1.0 Introduction

A long slit, laboratory- test spectrometer is being constructed for the evaluation of the Xspect computer model being used to determine the parameters for the NIR spectrometer. The objective is to do a preliminary design study on an inexpensive system where all the parameters can be controlled before committing to more exotic materials and designs for use in NIRIM.
 

2.0 Optical Design

The spectrometer is designed to operate in "direct view" mode using a prism to set the entrance angle to the transmission grating. The definition of the central wavelength and dispersion is set by the contrain of using easily obtained off the shelf components. This spectrometer emulates a grism by adding a wedge prism behind and tilting the transmission grating. The design of the system is shown below.
 
 

 

    A commercially avalilable wedge (10^o prism) was used as a base for the transmission grating. As a result the zero deviation wavelength has been shifted upward. In this version of the spectrometer I have opted to work at first order, but there is an excellent opportunity with this grism to operate at 2nd order with a zero deviation wavelength of 4000A. However, this is for another day!

    The spectrometer optical assembly is rather compact and measures about 3 inches long. A set of custom aluminum optical holders were machined from aluminum rod here at MWO and the collimating and focusing lenses installed.  The wedge prism was inserted into its holder and the 110 lines per mm transmission grating mated to the surface with optical oil with an index of refraction of close to 1.5. For the moment the spectrometer sports a pair of tape racing stripes that keep the assembly together and gives it a racey look.
 

Plate 2. Photograph of the prototype grism spectrometer.

    The optical path is begins in the lower left of the image in Plate 2 and follows a counter clockwise path. The path is defined by an alignment laser, shown in the lower left corner of Plate 2, and an optical fiber feeds the light from the neon source into the optical path. The focused light is folded using a first surface aluminum mirror onto the slits made from a pair of single edged razor blades. The expanding beam from the slit is folded into the collimating lens of the Spectrometer Optical Assembly which follows the schematic in Plate 1. The dispersed light is focused onto the CCD camera where it is imaged.

4.0 Bench Test Results

    The spectrometer is operating at the point where simple images can be made from bright objects using the low light ccd camera. A re imaging system in front of the slit allows several sources to be placed into the spectrometer path for calibration and testing.

    A 2 color LED was used to provide a simple calibration system. While the LEDs were specified as having wavelengths of 5650 and 6350A they were found to actually be 5700 and 6350A. These emission lines were used to determine the calibration of the system. The neon line source is a neon bulb whose light is fed through an optical fiber fed to the reimaging system to illuminate the slit.

The images recorded with the spectrometer and their line plots are shown.

Plate 3. Reference spectrum of a 2 color LED with both colors powered up. The leftmost is the yellow-green component with a wavelength of 5700A and the other is the orange-red component with a wavelength of 6350A. The LED reference spectra were used to determine the slope and offset of the linear calibration for the system.

Plate 4. Line cut through the LED reference spectrum showing that inspite of having a very narrow slit there is considerable width to the LED spectrum, but a well defined central peak. A better calibration source with wider separtion of the principal wavelengths would be desireable, but this appears to be sufficient for this test.
 

Plate 5. A single frame of the spectrum of the neon test source being used to determine the properties of the grism spectrometer. The neon test source that is being used is considerably weaker in intensity than the LED reference source so the spectrum is somewhat noisier.
 

Plate 6. Line cut through the neon test spectrum showing the grisms ability to resolve the spectral lines.

5.0 Predicted vs Actual

     The code Xspect appears to be working rather well and predicted that the spectrometer should be able to resolve lines on the order of 40 A separation. The neon spectra has two lines at about that separation (39A) and their separation can be detected. Other parameters such as the zero deviation wavelength and the spectral width appear to be in good agreement.

    It would seem that Xspect is providing a good agreement with the actual spectrometer and will provided a useful tool for the design of the near infrared grism to be used in NIRIM.

6.0 Astronomy Images

You must be joking! The clouds, the wind, the 2" of rain we just got ... okay I am just making excuses...

Back