlighted message when this transfer is complete The mosdea window is infrequently needed and can be positioned partially behind one of the others Fig 3 VNC window to Data Handling System rearranged by the observer Raw and Processed image directories have been set The zsuper window shows the acknowledgement messages 4 1 4 Set the image root image name and the data directories in the Supervisor window You need to enter the desired name or directory path in this window and check the super window for an acknowledgement message Click on th
19. individual pixel raw values that will result The S N in the averaged differential frame will be lower than for the other filters Table 1 Recommended lamp settings and integration times for dome flats Fiter_ Lamp setting Integration time Signal Tevel J Low 25 0 sec 3000 ADU H Tow 15 5 sec 3200 ADU Ks Low 15 T 2sec 3900 on 1700 off 1066 High 14 TO sec 3200 ADU 1187 2096 1644 Tow 25 EEA 1000 sec 2122 Low 20 15 sec 2600 on 500 off 2168 Tow 20 15 sec 3000 on 1000 off 1056 High 22 10 sec 2800 ADU 1063 High 22 T0 sec 3000 ADU J Tow 30 20 sec 2800 ADU J2 Low 30 10 sec 3000 ADU B Tow 30 65 sec 3000 ADU HT Tow 15 TO sec 3000 ADU H2 Low 15 Esec 3000 ADU Depending on temperature in the dome A value of I sec defaults to 1 24 sec the minimum readout time In warm weather when the dome flat screen itself is emitting a lot of thermal radiation it may not be possible to get unsaturated Ks flats 8 2 Taking dark frames You need to take dark frames with the same array parameters as your on sky science data integration time coadds Fowler samples and digital averages The Science Pipeline uses the on sky data and matching darks to generate source free scaled background sky frames for sky subtraction On your first afternoon you may have to estimate on sky times for the various filters from previo
20. night have Observing Assistant go to a bright star 3 6 mag Take a short integration 1 2 seconds in Test mode IF bright star is not seen in raw image it may be in the gap between arrays Offset the telescope by 100 aresec in RA and in Dec via the Telescope Control GUI and take another frame You should see the star now Use the cursor to determine the offset in pixels between star position and the inside comer of the array it s on Convert to arcsec at 0 40 arcsec pixel Add 17 arcsec in RA and in Dec to allow for half the gap between arrays Apply this offset to the telescope to place the star at field center Take a final Test frame to confirm positioning The star should not be seen Diffraction spikes on very bright stars may be seen They should radiate from the inside corner of the active area of each array Ask the Observing Assistant to zero the telescope at this position This sets input coordinates for the star current telescope position recommend verifying pointing in this manner whenever you go to a new target more than 30 degrees or so from the previous pointing Return to Table of Contents 2 57 58 10 How to focus Take a focus sequence using a FOCUS script Sec 6 2 4 The first frame duplicates the first focus value and is offset on the sky to get a frame for sky background subtraction It is easiest to focus in J and apply a known focus offset between J and other filt
21. no spaces The system will automatically load the Monsoon Supervisor Layer again Watch for the blue RECV and red ERROR messages Repeat this cycle until the load runs to completion with no red ERROR messages If they appear at all usually one or two nmslReset cycles are sufficient to get through this Once the startup has completed without errors and the observer newfirm kp prompt has appeared in the command line window bias the arrays with the command This applies a 400 millivolt bias to the array detectors making them electrically active Summary log in to mayall 2 as dmeter ssh connection to newfirm kp as observe or Start Newfirm in GUI in the event of red ERROR messages to activate the arrays Return to Table of Contents 2 3 Shutdown and logout At the end of the night protect the instrument by closing the warm entrance window cover and shutting down the array controller and user software In the Instrument Control System GUI close the Environmental Cover For more about this see Sec 5 This cover must be closed before the OA closes the telescope mirror covers Be sure to coordinate with the OA Use the Options button in upper left corner of the GUI to enable the button bars running across the bottom of the GUL Click the bottom button bar labelled Close From the command line window in the right hand monitor type noes stop all lt cr gt This debiases the arrays
22. patterns and parameters are the same as for Dither Configuration The same rules apply as to location of starting point how the telescope walks through a Map pattern and what RA Dec offsets mean depending on the chosen Map pattern Large 1600 arcsec Map offsets displace the telescope by nearly a full NEWFIRM field of view with a little overlap for registration The fixed half field offsets 800 arcsec of the 4Q pattern can be used to position a moderately extended target on each quadrant of the array mosaic in succession and then dither it using the Dither Configuration A large Map motion gt 2 arcmin will place a selected guide star out of range of the guide probe The Guider Wait option will pause data taking to allow acquisition of a new guide star at each Map position This will impact observing efficiency One generally thinks of Map motions as large and Dither motions as small However Map offsets can be any size Small Map offsets few arcsec can be used to quasi randomize the pixel sampling of the sky in an iterated dither pattern while guiding This is useful for very deep imaging of fields containing only pointlike sources 43 6 4 1 DeepSparse This script adds Map Configuration Map patterns and parameters are the same as for Dither Configuration The same rules apply as to the location of the starting point how the telescope walks through a Map pattern and what RA Dec offsets mean depending
23. procedures in Appendix A 2 75 Return to Table of Contents 2 53 3 Taking dome flats and darks Both the Quick Reduce Pipeline and the Science Pipeline require dome flats taken with the Dome Flat Sequence script to process your on sky data Be sure to get these during the afternoon Dark frames that match the science exposures are needed for the best quality Science Pipeline reduction Provision of dark frames is critically important for exposures with low background especially narrowband images This requires advance planning of science exposure parameters for the first night so that darks can be taken to match in the afternoon 3 1 Taking dome flats Ask the operator to request that the telescope be set up for dome flats at 4 00 p m Use a DomeFlatSequence script to take these data Sec 6 2 3 This is necessary for the pipeline processing This script allows you to choose which filters to use and to set the integration time for each filter It will take a set with lamps off in all filters and then a set with lamps on DomeFlatSequence will prompt you to tum the lamps on or off and to set them to the desired brightness at every filter change Take at least three dome flats in every filter The pipeline requires this I recommend a minimum of ten in every filter for adequate photon statistics in the flatfields Execution will stall if a filter wheel does not position properly and the script must be a
24. red ERROR messages try a hardware reset of the Master Control Boards from the keyboard these boards are in the Digital Head Electronics box on the Dewar a Issue the following command sequence from the shell window b Take a short single image with a STARE script to see if the system is now running normally See the note about debiasing and rebiasing the arrays above at A 2 6 2 A444 Fourth recovery procedure This is a reset of the Master Control Boards at the boards with a reset button 91 b From the shell window do Ask the Observing Assistant to perform a hardware reset on the DHE Master Control Boards This may require moving the telescope for access to the Cass cage where these electronics reside Instruc ns to OA EM for Master Control Boards reset 1 Be sure the observer has halted the data taking system steps a and b above 2 Go to the Cass cage and open the top cover of the DHE box 3 Identify the two Master Control Boards front panels labelled MCB with a reset button 4 Use a pen tip to firmly push the small black reset button on each panel you may not hear or feel a click 5 Various red lights should come on on these boards and on some of the other boards in the box 6 Securely close the top cover of the DHE box 7 Inform the observer that he she may continue with the recovery procedure d After the OA has done this restart the system from the shell window with d In
25. shut the NOCS down with ASSMAN this is very important and then execute then restart the system with 87 88 A43 Red ERROR message appears about a voltage out of range and the array does not integrate and read out this may be a telemetry glitch or an actual hardware voltage error This may occur during initial system startup or while taking data If the error appeared while taking data close the image file that s been opened in anticipation of a completed integration This will keep the DHS from hanging up From the command line window issue the command ditscmd nohs nohs_endobs lt cr gt Now proceed with the recovery procedures A43 1 First recovery procedure a In the command line window type b A process will run in the Monsoon Supervisor Layer Window If it completes with no red ERROR messages Take a short single image with a TEST script to see if the system is now running normally d If you still get red ERROR messages continue to the second recovery procedure A432 Second recovery procedure for an actual hardware error a From the row of iconized windows at the bottom of the lower left monitor foreground the panDaemon windows for PAN A and PAN B b Look for messages in these windows like ERROR detGetState NEWFIRM VssExt_Ary2 4 955 volts Out of Range requested was 2 471 Note down which voltage s on which PAN are identified as wrong for entry into the Service log d From the comma
26. these sources Displays a plot of source FWHM vs telescope focus in a graphics window 59 Displays its solution for best focus value from a fitting routine Figure 45 nffocus graphics with course upper and fine lower sampling Occasionally for unknown reasons the nffocus routine hangs leaving the IRAF cl window unresponsive to mouse or keyboard See Appendix A 2 7 75 for recovery procedures Visually inspect the plot of FWHM vs focus setting as a sanity check Your eye may do better than the curve fitting routine at identifying the best Focus 59 Note the focus value solution and eyeball then type q in the graphics window to quit this routine or explore the data further with keystrokes below Keystrokes in the graphics window dis in various ways and can be used to delete suspect data points lay the analysi plots FWHM vs focus setting default plot plots focus vs various spatial parameters radial distribution of FWHM magnitude distribution of FWHM locate cursor over bad point hit d to delete it deletes entire data set for given focus value reverses last delete action exits the graphics window may have to hold down firmly not merely tap fee agron It may be useful to use the d delete keystroke to remove discordant data points from the FWHM data sets near the fitted best focus If the fitted focus solution is stable it won t change much Ifthe displayed FWHM
