SMT Optical

Emergency Info
Systems Status
Observer Info
Science at ARO
ARO In The News
Internal Docs
Tucson Info
Photo Gallery
Search ARO


SMT Optical Pointing


System Overview

Block Diagram

Focusing the Camera

Preparing for Focusing

Starting the Focusing Process

Starting the Optical Pointing Run

Finishing up a Pointing Run


Starting the Full Frame Viewer

Searching for the Missing Stars

Serial Port Troubles

Appendix A




The radio pointing model contains 14 pointing parameters. Only six of these are relatively stable and are usually fixed to standard values. The remaining eight must be measured periodically. It would be best to measure these with the radio receivers, but there are relatively few strong pointing calibrators and it takes a least half a day to get a good enough sky distribution to get a reliable fit to the pointing parameters. If one doesn't have to solve for all the pointing parameters, one doesn't need as many points to get a good fit nor as uniform a sky distribution.

The optical telescope pointing is described by the same pointing model but some of the pointing parameters have different values than the radio receivers'. The zero points (constant terms independent of azimuth and elevation) for example are different, as they depend upon the mounting of the optical telescope relative to the axis of the radio telescope. And the two flexure terms are different since there is internal flexure in the optical telescope and the optical telescope pointing is not affected by the flexure that moves the radio telescope's sub-reflector off the axis of its primary.

The optical telescope however can measure the Azimuth encoder zero point "NulA", the orthogonally of the azimuth and elevation axes "Perp", and the tilts of the azimuth axis relative to the zenith "IncN" and "IncW". And the optical telescope pointing can measure the six stable pointing parameters much more accurately than the radio telescope pointing as these are small (about 2 arcsec or less) and a 14 parameter fit requires many more points than can usually be acquired with the radio receivers.


System Overview:

Like most all observing procedures done with this control system, optical pointing involves many processes running on many different computers. Here is a list of the main processes involved in optical pointing.

        OPD:	The optical frame grabber daemon. Runs on kueppers.
        OPB:	The data display program that fits the results. Displays its results on the X-Display smtoxdpy2. Runs on kueppers.
        Optical;	The "backend" control programs. Coordinates the data taking with the rest of the control system. Writes its results on file-server smtotux. Runs on smtocont.
        Exec:	Coordinates data taking with the main control system. Runs on smtocont.
        Servo;	Closes the servo control loop. Takes commands from exec and optical. Runs on smtots.
        Focuser:	Interface to the camera focuser. Runs on smtoxdpy2.
        Disk File:	Data stored on fileserver smtotux.

Block Diagram:

Below is a simple block diagram of the optical pointing system.



Focusing the Camera:


The optical cameras focus drifts with temperature so every time you use it you must refocus. He is a brief history of the micro focuser.

The new Fine Focus Controller was installed on November 3rd. 2003. So far as the software is concerned, it has a range of travel of about 8 inches but that is really 0.8 inches as there is a 10 to 1 gear box on the encoder. The controller mechanism isn't stiff enough to keep the focuser from drifting to the stops when the power is off and the telescope is at the zenith, especially if the subreflector is nutated. There is a "clutch" that is supposed to stiffen it up enough to hold it in place, but even tightening it hard with fingers isn't sufficient. We've been loath to use channel locks as that could wear out the mechanism prematurely when one was focusing the telescope against that tension!

The work-around to this drifting is to move the fine focus to 8.00 just prior to turning off the focuser power. That will put the focuser near its outer stop. (The power should definitely be turned off when the optical telescope is not in use as it is possible for the electronics to decide to move the focus indefinitely. This ruined the previous focuser.) When the power is restored, one is then free to move the focuser between 0 and 8.0 inches. (Actually the range of travel might be slightly less than 8.0 so one probably shouldn't go all the way to zero. This doesn't damage anything but if it hits the inner stop, the zero point changes.)

The following combinations gave a good focus on 3 Nov 03 with the primary unclamped.


Coarse Focus Step Fine Focus Position
-300 8.00
-1000 7.00
-1700 6.00
-2400 5.00
-2500 *4.00  
-3200 3.00
-3500 2.00
-1700 6.00

*>One of the two must slip around -2500/4.00, but it must have slipped back for the last 6.00 to agree with the first one.

The focus was then set back to 4.00 inches and the mirror was clamped (Buffalo finger tight). After clamping the best fine focus changed to 5.80 inches.  (It was 6.3 at first then something shifted as we moved from the zenith to the horizon. After that, the best focus was 5.80.) Temp = -5 C.

See History of focuses in Appendix A.

Preparing for Focusing.


The following procedure should help you focus the optical camera.


