The Hydra Bench Spectrograph with the 400mm Bench Schmidt camera and 2x4K SiTe CCD

The CTIO Hydra bench spectrograph (HBS) is located in the so-called "Small Coudé Room" of the Blanco telescope building, where the Argus spectrograph was located before Argus was retired. The HBS borrows design elements of the Argus Bench Spectrograph, the 1.5M BME Spectrograph and the WIYN Hydra Spectrograph.

The optical design of the HBS with the Site CCD is shown here. The GUI can be seen here.

The tips of the fibers are arranged in a fanlike array, at the focus of a corrector-less Schmidt collimator mirror of 1200mm focal length. The grating is placed at the center of curvature of the collimator. The tips of the fibers are set on the curved focal surface. This locates the pupil on the grating for maximum efficiency.

The grating is tilted about a vertical axis at an angle of nine degrees to the side in a "quasi- Littrow" configuration, directing the light eighteen degrees off-axis to the camera. The camera is a classical Schmidt with the CCD mounted in a cold finger holding it at the focal surface.

This spectrograph will normally be used with a Bench Schmidt camera of 400mm focal length and a SiTe 2x4K CCD. This combination has replaced the old 229mm Air Schmidt camera which used a Loral 1Kx3K CCD. This combination is now only used as an emergency backup. For more information on the old camera, please click here.

The SiTe CCD has 2048 x 4096 15u square pixels and good QE in the visible and near IR. Its efficiency in the UV is only fair. For more details see the CCD section of the CTIO WWW site. The Bench Schmidt camera gives an optical reduction of 3:1 in the spectrograph, producing a projected fiber size of approximately 6-7 pixels fwhm on the ccd for the 300u fibers. The fiber images have wings 10-12 pixels wide and are separated by approximately 14 pixels on the CCD. This camera/CCD combination permits using all of the 138 Hydra fibers. There is very little separation between the images of adjacent fibers.

Any of the gratings which can be used with the R/C spectrograph can also be used with the 400mm camera. This list of the Hydra gratings details the options available.

The 300u (2 arcsec) fibers are fully operational, though a few have low transmission, are broken or cannot be seen on the Loral CCD, making the actual number of good fibers approximately 130. Many of the 200u (1.3 arcsec) fibers are functional, but unfortunately the fibers used have proven to be quite fragile. About half of the small fibers are either broken or have low transmission. As a result, the small fibers have been decommissioned and are not available.

The 300u fibers project to a fwhm of roughly 7 pixels, limiting the resolution of the spectrograph. Higher resolutions are achievable by the use of slit plates which can be put in front of the line of fibers to reduce their diameter in the dispersion direction, with some loss of light. Ask Observer Support to install the slit plates if you want to use them. 200u and 100u slits are available for the large fibers. With the 200u slit plate, the fwhm of the image is approximately width of the small fibers, achieving 33% higher resolution with a loss of only about 20% of the light. In theory, the 100u slits will produce images with approximately 2.5 pixels fwhm, giving resolutions as high as 50K in the echelle mode (with a loss of 60% of the light), but this capability has not yet been tested.

The SITe CCD has an unusual property in that having high full well capacity increases the readout noise. For this reason, there are two readout modes, "High S/N" (Gain #1) has full well capacity of approximately 60,000 electrons/pixel and a readout noise of about 5.2 electrons. "Low Noise" (Gain #2) has a full well capacity of less than 15,000 electrons and about 3.0 electrons readout noise. The full well capacity of the pixels is significantly different between pixels. This causes vertical trails which emanate from pixels with low full well capacity.

There is only a small sacrifice in spectral resolution if the CCD is binned by a factor of two in the dispersion direction so this generally recommended. Readout along the dispersion direction should not be binned if the slit plates are used.

The fiber images are just barely separated on the CCD. If the chip is binned by 2 in the spatial direction, the fiber to fiber contamination increases significantly. It is therefore best not to bin in the spatial direction unless the decrease in readout noise obtained by binning is important. When the chip is spatially unbinned, the individual spectra will be about 14 pixels apart. An extracted spectrum generated by summing the intensity in the central 8-9 pixels will contain nearly all the photons received by the target fiber and have negligible contamination from adjacent fibers.

The HBS shutter can be seen in the optical diagram. It is a large rotating disc in front of the corrector of the Schmidt camera. Although the shutter rotates relatively slowly and exposures of less than one second are not possible, it cuts the entire beam evenly and symmetrically so that exposures are quite uniform and accurate.