27. vs focus setting doesn t have a clearly defined minimum you probably haven t gone through the best focus position Run the focus script again with a different START value to try to bracket the estimated telescope focus On nights of poor seeing the focus may have a very broad minimum or be undefined This may also happen if you have chosen a very star poor field for the focus test Focus stability can be monitored on the science frames as observing proceeds Both changing temperature and truss flexure will change the focus I recommend running a focus sequence if the telescope truss temperature changes by 0 5 deg C or more or if you move across the sky to a new target by more than 30 degrees 6l Here is a tabulation of the focus offset for each filter with respect to filter JX broadband D Filter Offset J 0 H 7100 Ks 3200 1066 1187 2096 1644 7100 2722 185 2168 1056 350 1063 50 Fil 0 2 3 HI GA Return to Table of Contents 2 61 11 Guiding The NEWFIRM guider is operated by the 4 m Observing Assistant OA from the OA s telescope control station The guider uses two cameras that access stars in two rectangular guide fields via X Y stages Each camera can access any position within its guide field However the guide star must stay within the field at all dither positions The guide field size is 26 8 arcmin RA by 4
28. we refer to it as the command line window in what follows 2 1 Startup from command line window In the command line window ssh to computer newfirm as user observer with ssh X observer newfirm kp lt c gt PW for observer is At the prompt observer newfirm kp type ASSAR lt er gt three words space between 2 2 Startup from i Alternatively on the lower left monitor right side click on the Newfirm icon This will open a startup GUI in the upper left corner of this monitor Click on the blue bar labeled Start Newfirm X Ne Ready Start Newfirm Fig 1 Newfirm icon upper and startup GUI lower After a command line or GUI startup several windows appear on all four monitors This takes about one minute Some windows appear briefly and are then iconized in the toolbar at the bottom of the lower left monitor A VNC window to the Data Handling System will appear centered on the four monitor system use the mouse to move this to the lower right hand monitor Along string will scroll through the command line window a preliminary to opening the next set of NEWFIRM operational windows Several small windows appear in the upper left hand monitor with some being iconized Several GUIs appear stacked on top of each other in the lower left monitor The startup is finished when the observer newsirm kp prompt appears again in the command line window of messag
29. 2 longer at more northerly Declinations and shorter near the celestial equator For many programs using broadband filters 60 sec or less integration time followed by a telescope dither is a good protocol One can also select a guide star but disable guiding and use it only as a telescope position check at the end of a dither sequence This is useful if observing the same field in multiple filters The periodic loss of the guide star when using a Map configuration can be finessed a bit For Map positions that will only be used to define blank sky for sky subtraction it doesn t matter if the images are a bit elongated so a new guide star doesn t have to be acquired Return to Table of Contents 2 63 12 Real time data inspection There are two ways to inspect your data and evaluate data quality One is to access the Taw data as they come in and perform simple image analysis with IRAF For example sky subtraction followed by image quality evaluation using the imexam task provides a running check on telescope focus The other way is to inspect the Quick Reduce Pipeline products These include sky subtracted linearized flatlielded shifted and added final frames from dithered observations This permits confirmation of detection of sources plus quantitative evaluation of signal to noise Both approaches can use the workstation to the right of the NEWFIRM observer s workstation so that inspection does not interfe
30. 4 arcmin Dec This constrains dithering in Declination N S If the chosen guide star is near the N S centerline of the guide field then dithers must remain within 2 arcmin of the starting position If the guide star is less favorably positioned then N S dither motions are further constrained The guide star display and selection tool at the OA station displays the science field of view and the two guide fields at the current telescope position from the Digital Sky Survey Consult with the OA to choose a suitable guide star balancing brightness and accessibility while dithering i i Daa aene Oa magan eA atr ny me Figure 48 Guide star display and selection tool 63 There are two overhead costs with guiding time spent manually selecting and setting up on a guide star before starting an integration sequence 1 2 minutes and time needed to automatically recover the guide star after a telescope dither a few seconds Use of a script that includes Map motions Sec 6 4 introduces another complication A large Map motion will move a selected guide star completely out of the guide field The Guider Wait option will pause data taking to allow manual acquisition of a new guide star at each Map position This further impacts observing efficiency The most efficient observing mode is to not use the guider The open loop tracking of the telescope will deliver round star images for exposure times up to 60 seconds at Dec 3
31. DHS Supervisor window click on Paths and Files tab and redefine the file root name and directory paths Take a short single image with a STARE script to see if the system is now running normally A445 NMSL task failure error message appears in the command line window or the Monsoon Supervisor Layer GUI There is no activity in the Monsoon Supervisor Layer GUL The Monsoon Supervisor has lost its connection to the router To recover In the command line window close any observation that may have been left open with and watch the DHS windows for activity indicating data transmission Then shut down and restart the system with the usual 91 A4 6 Red ERROR message appears during data taking about a timeout error and there are problems in the DHS z super window data capture process has stopped at Trigger Host and issuing the command ditsemd nohs nohs_endobs from the command line window doesn t produce a response in the z super window This is the same root problem as A 2 8 bogus lock files in the tmp directories of the nfdhs 01 and nfdhs 02 machines are confusing the processing See A 2 8 75 for the recovery procedure Return to Error Recovery Procedures list 69 93 AS Problems with scripts AS 1 How to abort a script This has to be done carefully or the Data Handling System will get tied up requiring an extensive recovery process a Ifthe script will run to completion in a minute or two just let it fi
32. NEWFIRM User Manual for KPNO Operating Procedures for Observers Ron Probst NEWFIRM Instrument Scientist August 2015 v 74 Contents Note Page numbers in brackets are clickable links to the corresponding section 1 Introduction Login and startup 4 2 1 Startup from command line window 2 2 Startup from icon 3 Shutdown and logout 9 Setup steps after login and startup 10 4 1 Setup actions in the DHS VNC window 4 2 Setup a 4 3 Other useful tools Instrument initialization 18 Data acquisition and scripts 20 6 1 Creating and using scripts 20 6 2 Calibration and telescope setup scripts 24 6 3 Simple observing scripts 28 6 3 3 Figures illustrating script dither patterns 6 4 Scripts with complex telescope motions 43 7 Using the DHS windows while observing 50 7 1 Image display and interactions 7 2 Monitoring data logging 8 Taking dome flats and darks 53 8 1 Dome flats 8 2 Darks 8 3 Standard calibration plan 9 How to check pointing 57 10 How to focus 58 11 Guiding 62 12 Real time data inspection 64 12 1 Raw data 122 Quick Reduce Pipeline data products Appendix A Error recovery procedures 69 1 Introduction This is a procedural guide to NEWFIRM observing with the NEWFIRM Observation Control System NOCS It covers system startup script creation and data acquisition daytime calibration frames pointing and focus checks on the sky observing tips and data inspectio
33. P has processed for your run so far with various information about each observing sequence including 67 image filenames UT at the start of the observing sequence observation type exposure times filter focus setting number of digital averages DA and Fowler samples FS image title In addition this log reports the ranges spanned by some data quality parameters that are measured by the pipeline such as sky brightness in ADU and mag per square arcsec o seeing FWHM o magnitude zeropoint estimated from 2MASS stars useful for monitoring relative transparency estimated 5 sigma point source sensitivity Clicking on the image filenames for a given sequence takes you to a pipeline review page with information about data from that sequence There are review pages for processed combined dark frames and dome flats as well as for nighttime on sky observations For nighttime science observations the QRP review pages provide thumbnail pictures of the data links to images at two different higher resolutions for quick visual checks basic header information about each image links to the full FITS headers data quality measurements seeing magnitude zeropoint depth etc see above For dithered on sky observing sequences preview images are shown for each individual exposure as well as for stacked mosaics of the observations The stacked images are shown in the second column of thumbnail images If during
34. aborted Any data taken up to that point are useless to the pipelines I recommend doing filters in batches J H Ks first then narrowband filters in a separate sequence Jmn RS ae ae meee O 1 E Fee oS mee TPA te Meme I Sa ealo ae See eS ter et PA ES Wine Voz e RoC Te fre aie a n E al lesen al al we Fig 13 Dome flats sequence script DFLATS This script will take 10 observations in each filter JH Ks The integration time s different for each filter The dome lamp setting for each filter must be manually set by the observer as the script proceeds 6 24 Focus Focus takes a series of images stepping the telescope focus in between This script adds Telescope Configuration to the configurations The script first offsets the telescope n 28 and acquires a sky frame then offsets back to the initial position and takes an image at each focus step These are analyzed with an IRAF routine For more information on the focus process see Section 9 below Telescope Configuration user variables are the parameters for the focus sequence RA Dec and Epoch are disabled Leave these parameters as is OFFSET RA Dec define the initial offset for the sky frame in areseconds RA 30 and Dec 0 are reasonable choices FOCUS Start Focus Step Size Focus Steps define the focus sampling Start Focus is the starting value of telescope focus Step Size is the focus increment for each step
35. antages of 4Q for this source are i 100 efficiency integrating on the source and ii the sky is well sampled by each array in the 2x2 detector mosaic for good sky subtraction on the source Next consider an area of sky with dispersed small sources observed with an RA x Dec pattern Fig 24 shows the DITHERSTARE script GUI that defines the observations Fig 25 shows the detector point of view how the field center of the initial telescope pointing moves around the focal plane Notice that the offsets between dither pointings are larger than the gap between detectors Fig 24 Script defining the dither pattern shown in Figs 25 27 This RA x Dec pattern uses 3 x 4 positions with different offsets in RA and in Dec S 2c gt tee Fig 25 Observation with an RA x Dec pattern detector point of view The X s indicate how the field center of the intial pointing moves around the focal plane This is a 3 x4 pattern with spacings of 105 arcsec in RA and 70 arcsec in Dec 41 The next two figures show how the sky is sampled and compares this to a single pointing The area of maximum source time on pixel is smaller than the single pointing FOV Over most of this area source time on pixel is uniform and equal to the total integration time of the dither pattern The gap between arrays introduces stripes of reduced time on pixel across field center But since the dither offsets are larger than the gap spatial