  1. Remove the optical telescope lens cap cover.
  2. If you haven’t already, select the “optical” receiver from the “receivers” screen in Rambo. This sets up the correct pointing model to be used in tracker. Be advised that you may have to reset or renter different Az and El pointing corrections than the ones that are restored when selecting the “optical” receiver.
  3. Turn on the little black box labeled “JMI Smart Focus 232”, located behind the rightmost monitor of smtoxdpy2. The power switch is the little white slide one. Slide it to the right.
  4. Start the JMI focuser program on smtoxdpy2. You will find it in the “background” menu by right clicking on the screen background and selecting “Optical Focuser”. NOTE: This program can only run on xdpy2 as that is where the serial port line is connected.
  5. Once the focuser program has started, select the ‘Operate’ tab. The top two right status fields should indicate good Activity and Status, provided the serial port communications has been established correctly. See Section on troubleshooting for more information. You will also notice that the position of the focus should be around 8.000”, the normal ‘parked’ position. See the section below on closing down the optical pointing session.  Click the [Go To Specified Position] button to manually set the focus. Inside this new window, enter the latest value from the table at the end of this manual, and press [OK]. You should see some activity on the red LED on the JMI focuser box.  NOTE: The actual focus position varies throughout the year and is mostly affected by temperature. At the time of this writing, it was approximately 2.8”. Please enter any new focus values in the table provided at the end of this manual.
  6. Start 'opd' from Xcontrol: This is the frame grabber control program and occasionally gets a segmentation fault and crashes. You may need to restart it, but first you might need to find the 'opb' display, program displayed on smtoxdpy2 and kill it as it seems to hang on to the socket port.
  7. Start the ‘Starfinder’ program from Xcontrol.
  8. Once the program is up, press the (Star Finder) button to bring up the star field display.
  9. Locate a star in the field that is red, 1st magnitude, or orange, 2nd magnitude, and located about 360º Azimuth and 45º Elevation. Click on it and a circle should be drawn around it. If you miss, try again.
  10. Once you have selected a star, then you can press the (Seek Selected) button to command the telescope to this star.



Starting the Focusing Process.

Once the telescope has settled on this star, see steps above, we need to create a reference frame in the frame grabber. A reference frame is subtracted away from the data frame and thus subtracts out the noise of the CCD and the background sky. This action results in a much higher signal to noise ratio.


  1. Open a terminal window and type the following: sock OPOINT<cr>. This window will be used to talk directly to the OPD daemon during the following steps.
  2. Press the  button on “Servo Control”. This will cause the telescope to stop tracking. The reason for this is to let the star you were tracking drift out of the field of view so we have an empty field. Watch the status display and when the azimuth tracking error is greater then 10:00’, type the following in the terminal window: ‘start –r –t 100’. This instructs the OPD daemon to start a frame acquisition in the reference buffer, -r, with a duration of 100 msecs, -t 100. The time for the reference and signal frames must be the same or the subtraction will not come out correctly. After this, though, you do not have to give a time specification as OPD remembers the ‘-t 100’.
  3. After the OPD daemon reports that it has finished the reference frame, something that will happen almost instantly, you are ready to start focusing.
  4. Turn the telescope tracking back on to allow the star to return to the field of view.
  5. Once the tracking error has returned to zero, type the following in the terminal window: ‘start’. This instructs the OPD daemon to take the 100 msec signal frame which OPD will then subtract the previously taken reference frame. It will then search for a star and if it finds one will fit a Gaussian and plot the results in the OPB display on the right most screen of smtoxdpy2. These two figures are representative of a good focus, , and a poorly focused camera, .
  6. If the focus results are bad, then you will have to increase or decrease the focus value in the JMI interface by using the (Step In) or (Step Out) buttons on the JMI interface . That will cause the focus to move the amount in the “Step Increment” setting. Remember that the focuser can physically only move 0.8” but the software in the JMI focuser indicated 10 times that amount. So, in order to move that focus 0.100” you need to increase or decrease the focus by 1.0”. The maximum range of the focuser is 0.0” – 0.8”. But, the JMI interface will indicate 0.0” – 8.0”. There is no way to tell which direction you should move the focus. Trial and error are your best bet.
  7. Repeat steps 5 and 6 until the results look more like the left figure in step 5.
  8. Once you have acquired a satisfactory focus, please email the date, pointing, temperature  and focus values to Tom Folkers and Bill Peters so the data may be recorded in the table in appendix A.
  9. Referring to the figures in step 5 above, notice in the upper right corner of each figure you will see a green dot inside the grid. That represents the location of the star within the full CCD frame. We want the star to be near the middle, like the left figure, not the right one. The lower half of the OPB frame, contains some analysis details two of which are the frame az: and frame el: offset of the star from the middle of the frame. In this example the corrections are 13.7" and 15.3". You need to add these amounts to the telescope pointing to bring the star into the center of the frame. Take another frame after you add these minor corrections to make sure you got the signs correct. When you are done, the star should be within 10" of the center of the frame. Email the date, pointing, temperature and focus values to Tom Folkers and Bill Peters so the data may be recorded in the table in appendix A.
  10. You probably want to leave the JMI focuser interface running during the entire optical pointing run as it will remind you to shut down the focuser when finished. Do NOT turn off the power to the JMI interface as that holds the focus in place while doing an optical pointing run.
  11. You are now ready to start the optical pointing run.


