Basic Guide to Operating the LBC

Please note: This document is replaced by the Basic Guide to Operating the LBC. This page should eventually be deleted.

Starting up LBC

The LBC Control system operates within a firefox browser. To bring it up, log into one of the workstations in the control room (lbtmu03, lbtmu04 or lbtmu05; _lbtmu01 is reserved for the telescope operator) and either click the LBC Control system icon on the top panel or start a firefox browser and open the URL http://lbccontrol or http://10.144.0.21 (note this link is internal to LBT only)

Click "Enter" and a new window will come up. This user interface (UI) has six displays, one for each of six sets of tasks or information: "House Keeping", "Power Control", "OB Execution", "Active Optics", "Log Analyzer" and "Info". Click on the icon above the name of the page you wish to display.

The UI comes up with the "House Keeping" page displayed. This has plots of the temperature of the thermal link inside the dewar (should be ~125-150 K and can vary with telescope position) and vacuum (should be from 2.e-5 -1.e-6 mbar). LBC will automatically shut down if the vacuum or temperature readings exceed safe values.

hk.jpeg

Switch to the "Power Control" page: To start up LBC follow the three steps below. Note the order, and wait until one step has finished before starting the next. Also insure that the "Use it!" boxes are checked - otherwise nothing will happen...
  • Turn on Housekeeping: Status will change from "enabled" (off) to "initializing" and finally to "ready" (on) after a few minutes.
  • Turn on Other Systems (camera & shutter, filters, rotator, trackers): If the temperature is below freezing, click the "warm-up" button about 30 minutes before doing this. This turns on the power to the technical and science chip controllers but does not initialize them. After clicking "Turn On", status of each of the 4 systems, first rotator and filters, then trackers and finally the camera, should change from "enabled" (means "off") to "initializing" and finally to "ready" (means "on"). If one of these ends in an "error" or "warning" state, see the troubleshooting section. Usually turning off and back on all systems will clear the error.
  • Connect to LBT: This connection to the Instrument Interface (IIF) of the TCS should be nearly instantaneous, but if it is taking a while, even after a 2nd click, see the troubleshooting section.

power.jpeg

Note that the "Scan Filters" button is not used in start-up and almost never used (don't even think about it). See the troubleshooting section for an explanation of what it does.

top

Executing an OB

The OB execution page is divided into 8 sections. The "OB Operations" control is at the upper left, "OB Info" is displayed in the upper right; the middle left shows the "OB Progress", and middle right the guiding offsets (rotator guiding is not enabled but RA and DEC offsets will be shown). The middle shows weather information. Finally along the bottom: the left panel displays telescope information and the right two panels display the status of the blue and red channels.

obexec.jpeg

Loading the OB: To load an OB, click on the "Browse..." button in the "OB Operations" section. Select the OB you wish to run and click the downward pointing arrow to upload it. Or, to create an OB, click on the "Create an OB" button within the same section. A new window will pop up and after you have finished filling in the fields in it, click the downward pointing arrow at the bottom of it to upload it to the OB execution page. This newly created OB will be called "fast". The "Create an OB" functionality is discussed in detail in the ObPreparation page.

When the OB is uploaded, the downward pointing arrow in the "OB Operations" section should change from black to grey and the "play", "pause" and "stop" buttons (right-pointing, double vertical dashes, box) should turn from grey to black. The OB name, target name, type and coordinates will be displayed in the "OB Info" section. If the observation should be carried out at the current telescope pointing, then current will be written instead of the coordinates. See the ObPreparation page for information on how to indicate that the current pointing should be used.

