The SST Adaptive Optics System A User Guide

Transcription

1 The SST Adaptive Optics System A User Guide April 8, 2005

2 Contents 1 Introduction 3 2 Optical Setup Mirror part Wavefront Sensor Optics Correlation Tracker Optics Science Optics Running and Calibrating the AO System Starting the System Interface Modes Major buttons Interface Options Checking and maintaining wavefront sensor alignment Calibrating the system Wavefront calibration Troubleshooting Control matrix Troubleshooting Offset voltages Offset Voltage Recovery Dark frames Troubleshooting

3 3.10 Flat fields Window Recovery Running AO Updating Telescope Focus Troubleshooting General AO Troubleshooting AO Configuration File Reference Running the Correlation Tracker Starting The Tracker Running The Tracker Taking darkfield calibrations Taking flatfield calibrations Mirror Calibration Running Cross Correlation Tracker Options Menu Window Menu System Menu Tracker Troubleshooting Configuration File Reference

4 News for year 2005 Changes since the 2004 observing season are the following: The correlation tracker has been upgraded to a run on a dual Xeon processor machine running LINUX with real time extensions. This gives improved performance while allowing for a hopefully more user friendly interface. The tracker DACs have been replaced doing away with the DAC reset switch. The tracker now uses the full field of view of the dalsa camera, thus doing away with the need to use the Physik Instruments translation stage. A new stepping mode has been added in software to allow spectrograph scanning. See the camera manual for information on using this mode. The tracker optics have been replaced leading to better image quality at the tracker ccd. It is also a shorter setup. The AO computer crashed during the winter and has been replaced. No major upgrade has been made though. 1 Introduction This is an introduction to and user guide for the Swedish Solar Telescope (SST) Adaptive Optics (AO) system. The main parts of the AO system are a Shack Hartmann (SH) wavefront sensor, a servo and a deformable mirror. The SH sensor consists of a microlens-array with 37 hexagonal microlenses within an image of the pupil plane, making images of a secondary, corrected focus on a Dalsa CAD6 CCD with a read-out rate of 955 Hz. The system is running on a dual AMD Athlon processor machine operating under Linux with RTAI extentions. The 37-electrode bimorph deformable mirror was constructed by AOptix, Inc. 1 The electrode pattern shown in Fig. 1 is deposited upon each of two thin, sandwiched plates of piezo electric material that are glued to the mirror. Voltage applied to an electrode expands one of the plates and contracts the other, which makes the bimorph bend

5 Figure 1: The electrode configuration of the deformable mirror with the microlenses superimposed. 2 Optical Setup This description is based on the setup used at the SST in 2002 and shown in Fig Mirror part The field lens reimages the telescope 1-meter pupil onto the deformable mirror (DM) via the tip-tilt mirror and the re-imaging lens (L1). The present re-imaging lens is a 55mm diameter cemented apochromatic triplet which, because of its short focal length, provides a nearly perfectly flat focal curve throughout the visible part of the spectrum. The tip-tilt and adaptive mirrors fold the beam from nearly vertical to horizontal, or approximately by 90. This is arranged such that the tip-tilt mirror deflects the beam 60 and the adaptive mirror by 30. With an angle of incidence of 15 on the adaptive mirror, the pupil image is nearly circular (since cos 15 = 0.97). The re-imaging lens magnifies the image by approximately 2.2 times, producing an image scale of arcsec per pixel on our Kodak Megaplus 1.6, 4.2 and 6.3 science CCDs. 4