Starting the Optical Pointing Run.


  1. Start the 'optical' program from Xcontrol.  It runs in a separate window and displays on "smtoxdpy1:0.0".
  2. Under utilities on Rambo, select the (Optical Pointing) screen.
  3. Press the (Start Optical Pointing Run) button to start an optical pointing run.
  4. The ‘exec’ will initiate a pointing run that takes about 2 hours to complete.
  5. If during a optical pointing run, the pointing run seems stuck and the ‘optical’ window says waiting for a TS ‘ack’ and it got one, then just type ' ni <cr>'. That should get things going again.
  6. The results should go into /home/data1/optical/results.`date`. Be sure to rename any result files to different names before starting another as it will append the next run to the end of the last, unless that's OK.


Finishing up a Pointing Run:


Please take the following steps when you’re done doing optical pointing.


  1. Make sure you email the date, pointing, temperature and focus values to Tom Folkers and Bill Peters so the data may be recorded in the table in appendix A.
  2. ‘Park’ the focuser at 8.00”. Wait until the JMI interface reports that the focus is truly at 8.00 or there about. NOTE: the new focuser unit installed is slower then the electronic expects so you most likely will receive a time out error on long focus moves. But, if you wait long enough, the interface will finally report the true position. During the move the popup window will stay up until it times out. The JMI box will start reporting back in when it reaches its destination. You just need to wait.
  3. Turn the power off of the JMI Black Box by sliding the switch to the left.
  4. Kill the JMI interface by clicking on the top rightmost [X] button on its frame.
  5. Replace the lens cap on the optical telescope.
  6. Under the ‘Utilities’ menu in Rambo select the ‘Receivers’ window and reselect a radio receiver.
  7. Select OPD, Starfinder and Optical in Xcontrol and press [Stop] to end these programs.
  8. Type Control-c in the "sock' window, then Control-d to exit that window.
  9. Notify Bill Peters that you did an optical pointing run.



Q: What if I don’t see a star in the steps above.

A: This can be caused by many things.


  1. Forgot to select the ‘optical’ receiver?
  2. Didn’t remove the lens cover.
  3. The pointing is too far off.
  4. The OPD programs may not be able to pick out the star because it’s too out of focus. 

The first two are relatively easy to fix, the last two, not so easy.

The first thing you will want to do is get a good look at the full frame as seen by the frame grabber. To do this, you will need to do the following:


Starting the Full Frame Viewer:


  1. Open a terminal window.

  2. Type: “cd to /home/opdata/raw”  in the window.

  3. Type: “./makelink &” ( This starts up the automatic symbolic link generator)

  4. Type: “xv –poll newest &” (This will start up the full screen frame viewer.


Searching for the Missing Stars:


If you still have that ‘sock OPD’ session open that you used in section Starting the Focusing Process, step 1 then you can type “start” again to command the OPD to take another frame. After the frame is complete, OPD will write out the frame into the /home/opdata/raw directory. That makelinks process you started above will notice the new file and create a new symbolic link called ‘newest’ and then the xv process will notice the new file, that’s why we included the ‘–poll’ switch, and then it will display the new frame. If the telescope is pointed correctly but out of focus, then you will most likely see a large, faint doughnut instead of a point of light. You just need to adjust the focus in or out to bring the starlight into focus.

If you still do not see anything, doughnut or otherwise, then you might be off in pointing.  Make sure you have selected the optical receiver and removed the camera lens cover. Then you need to increase or decrease first the Az then El pointing in some sort of grid fashion. After the telescope settles down with each pointing change, type ‘start’ again inside that sock window. Wait for the frame to finish and for xv to repaint the frame before trying another grid position. Repeat until you find the star.

NOTE: The optical frame is only 9' 48” by 6’ 32” so limit your pointing changes to 1' of arc.

If all else fails try a brighter star or a bright planet. If you are really striking out, try the Moon!

If doing bright planets or the Moon remember to lower the frame time to 10 msecs to avoid saturation of the CCD chip. (see section on Starting the Focusing Process above, about creating a reference and signal frame)

Serial Port Troubles:

This section still to be written...



Appendix A.

Focuser Position Log:


Temp C.


Az Correction

El Correction



5.0 C




First Measurement.


-10.0 C






0.0 C




Doors where close before start of run.


0.0 C




Refocused after doors opened 1 hour.






Installed new pointing model.












Spring Added to CCD 






Refocus and re-center












Focus & pointing jumped.






Spring attached to camera grille.












































































































































 Copyright Arizona Radio Observatory.
For problems or questions regarding this web contact [tfolkers{at}email{dot}arizona{dot}edu].
Last updated: 11/08/11.