Review Settings: Before executing the OB, review the settings for Telescope, Guiding, Channel, Mirror and Scaling.
  • A check mark to the left of Telescope indicates that:
    • the telescope will preset (slew) to the coordinates in the OB (or remain where it is if the Coords are listed as current),
      • even if "Preset LBT" is indicated in OB, telescope will not slew unless "Telescope" is checked in user interface
    • track the target at the sidereal rate according to the trajectory calculated by PCS and
    • that the LBC rotator will track (rotate) according to the trajectory calculated by PCS.
      • if "Rotator Tracking" is not checked in OB, but Telescope is checked, rotator will track
  • If Guiding is checked, then guiding offsets (corrections to tracking) measured from the technical chip images will be sent to the telescope and used. Leaving it unchecked will not prevent technical chip images from being taken and offsets computed, only they will not be sent to the telescope.
  • A check mark to the left of Channel indicates whether to use the Blue, Red or both channels; and
  • A check mark to the left of Mirror indicates that the instrument will be able to command adjustments to the primary mirror position (necessary for focussing).
  • Insure that the exposure time scaling factor is what you want (normally 1.00).

Executing the OB: To execute the OB, click the "play" button. First the telescope slews to the desired coordinates, then the mirror is adjusted for focus and collimation. During this time (seconds to a minute or two), the status written below whichever channel was checked will remain "ready". Once the mirror is in position and the filter and rotator are being moved, the status is shown as "preset/read/save". Look at the Log Analyzer page if you wish to find out just what is happening at the time. The OB will run to completion unless the "stop" or "pause" buttons are pressed.

log.jpeg

Stopping an OB: Pressing the "stop" button will interrupt the current exposure or readout and will NOT save it. Pressing "play" will resume the OB, and retake this exposure. Importing the OB again before pressing play will start the OB from the beginning.

Pausing an OB: Pressing the "pause" button will cause the OB to pause after the current exposure finishes. It will be written to disk. Pressing the "pause" button a second time will resume by taking the next exposure in the OB.

top

Looking at the Data

  • Repository: Images, both from the science array and from the technical chips, are sent to and stored in the LBC Archive and they are also sent to a disk which is mounted on the four Linux workstations in the LBT control room, lbtmu03, lbtmu05, lbtmu04 and lbtmu01 (though lbtmu01 is the telescope operators' workstation). The observer, logged into either of the 3 others, can find the images in /Repository (the Repository). A set of quick-look & reduction tasks as well as tasks useful for target acquisition and setup are in the IRAF packages LBTreduce and LBTtools. These are described in LbcIrafPackage.

  • quicklook or Science: Images will be automatically displayed in ds9 as they arrive in the Repository when the script /home/lbcobs/supportscripts/Observerscripts/quicklook.py is running. Use IRAF imexam to interact with the display. If the script aborts with an error, or does not display the image, stop and restart it. If there are still problems, note the messages and let us know. Use /home/lbcobs/supportscripts/Observerscripts/Science.py (aliased as Science) to display only science array images and not technical chip images.

  • quickfile: A cousin of quicklook.py, /home/lbcobs/supportscripts/Observerscripts/quickfile.py (aliased as quickfile) looks at /Repository once per second and tells you when a new file has arrived. Small images such as extrafocal images and test flats arrive in /Repository within 5 sec of the end of readout, while full images take about 25 seconds from the end of readout.

  • Image Quality plots (also under development) On the mountain machines, there is a c-shell script sstat1.csh in /home/lbcobs/ImQual which, given a list of images will run Source Extractor on these and produce catalogs. Following is the procedure for how to run this script and plot the results:
    • cd /home/lbcobs/ImQual
    • make a listing of the files you wish to examine. This should have only the filename (no path), e.g. lbcb.20061222.012345.fits (make sure to include the .fits). Usually these will be a series of images in a single filter.
    • run the script: ./sstat1.csh /path/to/images filename where the full path to the images should be specified (e.g. /Repository/) and filename is the name of the file created in the previous step. The script then runs through each extension of each image in the list and creates a catalog (one per image) and a coordinates file for each extension. The coordinates files are ready for input to tvmark and useful only to see which stars it selected. There is an IRAF script mstars.cl in the same directory which can be run on the imagelist to display and run tvmark on these one-by-one (only for a single chip at a time) Type task mstars.cl = /home/lbcobs/ImQual/mstars.cl to load it and then mstars @filename 2 where 2 is the chip in this example. The default files (default.sex.save and default.param.save) are in this directory and set up specifically to find only the brighter stars (saturated ones are discarded in the plotting program). The threshold is set to 300. The parameter file designates the output data and hence also the number of lines of comments at the top of the output catalog. If this is changed, the script will no longer work as is, but must be changed to skip fewer or more lines. If you plan to change it, please save a copy of the original first.
    • Next plot the results using the SM macro ellipt.mac
      • sm X11
      • macro read ellipt.mac
      • hist lbcb.20061222.012345.cat red 100 (generally, hist catalogname color snum where the catalog name is the name of the *cat output file you wish to view, color is an plot color (white, blue, red, cyan, magenta, green, yellow) and snum controls the shading of the histogram (1000 is sparse, 100 more densely shaded).

