TEXES is an instrument used at Cassegrain focus, cooled in a LHe Dewar

Optical Design:

TEXES has two spectrograph chambers: One chamber contains the high-resolution echelon grating. The other chamber has the cross-dispersion gratings and detector.


In the picture above, the light path is shown as a dotted line as it enters the spectrograph. The lenses are used situationally. ZnSe lenses are used if the longest wavelength of interest is less than 14μm. KBr lenses are used for longer observations.

The lighter passes into the echelon chamber after it goes through the slit. From the echelon chamber, the light goes to the cross dispersion chamber, where it is finally reflected towards the detector

A 31.6 groove mm-1 R4 echelle is used for cross-dispersion. The spectral coverage is usually λ/200. This contains 5-10 echelon orders depending on the wavelength (a full echelon order lands on the detector if λ<11 μm), and gaps can occur at larger wavelengths.

When the assembly in the echelon chamber is rotated 45°, an Offner relay is formed. This reimages and rotates the slit 90°, making it perpendicular to the dispersion from the echelle grating.

The low resolution grating can be used by the echelon as a cross-disperser, which gives more coverage but needs a slit to separate the orders.

TEXES can also be used as a camera with reasonable image quality, but low efficiency compared to other cameras. The FOV is approximately 25" by 25" with 0.33" pixels on a 3m telescope.


The K-mirror shown in this figure can be rotated 45° for a pupil imaging mode that simplifies alignment with the telescope.

Echelon Grating:

The first echelon was created in 1898 by Albert Michelson by stacking multiple glass plates on each other to make a grism, which he called an "echelon" because of the stagger effect of the plates.

In 1993, we attempted to make an echelon with 100 diamond-machined aluminum facets spaced 0.4" apart on a 40 inch aluminum bar. This echelon had to stay aligned during cycling between 300 and 4 K for use in thermal infrared, but it moved too much during cool down.

After the failure of the 1993 echelon, a 36 inch long, 3.4 inch wide, R10 grating with 0.3 inch groove-spacing was ordered in 1999. Before the machining process, several steps were made to ensure that no heat-stressed material was removed from the grating. The echelon has been cycled about 15 times since its diamond-machining without changes. The echelon is supported only at its ends and is very stiff.


The detector is a 256x256 pixel Si:As array optimized for low backgrounds.