36. arameters GUI can be reopened from the script generating GUI Sec 6 1 to make changes for example if two different programs are splitting a night 4 2 2 Reposition other GUIs Look at the larger GUIs on the lower left monitor These are stacked or overlapping and need to be positioned for better access They are NOAO Extremely Wide Field Infra Red Mosaic Imaging Camera called NGUL for short This is the script generating GUL Its use is discussed later in this document NEWFIRM Telescope Control System Allows user control of telescope offsets Other capabilities are disabled NEWFIRM Instrument Control System Allows user positioning of filters and warm dust cover NEWFIRM Monsoon Supervisor Layer Shows progress of integration and readout NEWFIRM Observation Header System Tracks images and header information as they are assembled into FITS files on disk The following rearrangement has proven useful Minimize the Observation Header System GUL to the toolbar across bottom of screen Observers rarely use it Move the Monsoon Supervisor Layer GUL to the upper right monitor You will be watching the integration countdown and other activity scroll past in this window Position the Telescope Control System and Instrument Control System GUIs across the top of the lower left monitor They will overlap slightly In these two Control System GUIs activate their button functions Click on Options in the toolbar upper
37. borted Any data taken up to that point are useless to the pipelines I recommend doing filters in batches J H Ks first then narrowband filters in a separate sequence The DomeFlatSequence script sets the integration time in each filter However lamp brightness is set manually so taking a set of dome flats is an interactive process Lamp control is via a group of buttons on the telescope console below the windows to the telescope labeled LAMPS The upper button indicates which set of dome flat lights is tumed on The middle row of three buttons select high intensity lamps low intensity lamps or off no lamps The bottom row of two buttons let you make the selected lamps brighter or dimmer A large red display shows the brightness setting A toggle switch below the lamp control buttons allows the same buttons to be used to drive counterweights for balancing the telescope Be sure this switch is set to Lamps Here are recommended lamp settings and integration times for each filter Integration times were chosen for filters in the J and H bands to give about 3000 ADU of signal This is 24000 electrons or 240000 electrons total in ten frames for a photon statistics S N 500 Integration times were chosen for filters in the K band so as not to exceed 4000 5000 ADU in the lamps on frame lamp plus background thermal emission This keeps 53 54 the signal in the linear regime of the detector response curve over the range of
38. cedure in A 3 2 has connection refused We ve seen this Ask the OA to cycle AC power to the Instrument Controller This requires a halt and subsequent restart of the OA guider software The data system can be left running there is no need to stop or log out This may require moving the telescope to enable access to the cage This process takes 10 15 minutes Instructions to OA EM for power cycling the NEWFIRM Instrument Controller This is the crude way to reboot the Instrument Controller when itis refusing an ssh connection 85 Guider must be powered off first and remain powered off until instrament controller has rebooted and reconnected to the network 1 Exit the NEWFIRM guider software in the guider VNC window see separate NEWFIRM Guider Manual 2 Tum off power to the guider using the guider circuit breaker on the instrument interface panel facing the Cass cage door 3 Tum off power to the instrument controller at its circuit breaker on the instrument interface panel 4 Wait 60 seconds 5 Tum on power to the instrument controller 6 Wait 5 minutes for reboot of its self contained computer 7 Reconnect with ssh from mayail 2 following procedure A 3 1 8 start GUI menu with i fust local nice start menu lt er gt 9 Click on menu items to display GUIs for Temperatures Filters and Environmental Cover 10 At this point turn on power to the guider at the instrument interface panel 11 From the OA s
39. ction pipeline will apply linearity correction flatfielding sky subtraction image registration and stacking to data taken with these and subsequent scripts to produce science ready images Pay attention to the dither geometry discussion the same choices recur in all subsequent scripts If the NEWFIRM guider is used dither patterns must remain within bounds for guide star acquisition 6 3 1 Standard Standard adds Dither Configuration for telescope dithering capability It is used for standard stars The data are flagged as standard star in the archive and are not proprietary Other users can retrieve standard star data Telescope Configuration present but inactive inputs are ignored Observation Configuration user variable is NumObs NumObs is the number of observations taken at each telescope position before the dither motion to the next position An observation may be a single integration or a coadded sum of several integrations Dither Configuration sets the dither pattern and number of repeats The initial telescope pointing will be the center of the dither pattern The telescope returns to this position when done This Configuration appears in all subsequent scripts and the same rules apply RA and Dec set offsets or bounding box sizes for dithered offsets The meanings of these parameters depend on which dither geometry is chosen Iterations sets the number of times to repeat the dither pattem
40. dge i e background is brighter in the center Then clear the window by evaporating the condensate Ask the OA to turn up the dry N2 gas flow across the NEWFIRM entrance window It may be necessary to move the telescope to allow access to the Cass cage to do this There are two flow control valves in series Both need to be turned up One is the gas flow panel mounted on the east side of the cage The other is on a panel mounted on the yellow NEWFIRM truss on its northwest side near the instrument controller NICC 97 d Take a Ks frame on sky from time to time to monitor the background as the condensate is dried off the window Clearing the window may take anywhere from a few minutes to an hour or more depending on how badly it has fogged and the ambient relative humidity 1 Clearing will proceed more rapidly if the Environmental Cover is closed However open it whenever you do step d then close it again Return to Error Recovery Procedures list 69 Return to Table of Contents 2 97 AT Reboot instructions for NEWFIRM computers ATL Reboot newfirm kp 1 Be sure the observer has halted the data system with user commands debias and nocs stop all and done an exit from observer newfirm kp 2 Have the observer exit any other observer newfirm kp accounts he may have opened 3 Open an xterm window on Desktop 3 of mayall 2 Alternatively use the monitor and keyboard pro
41. distances between successive points Setting m 1 and n will result in no dithering one telescope position only at the initial telescope pointing e 4Q moves the target to the center of each quadrant of the mosaic in succession This is a fixed and very large dither pattern RA and Dec offsets do not apply to this geometry o Quadrant to quadrant motions exceed bounds for guiding Random generates a bounded irregular pattern At each successive step an RA and a Dec motion are chosen randomly with a value between zero and the RA and Dec parameters o RA and Dec also set the boundaries within which the telescope randomly pointed RA and Dec from the initial telescope pointing o Selecting Allow Brownian motion in the NGUI options will choose an unbounded random walk pattern This is not recommended From File moves the telescope using a user created list of successive relative offset positions Consult with the Instrument Scientist or observer startup assistant for use of this option o RA and Dec offsets do not apply to this geometry Setting Iterations gt will repeat the patter If using the guider the pattern will be fixed on the same detector pixels pretty closely which may not be desirable If not guiding telescope tracking drift will offset the pattern on the detector from cycle to cycle by a slight amount To introduce deliberate offsets use a mapping script such as DeepSparse The total number o
42. do in each intTime fSamples 1 is recommended for sky background limited observations with broadband filters Larger values of Samples are for use with narrowband filters to reduce readout noise when it isa significant noise contribution 26 digAvgs sets the value of a digital high frequency filter in the readout electronics The default value of 4 is recommended for broadband and most narrowband filters Larger values may help with noise reduction with very narrow bandpass filters Fig 11 DARK script GUI configured to take 50 10 sec individual dark frames coadds 1 Samples 1 and digAvgs 4 are the default values for broadband filters 6 2 2 Test Test is a very simple script It enables filter selection in the Observation Configuration so external photons can be detected Test is used for quick snapshots to check dome flat or sky signal levels confirm telescope pointing etc The science pipeline will ignore any data taken with a Test script Raw images are archived but not further processed Observation Configuration filter selection is enabled The button is to the LEFT of the filter name Fig 12 TEST script GUI configured to take one 5 see integration in the H filter 6 2 3 DomeFlatSequence This script is used to take dome flats for the Quick Reduction and Science pipelines These flats are essential to the reduction process This is not entirely automatic The observer must be pr
43. duce higher dark current or other transitory phenomena for tens of minutes Running a set of short integration time junk frames may help stabilize them It s not advisable to take low background data e g narrowband filters immediately after a debias rebias cycle A25 Data are stuck in the DHS shared memory cache indicated by many lines of entries in the DHS Supervisor Shared Memory Cache display and no processing activity in the z super window However Shared Memory Cache has not tumed red A 2 4 above The goal is to flush the DHS memory preferably saving the data There are several procedures to try A 2 5 1 First recovery procedure a Click on Update Status button on Shared Memory Cache page b Watch for activity in the z super window This indicates data transfer Click on Update Status again to see if the memory is empty if it isn t click a couple more times to see if this procedure flushes it d If no entries are seen in the Shared Memory Cache you have succeeded If data don t entirely clear go to next procedure A 2 5 2 A 2 5 2 Second recovery procedure a Stop any data taking script that is running with CNTRL C in the command line window Wait for the integration to complete and return observer newfirm kp prompt If no script is running go to next procedure A 2 5 3 b In the command line window issue an end of observation command Watch for activity in the z super window endi
44. e CL session is killed restart it with elses and reset the path to your data directory A 2 7 2 Second recovery procedure a Kill the IRAF window with the window manager button click on the left hand icon in the top bar to open a menu and then click on close b Open a new xgterm window from the row of icons along the bottom of the VNC window display Type el lt cr gt to start a new CL session and reset the path to your data directory This leaves the hung process running which does no harm If your comfort level allows you can clean things up as follows d Open another xgterm window from the row of icons Log onto machine newfirm with neers z 6 Type psx lt er gt 80 s to generate a process list Look for an x_ecl e process that is running and an x_ace e process that s defunct ecl is the IRAF CL and ace is the stuck cataloging and analysis process 2 Kill the stuck ect process with aILS9 fpigf 2ES where pid is the process ID number from the list h This will return a Unix prompt in the window with the hung CL session i If you had multiple IRAF sessions running perhaps to do some image analysis care is needed not to kill the wrong one But you can always just restart any that you kill by mistake A23 Data transfers to the DHS processing computers appear to freeze bogus lock files in the tmp directory are confusing the processes and must be removed We ve seen th
