Welcome to the new version of CaltechAUTHORS. Login is currently restricted to library staff. If you notice any issues, please email coda@library.caltech.edu
Published November 1993 | Published
Journal Article Open

A multi-object fiber spectrograph for The Hale Telescope


A new faint-object spectrograph has been designed around the capabilities of fiber optics. This instrument, the Norris Spectrograph, is for exclusive use at the Cassegrain focus (f/16) of The Hale Telescope and is optimized for faint galaxy spectroscopy. There are 176 independently positionable fibers that are serially manipulated by a single robotic system. Each of these fibers sees 1.6 arcsec on the sky and the total positionable area is in excess of 300 arcmin^2. Unlike most multiobject spectrographs which utilize fibers that are several tens of meters long, the philosophy of the design of the Norris was quite the antithesis, i.e., to minimize the fiber lengths; hence, it is an entirely self-contained telescope-mounted instrument for the Cassegrain focus. The instrument consists of an integrated xy stage, for the fiber positioning, and an attached optical spectrograph. The design of the spectrograph is basically classical: spherical collimator mirror, standard reflection grating, and a newly designed all-transmissive-optics camera lens. The detector currently used is a thinned, AR-coated 2048 X 2048 Tektronix CCD. Fibers are arranged in two linear opposing banks that can access the 20 arcmin diameter field-of-view (FOV) of the instrument. The accuracy of fiber placement (assuming errorless coordinates) is less than 0.1 arcsec over the entire FOV. Fibers may be placed as close as 16 arcsec. This permits close pairings of fibers for very faint-object spectroscopy. Beam switching between paired fibers, as was done with two-channel spectrographs of yesteryear, will help average out temporal and spatial variations of the light of the night sky. Actual observations performed in this mode of operation indicate that the quality of the sky subtraction improves, as would be expected. The density of paired fibers within the Norris FOV matches the approximate density of faint field galaxies expected to a blue magnitude of 21. Software exists to take object lists (α,δ) and convert them to rectilinear (x,y) values (mm) on the xy stage by gnomonic projection and to assign fibers. This software also corrects for precession of the equinoxes, proper motion if epoch differences exist, and corrects for differential atmospheric refraction. To place a single fiber takes approximately 5 s on the average. A lower limit to the efficiency of the spectrograph plus telescope has been estimated to be 6.8% at 5500 Å. In order to derive the throughput of the instrument, the efficiency of the telescope, estimated to be approximately 56%, must be divided out. This value is consistent with the expectation that the reduction in efficiency from that of a standard CCD spectrograph such as The Hale Telescope's Double Spectrograph will be about a factor of 2. This results from the 60%-70% transmittance of the fibers and other losses. The spectra produced are linear with little distortion. With 10 A spectral resolution, fitting residuals on the order of 100 km s^(-1) are easily obtainable by modeling the dispersion by a third-order polynomial. The resolutions currently available range from 1 to about 20 Å. The spectra have a FWHM in the direction perpendicular to that of the dispersion of about 90 µm, or equivalently about three 27 pixels found in the older Tektronix 2048 CCDs. The interorder spacing of 250 µm is large enough to permit clean spectrum extractions. The instrument has been in use for several years. The scientific programs vary from high resolution (1 Å resolution) spectroscopy of stars in nearby globular clusters to a low spectral resolution (10 Å) survey of faint field galaxies. In this latter survey, with typical 2-hr exposures, absorption-line redshifts as high as z ~ 0.5 have been routinely measured. Several heretofore unknown quasars with redshifts around three have also been discovered serendipitously.

Additional Information

© Astronomical Society of the Pacific · Provided by the NASA Astrophysics Data System. Received 1993 March 1; accepted 1993 August 5. We would like to thank The Kenneth T. and Eileen L. Norris Foundation and especially Kenneth Norris for their kind support in the creation and use of this instrument. Not only did they make this instrument possible, but their interest in the science derived with it is much appreciated. The success of this project was due to the many people who made important contributions to the instrument's completion. In the area of software, D. C. Oke wrote the CCD control software for the Amiga; M. Strauss, A. Picard, T. Small, and Professor P. Seitzer all contributed to the offline software. V. Nenow and G. van Idsinga both made valuable contributions to the electronics of the instrument. We would like to thank Professor Harland Epps and Gerard Pardahlan for their work with the camera lens of the Norris and Professor James Westphal for the use of his laboratory during the assembly and early testing of the instrument. Donald Loomis of Custom Optics did the optical work on the spherical collimator and the entrance window. Gaston Araya helped out in countless ways. We thank Professor James Gunn for illuminating discussions regarding flat-fielding techniques. Michael Doyle and John Henning of the Palomar Observatory staff provided uniquely invaluable service during the commissioning of the instrument. Finally, no Palomar paper would be complete without acknowledging the excellent assistance that Juan Carrasco has provided at The Hale Telescope over the decades.

Attached Files

Published - 1993PASP__105_1308H.pdf


Files (2.1 MB)
Name Size Download all
2.1 MB Preview Download

Additional details

August 20, 2023
October 20, 2023