6 . Dalsa camera : wavefront sensor (AO) DM TTM side view field lens (L1) (prime focus) Dalsa camera : correlation tracker L5 deformable mirror (DM) tip tilt mirror (TTM) Science filter (487.7 nm) re imaging lens (L2) L micro lens array re imaging lens (L3) secondary focal plane (FS3) phase diversity beam splitter dichroic beam splitter transmission > 525 nm Figure 2: Optical setup. Measurements in mm. G band filter (430.5 nm) continuum filter (436.4 nm) optical table (3 m x 1.25 m). 2.2 Wavefront Sensor Optics A field stop, FS3, at the secondary focus defines the field of view, so that the microlens images do not overlap. The field lens, L3, re-images the DM onto the microlens array. The microlenses re-image the secondary focus onto the detector, a Dalsa CCD. 2.3 Correlation Tracker Optics Need notes about the new optical setup... The tracker uses a 400 mm collimator followed by a 50 mm re-imaging lens to de-magnify the science image by a factor of 8 2. The field of view of the correlation tracker is smaller than the field of view of the science CCD s and the correlation tracker CCD is therefore mounted on a Physik Instrument 3 x-y translation stage to allow the tracker to operate on any part of the telescope field of view. 2 Measured??? 3 5

7 2.4 Science Optics A dichroic beam-splitter reflects wavelengths greater than approximately 470nm back into the AO and tracker optics. A beamsplitter is directs half of the reflected light onto a third science track. The blue light transmitted through the dichroic is available for science cameras. 3 Running and Calibrating the AO System 3.1 Starting the System 1. Log in to royac28 as user obs 2. Use the AO icon on the toolbar to start the AO system. Figure 3: The AO graphical user interface main window. 3.2 Interface Modes There are two 4 modes selectable through the Mode menu, observer, and recovery. Observer (figure 3) mode is the mode that should normally (and hopefully always) be used. Recovery 4 Actually three modes if the program is started with argument sys 6

8 Figure 4: The AO recovery mode main window. (figure 4) mode provides a few options which are meant to allow the user to recover the system if one cannot get a good control matrix or the window positions become messed up. Recovery mode options Wavefront Calibration, Control Matrix, and Offset Voltages are exactly the same as those in observer mode except that a centroid algorithm is used instead of the usual cross correlation. This provides a larger lock range but will only work on a point source. So one should only ever use recovery mode with the pinhole in the primary focus. The centroiding algorithm will not work at all with the sun as target. Window Recovery mode is described in section Major buttons Start: Prepare to start the action given by the selected Modes button. Pop up the image display. The actions are actually started by moving the mouse pointer into the display and hitting the return key. Stop: Stops the action prepared for with the start button or started by also moving the mouse pointer into the display and hitting the return key. load defaults: This is actually an option Load default setting, valid for the present optical setup. These values cannot be modified by the observer and can be used to restore correct settings, in case a bad choice of parameters were saved. Load: Load previously stored settings. From the file cfg/ao.default. 7

9 Save: Save current settings. To the file cfg/ao.default. Exit: Exit program. Copy Data: Copy servo statistics data etc. to a directory where it will not be overwritten. The directory name is made up of today s date and the current time and is created under /AO/AO DATA/save/. Display Gamma Changing gamma changes the contrast of the displayed image only, but has no effect on the performance of the system. A gamma of about 0.7 (low contrast) can be used to see the relatively dark pinhole images during calibrations, a gamma of 1-3 is more adequate for solar fine structure. 8

10 3.3 Interface Options Save: Save current settings. To the file cfg/ao.default. Save As: Save current settings to a user defined filename. Load Settings: Load settings from a user defined filename, a file dialogue will appear allowing the observer to select the settings to load. Load Defaults: Load system defaults. This will load settings system defaults. It can be uses to reset options if you suspect that the settings have become corrupted. Quit: Exit program. The following options, shown in figures 5 and 6, should not be modified and the corresponding instructions can be ignored by normal observers. Reverse Sign X Reverse the sign of the WFS measured x term. This may change if the optical setup is changed. Reverse Sign Y Reverse the sign of the WFS measured y term. This may change if the optical setup is changed. Figure 5: The AO options menu. Light Normalisation Select whether of not to normalise the intensity of the WFS subimages before processing. We believe light normalisation improves the performance of the system but we have limited experience from running the system with this option enabled and do not reccomend it. Servo Parameters Servo gains, should not be modified by the observer. DA Limits The maximum DA count to apply to the mirror. Reference Update How often to update the reference subimage. We have sofar used 15s. Longer time intervals may improve frame selection but the downside is evolution of solar structures that will degrade performance of the cross-correlation algorithm. Figure 6: The AO debug options menu. Save images If selected, the last 10s of WFS images are saved to disk. The observer should never run with this mode selected since it takes a long time to copy the images in the running system which could degrade performance, and it takes a long time to write the images to disk after each run. 9

