Ron Probst (rprobst@noao.edu), Maxime Boccas (mboccas@noao.edu), Patrice Bouchet (pbouchet@noao.edu), Eduardo Mondaca (emondaca@noao.edu), Ricardo Schmidt (rschmidt@noao.edu), and Rolando Cantaruti (rcantaruti@noao.edu)
The IR f/14 tip/tilt system has been in use for about a year and a half now. It has performed reliably and has delivered an improvement in IR image quality. We have continued to improve its performance, reliability, and ease of use, within a resource environment constrained by work on SOAR, Hydra, and MOSAIC.
We have just upgraded the optical tip/tilt sensor to one that is better
suited to our small-field, high-speed application. This is an EEV CCD 39-02
with 80×80 pixel format, 4 e- read noise, and QE in excess of 0.7 from 0.4
to 0.75 m; 0.9 in the V band. We use reducing optics to optimally match its
24
m pixels to the reference star FWHM for use with a quadrant-sensor
algorithm. With lower read noise and higher QE than its predecessor, this
sensor should give better performance on stars of intermediate brightness (V
~10-15) and permit guiding on fainter stars than previously. In initial
tests under full moon and a hazy sky, we were able to guide robustly on a V
= 16.5 star at a correction frequency of 20 Hz, a factor of two gain in
speed. Correcting faster on a brighter star (V ~11.5, 130 Hz) produced an
rms error of the star centroid of 0.01". This represents significant
improvement in the minimum necessary speed, reference star faintness, and
achieved correction with respect to previous performance.
Optimizing system operating parameters for best performance under different conditions requires experimentation with actual seeing-degraded images. Since the 4-m telescope is a very expensive optical bench, we have set up an atmospheric seeing simulator, or "turbulator," for lab tests in La Serena. This is based on a simple design developed by a French group (Masciadri and Vernin, Applied Optics 36, 1320, 1997). It has allowed us to use our telescope engineering time to best effect as we work on improving the hardware and the algorithms.
Several background tasks to address system maintenance issues, invisible to users, have been accomplished or are ongoing. The piezo controller electronics from Physik Instrumente have been replaced and duplicated to provide a backup. Telemetry circuitry developed in-house has been added to the controller for easier setup and troubleshooting, from the control room, of piezo control behind the secondary. Printed circuit boards, which had been modified extensively by hand during system commissioning, are being replaced with final versions to avoid subtle failures later. Mountain personnel have been systematically trained in setup, operations, maintenance, and troubleshooting at the telescope, and in La Serena during lab tests. A second dedicated instrument PC was purchased to provide a backup, but unfortunately some internal components necessary for network communication on the mountain were damaged beyond repair in last April's lightning storm. This PC is now used in La Serena with the laboratory simulator. Finally, recovery from a severe mechanical miscollimation following removal and replacement of the entire system (for work on the prime focus pedestal in support of MOSAIC II) resulted in some mechanical modifications to prevent a recurrence.
While the job of a tip/tilt sensor is to stabilize an image centroid (and our system does this very well), the ultimate goal is to deliver a tighter image to the science sensor. Other aspects of the wavefront error that we can control relate to optical alignment and the telescope thermal environment. Optical engineer Max Boccas has taken the lead on these issues. Telescope alignment, diagnosible in real time with the IMage ANalyzer, has been improved. The mechanical modifications mentioned above also addressed a small random decenter of the f/14 secondary. Once we are satisfied that this is under control, we can institute a lookup-table active correction for the secondary similar to what is done now for the 4-m primary. We are also investigating alignment issues between the tip/tilt optics box and the IR cameras, and improving thermal control of the Cass cage environment and the primary mirror. These are now the areas where we expect improvements to impact FWHM with our stabilized images.