45. e NEWFIRM DHS Supervisor window to foreground it Click on the Paths and Files tab Find the root filename and image directories in the boxes near the top of the page Set the raw image directory path and hit enter Check the z super window for an acknowledgement message Repeat for the processed image directory path To reset the root name delete the default type in your preference and hit enter Image root name changes do not generate an acknowledgement in z super window You can change the image root name at any time for example different root names for different targets The appended six digit number just keeps incrementing so you can t accidentally overwrite data 4 1 5 Start automatic flushing of the DHS memory In the DHS Supervisor window click on the Shared Memory Cache tab Click the Continuous Polling box at page bottom to activate this feature This helps to keep data flowing smoothly from memory to disk ial Fite Heap Quit Process Status Shared Henory Ceche Real Tine Display Pathe amp Files shored Henory Lache Status OO M i meira porsche nfparite Se Page Usage nfehe 02 o nfdheolr o Diek Uea Fig 4 DHS Supervisor window Shared Memory Cache tab with Continuous Polling bas at bottom center Summary Open IRAF window from the row of icons at bottom of display Arrange desktop to your liking Go to NEWFIRM DHS Supervisor gt Paths a
46. er shm d set to 1 it will be necessary to g0 through a debias nocs stop all nocs start all 400mvbias cycle to get the status window to show as empty 411 609138 A26 Data aren t being written to disk as fits files This probably indicates that a data disk is full There are two disks that can produce this error One is on newfirm kp nfdata the other is on the Data Capture Agent computer nfdea kp home2 data You need to check both disks and clear off old data to make room nfdata and Jhome2 data each have the capacity to hold a month s worth of broadband data so anything that you delete has long since been archived and checked A2 6 1 First recovery procedure to clean off data disks a b d In the NEWFIRM DHS Supervisor window Shared Memory Cache page check the Disk Usage Bar in bottom right corner If it reads 100 then nfdata is full In the IRAF window cd to nfdata observer You will likely find many nights worth of data in subdirectories Delete subdirectories beginning with the oldest while watching the Disk Usage Bar Get it below 90 The delete command you need is Now go to an unused desktop and open a terminal window on mayall 2 prompt will be 4meter mayall 2 Log in to the Data Capture Agent computer nfdea as user monsoon with 78 9 f Check the available space on home If it s full you need to delete some data From home monsoon enter g ed to the
47. ers Ten seconds x 1 coadd is a good integration time for J This will average out any rapid variations in seeing For H and Ks use 5 sec x 2 coadds due to the higher sky backgrounds Focus is a function of telescope truss temperature T and wavelength Use these formulas to get reasonably good starting values Set up the focus sequences to be centered on these values J Focus 12000 100 x T C H focus J 100 units K focus J 200 units Or check with the previous observer or the telescope operator for best focus at the start of the previous night A good step size is 100 units for an initial search followed by 50 unit steps around the best resulting focus value to refine the result Set starting value and number of steps to bracket the estimated focus with three or four steps on each side The truss temperature is obtained from the Truss Temp GUI Sec 4 3 Afterall the images in the sequence have been taken run the IRAF focus analysis In IRAF el window Sec 3 1 1 type file number is the running image number appended to the root filename for any image in the focus sequence The IRAF routine will find all the images in the sequence given any one of them This routine Subtracts the initial sky background frame from each focus frame Generates an object catalog for each array at each focus value Uses the catalog to identify round unsaturated sources at each focus setting Derives the FWHM for each of
48. esent to turn the flatfield lamps on and off and to set lamp levels The script includes pauses and advisory prompts for these purposes You must use DomeFlatSequence for all dome flats even if doing flats in only one filter See 8 below has instructions for how to take dome flats and a table of recommended lamp level and exposure time for each filter 26 2 Observation Configuration select the filters for which you want to take flats NumObs is the number of images recorded to disk for each filter typically 10 The GUI expands to allow you to set parameters for each filter selected intTime coadds foamples digAvgs intTime is selected to get a good exposure level 2000 4000 ADU coadds is set to 1 fsamples is set to 1 for broadband filters higher values for narrowband filters digAvgs is set to 4 for broadband and most narrowband filters higher values for extremely narrowband filters The values for fsamples and digAvgs should match the values for on sky science data taken with the same filters Read noise reduction is an issue with narrowband filters hence the higher values for these cases When the script executes first it goes through the filter sequence taking images with the dome lamps off Then it repeats with dome lamps on prompting the user for each filter to set the lamps at the appropriate level Execution will stall if a filter wheel does not position properly and the script must be
49. f coadded integrations taken of the standard star will be NumObs x number of positions in dither pattern x Iterations Figures 17 21 below illustrate these dither geometries Each figure shows the STANDARD script GUI with appropriate entries for a chosen geometry and a drawing of the pattern traced out by the standard star on the NEWFIRM focal plane a1 t aa ae lt samo Fig 17 STANDARD script GUI with SPX dither geometry Four 3 second observations at the comers of a box 400 arcsec RA x 800 arcsec Dec plus a fifth observation at box center starting telescope position Settle time 0 reduces overhead at the expense of some image blur from telescope shake The on sky pattern shown assumes a manual telescope offset NE from the initial telescope pointing before starting the script so the box center position doesn t fall into the gup between arrays 33 Cilia O em Fig 18 STANDARD script GUI with RA x Dec dither geometry With 70 aresee spacing between Points the 3 x 4 dither pattern defines a 140 x 210 aresee box The telescope has been manually offset NE from the initial telescope pointing so all dither positions lie in one quadrant one detector of the focal plane 3 sree gt taco Fig 19 STANDARD script GUI with 40 dither geometry Predetermined offsets put the star in the center of each quadrant of the
50. f telescope motions around the initial pointing with the Dither Configuration Now we introduce a second level of telescope pointings via the Map Configuration Dither motions and Map motions are nested Starting from the initial telescope position the telescope goes to the first pointing specified by the map parameters It executes a dither pattern around that pointing as specified by dither parameters ending back at the first map pointing Then the telescope moves to the second map position and executes a dither pattern and so on In some scripts below Telescope Configuration is used to introduce a third level of motions for observing blank sky positions free of extended emission Yes this is getting complicated Scripts with both Dither Configuration and Map Configuration are used to i observe large contiguous areas and ii to subsequently treat the image set as a whole in the science reduction pipeline There are two types of usage Efficient observation of large areas containing only point or very compact sources also yielding a well defined mean sky image for sky subtraction Observation of extended sources The Map Configuration is used together with Telescope Configuration to define sky pointings chosen to be free of extended nebulosity to observe them intermixed with target pointings and to flag them for definition of background mean sky image in the pipeline reductions Map Configuration
51. focal plane mosaic in succes 34 35 _ corres O lt a a Fig 20fa STANDARD script GUI with Random dither geometry With RA Dec 120 arcsec the specified 11 steps will lie within a 240 x 240 aresee box The telescope has been manually offset NE from the initial telescope pointing so this bax will be in one quadrant one detector of the focal plane 35 36 a i _ 0 aren gt Fig 20 b Enlargement of the sequence of telescope positions generated by the Random seript generator The open circle marks the starting point after manual telescope offset which is also the center of the pattern Integrations are taken at the red positions The first and last dither moves dotted lines are from and then returning to the pattern center 37 o OO oe Fig 21 a STANDARD script GUI with dither pattern From File The first telescope motion moves the star out of the central gap between arrays The last motion recenters it The pattern has been defined so thatthe star position does not repeat any row or column and the separation between images adjacent in time is many tens of arcsec 37 38 File stdpattern txt 123 158 105 70 10 70 140 35 35 140 263 123 Fig 21 b The file of telescope motions that generates the pattern shown in Fig 21 a The figures used above to illustrate dither pattem
52. h each of these descriptions and examine the associated script GUI Each successive script description assumes knowledge gained by reading its predecessors The Instrument Scientist or startup person can also be consulted about which scripts and what parameter values within them are best to use for your observations 6 1 Creating and using scripts All NEWFIRM images are taken by executing a script even test exposures Scripts are created using the script editor NGUI For compatibility with the NEWFIRM science reduction pipeline scripts are limited to the choices defined by buttons in NGUL There are scripts for darks test exposures focus sequences dome flats single telescope pointings including dithers and multi pointing maps Clicking on any of the script selection buttons at the bottom of NGUI launches a script specific GUI divided into subsections called Configurations with user selectable options and user set parameters The button What s This at the bottom of each script GUI pops up a short description of the purpose and operation of the script Above the script selection buttons in NGUI are two rows of options flags that enable or disable certain actions in all scripts The default options have been chosen from observing experience and should only be changed in consultation with your support scientist They will be referred to in the descriptions of individual scripts below Many scripts contain a section for
53. h the Instrument Control System GUI by clicking on the filter button Since it can take up to 60 seconds to position the filter wheels this saves some time at the start of an observing script Filter selection is also commanded automatically from within a script so no harm if you forget Fig 8 Instrument Control System GUI with J filter selected and Environmental Cover open 5 3 Check Dewar temperatures Look at the TCS VDU Sec 4 3 Figure 7 Two critical instrument temperatures are displayed at bottom right These are the temperatures of the internal cold Optical Support Structure OSS and the array detectors These should read OSS 65 0 K ARRAY 30 0 K 18 19 I either of these values does not match the above value contact your observing support person or the Instrument Scientist before proceeding with observations If the array temperature is more than 0 5 K out of range safety features in the software will not allow you to take data Summary Open Environmental Cover via Instrument Control System GUI Check critical Dewar temperatures on telescope status display Return to Table of Contents 2 19 20 6 Data acquisition with scripts Sec 6 1 20 has information and instructions that apply to all scripts Secs 6 2 24 6 3 28 and 6 4 43 describe the observing scripts from simplest to most complex New users are strongly urged to read throug