11 Use Mirror If not selected then no output voltages are made to the mirror. Mirror Feedback If the system is run without tip/tilt correction and mirror feedback selected then the system will measure tip/tilt from the output voltages and feed them back into the servo. This compensates for tip/tilt modes building up in the mirror voltages over time and thus driving the system out of lock range which can happen during the course of a day s observing if not selected. Run Time The number of seconds to run the system, if zero the system will run forever. Focus Z to mm Constant factor to be used to convert a measured WF mode into mm on the optical table. This is used to report the focus error to the observer. 3.4 Checking and maintaining wavefront sensor alignment Proper alignment of the wavefront sensor is crucial for successfully operating the AO system. Minor drifts due to heating of optics and slight movements of the vertical vacuum tube relative to the optical tabe will occur at some level during the course of one or several days of observations. Here is what to watch out for and how to fine tune the alignment of the wavefront sensor. 1. Start up the live image of the correlation trackerto ensure that the tip-tilt mirror is in its home position. 2. Remove any pinhole from the setup and insert the neutral density filters in the wavefront sensor beam. Make sure that the ND filters are perpendicular to the beam, if they are at an angle, this will cause misalignment. 3. Next watch the live display, paying particular attention to the red, large reference box enclosing the reference image. This reference image must be precisely aligned with the reference box, else the cross-correlations (which are made with all the smaller 16x16 pixel sub-images relative to this 24x24 pixel reference image) will degrade. If the red-box is misaligned in the x or y directions, modify the X offset and/or Y offset values in the menu, stopping and restarting the live display after each change, until alignment is as good as possible. 4. Check that the entire reference box is illuminated by the corresponding sub-image. If the reference image appears to be slightly too small for that box, make minor adjustment of the left-right and up-down adjustment screws of the field stop to maximize the size of the reference image while not causing overlap of any of the sub-images (overlap of the subimages is indicated by strong increase of intensity at the boundaries of the sub-images). 10

12 This adjustment should not be needed unless something has been bumped. Also, do not use the adjustment screws of the field stop to move the sub images up-down or leftright since any large or repeated small such movements will cause overall degradation of the optical setup. The adjustment screws of the field stop should only be used to adjust the size, not position, of the reference image. 5. The illumination of the pinhole images is controlled by the field lens that re-images the DM pupil onto the micro-lenses. By displacing the field lens in the x or y directions, the pupil image is moved across the micro-lens images. For small movements, this is most easily seen as intensity changes of the darker surrounding sub-images (corresponding to microlenses that are close to the edge of the pupil and only partly illuminated) that are not used by the AO system. When the field lens is in the proper location, i.e. centered on the micro lens array, the surrounding darker sub-images will all have the same intensity. Set gamma to about 1.2 and adjust the field lens until all surrounding sub-images appear equally bright. Note: This misalignment is possibly due to drift from heating of the dichroic mirror or its mount. We believe that this mounting has been improved in 2004 and should now be more stable but the observer should watch out for this misalignment. Also note, that only very small adjustments of the field lens must be made. If large adjustments are made, this will cause the pupil to skip over to adjacent microlenses, such that the sub-images within the red boxes are no longer illuminated. Never compensate such a mistake by moving the red boxes with the X and Y offsets on the menue, always correct by re-adjusting the field lens. 6. Remove the ND filters and insert the pinhole marked AO at prime focus. The image of the pinhole should appear right at the center of the red boxes marking the sub-images. If there is a systematic displacement of the pinhole images relative to the red boxes, use the dials on the physik instruments box to move the pinhole images by changing the offset voltages on the tip-tilt mirror. Note that for proper calibration of the control matrix and for proper focusing of the science CCD cameras, the pinhole images must be well centered in the field stop, else light from part of the pupil (deformable mirror) may be vignetted, causing bad control matrix. Note: This is the most frequently occuring misalignment and adjustment may be needed every time you calibrate the AO system. The reason for this misalignment is most likely small movements of the vacuum tube relative to the optical table, causing the image of the pinhole to move slightly at the wavefront sensor focus. 7. Remove the pinhole at prime focus and fold in the pinhole at the front of the wavefront sensor. Press slightly the pinhole holder such that it is in proper contact with the field stop. The pinhole images will be darker, and you may want to reduce gamma to 0.7 or even 0.5 to see them well. Now check that the pinhole images are exactly at the center of the red boxes. If not, adjust slightly the x-y screws on the pinhold holder.this misalignment is unlikely to occur. If it does, it is probably due to the screw allowing rotation of the pinhole (allowing it to be folded into the beam) being loose. If so, tighten it. 11

