Tip-tilt wavefront correction, a project conceived and initiated by Richard Elston (Florida), has been successfully implemented on the Blanco telescope. This is now the standard facility installation for image stabilization and telescope guiding with the IR imagers and IR spectrometer.
In the CTIO realization, a dichroic above the Cass focus reflects IR radiation to the science instrument and transmits the optical beam to a high speed CCDTV. The digital signal is passed to a dedicated PC for centroid determination. Error correction is done at the f/14 IR secondary mirror via a commercial piezoelectric actuator system. Setup and control is the responsibility of the telescope operator. The CCDTV is mounted on a translating stage to allow access to guide stars over a 5' × 5' field centered on the science target. We have successfully guided on stars as faint as V = 18, although brighter stars are preferred for optimal correction.
K band images generally show an improvement in FWHM of several tenths of arcsec with respect to the V band free air site seeing reported by the mountain seeing monitor, and similar improvement over the K band open-loop telescope performance. Available IR imagers are CIRIM (50" FOV, 0.2 "/pix) and COB (50" FOV, 0.1 "/pix). COB offers better sampling and a wide range of filter options, while CIRIM has lower system read noise and much better detector cosmetics. (CIRIM is also available at f/8, with 0.4" pixels and 100" FOV, but without tip-tilt correction.) The best images obtained to date have been with CIRIM, due to vagaries of weather and site seeing. Background-limited exposures on science targets have given 0.3"-0.4" FWHM at K. On nights of good but not superb site conditions, we have maintained < 0.5" FWHM for lengthy periods.
Spectroscopic comparisons with previous performance are made somewhat uncertain by optical changes implemented inside the spectrometer to accept the tip-tilt beam. Our best present estimate is a throughput gain of 1.5 to 2 in point source flux with the two pixel (0.64") slit in the K band (further information on all IR instruments can be found on the CTIO Web pages). In the coming months we expect to characterize and optimize system performance further. We will change the CCD sensor to a special purpose, high speed low-noise sensor to extend our guiding range to fainter stars. Fixed overheads in the PC hardware presently limit correction frequency to 100 Hz, even on very bright stars; we expect to be able to operate significantly faster. Finally, auxiliary f-ratio conversion optics for the 4-m Image Analyzer will enable us to determine primary mirror support corrections specifically for f/14 and to "tweak" the primary in real time, which may produce some further improvement in image quality (we presently use the f/8 corrections).
Together with the image quality improvements resulting from telescope and dome modifications over the past few years, tip-tilt image stabilization offers significant gains in IR spatial resolution and sensitivity; however, it is not a cure-all. Some gain is realized over a wide range of site conditions, but bad seeing "in" will produce, at best, mediocre images "out." Neither our telescope's traditional design nor our present scheduling practices permit rapid real-time instrument or program changes in response to varying conditions, so a certain "luck of the draw" element will continue to affect users' results. Historically, the best site seeing occurs in the southern summer (November-February); winter can be significantly poorer. A conservative approach is suggested for planning science programs. Programs that must have the very best conditions to achieve any useful result will have a significant element of risk.
Attendees at the SPIE Symposium on Astronomical Telescopes and Instrumentation in Kona, Hawaii, can see two poster papers about this system on Monday, 23 March: Pérez and Elston, "Fast tip-tilt secondary for the CTIO Blanco 4-m telescope," and Probst et al., "Performance of a novel tip-tilt secondary on the Blanco 4-m telescope."
Ron Probst (rprobst@noao.edu),
Patrice Bouchet (pbouchet@noao.edu),
German Schumacher (gschumacher@noao.edu)