54. in the repetitive metadata blocks in the script used by the reduction pipeline 6 1 5 To create a script of scripts GUL created scripts can be strung together as a sequence to make observing more efficient You have to create a new script which is just line by line listing of existing script names The resulting script must be in home abserver exec and be executable For example to automatically run a set of dark scripts at various exposure times gather the script names into a list file denominated as a shell script e g DarkSequence sh Execute this script to run all the dark scripts sequentially You can do this in the command line window using your favorite text editor For example 4 vi DarkSequence sh dark5s sh dark30s sh dark100s sh ddark300s sh wq will create the script DarkSequence that runs existing scripts for 5 30 100 and 300 second darks in succession NOTE The telescope needs to be refocused when changing filters and this cannot be done from a script The telescope operator does focus changes So it s inadvisable to create a sequence of scripts in different filters 6 1 6 To download a copy of NGUI New NEWFIRM observers may find it useful to download NGUI to practice creating observing scripts The latest version of NGUI may be downloaded from the NEWFIRM web site www noao edu et newfirm Scroll down the left hand table of contents to Observing with NEWFIRM Click on NOCS Script Generator
55. ions for the upper filter wheel upper row and the lower filter wheel lower row Any commanded filter position corresponds to a pair of wheel positions one in each wheel The filter wheel GUI will probably show a position button lit up green in one row This indicates that it has arrived at the correct position for that wheel It will probably show a position button lit up white in the other row indicating that it has not arrived at the correct position for the other wheel Click on the next bution in sequence after the one lit up in white This wheel will move and attempt to find this position If it succeeds the button for this position will turn green If it fails an error message about failure to find position will show up in the filter wheel GUL If this happens try again If it succeeds you now have one wheel correctly positioned for the requested filter and the other wheel the stuck one in a known but likely incorrect position Once the stuck filter wheel has succeeded in finding a position button lights up in green then try commanding the desired filter position from the bottom row of buttons Ifthe wheel can t figure out where it is and position itself properly after three tries call the Instrument Scientist A34 Can t move filters and cannot ssh connect to nice The Environmental Cover responds to open close command buttons but filter wheels don t move Attempt to ssh to nice A 3 1 or OA reboot pro
56. is happen very rarely Data seem to be coming from the PANS normally but one or both of the nfdhs computers fail to process it One error condition is that only half of the image is displayed the final file is never written to disk and the DHS becomes unresponsive Another error condition is that the messages in the DHS z super window show the processing stopped at Trigger Host and issuing a ditscmd nohs nohs_endobs command from the NOCS command line window does not produce any response in the z super window First check the data disks A 2 6 above to be sure this isn t a full disk problem Lock files are created in the system tmp directory for use by the Shared Memory Cache SMC system to describe how to attach to the current SMC This file is named ump sme501 In rare cases multiple versions of this file have been created of the form Amp sme501_1 that cause confusion about how to attach to the shared memory cache on the machine Recovery consists of deleting all existing lock files then restarting the system a Shut down the data taking system with noes stop all lt er gt b Open a new terminal window on mayall 2 s d C h From this monsoon login on newfirm kp execute the command ndhs dev zzkpvm lt er gt This should remove all lock files From this monsoon login on nevfirm kp verify that the lock files are removed with the commands Neither of these commands should list any fi
57. is is a web page showing a log of the data that the QRP has processed for your run so far Initially it will be empty until you take data for the QRP to process such as dome flats darks or your first nighttime observing sequences You will need to reload refresh this web page in your browser periodically It will be updated as new data are processed and added to the QRP review log Open one xterm window and log in to the newfirm computer via ssh observer newfirm kp From there you can run the plstatus command to check that the pipeline is running properly Open another xterm window and log in to the nfpipe 01 computer via ssh observer njpipe 01 kp kpno noao edu using the same password as for observer newfirm kp From there you can run the dqquery command which provides data quality information for images that the pipeline has processed recently Open an xgterm window and log in to the newfirm computer via ssh X observer newfirm kp Open an image display ximtool or ds9 in the background and then start an IRAF session from the home observer directory by typing cl Change directories to go to the location you specified for the QRP FITS output products Sec 4 1 2 You can then display and otherwise analyze these images as they appear during the course of the night The QRP web page first bullet above provides access to a wealth of information and to the QRP images You will see a log of data that the QR
58. ke DeepSparse this allows two levels of telescope motion Dither and Map which can be used in the same ways Telescope Configuration is now active and allows a third level of motion with the Offset parameters in RA and Dec Telescope Configuration user variables are RA Offset and Dec Offset RA Offset in arcsec specifies how far to move the telescope in RA from the current map position to take sky frames east west Dec Offset does the same in Declination north south The offsets are applied between Map positions during the dither map cycle to move off of an extended target for sky frames This produces a dithered sky frame set for every Map position Note that this is not the same sky pointing every time The offsets to sky mimic the Map pattern offset in RA and Dec The routine begins and ends with a sky position This has proven to be a time inefficient script when used full up for dithering plus mapping plus sky The repeated telescope offsets to sky are time consuming For many purposes they also result in far more sky frames than are necessary DeepRich can be useful for observing an extended source that fits within a single NEWFIRM field of view In this case choose RA x Dec as the Map option and set m n 1 to get a single Map position Point the telescope to the target The telescope will then offset to the sky position specified in Telescope Configuration execute a dither pattern there
59. left Click on Enable in the menu that pops up The buttons across the GUI bottom change from grey text to black text Position the large script creation GUI NGUT in the lower left corner of this monitor The command line window is already in the lower right comer Leave it there The four windows that you will interact with while observing are now readily visible and accessible on the lower left monitor Summary Enter and save project parameters in the small GUT Enter G68 SGU pRO SEE lt EES in terminal window Rearrange GUIs grouping the ones you will be using in the lower left monitor 14 Fig 5 Lower lefi monitor with GUIs and command line window positioned for observing Next look at the upper left monitor It has several small windows stacked up These are engineering status displays with scrolling text They are only used for troubleshooting These can be left as is or minimized to reduce clutter in this monitor 4 3 Other useful tools There are two other useful displays that are not part of the NOCS They may already be opened on one of the four monitors If not start each one by double clicking on its icon found on the lower left monitor Truss Temp telescope temperatures GUI VDU Video Display Unit telescope and instrument status Position these displays in the upper left and upper right monitors for quick visual checks The Truss Temp GUI displays telescope truss p
60. les Restart the system normally from the existing observer newfirm kp login with This will create a single new lock file on each of the nfdhs machines Repeat step e commands from the monsoon newfirm kp login and verify that only one file of the form tmp smcS01 exists on each machine To further verify that the SMC is operating properly from your monsoon login issue these commands On each machine only one entry of type SMCache should be listed Exit from monsoon newfirm kp and close this window A29 Integration times out and nohs nohs_endobs returns Broken Pipe error message something in the DHS Shared Memory has gotten corrupted This error is confusing because the first symptom appears as a timeout error in the Monsoon Supervisor Layer GUI The integration doesn t finish This looks like a Monsoon error Sec A 4 but it isn t The key symptom is that attempting to close the open fits file with command s3 results in an error message Broken pipe in the command line window and no activity in the DHS z super window Also the DHS Supervisor Window in the VNC window to DHS becomes unresponsive to the mouse Recovery consists of halting the system issuing a DHS clear all command then restarting the system a Halt the user interface with noes stop all lt er gt b Open a second terminal window on mayall 2 In this window connect to newfirm kp as user monsoon
61. mp lt er gt to restart the Ximtool Or alternatively Select New Ximtool window from the window manager menu A24 DHS Supervisor Shared Memory Cache Status window suddenly shows a red background Data Collector Agent process DCA has crashed a Check the VNC screen to verify that the z mosdca window has disappeared if not something else has crashed and we have a different problem b Close the pending observation file for the last frame From the command line window issue the command This moves data to DHS memory but not onto the disk Restart the system with d In DHS Supervisor window click on Paths and Files tab and redefine the file root name and directory paths If you want to save the data in memory click Update Status in the Supervisor window to list the exposure ID s ExpID of images in the memory These are numeric e g 245441 1 609138 f Flush these to disk using repetitively if necessary ASNES for example dhsflush 245441 1 609138 2 This should save the data although the file numbers will be lt previous image filenumber gt lt appended letter a b c ete gt h The file numbers will begin incrementing normally when the next script starts i If you don t care to save the data use instead 16 1 j See A 2 5 3 below for more on dhsflush and dhsclean if these commands don t seem to work from the command line window Debiasing and rebiasing the arrays may pro