13 After following the above instructions, the wavefront sensor should now be well aligned. The observer will normally not need to go through the above steps of checking the alignment in the order indicated on a regular basis, it will be enough to check the appropriate alignments during calibration of the AO system and when running the system in closed loop. However, if any indication of misalignment is detected during calibration of the AO system, the calibration of the AO system should be interupted, the alignment procedure above should be followed in the order indicated above and the entire calibration be redone. When doing the control matrix calibration, it is easy to detect evidence of misalignment: If the partly illuminated and darker pinhole images do not appear equally dark and rougly equally elongated (in the radial direction), this indicates bad pupil illumination that may need correction by adjusting the field lens. If the pinhole images are not exactly at the center of the red boxes, this indicates that the correlation tracker offset voltages must be adjusted. After the calibration of the control matrix, watch the displayed wavefronts from the 37 electrodes: The wavefront from the central electrode should look symmetric and the wavefront maps from the three rings (6 in the inner, 12 in the middle and 18 in the outer rings) should appear virtually identical within the group of 6, 12 or 18 from each ring. If not, the field lens is misaligned or the AO pinhole is not centered in the red boxes. The plot of KL coefficients from the center electrode also gives a good indication. This plot should show strong KL4, Kl11 and KL22, but KL2 and KL3 (wavefront tilt in the x and y directions) should be small. While running the AO system in closed loop on the sun, check the reference image such that it is well aligned with the red box surrounding it. 12

14 3.5 Calibrating the system We suggest the following steps as a routine calibration round each day and whenever the system does not perform well. Note that you should start the correlation tracker s live display to initialise the mirror offsets when you do these calibrations. 1. Wavefront calibration (Sect. 3.6) 2. Control matrix (Sect. 3.7) 3. Mirror offset voltages. (Sect. 3.8) 4. Darks frames. This is not needed every day. (Sect. 3.9) 5. Flats fields (Sect. 3.10) Each step is explained below. 13

15 3.6 Wavefront calibration This calibration defines the relative positions between subfields that correspond to zero wavefront error by use of a pinhole placed immediately in front of the wavefront sensor. The AO system will try to keep these relative positions when in closed loop. This is equivalent to driving the mirror such that the wavefront from the telescope and re-imaging system combined is flat. To make a wavefront calibration; 1. Fold in the pinhole at the secondary focus, located at the front of the WFS. Press slightly, such that the pinhole is in its pre-defined position. 2. Select Wavefront Calibration from the observer mode action list. Press Start. You should now see a live display of the image with the WFS subimages superimposed. 3. Once the pinhole images are located within the subimages press the Return key in the live image display to calculate and store the sub-pixel shifts. 4. Press Start again to see the live image and verify that the images are all located within the same relative position within the windows Troubleshooting The subimage positions are completely wrong after wavefront calibration. This may happen if you run the calibration with the pinhole images out of correlation lock range in some sub images. I.e. if the pinhole is located too close to the edge in some sub images. One needs to do window recovery (Sec. 3.11) and then do wavefront calibration again. 3.7 Control matrix Find the relationship between changes in mirror voltage and changes in wavefront. This is done by measuring wavefronts while varying the voltage on individual electrodes. The relationship is then inverted, making a DA matrix that converts from wavefront modes to electrode voltages. To do a control matrix calibration; 1. Put the pinhole marked AO in the primary focus (the other pinhole is much bigger and is used for aligning optical components only). Ensure that the filters labelled for focusing on the pinhole are in front of the WFS. 14