top

Active Optics: Focus and Collimation

At the start of the night, after each slew, and periodically while tracking a target, the telescope focus and collimation must be adjusted. The procedure for doing this is as follows:
  • Obtain an extra-focal image, with the primary mirror -0.8 mm below focus, using a subregion of chip 2. Generic OBs for this purpose are in /home/lbcobs/OBlib and are called ?fastextra.ob where ?=U,B or V; these use the current pointing (through the dec=-90 convention). Best results are obtained by measuring the wavefront with Vfastextra.ob rather than at the wavelength of your science filter. See ObPreparation#ObLib and the README file in /home/sdtobs/OBlib for more information. OBs with target coordinates, but the same instrument configuration as ?fastextra.ob, can be made using the program mk_fpia_ob. At the prompt, type mk_fpia_ob . This will create a new focus OB called target_OB_FPIA.ob More information on generating OBs for focus and collimation can be found on the OB Preparation page at ObPreparation#AoObs. See the ImageGallery for examples of pupil images.
  • Run the IDL program lbcfpia to analyze this image To run up lbcfpia at the start of the night:
    • Open a new xterm window.
    • Start IDL by typing idl
    1. Type lbcfpia. This will pick up the latest fits image in /Repository, display it and analyze it. The output will be a set of Zernikes: Z4 (defocus) Z5 (astig) Z6 (astig) Z7 (coma) Z8 (coma) and Z11 (spherical). followed by the prompt: "Do you want to send?" Note that anything other than "n" will send the corrections to the mirror. Wait until the IDL prompt returns before taking another image.
      • If the focus is way off, lbcfpia will only send an offset in Z4 (focus). This is often the case that the start of the night. Send the focus correction and take another extra-focal image.
      • When you send the results to the TCS* by saying "y", you will see several lines of messages. When the prompt returns, the mirror is done moving and you are ready to play your next OB. Occassionally you will see divide by zero errors that can be ignored.
    2. Repeat these steps until corrections to be sent are within the atmospheric errors (~100-450 nm). In good conditions, 1-2 iterations may be required. In poor seeing or otherwise bad conditions, convergence may be slow without manual intervention. The support astronomer or telescope operator should be familiar with this procedure. See the ByEyeAO page for guidance on the type, magnitude and sign of the corrections to be entered to the PSF.
  • If the seeing is poor, lbcfpia may not converge on spherical aberration. Lowering the gain (default=1.0) for sperical aberration in either lbcfpia or PSF will make the convergence more stable. Often for seeing greater than 1 arcsecond, it is sufficient to rely on the telescope collimation lookup table. Ask the operator to "Clear Active Optics" to remove any residual active optics corrections (this clears both mirror shape and collimation corrections).
  • If there has been a significant temperature change, then it the mirror may take a while to reach equilibrium and a radial temperature gradient may appear. See the ImageGallery for examples of the pupil images that result when there is a radial temperature gradient. Often a manual adjustment to spherical aberration (Z11) will be needed to bring the pupil into the regime where lbcfpia can work. Ask the operator to add +500-1000 nm of Z11.