62. n This guide supplements hands on training at the telescope Recovery procedures from various error conditions are given in Appendix A This guide is site specific to the Mayall 4 m telescope on Kitt Peak Users familiar with NEWFIRM will notice differences in detail for operations at the Mayall or the Blanco 4 m telescope Throughout this guide look for ext highlighted witha yellow background This denotes a command string to be entered by the user followed by the return key denoted as ZEB in the command line window This manual also includes actions to be taken only by OAs or EM personnel These are highlighted in bold blue font These actions usually involve reboots with root permission or resets of electronic hardware Observers should not be doing this 2 Login and Startup The NEWFIRM observer operates the instrument from the workstation mayall 2 near the telescope operator This station has four monitors which the NEWFIRM software regards as one big sereen Log into mayall 2 as 4meter and open a terminal window Login window will be displayed on mayall 2 Login as dmeter Password is also posted on one of the monitors Atdmeter mayall 2 prompt type HOF lt cr gt ignore the disabled message it s not an error condition Open a second terminal window in the lower left monitor This window will be connected to the instrument control computer newfirm and used to issue commands to it So
63. n see note below on power buttons on computer nfdea to cycle power back on and reboot This machine is in the NEWFIRM rack in the 4 m computer room 9 Wait 5 minutes for reboot to finish 10 Inform the Support Team person or observer that he she may continue with troubleshooting or recovery procedures A 74 Reboot nfdhs 01 and nfdhs 02 1 Be sure the observer has halted the data system with user commands debias and noes stop all and done an exit from observer newfirm kp 100 Have the observer exit any other observer newfirm kp accounts he may have opened 3 Open an xterm window on Desktop 3 of mayall 2 Alternatively use the monitor and keyboard provided in the computer room next to the NEWFIRM computer ecting the desired machine with the rack mounted switcher box to nfdhs 01 kp i ssh X newfim nfdhs O1 kp ii pw 5 Log in as root to perform a shutdown i suce ii PW is 6 From root login give shutdown command 1 sbin shutdown h t 0 now T Wait a couple minutes for the ssh connection to drop out 8 Open a second xterm window on mayall 2 and repeat steps 4 7 for nfdhs 02 kp 9 Push front panel power button see note below on power buttons on each computer nfdhs 01 and nflhs 02 to cycle power back on and reboot 10 Wait 5 minutes for reboot to finish 11 Inform the Support Team person or observer that he she may continue with troubleshooting or recovery procedures
64. nd Files Create root image name hit enter Click on raw image directory path hit enter verify path in z super window Click on processed image directory path hit enter verify path Switch to NEWFIRM DHS Supervisor gt Shared Memory Cache Click Continuous Polling box at page bottom to activate this feature 4 2 Setup actions in other monitors The lower left monitor has the terminal window command line window and a large number of GUIs stacked up Fig 2 These need to be rearranged for more convenient 4 2 1 Set and save Project Parameters Locate the small Set Project Parameters GUI Fill out the entries in this GUI and click Save The GUI will close This information is used for data archiving including access permission for the astronomer s on the proposal Actual Observer means the astronomer who is at the telescope doing the observing this is not always the PI Proposal Identifier is your proposal number e g 2008A 0035 This can be found on the online telescope schedule at utp Avww noao edu kpno forms tel_sched VERY IMPORTANT After filling out and saving the Project Parameters GUI type in the command line window This command uses the information entered in the Project Parameters GUI to actually set access permission and synchronize what the NEWFIRM system is producing with what the NOAO Science Archive is expecting from the telescope schedule The Project P
65. nd line window do 88 89 Ask the OA to power cycle the Digital Head Electronics DHE also known as the Monsoon array controller This may require moving the telescope for access to the Cass cage where these electronics reside Instructions to OA EM for DHE power cycling 1 Be sure the observer has halted the data taking system step d above 2 Go to the Cass cage 3 Power off the array controller aka Monsoon DHE at the circuit breaker on the instrument interface panel attached to yellow truss faces Cass cage door 4 Wait 30 seconds then power it back on 5 Inform the observer that he she may continue with the recovery procedure f After the OA has done this restart the system with 2 In the DHS Supervisor window click on the Paths and Files tab and redefine the file root name and directory paths h Take a short single image with a TEST script to see if the system is now running normally Debiasing and rebiasing the arrays may produce higher dark current or other transitory phenomena for tens of minutes Running a set of short integration time junk frames may help stabilize them It s not advisable to take low background data e g narrowband filters immediately after a debias rebias error recovery A44 Monsoon Supervisor Layer window shows a red ERROR message during integration and readout cycle while DHS activity appears normal a problem has occurred with communications to from the array controller
66. ng with a Postproc DONE message d Click on Update Status again to see if the memory is empty if it isn t repeat steps b c for a couple more times to see if this procedure flushes it If no entries are seen in the Shared Memory Cache you have succeeded f If data don t entirely clear go to next procedure A253 Third recovery procedure a Switch to an unused desktop on mayall 2 and open a terminal window b Log in to newfirm kp with 1 o d e 5 D h dD rs 78 Get the exposure ID number of stuck data from the field ExpID in the Shared Memory Cache window The ExpID is numeric e g 245441 1 609138 There will be several lines with the same ExpID In the monsoon login window you just created type GRMGSHZERPIDSZES for example dhsflush 245 This should save the data although the file numbers will be lt previous image filenumber gt lt appended letter a b c etc gt The file numbers will begin incrementing normally when the next script starts Watch for activity in the z super window and the Shared Memory Cache window Click on Update Status again to see if the memory is empty If it isn t repeat steps d and h a couple more times If data still don t entirely clear from monsoon newfirm kp type You will lose the data in memory with the clean procedure Click on Update Status to be sure the memory has cleared If the exposures are still shown as present but with paramet
67. nish b To abort type CNTRL C lt er gt in the command line window During an integration while the countdown is proceeding type CNTRL C lt cr gt then let the integration run to completion with DONE in the Monsoon Supervisor Layer window ii The script will stop when integration has finished but leaves the last FITS file open in the data system iii Issue this command in the command line window to write the last integration to disk and close the FITS file iv Inthe zsuper window within the VNC DHS window check that an image processed to disk v You are ready to take data again A 5 2 Permission denied error message when executing a script for some reason script was created as non executable a Check to see if it permissions are rwx executable or rw non executuble b To make it executable enter ass Should be OK to execute now A523 File not found error message when executing a dither or map pattern from a file text list of offset commands NGUI looked in the wrong place for the file when creating the script a Go to the terminal window in which you created the text file and check the path to the file b Insert the full path plus filename in the GUT in the filename box 93 94 The most common problem is that NGUI was started in home observer but the file was created in the directory where the scripts are stored home observerlexec Return to Error Recovery P
68. ns Computer power buttons are indicated by the symbol Q PAN A and PAN B rack positions 3 4 Button is on the left hand side of the computer front panel next to the DELL logo Symbol is on the button Symbol glows green when power is on nfdhs 01 nfdhs 02 rack p nfdea kp position 7 newfirm kp position 8 Button is in upper right hand corner of the computer front panel Button is a dull red color with the symbol engraved on it Button is slightly recessed may need a pencil ete to push it It is NOT the small button labelled RESET jons 5 nfguider rack position 9 Button is a square grey button in the middle of the computer front panel There are two square grey buttons side by side Power is the one on the right The power symbol is directly above the power button Return to Error Recovery Procedures list 69 Return to Table of Contents 2 102
69. odMapRich will do a blank sky set the first three Map pointings another blank sky set the middle three Map pointings a sky set the last three Map pointings and a sky set to end the script This results in less clock time spent on large telescope motions and a more optimal number of sky frames for definition and subtraction of a mean sky frame compared to DeepRich Consult with the Instrument Scientist or startup support person for setting the parameters of this rather complex script Figure 30 is an example of a ModMapRich script The map on source pink is a 3 x 2 pattern in RA x Dec as in Figs 28 and 29 With parameter SkyMod 3 the script takes an offset sky frame at the start sky 1 executes the three position southern row of the source map in RA takes another offset sky sky 2 executes the second northern row of the source map and finishes with a final offset sky position sky 3 Since the Map offset 1600 and the Sky offset 3600 in Dee there is a 400 arcsec gap in Dec between the source pattern and the southerly sky pointings The script also executes a 5 point dither pattern at each source and sky pointing 48 49 RLEOrE RY COTTON vem ome ivan m Fig 30 Example ModMapRich script GUI top and resulting sky positions bottom Pink is the map on source blue isthe map on sky Return to top of Sec 6 20 Return to Table of Contents 2 49
70. offset back to the target and execute a dither pattern there and then repeat the sky observations The reduction pipeline will use only the sky position images to define and subtract the sky background from the extended source images Figure 29 is an example of a DeepRich script The map on source pink is a 3 x 2 pattern in RA x Dec as in Fig 28 Each source map pointing is accompanied by a pointing on sky blue The sky pointings build up a similar pattern The map offsets are 1600 arcsec in Dec x 2 Dec positions The offset to sky from each source pointing is 3200 N in Dec As a result the pattern on sky is contiguous to the pattern on source in this example 47 semer courcumanion wwe nlai alo slane Fig 29 Example DeepRich script GUI top and resulting sky positions bottom Pink s the map on source blue isthe map on sky 47 48 643 ModMapRich Like DeepRich this allows three levels of telescope motion dithering mapping and offset to a blank sky position However it takes fewer blank sky positions Telescope Configuration new user variable is SkyMod SkyMod sets the frequency of offsets to sky while executing the Map This is best illustrated by examples Fora 2 x 2 raster Map and SkyMod 4 ModMapRich will do a dither pattern at the blank sky offset position all four Map positions and another sky set Fora3 x 3 raster Map and SkyMod 3 M
71. on the chosen Map pattern Telescope Configuration is present but inactive inputs are ignored This script is used for efficient observation of large areas containing only point or very compact sources It also produces a well defined mean sky image for sky subtraction Map parameters are typically set to offset the telescope by large amounts to cover the desired total area Dither parameters are set to do a dither pattern at each map pointing The pipeline reduction uses all images together to define a mean sky image for sky subtraction Figure 28 is an example of a DeepSparse script The map is a rectangular pattern 3 x 2 positions in RA x Dec At each map position a 5 PX dither is executed E W shading is used to denote the individual Map pointings Heavier lines are the array gaps The 1600 arcsec map offsets allow 70 arcsec of overlap between positions for registration into a composite image using stars in common This overlap is also slightly larger than the 60 x 60 arcsec box covered by the dither patterns This dither pattem in turn is chosen to fill in the 35 arcsec wide array gaps 45 SoRPT CONAGURATION THLESCOPETSKY OmNGURATION Fig 28 Example DeepSparse script Top Script GUI entries Middle Large seale Map pattern Bottom Small scale Dither pattern at each Map pointing 45 46 6 4 2 DeepRich Li