16 2. Select the Control Matrix action from the interface and press Start. You should see the spot images moving in and out as the voltage on the central electrode is varied. The spot in the centre sub image should not move more than a pixel in any direction. If it does, it indicates that the WFS is not well aligned. 3. If the pinhole images are not centered in the subimages change the offset of the tip/tilt mirror. 4. With the live image display window press Return to run the calibration. It takes about 37s to run the calibration after which post processing is done. 5. During post processing, two plots are displayed for inspection. The first, shown in figure 7 shows a visualisation of the wavefront for variation of each electrode. On the left hand side are the original wavefronts and the right hand side shows rotated wavefronts which should show that the effect of all the electrodes in a given ring is the same after rotation. Note in particular the wavefront from the center electrode (upper left sub-image). This wavefront should look neraly perfectly rotationally symmetric, else this is an indication of optical alignment problems or possibly bad offset voltages. 6. The user will be prompted for whether they want to accept the new calibration or not. Figure 7: Electrode response wavefronts Troubleshooting The electrode wavefront responses don t look anything like the ones in the manual. Make sure the full range of the spot motion in the live image is well contained within the windows. 15

17 If they are not then you should either move the sub images using the offset dials on the physik Instruments amplifier. It may also be that you have bad offset voltages so you can also try try again with Zero Offset Voltages selected or take new offset voltages with the old matrix. The electrode wavefront responses look ok but when rotated do not line up together. The control matrix program has the orientation for the imaging table hardcoded and so when running on the spectrograph they do not line up. So if you are running on the spectrograph table you should ignore this until we can fix it. Pinhole images do not move in the live display First check that the debug option Use Mirror is selected in the options menu. Otherwise check 5 that the high voltage amplifier has power and that the cables from the computer to high voltage amplifier and high voltage amplifier to mirror are connected properly. The system seems to stop running in closed loop but I am never shown the results. The AO command interface must be run in the foreground mode. If not then the ANA 6 post processing code will not run. 3.8 Offset voltages Find the mirror voltage offsets that make the mirror flat, or more precisely, gives the mirror a shape that compensates for aberrations between the pinhole and the wavefront sensor. Note that these offset voltages are only used when the AO systems starts in closed loop. Once the AO system is locked, the offset voltages are of no importance. However, proper offset voltages help the system lock initially by positioning the sub-images in their proper locations, such that the cross-correlation algorithm (that has limited lock range) can measure the sub-image displacements. To take offset voltages; 1. Ensure pinhole is in primary focus, and the filter labelled for focusing on the pinhole is in front of the WFS. 2. Select Offset Voltages from the action list and then press Start 3. If the pinhole images are not centered in the subimages change the offset of the tip/tilt mirror by using the arrow keys in the ocrrelation tracker live display. 4. Check that the sub images look ok. 5 ask one of the staff if you re not sure about this