There are also OBs in the ObPreparation#ObLib to take a series of 5 images, between which the mirror Z position is moved by 0.1mm (?focus.ob) or 0.04 (?finefoc.ob). These are not used often, because they take alot of time. Mainly they are used to measure or check filter focus offsets. The procedure for obtaining these data and measuring best focus from them is as follows:
  • Select and load the OB for the appropriate filter and focus step (coarse or fine). The OBs are set up to pass through best focus on the 3rd of 5 images.
  • Run the ?focus.ob or ?finefoc.ob OB. This will take the 5 5-second images.
  • To analyze the results:
    • Load the LBTreduce package (see LbcIrafPackage) which will load the foclist task.
    • epar foclist and enter the UT date, name of output file, number of images in focus sequence, extension number (enter 2) and first letter of filter.
    • Run foclist. It will instruct you to run the IRAF task starfocus (in the nmisc package) and provide a command line.
    • Cut/paste the starfocus command line to run starfocus. This will fit a parabola to the fwhm vs focus based on stars that you select interactively.

Shutting Down LBC

To shut down LBC at the end of the night, follow the steps below. Note the order, and wait until one step has finished before starting the next:
  • Turn off Other systems: Camera, Filters, Rotator, Trackers status will revert from "Ready" (on) to "Enabled" (off).
  • Disconnect from LBT
Leave Housekeeping on and the temperature/vacuum plots displayed. If the instrument is losing vacuum or it is planned to warm it up, be sure to shut it down properly (the camera controllers must be shut down if the vacuum exceeds > 5 x 10-4 mbar). LBC should always be powered down cleanly. If the UI or LBC server program crash when the power is on, LBC should be restarted and then powered on/off in the normal fashion.

top

Getting your Data

  • From the Repository:
    • The images are sent to the LBC archive immediately after readout. From there a copy is sent to the user workstation, lbtmu0{3,4,5} in the directory /Repository/.
    • At the end of the night or the end of the run, the images can be copied directly from /Repository to a storage medium: portable hard disk or DVDs.
    • Presently (22-Dec-2006) this is the preferred way to get your data.
  • From the LBC archive on the mountain:
    • open a firefox web browser to the URL: http://lbcarchive (http://10.144.0.23) (this link is not available from outside the LBT)
    • In the LOGIN box, enter “common” (if OBSERVER is left blank in the OB), but otherwise enter “lbtsdt” to view and retrieve science images, and enter “tech” to view and retrieve images from both technical chips. Password for both is blank.
    • Click “Reload” - the top link of the page that is displayed.
    • Then click either “view last image” or “view image list” to see the last image taken or the entire list. In the latter case, click the dot to the left of the image name to select and view that image.
    • Once the image is displayed, you can download it to /tmp/ by clicking on the image itself. This will display the image in ds9, using the script: “/home/lbcobs/supportscripts/mosaic-ds9.sh” and download the FITS file to /tmp/common.last-n.fits or /tmp/tech.last-n.fits where the n is a running index number that increments as you download image by image. The first image downloaded to /tmp will not have the “-n” and if you reload an image it will contain the string “reload” rather than “last” in its filename. If the image does not appear in the archive interface, check that it is on the archive machine (ssh root@lbcarchive and look in /data/store/full/{common or tech}/). If so, then check the headers. We have seen that images with odd values in the FITS header values, (e.g. airmass = \xB1inf as occurs when the value of the airmass reported on the OB execution page is unreasonable – as seen when the telescope is horizon-pointing) are written to the archive machine but do not appear via the interface.
  • To retrieve images directly from the archive machine: scp from the archive machine: scp root@lbcarchive:/data/store/full/common or scp root@lbcarchive:/data/store/full/tech for the tech chip images (passwd is written in the LBC memo by R. Speziali that is in the black LBC binder in the control room.)

top

Troubleshooting

Specific errors:

  • Image does not appear in lbcarchive or Repository. There are 2 flavors of this error:
    • The full image may not have been downloaded to the windows PC. This happens mostly when taking a series of biases. 1-2 of the series of 5 are not saved to the lbcarchive and hence are not transferred to the Repository. In the Log Analyzer, there will be a red error message which says:
      2006/12/21 23:44:48.812000 E B CAMERA   CCDCTRL  too few downloaded pixels (42434560 instead of 42467328) [ccdctrl.c:2168]
      
