Pathfinder Multi-Star Campaign – March-April 2014

After a three month break, the Pathfinder team returned to Mt. Graham in late March 2014 with the goal of closing the adaptive optics loop on multiple, rotating, off-axis natural guide stars. This mode is central to the operation of LINC-NIRVANA's multi-conjugate adaptive optics (MCAO). Centering the wavefront sensor guide probes on each star while the sky rotates overhead is a nontrivial exercise. We must also provide the facility adaptive secondary mirror with a continuously evolving calibration. The procedures and software routines used during T7 will feed directly into the final instrument.

Daytime Activity (5 days, 29 March – 2 April)

Work began on 29 March, with two of the team on the mountain, unpacking, verifying, and assembling test equipment for the run. On 30 March, the Pathfinders, along with Wolfgang Gaessler and Diethard Peter of the ARGOS team, set about trying to see the newly installed ARGOS calibration fiber with the LINC-NIRVANA (LN) ground-layer wavefront sensor. The test was successful, with Pathfinder able to close the adaptive optics (AO) loop using 100 Zernike modes, corresponding to modestly high wavefront correction. This is important, since the optical configuration of LN requires an external reference for calibration of the two ground-layer wavefront sensors (GWS - LN's second pair of "high-layer" wavefront sensors have their own calibration fibers). Taking advantage of the ARGOS source means not having to expend valuable observing time calibrating on bright stars. The LINC-NIRVANA team is now in discussion with ARGOS about adapting their fiber feed to better match LN's optics.

Diethard Peter and Wolfgang Gaessler of the ARGOS team (left) work with Carmelo, Maria, and Valentina on Pathfinder calibration.

Calibration and verification of the AO control loop continued on 1 April, while a subset of the Pathfinder team worked to install and test a system of cameras and screens to assist in finding and centering the reference stars on the movable probes. The automatic software to do this is currently in a preliminary state, and hence visual feedback will help in debugging and optimizing. The team installed a total of three cameras which can focus on the screens mounted to the tips of the star probes. As a final check, the Pathfinders imaged a faint calibration fiber on a probe screen and used the control software to bring the artificial starlight into the wavefront sensor.

Maria and Carmelo optimizing the adaptive optics control loop.

Because LINC-NIRVANA must remain fixed with respect to the telescope while the sky rotates, all of the wavefront sensors must turn continuously around a common optical center. The deformable secondary mirror (DM) of the LBT applies corrections to remove ground layer turbulence, and this mirror is also fixed with respect to the telescope. This means that there is a constantly changing relationship between the wavefront sensors and the geometry of the actuators in the DM. To compensate for this, the adaptive optics calibration must also be continuously updated, using either a large series of measurements at many different rotation angles, or with a more sparsely sampled calibration artificially rotated to the correct angle. On day 4 (April 2) of T7, the team successfully closed the AO loop on a reference fiber for a full 60° of rotation using a synthetic interpolation of the calibration. Since the actuators in the deformable mirror are six-fold symmetric, this important milestone means that Pathfinder, and eventually LINC-NIRVANA, should be able to operate successfully, independent of the current relative orientation of telescope and sky.

The bright ARGOS alignment source produces a psychedelic display on the Pathfinder on-axis camera. The fainter calibration fiber (see below) looks considerably less interesting but will be very valuable.

The "high view" CCD camera mounted (and well secured) above the Pathfinder focal station.

The high view from one of three remote cameras monitoring the positions of the star probes. This frame includes the other two cameras (yellow arrows). Four of the twelve star probes carry small white screens to help in locating and centering reference stars.

The faint ARGOS calibration fiber produces a well-focused spot on the star probe screen (arrow), mimicking a moderately bright star.


Nighttime Activity (2 half-nights, 2,3 April)

The T7 nighttime activity did not begin auspiciously. The LBT had been closed for days due to high wind, and when the Pathfinder team arrived to execute the second-half-night of observations on 2 April, it was snowing lightly and the mountaintop was encased in cloud. Despite this continued bad luck with the weather (see the T5 and T6 reports), the team forged ahead, working with the LINC-NIRVANA team in Heidelberg in refining the telescope preset process and in optimizing the wavefront sensor derotator. Although the weather began to improve toward sunrise, the presence of accumulated snow on the dome meant that the night was over.

April 3 showed much improved conditions, with crystal clear skies, little or no wind, and good (0.8 arcsec) seeing at midnight. We sent our first preset to the telescope at 00h50, after a late handover from the LBTB observers, and we finished at about 5h20, well into morning light. Sunrise showed us that light cirrus had moved in toward the end of our session, but it did not affect our tests. Accounting for lost time due to the AO secondary mirror going "pie-shaped" and other minor glitches (see below), we had approximately four hours total on sky.

Sunrise over LBT on 3 April 2014 - Looks Promising!

With the dome closed due to fog, Derek, Maria, and Al perform nighttime tests of telescope presets and derotator trajectories. Night assistant David Gonzalez (far right) provides valuable help.

An adaptive secondary “pie event” causes one sixth of the actuators to cease functioning.

During this period the team accomplished the following:

  • We unambiguously identified the "magic angle" resulting from the off-midline location of Pathfinder (and specifically the rotated tertiary mirror and four reflections)
  • We determined the parity / chirality of our focal plane (multiple reflections and focal planes affect this).
  • We verified the image scale to be 613 um per arcsec at our focus, which is 300 mm further from the secondary mirror than the “standard” LBT focal plane. This proved more challenging than expected, likely due to uncorrected telescope drifts and sky rotation.
  • We began using the small "cheerio" star probe screens attached to SE 1, 4, 6, 7, and 10 (see above). They worked well on bright stars, but it soon became clear that they would not suffice for the 9-10 mag stars in our target multi-star asterisms.
  • We repeated the T6 success in moving a star from the on-axis camera to the off-axis SE, this time with four star probes, not two. This was in telescope Parallactic Mode, as for T6. Note that we must operate Pathfinder (and LN) in the other telescope mode, called Position Mode.
  • We identified the GWS bearing axis on the DSLR camera by placing SE 1, 4, 7, and 10 symmetrically about the center and executing identical offsets of a bright star to each, noting systematic displacements on the cheerio screens.
  • As the sun was rising, we finally managed to understand how the various angles and symmetry inversions map to the Pathfinder focal plane. Despite increasing background light problems, we finally managed to move directly and accurately from the newly measured DSLR center to SE 1, 4, 7, and 10 in telescope Position Mode.

In the end, we did not manage to acquire multiple stars as we had planned. There are multiple reasons for this, including the aforementioned uncertainties in image scale, as well as some confusion about how various rotation angles add together in the final focal plane. Although we solved and now understand both of these issues, we simply ran out of time. Although this is disappointing, we accomplished a great deal in the four hours available, and we are eagerly looking forward to continued success in upcoming Pathfinder runs.

Tom attempts to sort out all the rotations and symmetry flips...

...while Valentina searches for the faint star on the “cheerios”

Team Pathfinder in the heat of observing.