18 5. Press Return in the live image display window to start the offset voltage calibration. The system will run in closed loop for 10 s and then display the results for the user to acknowledge. 6. Check the offset results, an example of which is shown in figure 8. There are several error conditions in which the offset voltages will not be updated. Normally, the maximum offset voltage is less than 350 digital units, larger values could indicate optical misalignment. Note: The science cameras must be focused with the AO system locked on the pinhole at primary focus. It is NOT sufficient to have the offsets voltages statically set. This is because sunlight heats the optics which causes focus change. When the pinhole is inserted, the heatflow is blocked and the optics cools off, changing focus. Cameras must always be focused while the AO system is locked in closed loop on the pinhole. See separate instructions in How to focus CCD cameras at the SST Offset Voltage Recovery If the offset voltages are too far from flat then the system will have a very hard time doing control matrix calibrations. This should not happen since the offset voltage process is supposed to check the results to ensure that it does not save unreasonable results. However it is still possible and the following procedure may be of help when recovering the offsets in such a situation. 1. Do a wavefront calibration with the pinhole at the PRIMARY focus, do a window recovery first if the pinholes images are too far off. 2. Control matrix with DA voltages zeroed. 3. Offset voltages with DA voltages zeroed. Repeating the offset voltage calibration after doing the wavefront calibration should improve the offsets. 4. Put the pinhole back in the secondary focus and redo wavefront calibration. 5. Offset voltages with DA voltages selected. 6. Control matrix with DA voltages selected. 7. Offset voltages with DA voltages selected. 17

19 Figure 8: Offset Voltage Display. 3.9 Dark frames Record dark frames to compensate for bias in the CCD. To make a dark level calibration; 1. Block the beam somewhere near the prime focus. 2. Select Dark Fields from the action list and press Start 3. Verify that no light source is present in the live image. 18

20 4. Press the Return key in the live display window Troubleshooting The offset display (Fig 8) never appears. The AO command interface must be run in the foreground mode. If not then the ANA post processing code will not run Flat fields Record flat fields to compensate for differential pixel sensitivity in the CCD. To take flat field images; 1. Set the telescope to scan the solar surface near solar center, but avoiding sunspots. (Flatfield option on telescope GUI.) 2. Select Flat Fields from the action list and press Start 3. Press Return in the live display window. 4. Wait until the program finishes. It samples over a 20 second period to ensure that no solar features dominate Window Recovery After an optical realignment or a failed wavefront calibration the relative window positions can be so incorrect that it is not possible to recover using the usual process of wavefront calibrations and control matrix and offset voltages. In this case one should use the recovery mode; 1. Put the pinhole in the primary or secondary focus, it should not make a difference. 2. Select Recovery from the Mode menu in the interface. 3. Select Window Recovery recovery from the action list. 4. Check that you can see all the required sub images. The windows will be incorrectly placed but you need to check that the pinhole images look ok. 5. Press return in the live image display window. 19

21 6. A CCD frame is now shown with the first subimage drawn in blue. Click on the real location of the first subimage, i.e the centre of the appropriate pinhole image. A red box will be drawn showing the new, revised location of the sub image. Repeat this procedure for all sub images. Note that at any time you may hit the Stop button to stop the process before writing the sub images if you make a bad choice Running AO Follow the following steps to close the servo loop on the solar fine structure. 1. Close the loop on the correlation tracker. The AO system will not be able to lock unless the correlation tracker is locked. 2. Select the Closed Loop option from the action list and press Start 3. Inspect the live image to make sure everything looks ok and that no intensities saturate (red pixels). If possible, slightly adjust the telescope such that small high-contrast solar features, such as a sunspot structure or a pore, is present within the wavefront sensor field of view. Such features greatly improves the AO system ability to lock. 4. Press return in the live display window and the system will be running in closed loop. 5. Check that the system is able to lock on the target by looking at the title string of the live display image. If it has locked then the string will read Closed Loop otherwise Open Loop. It may switch also between the two, indicating that the system is not locking 100% of the time. Once the AO system is running the program prints to screen a summary of performance every 30s. This summary will be something like Focus: -1.1 mm Loop Closed: 98.7% where; Focus This is the amount of focus change the DM is applying. It is the value shown as KL4 translated into mm to allow adjusting the position of the WFS. Loop Closed The percentage of time for which the system was running in closed loop during the previous sample period (30 seconds). 20