      We believe this problem was cured with new CCD controller firmware in May 2007. See Issuetrak #375. This happens mostly during bias sequences and very rarely during the observing.
    • The LBC archive may be down. In the LBC log you will see warning messages such as:
      2007/01/25 01:12:10.159000 W B CAMERA   CCDCTRL  issuing >tftp -b32768 -o 192.168.0.201 put lbcb.20070125.005307.fits d:\lbcb.20070125.005307.fits > nul 2>&1< upload command failed with retcode:6
      Check whether the two LBC archive machines: lbcarchive (lbcdh2) and lbcarchive2 (lbcdh1) are up and running by typing from your sdt account ping lbcdh1 or ping lbcdh2. If one or the other is not responding, you may need to reboot it. In the computer room to the right of the operator, the LBC archive machines are on the left and labelled. Press the reset button. Ask the operator or call software or observing support in case of any doubts. After the machine is up and running, the data should be automatically transferred and should appear in the Repository. See Issuetrak #454 for a recent occurence. If the machine does not come back up, you can still take data, but some LBC configuration files will need to be edited to store the data to the CMU from which you will need to manually transfer them via _scp. You should call observer support for help with this - some notes appear in the EngInfo section below.

  • Image is slow to appear in /Repository: See the discussion about quickfile above.

  • LBC User Interface Hangs: This happens sometimes and it comes in several flavors. Sometimes the UI browser seems to be hung waiting for the CMU, while other times the UI seems to be refreshing without updating the important parameters related to the present observation. This problem is documented in Issuetrak #180. The following actions are recommended to restore full functionality:
    1. Try changing from the OB Execution page to the Power Control page and back. This will not interupt the on-going observation.
    2. Try simply waiting until the end of the on-going exposure. Often the UI will recover when the image begins to readout.
    3. Try killing your Firefox browser window on the local workstation and starting a new one. This will not interupt the on-going observation.
    4. If LBC loses contact with TCS during a preset, it will wait 5 minutes before timing out. Always wait at least 5 minutes before proceeding to more drastic action. You should also spend the 5 minutes trying to understand why the TCS failed to respond (e.g. there could also be a TCS shared memory problem or other telescope issue).
    5. Have a root user log into CMU and restart the web server with: apachectl restart , and then restart your Firefox browser again. Your observation could still be running.
    6. Have a root user log into CMU and run the lbckill.sh and lbcstart.sh scripts in /home/lbccontrol/ (see below). This requires going through the full power-up cycle (10 minutes) on the LBC power control webpage. Any observation in progress will be lost.
    7. Have a root user log into CMU and reboot (see detailed instructions below). This also requires going through the full power-up cycle on the LBC power control webpage.

  • Main Mirror Fails to Focus: When the temperature gets down to -5 degC, the SX Primary Mirror gets a bit sticky in its hardpoint motions and sometimes fails to meet the position tolerance. When you get the LBC pop-up window that says "Left Main Mirror Fails to Focus", the observer should click "OK" in the LBC pop-up window and allow the OB to continue. Then it should be verfied on the PSF GUI that the mirror is only out of tolerance by 10 microns or so. Compare the "total collimation" XYZ with the "platform position" XYZ. (remember that XYZ are in millimeters and RXRYRZ are in arcsec) This check is to make sure that you aren't observing with some actual mirror support failure which would give similar symptoms except with position errors of hundreds of microns or more. This problem is documented in Issuetrak #383.

  • LBC User Interface goes black & white: In Firefox, pull down the Tools menu, and click on Clear Private data incl. cookies.

  • LBC spontaneous Shutdowns: Suddenly the status of LBC reverts from "ready" to "enabled". Issuetrak #298 reports one such instance when there was a CCD Controller error. The only thing to do is restart LBC. Do not leave LBC in this state as the Windows machines and CCD Controllers are in an unknown state. Even if you are at the end of the night, LBC should be turned on and back off again. A close relative of this failure is the "hinibit" (inhibit) failure where the LBC safety software shuts down the camera when it fails to read the dewar temperature or vacuum 10 times in a row.