72. on are highlighted in blue Hardware procedures are embedded in the troubleshooting instructions Instructions for rebooting the NEWFIRM computers are given separately for each computer in Section A 7 AO Kitchen sink error recovery 72 If symptoms are unclear help is unavailable and or a return to science readiness without diagnosis is desired this sequence of steps will clear up many errors with data acquisition transfer and display A L Global window management errors 73 A L L No windows or GUIs appear after doing nocs start all A L2 All NEWFIRM windows suddenly close A2 DHS errors VNC window to DHS lower right monitor 75 A21 Images don t appear in image directory The IRAF el window in the DHS display has disappeared A23 The Ximtool in the DHS has disappeared or been accidentally closed A24 Shared Memory Cache Status window shows a red background A25 Data are stuck in the DHS shared memory cache A26 Data aren t being written to disk as fits files A27 ntfocus IRAF routine hangs A28 Data transfers appear to freeze A29 Integration times out and nohs nohs_endobs returns Broken Pipe error message 69 70 A3 Instrument control errors GUT in lower left monitor or command line window 84 A3 1 Instrument Controller GUI not seen during instrument startup A3 2 No communication with Environmental Cover or filter wheels A3 3 Filter wheel doesn t reach commanded position A34 n t move filter
73. r mayall 2 prompt type d ssh to observer newfirm kp and type Now all the windows will appear and you can proceed A L2 All NEWFIRM windows suddenly close leaving only the 4merer login prompt to mayall 2 a high level network communications glitch closed all the connections We ve seen this happen once at the Mayall telescope and never at the Blanco The communications dropout appeared to spontaneously recover However this left several NEWFIRM computers in a strange state A L 2 1 First recovery procedure observer restart of PAN computers a Exit from any other observer newfirm kp logins on mayall 2 and mayall 3 b Ask the Observing Assistant to shut down and reboot newfirm kp Log back in to newfirm kp from 4meter mayall 2 ssh X observer newfirm kp usual password d Ask the Observing Assistant to restart the PAN computers from the command line window The commands are VERY IMPORTANT be sure observer has issued the nocs stop all command first Otherwise nuke can have catastrophic consequences If this makes you nervous don t do it Move on to A 1 2 2 below Continue with a normal system start f Reset the filename and directories in the NEWFIRM DHS Supervisor window g Take a short test frame with a TEST script to verify that all appears normal B 4 A 1 2 2 Second recovery procedure complete shutdown and reboot of PAN computers This takes about 15 minutes Try if the fi
74. re with data taking 12 1 Inspection of raw data To the right of mayall 2 is another workstation with multiple screens mayall 3 where you can set up a separate data analysis station This is also the observer s station for the MOSAIC CCD camera Login as password Let the VNC session come up it does so automatically then kill it Open a shell console window At the mosaic prompt open an xgterm In the xgterm window ssh to newfirm kp password Then start ira Once in iraf load the mscred and newfirm packages and open up a ds9 display tool Now you can cd to your data directory and access your data 65 65 66 12 2 Accessing the Quick Reduction Pipeline products The NEWFIRM Quick Reduction Pipeline web pages are under the purview of Rob Swaters At present they may be accessed at tip Avww noao edu staft med mewfirm QRP 1 especially recommend that observers read the section on How to observe in a pipeline friendly manner Questions about the QRP or reporting of problems with the QRP at the telescope should be reported to nfpipe noao edu To access the pipeline products I recommend using the other multisereen workstation in the 4 m control room mayall 3 or you could work from a laptop You need to open several windows in order to connect to or view several processes simultaneously Open a web browser and point it to ttp nfpipe 01 kp kpno noao edw NEWFIRMpipeline observer Th
75. rimary mirror and ambient temperatures Any of these will change from green to red background if that temperature changes by more than 1 degree C since last the observer reset This is useful for tracking temperature changes that affect telescope focus 16 The VDU display shows much useful information about the telescope and instrument see the Figure X Truss temp Saal 7 Mar 26 11 38 31 2014 11 Trs change gt 1 deg 1 3 Click to Resot Fig 6 Truss temperature GUL Trs is the steel telescope truss Mir the primary mirror and Amb ambient air temperature in the dome 18 46 55 3 26 2014 LS D 0 00AM 1 000 OFFSET 0 0 0 0 5 0 00 0 0 Focus 10700 DOME PARAL 360 0 ERI INST NEWFIRM 8 OSS 65 0K a Fig 7 VDU display of telescope and instrument status information Summary Open other tools from icons and position for easy access 16 Return to Table of Contents 2 5 Instrument startup 5 1 Open the Environmental Cover warm dust cover over Dewar window Locate the Instrument Control System GUI in the lower left monitor This GUI has a row of buttons across the bottom with filter names and a second row below this one with two large buttons labeled open and close Click the large button on bottom row labeled open It will turn green when the Environmental Cover is open 5 2 Select filter You can select a desired filter throug
76. rocedures list 69 94 95 A6 Problems with appearance of raw images A6 1 Raw image has identical count in every pixel for example all zero s or all 10000 ADU or all 25000 ADU The array control computers need to be restarted These are the Pixel Acquisition Nodes or PANS AG 1 1 First recovery procedure observer restart of PAN computers a In the terminal window command line window issue this command string mocs stop all lt er gt b Ask the OA to issue the PAN computers clean all command The command is issued from the observer newfirm kp prompt VERY IMPORTANT be sure observer has issued the noes stop all command first Otherwise nuke can have catastrophic consequences If this makes you nervous don t do it Move on to A 6 1 2 Restart the system with the usual commands d Reset the filename and directories in the NEWFIRM DHS Supervisor window Take a short test exposure using a TEST script to verify that all is wel A6 1 2 Second recovery procedure complete shutdown and reboot of PAN computers Try this if the first procedure doesn t restore normal operations This takes 15 minutes a Exit from any other observer newyirm kp logins on mayall 2 and mayall 3 b Ask the Observing Assistant to turn off the Monsoon power in the Cass cage This may require moving the telescope for access to the cage Ask the Observing Assistant to shut down and reboot PAN A and PAN B computers
77. rst procedure doesn t produce normal operation a Exit from any other observer newfirm kp logins on mayall 2 and mayall 3 bb Ask the Observing Assistant to turn off the Monsoon power in the Cass cage This may require moving the telescope for access to the cage Ask the Observing Assistant to shut down and reboot PAN A and PAN B computers This takes about ten minutes See A 7 2 98 Reboot nfpan a and nfpan b Note requirement to turn off the Monsoon power DHE power in the cage first and turn it back on after the reboots d After both have been rebooted ask the Observing Assistant to tum on Monsoon power in the cage Ask the Observing Assistant to shut down and reboot newfirm kp See A 7 1 98 Reboot newfirm kp f Log back in from 4meter mayall 2 to newfirm kp and do a normal system start followed by the normal setup activities in all three monitors Take a short single image with a TEST script to see if the system is now running normally it should be Return to Error Recovery Procedures list 69 mW 75 A 2 Data Handling System DHS errors Error symptoms generally appear in the VNC window to the DHS Some DHS error conditions also produce error messages in the Monsoon Supervisor Layer GUI and or the command line window A21 Images don t appear in image directory probably you forgot to click on the directory path in the NEWFIRM DHS Supervisor window when the DHS desktop was brought up a b
78. s and cannot ssh connect to nice A4 Monsoon Supervisor Layer errors NMSL GUI or command line window 87 A41 Red ERROR messages appear during startup A42 Red ERROR message appears during startup about a timeout error A43 Red ERROR message appears about a voltage out of range A44 Monsoon Supervisor Layer window shows a red ERROR message during integration and readout cycle while DHS activity appears normal A4 5 NMSL task failure error message A4 6 Red ERROR message appears during data taking about a timeout error AS Problems with scripts command line window 93 AS 1 How to abort a seript A5 2 Permission denied error message when executing a script A5 3 File not found error message when executing a dither or map pattern from a file text list of offset commands A6 Problems with appearance of raw images image display in DHS window 95 A6 1 Raw image has identical count in every pixel A6 2 Raw image has negative counts and strange appearance A 6 3 Raw image shows a strong center to edge gradient or a vague shape covering the central part of the field 0 n AT Reboot instructions for NEWFIRM computers 98 ATL Reboot newfirm kp boot nfpan a kp and nfpan b kp Note power off and power on of DHE in the cage AT Reboot nfdca kp A74 Reboot nf hs 01 and nfdhs 02 A 75 Reboot NEWFIRM Instrument Controller aka NICC AS Locations of rack mounted NEWFIRM computer power buttons 102 n
79. s show these from the point of view of the detector where on the focal plane mosaic does a single point in the sky in the examples a standard star lie at any point in the pattern For science targets which may be extended sources or a defined area of sky with many small sources itis useful to ask what the coverage looks like on the sky how is a given area of sky being sampled First consider an extended source of moderate size observed with a 4Q pattern Fig 22 shows this from the detector point of view the source moves from quadrant to quadrant Fig 22 Integrations on an extended source with a 4Q dither pattern from the detector point of view The source moves around the 2x2 mosaic of individual detector arrays i 2 1 2 2 1 2 1 Fig 23 Integrations on an extended source with a 4Q dither patter in terms of sampling on sky Red dashed lines show the actual 2x2 detector mosaic The pattern on the sky is 3x3 with quadrant sized pieces of sky observed 1 2 or 4 times as indicated The fixed offsets of the 40 pattern are defined so that the gaps between arrays are filled n the final composite image Fig 23 shows the sky coverage of this pattern Total sky coverage is 2X the single pointing FOV by area but the time spent at any given point on the sky and resulting SNR have pronounced spatial variations The adv