22 Figure 9: The AO live image display. Note the 37 microlens images, red boxes showing the positions of the cross-correlation windows, the Max value of intensity. We refer to the central, larger box as the reference box. Positive X runs to the right, positive Y downwards Updating Telescope Focus If the system is locking the focus error should be consistent and can then be used to adjust the focus of the telescope. A focus error of +5mm (this value is not calibrated and probably too large by about a factor of 2.5) corresponds to one short step to the left at the telescope focus program. One should then verify that the reported focus error is reduced Troubleshooting 1. I have high-contrast structure in the sub images, seeing looks good but the system will not lock. First, check that the correlation tracker is locked on its target. If the correlation tracker is locked properly, this problem is probably due to a telescope focus error. Focus errors cause a displacement of the sub-images radially with a movement that is proportional to their distance from the center. If the focus error is too large, the cross correlation algorithm will not work because of its limited lock range (in principle +/- 4 pixels in each direction, in practice +/- 3.5 pixels). Watch the live image on the science CCD and try to improve its focus. Once that image looks good, the AO system should start to lock. When the AO system locks, it will print out focus information that will help you improve the telescope focus further, improving the lock rate of the AO. 21

23 3.13 General AO Troubleshooting 1. Sometimes the display is very slow to update. We are running the critical sections of the AO system using RTAI real time linux and our tests have shown that the AO performance does not suffer during these periods of sluggish display, and poor user response. We need to investigate this further to see if we can do anything to improve this. 2. Sometimes the Stop button on the interface doesn t stop the AO system. The synchronisation between the interface and the AO program sometimes get mixed up and the user is required to kill the ao program manually. To do this use the command ps aux grep ao to find the process id (PID) of the ao program, there will probably be 3 of them. Choose the lowest PID and issue the command kill -KILL PID to kill it. This will be fixed in the future. 3. The program doesn t start but prints CANNOT INIT TASK. This can happen if the ao program locks itself and does not respond to the stop command. If you observe this then log it as a bug at citing as much information about how you produced the problem. Follow the previous items instructions to manually kill the program. If things remain unstable then reboot the computer, call for help from the technical staff and/or pray AO Configuration File Reference The configuration file is not intended for observor use but it has some things which might are useful to be aware of. (It is located in cfg/ao.cfg) The format follows the standard SST configuration file format of key value pairs declared as key:value. 22

24 Key PID FILE LOG FILE ANA EXE DA OFFSET FILE OFFSET ANA CONTROL MAT ANA DARKS FILE DARKS SUM FILE DARKS ANA FLATS FILE FLATS SUM FILE FLATS ANA C MAT CALIB C MAT RUN WIN POSI FILE WF MODES Description The location and filename where the AO process id will be stored when running. The location of the log file. This will have the current date appended to it to make it unique for each days observing. The location of the ana executable to run. The mirror voltage offset file. Location of the ANA program for proecssing offset voltages. Location of the ANA program for processing the control matrix data. The darkfields file. The darkfields sum file. The ANA program for processing darkfield data. The flatfields file. The flatfields sum file. The ANA program for processing flatfield data. The control matrix to use when calibrating. i.e. when taking control matrix calibrations. The control matrix to use when running the AO system The window positions file. The number of wavefront modes corrected. This must match the number generated by the Shack-Hartmann configuration code. 23

25 4 Running the Correlation Tracker 4.1 Starting The Tracker 1. Log in to royac15 as user obs 2. Start the tracker by clicking on the CT icon on the toolbar. 4.2 Running The Tracker When the tracker is started 4 windows will appear; An interface for controlling the action of the tracker (Fig 10). A full field view of the dalsa with the current tracking window shown. (Fig 11). A view of the current reference image (Fig 12). An information window with current status of the tracker (Fig 13). Figure 10: The user interface for controlling the correlation tracker Taking darkfield calibrations 1. Ensure that light to the tracker is blocked. 2. Select Darkfield calibration from the interface option list. 24