  • CCD Controller Error: LBC occasionally reports a CCD Controller Error on start-up. In the Log Analyzer you will see:
    2006/12/25  08:29:36.874827  TRACKERS	 	CCD Controller error [src/trackers/trackers.c:799]
    2006/12/25  08:29:36.921000  TRACKERS	CCDCTRL  "PCI board PCI-A300 not found (or driver not installed)" controller error [ccdctrl.c:612] 
    This means that the CCD controller firmware was unable load into the controller. The normal fix is to power off the camera and start it again. This could also indicate a bad fiber connection between the Windows machine and the controller. This is sometimes seen if the cryostat has recently been remounted. The fiber errors typically come in two flavors:
    • This one is typical if the transmit/receive fibers to the CCD controller are swapped or disconnected.
      2007/09/15 10:20:56.568 TRACKERS    CCD Controller error[src/trackers/trackers.c:795]
      2007/09/15 10:20:56.569 TRACKERS    CCDCTRL "PCI_A300 RX link not work" (code:10) controller error [ccdctrl.c:676]
      
    • This one is typical if the fiber pairs have been swapped between the Science and Tech controllers. It also occurs if there are other hardware based communication issues with the CCD controllers. See Issuetrak #839 The communication issues are usually solved by cycling the camera power.
      2007/09/15 10:42:15.996 TRACKERS   CCD Controller error[src/trackers/trackers.c:795]
      2007/09/15 10:42:16.000 TRACKERS   CCDCTRL   load LCA+on SPC-A200 error (SPC-A200 status register read fail) [ccdctrl.c:696]
      

  • CCD Communication Error: Sometimes an executing OB will crash with a CCD communication error, Issuetrak #380. In the Log Analyzer you will see:
    2006/12/24 02:17:31.462463 CAMERA RPC Client/Host communication failed retval:4 [src/camera/camera.c:1056]
    Rebooting the CMU seems to cure this problem every time. An analogous problem can occur with the Tracker controller, and has the same solution [nb: stopping/restarting the LBC control system using lbckill.sh/lbcstart.sh often accomplishes the same fix without having to reboot the CMU proper]

  • Profibus Error: Profibus is the serial bus protocol used to communicate between the CMU and the various encoders on the rotator and filter wheels in the camera. If the protocol gets confused, you will see in the Log Analyzer:
    2006/12/25  08:16:43.825447  ROTATOR 	 Encoder problems with profibus encoder reading [src/rotator/rotator.c:432]
    2006/12/25  08:16:43.815566  FILTERS 	    problems with profibus encoder/s reading [src/filters/filters.c:514]
    
    The most reliable fix to this problem is to reboot the CMU in order to reset all the serial port parameters.

  • Profibus Error: Another Profibus error is an indication of a hardware problem. If you get this error, it likely means that the little white ribbon cable inside the encoder for the filter wheel has come loose. This sometimes happens during cryostat mounting when the encoder cap has been moved. The solution is to plug the cable back in and restart the camera. Rebooting the CMU may be required.
    2007/03/10 03:47:15.355606  FILTERS         Profibus board 1 device 10: equipment status 33
    2007/03/10 03:47:15.355680  FILTERS         problems with profibus encoder/s reading [src/filters/filters.c:1219]
    

  • Shutter gets stuck open: This problem first arose in Dec-2006 Issuetrak #352 and is apparent in the image which will have vertical trails in the +y direction from all of the stars cause by the readout clocking and fainter trails in the -y direction from the preflush clocking. (depending on image scaling, you may only see these trails for the brightest stars). An example image is included in the #ImageGallery. Be sure that the shutter test switch on the cryostat is positioned to "Auto". The faulty rolling ball bearing in the shutter was replaced on 10-Jan-07, so we do not expect this problem to recurr. Reinitializing the shutter by cycling the power may help the situation. This may be accomplished through the testpower program (support staff can run this) or by simply shutting down LBC and restarting it via the user interface (#ShutDown).

