Previous Article
Next Article
Table of Contents

Nicer, Tinier Images at the 0.9-m:...(1Dec93)
Image Corrector Installed
(from KPNO, NOAO Newsletter No. 36, 1 December 1993)

The field accessible with Tektronix 2048 x 2048 CCD at the 0.9-m f/7.5
focus is an impressive 23 arcmin on a side.  The Mini-Mosaic, currently
under development, will have an even larger field with the added bonus
of smaller pixels to sample better the observed point spread function
(PSF). However, it has been known for some years that the images at this
focus degrade far from the optical axis, becoming enlongated blobs in
the corners of our Tek 2048 chip. As the f/7.5 is a Richey-Chretien
design, we have been fighting two effects: (1) the focal plane is
curved, but the chip is flat, leading to position-dependent "best"
focus; and (2) the images are highly astigmatic.

In order to deal with both of these problems, a simple, two-element
corrector was designed and has now been installed and fully tested.
Although the corrector is easily removed from the beam, we expect that
most, if not all, observers will benefit from the use of this
corrector, and will want to use it.

                      [Figures not included]

We are pleased with the performance of the corrector.  The PSF
variability has been greatly reduced, particularly in the corners of
the chip. This is illustrated in the attached figure which shows images
in the center and corner of the chip with and without the corrector.
As an extra bonus, the images used for guiding by the off-axis TV are
greatly improved. Some residual PSF variability remains, but due to the
non-symmetric nature of the variability, we do not believe the
corrector is responsible.  Our leading contender for the origin of this
problem is the support of the f/7.5 secondary.  We plan further
investigation and hope to cure this problem this year. The PSF behavior
with the corrector is straightforwardly modeled using DAOPHOT II, which
was not true for images taken in good seeing without the corrector.

The scattered light performance of the corrector appears to be good. We
have observed fields containing bright stars with and without the
corrector, and no scattered light problems are apparent. Surface
photometry of galaxies has also been performed with the corrector in
and out of the beam, and identical results were found.

The one and only drawback we have found to the corrector is that the
corrector U throughput is only 65% relative to the uncorrected
throughput. The relative throughputs at other wavelengths are much
better: 87% at B, 94% at V, 98% at R, and 94% at I. We nevertheless
believe that most observers will still gain even at U from the use of
the corrector if attempting stellar photometry in anything but the
center of the chip, as more light will be concentrated in fewer
pixels.  Even for non-stellar sources, the photometric accuracy at U
will be improved through finite-sized apertures.

The corrector is not designed to be used with the f/13.5 secondary and
will be removed from the beam when that secondary is used. It may also
prove desirable to remove the corrector for f/7.5 projects which make
use of only the central part of the field (particularly at U), and
where paranoia over scattered light outweighs PSF constancy
considerations.

We are very grateful to the many staff members who have contributed to
the implementation of the corrector, particularly Liang Ming, Jorge
Simmons, Tony Abraham, and Skip Andree.

Taft Armandroff, Phil Massey,
Dave Silva, Ata Sarajedini

Previous Article
Next Article
Table of Contents