26 Figure 11: The live display window of the tracker. 3. Hit the start button, or press return with focus in the live display image. Darkfield calibration, takes a fraction of a second. Afterwards thedarkfield image will be shown to the user who can then either accept the calibration or decline Taking flatfield calibrations 1. Ensure that the telescope is in flat field mode. 2. Select Flatfield calibration from the interface option list. 3. Hit the start button, or press return with focus in the live display image. Flatfield calibration, takes sample images over a period of about 20 seconds. Afterwards the flat field image will be shown to the user who can then either accept the calibration or decline Mirror Calibration. This mode is a simpler version of the control matrix calibration of the AO. It outputs triangle wave voltages to the tip/tilt mirror and writes the data to disk. Unlike the AO control matrix however, this calibration does not need to be done very often. The DA counts per pixel does not change very much. Accepting an incorrect calibration will render the tracker useless so unless you know what you re doing leave this option. 25

27 Figure 12: The reference display window of the tracker. 1. Ensure that the large pinhole is at the telescope focus. 2. Select Mirror calibration from the interface option list. 3. click on the pinhole image in the live display to move the correlation window over the pinhole. 4. Hit the start button, or press return with focus in the live display window. The calibration takes about 4 seconds after which a window shows the counts per pixels in X and Y directions allowing the user to acept or decline them Running Cross Correlation 1. Using the telescope handpaddle move the desired observing target into the field of view of the tracker. 2. Select Correlation Tracker from the control interface. 3. Click on the target area in the live display window to move the correlation window to the correct place. 4. Hit the start button or press enter with the focus in the live display window. The tracker will start running and information about the performance will be displayed in the information window Tracker Options Menu Disable flatfield Do not perform flatfield calibration. This can be usefull if you want to see a raw image for some reason. For example if you suspect bad flatfield calibration. 26

28 Disable darkfield Do not perform darkfield calibration. Tracker Options Shows a dialogue which allows changing the time between reference image updates. This is typically set to 20 seconds but can be shortened if there are high clouds. Display Shows a dialogue which allows changing the display settings, gamma, max and min. These controls can be safely changed while the tracker is running in closed loop Window Menu All of these windows are displayed by default but if they closed by the user they may be reopened from the windows menu System Menu None of these options are intended for use by observers under normal circumstances. Disable mirror Do not about voltages on the mirror. Save Images Store and save correlation image raw data for debugging and verification. Writing this data takes a long time and so should not normally be done. Servo Options Shows a dialogue allowing setting of servo parameters. Figure 13: The information window of the correlation tracker. Drift Correction Shows a dialogue allowing setting of drift correction parameters. These option need to be saved and the tracker restarted. They do not take immediate affect, unlike all the other options Tracker Troubleshooting 1. The system takes a long time to exit after I press stop. Check the debug menu to see if debug data is being saved. Writing this data to disk takes quite a long time. 27

29 4.3 Configuration File Reference The configuration file is not intended for observor use but it has some things which might are useful to be aware of. (It is located in cc/cc.cfg) The format follows the standard SST configuration file format of key value pairs declared as key:value. Key Description COUNTS PER PIXELX The numbers of DAC counts required per pixel of movement in x. (Measured in mirror calibration mode.) COUNTS PER PIXELY The numbers of DAC counts required per pixel of movement in y. ARCSEC PER PIXEL The number of arcseconds per pixel on the dalsa. DRIFT UPDATE The number of seconds between updates. DRIFT PORT The port to communicate drift to (default 11125). DRIFT HOST The hostname of the turret control computer. STEP PORT The port to listen for stepping commands on. WIN X POS The saved x position of correlation window. WIN Y POS The saved y position of correlation window. DISPLAY GAMMA Gamma of dislay. DISPLAY LOW Lower limit of colourmap. DISPLAY HIGH High pixel of colourmap. REFERENCE UPDATE Reference image update period (s). USE DARKS Dark calibration enabled/disabled. USE FLATS Flatfields enabled/disabled. SERVO IGAIN Servo loop Integral (I) gain in percent. SERVO PGAIN Servo loop Proportional (P) gain in percent. 28