  • Filter wheel gets stuck: This hardware problem arose in Jan-20007 and has been repaired. Issuetrak #470 provides the details with the following key messages in the Log Analyzer:
    2007/01/28 13:58:37.782936 E B FILTERS WHEEL#2 timeout error on stop [src/filters/filters.c:2187]
    2007/01/28 13:58:37.993040 FILTERS WHEEL#2 hardware failure [src/filters/filters.c:1099]
    Powering down the camera and rebooting the CMU would be a logical debugging step, although this was a hardware problem. Note that a much milder problem with a similar message happens when the wheel fails to converge in 10 iterations.
    2007/02/10 10:48:34.507203 FILTERS WHEEL#1 filter is more than 10 times the maximum allowed distance from it nominal position [src/filters/filters.c:2136]
    This can often be recovered by simply pressing the Play button to start the OB again.

General instructions and information:

Rebooting the CMU:
  • Turn off the “other systems”, then “housekeeping”, then “disconnect LBT”.
  • Reboot the CMU:
    • “ssh root@lbccontrol” (login to CMU)
    • “who” (in order to verify that there are no other users or developers logged in)
    • “reboot” (It will take about 3 minutes for the full reboot. When the UI returns to life, proceed.)
  • Turn back on the “housekeeping”, “other systems”, then “connect LBT”.
Be aware that after rebooting the CMU, the user interface may come up with exposure time scaling factor (OB execution page) equal to 0.0 rather than 1.0.

The Commands lbckill.sh, lbcstart.sh and lbcexit.sh
  • Each time LBC control program is started up in the CMU, a pair of daemon processes is started. /home/lbccontrol/lbckill.sh kills the daemon processes and /home/lbccontrol/lbcstart.sh starts them.
  • Use lbckill.sh followed by lbcstart.sh when it is necessary to stop the LBC program running on the CMU, but a full reboot of the CMU is not needed, e.g. if you want to switch to the engineering software or stop & restart the software after a modification.
  • lbcexit.sh is a softer stop than lbckill. lbcexit.sh was once used to successfully recover from a hung user interface but it is very infrequently used.

top

Engineering Information (Observers should not need this)

Focus offsets
Focus offsets between the filters based on data taken 28-Jan-07
Filter offset[mm] from V
V-Bessel 0
B-Bessel 0.05
U-Bessel 0.32
r-SLOAN 0.07
g-SLOAN 0.08

Focus offsets between the filters based on data taken 17-Mar-07 (Better measurements, but their V-zeropoint may be less consistent.) We believe that the B filter was inverted during cleaning on 9-Mar-07.
Filter offset[mm] from V
V-Bessel 0.00
B-Bessel 0.072
U-Bessel 0.32
r-SLOAN 0.033
g-SLOAN 0.044
SDT_Uspec 0.486, updated to 0.450 on 11-May-07

Focus offsets between the filters based on data taken 10-Jun-07. These had the V-zeropoint adjusted on 9-Jun-07. (JMH)
Filter offset[mm] from V
V-Bessel 0.000
B-Bessel 0.053
U-Bessel 0.353
r-SLOAN 0.086
g-SLOAN 0.083
SDT_Uspec 0.454

Changes in the filter offsets need to be updated in:
  • LbcUserManual (above)
  • fastextra and finefoc OBs in OBlib
  • /home/lbccontrol/conf/leftchannelfocus.dat
  • mk_fpia_ob

Platescales
Reported by Vincenzo Testa 14-May-2007

Filter Platescale (arcsec/pixel)
V-Bessel 0.2253
B-Bessel 0.2253
U-Bessel 0.2255
r-SLOAN 0.2254
g-SLOAN 0.2254
SDT_Uspec 0.2255

