The instrument dedicated to the TSU 2-m Automatic Spectroscopic Telescope is a bench-mounted fiber-fed echelle spectrograph. A diagram of the optical system shows the major components. Light comes into the spectrograph from the guiding head in the telescope through one of three 200-micron optical fibers. The active fiber is selected by a motorized slide that places that fiber at the focus of a 56-inch collimator mirror, which colimates it into an 8-inch beam off-axis by 6.5 inches. The beam falls onto an R2 echelle grating used in quasi-litrow mode (theta=0, gamma=1), which provides the primary dispersion of the light. The light is then reimaged with three mirrors (the collimator, a folding flat and another concave mirror identical to the collimator) onto the cross-dispersion grating. The doubly dispersed light is then focused onto a CCD by a special camera designed by Harland Epps with mount by Allan Schier. The components are mounted on an optical bench isolated on pneumatic pads. (You can only guess the relief we felt after jacking it into the observatory.) This drawing shows the rough layout of these parts with the camera represented as a cylinder; the final assembly has all the parts in place (Feb. 2003), and here we see the finished spectrograph in the observatory.

The optical system is a variant of a now standard design used for the high-dispersion spectrographs in the ESO Very Large Telescope and in the Hobby-Eberly Telescope (see Tull). A map of the orders (with associated list) shows how the spectrum requires two placements on the detector (CCD) to get all the astrophysically realistic lines (blobs show Li I, H-alp, Na I D, Mg I, Mg II, and Ca II). We have simulated spectra for the short- and long-wavelength regions. A typical calibration spectrum for the long-wavelength region and a first-light spectrum of HD 36167 (K5 III, V=4.7) of 3 May 2003 UT, with a detail showing H-alpha and the Sodium D lines, are shown here.

Harland Epps designed a special camera for this instrument, which looks very similar to his designs for other instruments of this type, but which incorporates slightly different types of glass to extend the wavelength range it covers. Dealing with Epps was one of the highlights of the whole project.

We will record the spectrum with a SITe ST-002A CCD (2Kx4K 15u pixels) housed in an Infrared Labs dewar with mechanical refrigeration. The CCD will be run by a controller from Astronomical Research Cameras, Inc. (Bob Leach at San Diego State University). The final setup is shown here.

We will feed the calibration light into the spectrograph through a 600-u fiber that takes light from a calibration bench in the spectrograph enclosure, up to the guiding head in the telescope, then back through the active fiber to the spectrograph input. The three fibers for feeding light into the spectrograph are on a slide that will position them under the calibration feed. Thus we can feed calibration light into the spectrograph at roughly the f-ratio and from the same direction as light from a star. The calibration light comes in through the red fiber to the right in this picture of the guiding head and is projected onto the fiber to the spectrograph with a lens in the black tube to the left of center in the picture. The calibration bench has a white light (10W, 3200 K incandescent bulb) for flat-field (intensity) calibrations and a Th-Ar hollow-cathode lamp for wavelength calibration. They may be selected with a flip mirror. The flat-field source requires a neutral density filter (D=2) for reasonable exposures in the red and a color-compensating filter (BG-1 ?) in the blue.

We are controlling the spectrograph with a separate computer which will be housed in a different building (computer t13c to the left in Figure 3). It will (1) configure the spectrographi, control its CCD detector, and take calibration images, (2) receive commands from executive computer t13x to take spectra of stars, and (3) monitor the status of the spectrograph.