80. shuts down all processes and closes all GUIs Alternatively click on Stop Newirm in the Newfirm startup GUL Afr all the GUIs have closed in the command line window type exit lt c gt This closes the connection observer newfirm kp and takes you back to the Ameter mayall 2 prompt From here you can log out of mayall 2 if you like Summary Close button in Instrument Control System GUI OO Newfirm in GUI Return to Table of Contents 2 4 Setup steps after login and startup 4 1 Setup actions in the DHS VNC window This VNC window could be moved to any of the four monitors The lower right one is convenient for most users 4 1 1 Start IRAF Locate the toolbar of icons across the bottom Left click on the CL icon the one furthest to the right in the toolbar This opens a menu Click on Restore in this menu This opens an xterm with IRAF running in it set to the mscred package This is called the IRAF cl window This is primarily useful for image display to the DHS ximtool with msedisplay and interactive image querying with mscexamine tools The msc prefix denotes IRAF tasks that operate on multiextension FITS files originally developed for the Mosaic optical imager 4 1 2 Change directories to your data directory The startup process automatically sets the image root name and creates a directory for raw and processed images In the IRAF el window just opened the path
81. t the integration countdown is proceeding Then follow these steps i During integration while the countdown is proceeding type CNTRL C lt cr gt in the command line window 23 ii The window will display instructions about how to proceed These are the same as what is written here below iii Let the integration run to completion until you see the message DONE in the Monsoon Supervisor Layer window iv The script will stop when integration has finished but it leaves the last FITS file open in the data system This has to be closed vy Issue this command in the command line window to close the last FITS file vi Inthe z super window within the VNC DHS window check that the last image was written to disk vii You are ready to take data again 6 14 To inspect or edit a script A script can be inspected in the command line window or in a separate xterm window logged in as observer newfirm kp using your favorite text editor This capability can be used to check input parameters near the start of the script If the pattern of telescope positions on the sky is not what you expected you can scroll through the script and check the sequence of offset commands to see how your input was translated into telescope motions Scripts are text files so a text editor can also be used to change parameters without creating a new script via a GUI However using the GUI is likely to be faster and less prone to error or to omissions
82. tation in the control room restart the guider per separate Guider Setup Instructions Return to Error Recovery Procedures list 69 86 87 AA Monsoon Supervisor Layer array controller errors A4 1 Red ERROR messages appear in the Monsoon Supervisor Layer window while noes start all process is running need to run this window process again a Wait for the loading process to stop and observer newfirm kp prompt to appear in shell window b In the command line window type ARREZ that s en em ess ell Reset Process will run again in the Monsoon Supervisor Layer window d Watch for red ERROR messages Repeat this cycle until the process runs to completion with no red ERROR messages If they appear at all usually one or two nmslReset cycles are sufficient to get through this A42 Red ERROR message appears in the Monsoon Supervisor Layer window upon startup about a timeout error the NOCS may have reconnected to a ghost memory page and gotten confused about how many PAN computers it s talking to Recovery procedure is In the command line window terminal window type grep PANS home monsoon log lt er gt A normal response would be something like 2454753 6885068 165138364 mslSuperConfg status OL gt PANS 2 The abnormal response is something like 44245475 8555645765736699 mslSuperConfg status OL gt PANS 4 Where 47 rather than 2 indicates the confusion If you see an abnormal response
83. the course of an observing sequence the telescope motions exceed a particular hardwired tolerance currently 8 arcmin but this may change the QRP will break the sequence into multiple independent stacks that will all appear in the 2nd column of thumbnail images FITS data products When processing is complete for a sequence FITS files of the combined image the cks will be delivered to the directory that you specified for processed images see Sec 4 1 2 above The names of the files start with NFOR for NEWFIRM Quick Reduce FITS files are delivered only for on sky observations not for dome flats or dark calibrations Processed image filenames are long and opaque The key is to look for an image sequence number This will be the first image in a data sequence a set of dithered 67 68 exposures for example This can help you trace a processed image file back to the raw image sequence from which it originated The QRP products are not archived so if you want to save them move them to your local hard drive or otherwise retrieve and store them before your run is over The QRP log and associated data products will automatically reset for the next observer Return to Table of Contents 2 68 69 Appendix A Error recovery procedures Here is a quick functional symptomatic index to the error recovery procedures In each section the more common problems are presented first Procedures requiring OA EM interventi
84. up person about the instrument and telescope configuration in which to take Dark frames If in doubt take darks with the Environmental Cover closed telescope mirror covers closed and the dome darkened Script Configuration user variables are Object Name and Script Name Object Name is the title keyword in the image FITS header It should be short descriptive and meaningful to you and that observation Use only alphanumeric characters and the space bar Script Name should be something short that readily identifies the script e g DarkSs for a five second dark integration Observation Configuration user variable is NumObs NumObs is the number of images to be recorded to disk The Filter selection in this Configuration is fixed as Dark Monsoon Configuration user variables are intTime coadds fSamples and digAvgs These define parameters for array operations in the Monsoon array controller Since dark frames are usually taken for dark subtraction from on sky data the user entered parameters under Monsoon Configuration should match the on sky data configuration intTime is the basic integration time the amount of time that will be spent collecting photons before the arrays are read out coadds is the number of integrations to take in sequence and sum into a single image which will be written to disk A coadded image is written as a sum not an average Samples is the number of Fowler samples to
85. us experience and or advice of your setup person For the broadband filters darks taken at 10 15 20 30 and 60 seconds would cover the normal range of integration times on the sky Set coadds to match your planned science observation protocol For example 15 sec integration x 4 coadds to get 60 sec total exposure on source before dithering Close the Environmental Cover in the Instrument Control GUI to take dark frames Consult with your Instrument Scientist about the advisability of further closing the telescope and turning off lights in the dome 54 55 Figure 38 Instrument Control GUI showing configuration for dark frames Use the Dark script to take these data Take at least three frames at each combination of integration time coadds Fowler samples and digital averages that you plan to use To automatically run a set of dark scripts at various exposure times gather the script names into a list file denominated as a shell script e g DoDarks sh Execute this script to run all the dark sequences see also Sec 6 1 5 8 3 A typical calibration plan Atypical calibration plan for a program doing both broadband and narrowband imaging would be this Dome flats in J H and Ks filters using a DomeFlatSequence script with lamp settings and integration times from Table 1 above Dome flats in 1644 1 64 jum Fell 2122 2 12 um Hz and 2168 2 17 pm Br Y filters using a separate DomeFlatSequence script
86. utes See A 7 5 98 Reboot NEWFIRM Instrument Controller aka NICC A33 Filter wheel doesn t reach commanded position One of the filter wheels failed to sense its position Ifthe filter positioning was part of a script the script will halt and generate an error message This is potentially very serious The filter wheels are inside the Dewar and not accessible for adjustment or repair on the telescope AB 3 1 First recovery procedure Command the filter wheels to the desired filter position from the Instrument Control System GUL Click the button for the desired filter This will take a minute or two The wheel with sensor failure goes into a recovery mode in which it creeps along slowly repeatedly querying its position sensors until it finds a known position and detent Then it completes the desired motion Numerous status messages appear in the GUI message window If you still get a failed to reach filter position error try this procedure one more time It may take the wheel a couple of tries to figure out where it is s4 85 A3 3 2 Second recovery procedure Command the filter wheels individually from the NICC Filter Wheel GUI on Desktop 2 Go to Desktop 2 From the menu for NICC GUIs click on Filter Wheels near the top of the menu A GUI will open that has a row of filter position buttons J H K etc Above them will be two rows of buttons labelled 1 through 8 These are the physical wheel detent posit
87. vel without notification to clean up the directory from time to 2 2 time If you wish to save scripts for future use create a subdirectory with your name e g PProbstScripts and copy your scripts into this directory OBSERVATIONCONFIGUNATION ane od Jeke ke ke ka ja WoROON CONFATION fetes in niin SS oj tmnes a ale ala E e mel Saale ee C3 T mn nea I Fig 10 Simple script GUI showing Object Name and Script Name entries top and OK What s This and Cancel buttons bottom 6 1 2 To execute a script After creating a script and clicking OK in the script GUI type rehash in the command line window To execute the script make sure that you are in the home observer exec directory and type its name as scriptname sh in the command line window Sipassi for example Dark5s sh lt er gt If a script does not execute command line window returns an error message check to see that You are in the correct directory A script exists with that name The script is an executable file 6 1 3 To abort a script in progress This has to be done carefully or the Data Handling System will get tied up requiring an extensive recovery process F the script will run to completion in a minute or two just let it finish To abort the command must be issued while an integration is in process Examine the Monsoon Supervisor Layer GUI upper right monitor and verify tha
88. vided in the computer room next to the NEWFIRM computer rack selecting the desired machine with the rack mounted switcher box 4 Connect to newfirm kp ssh X newfirm newfirm kp i pw 5 Log in as root to perfor iii su lt er gt iv PW is __ 6 From root login give shutdown command sbin shutdown h t 0 now 7 Wait a couple minutes for the ssh connection to drop out 8 Push front panel power button see note below on power buttons on computer newfirm to cycle power back on and reboot This machine is in the NEWFIRM rack in the 4 m computer room 9 Wait 5 minutes for reboot to finish 10 Inform the Support Team person or observer that he she may continue troubleshooting or recovery procedures down Reboot nfpan a and nfpan b 1 Be sure the observer has halted the data system with user commands debias and noes stop all and done an exit from observer newfirm kp 2 Have the observer exit any other observer newfirm kp accounts he may have opened 3 Go down to the Cass cage Power off the array controller aka Monsoon DHE at the circuit breaker on the instrument interface panel attached to truss faces Cass cage door 4 Open an xterm window on Desktop 3 of mayall 2 Alternatively use the monitor and keyboard provided in the computer room next to the NEWFIRM computer Tack selecting the desired machine with the rack mounted switcher box 5 Connect to PAN A i

Download Pdf Manuals

Related Search

Related Contents

W38L100-CHA1 .cdr  9106 9102 AK Ins TBD new  取扱説明書 S-P3 保証期間 1 ヵ年    2 - 取扱説明書ダウンロード  Istruzioni d`uso Bilance contapezzi KERN CXB/CXP  取扱説明書 / INSTRUCTION MANUAL  平成21年 新春号 - 中村健一郎税理士事務所  

Copyright © All rights reserved. Failed to retrieve file