Guiding parameters
Guiding parameters are set on the lbccontrol computer (root access only). Best results were obtained for a minimum exposure time on the guide CCDs of 16sec and azimuth and elevation gain factors = 1.0. The guide.dat file currently contains the lines (07-Feb-07):
TypicalExposureTime=8 (was 16 in Dec 06)
MaximumExposureTime=32
MinGuidingCorrection=0.01
MaxTrackingErrorPerSec=0.25
MinimumTrackingStars=1
AzTrackingGain=1.0
ElTrackingGain=1.0
Note that the image taken immediately after this file is edited and saved uses previous values; the changes are made only for the subsequent image and onwards. LBC does not need to be run down and back up to accept changes made to guide.dat. We shortened the TypicalExposureTime to 8sec in late December in order to better fight the problem of the elevation jumps.

The MinimumScientificExposureTime on the science array for which guiding with the technical chips will be activated is a parameter set in the bluechannel.dat file (root access only). This is currently (21-Dec-06) 24 seconds. LBC must be run down and back up to accept changes made to the bluechannel.dat file.

Directions of Dither Offsets
Measured at position angle = 0: Dither offsets in the +X direction move the star image right (+X) on Science chips 1-3, and down (-Y) on Tech chip 1. Dither offsets in the +Y direction moves the image down (-Y) on Science chips 1-3, and left (-X) on Tech chip 1. Tech chip 1 is right of science chip 1.

Pixels are 13.5 microns, and the blue plate scale is 0.227 arcsec per pixel. The dither offset onto Tech chip 1 is +860, 0 arcsec from the field center.

Rotator center
The center of rotation for LBC-Blue was last measured on 12-Dec-06 and is, on chip 2, including the 50 pre-scan pixels (measured from a raw, untrimmed image with 2304 columns): (x,y) = (1048.5, 2957.8). An updated value measured from data taken on 16-Sep-07 is: (x,y) = (1083,2951).

Rotator velocity scale
A scale factor in rotator.dat will adjust the rotator velocity relative to the trajectory calculated by PCS. This should be 1.00.

Reconfiguring LBC to store data locally rather than transfer images to the archive.
If the LBC archive is down and does not come back up after a reboot, as a last resort, you can take data and store it on the the CMU. You will need to login to the CMU and edit bluechannel.conf and lbc.conf. In bluechannel.conf, look for Camera.Upload.Address and Trackers.Upload.Address. These normally are set to 192.128.0.201 to send data to the LBC archive. To send data instead to the CMU:/images/tftp change the "201" to "1". In lbc.conf, look for UploadAddress. Again, change the IP address from 192.128.0.201 to 192.128.0.1. Remember to change these back once the LBC archive machine is up and running again!!!

Pre-requisites for collecting LBC pointing model data

0) Have the operator reset the Az and El encoders before the pointing run.

1) You need to have a good collimation lookup table for the primary mirror(s). You do not want to run active optics much in the middle of the pointing data. That could change the pointing, adding to the rms residuals of the model.

2) You need to turn on the mount logging on PCS Pointing page. It logs continuously once on, every 2.5 seconds, so all you need to do is turn it off at the end of the night and copy the logfile to somewhere.

3) Take at least one pair of images rotated 180 degrees apart to re-determine the rotator center. It is better to take the rotated pair off meridian.

4) You really only need to save data from chip2. That will also make it easier to download the data to Dave's laptop. The field of view is large enough that we can always do a decent WCS on the available stars. I need at least ~21 well-distributed fields to make a "low-order" pointing model (just mount and tube flexure). Use 10s exposures to make sure there are a couple of valid mount snapshots during the exposure. You can point to anywhere on the sky (it does not have to be on pointing stars), I'll derive the PM for a virtual star on the detector.

top

-- OlgaKuhn - 19 Dec 2006; -- JohnHill - 22 Sep 2007

  • obexec.jpeg:
    obexec.jpeg
Topic revision: r39 - 01 Aug 2012, KelleeSummers
This site is powered by FoswikiCopyright © by the contributing authors. All material on this collaboration platform is the property of the contributing authors.
Ideas, requests, problems regarding Foswiki? Send feedback