Home    Overview    Equipment    Operation   Processing   Images   Spectroscopy   Links

                        Spectroscopy

A  self guiding spectrograph (SGS) manufactured by Santa Barbara Instruments (SBIG) is used for obtaining spectral images.  It uses a slit and a grating to obtain spectra with a dispersion of 1.07 angstroms per pixel in the high resolution mode.  The full width half maximum value for the high resolution mode is 2.4 Angstroms.  Using the supplied software (a program called SPECTRA) the center of lines can with care be located to +/- 0.1 Angstroms.  This corresponds to a Doppler shift of approximately 5 Kilometers/second at 6500 Angstroms.  Using the low resolution grating, which has a dispersion of 4.3 Angstroms/pixel a spectra 3200 Angstroms in width can be obtained on a single image.   Using the self guiding feature exposures up to one hour in length can be easily made.  This allows the spectra of stars as faint as magnitude 12 to be obtained with a reasonable signal to noise ratio.  A detailed description of the instrument is available at the spectroscopy page on their web site.  The SGS was selected because it can take high resolution spectra of dim objects with long exposures and is capable of being operated remotely.  The relatively simple additions that were made for remote operation are described in the modifications section on the equipment page.  The micrometer setting is the one adjustment that has to be made at the telescope but since in practice one wavelength setting is used for a series of images this has not proved to be a major hindrance in practice.  In the future a motor controlled from the operating room may be added to the micrometer for complete automation. For use at wavelengths longer than the specified 7500 Angstroms a IR band pass filter is inserted at the entrance of the spectrograph to eliminate the possibility second order reflections from the grating.  The spectrograph camera unit is attached to the telescope with JMI focuser which allows precise adjustment of the focal plane.

Spectrograph Alignment-  The spectrograph comes factory aligned but must be fine tuned for the individual camera.  A coupling plate is attached to the camera in place of the D-block and the barrel of the coupling is then inserted into a clamp on the spectrometer to attach the units together.  This aligns the slit and the grating parallel to the CCD.  There are two independent optical paths in the spectrograph, one for the light that goes to the grating and the second for the light that goes to the tracking CCD.  When in use the slit is at the focal plane of the telescope.  This means that the tracking CCD must be focused on the slit.  This is accomplished by illuminating the slit with the internal LED and adjusting the position of a focusing lens which is only in the optical path of the tracking CCD.   This takes a number of iterations since the distance of the lens from the slit is critical and the lens must also stay perpendicular to the optical path to keep the entire image of the slit in focus.  Thankfully once this adjustment is done it does not have to be repeated.  When in focus, the image of the slit is approximately 2 pixels wide on the tracking CCD.  If necessary the position of one of the mirrors in the tracking CCD optical path can be adjusted to position the slit image in the center region of the tracking CCD.  In our case the factory alignment was sufficient.  Next the spectrograph is illuminated with  light source that has lines. We use a neon bulb attached to the diffuser port of the spectrograph.  Using CCDOPS the imaging CCD is selected with the 1XN imaging mode with a vertical binning of 4.  This results in an image that is 765 pixels long by 128 pixels high. The camera is then rotated with respect to the spectrograph until the lines are vertical on the imaging CCD in the high resolution mode.  The next to last alignment step is to adjust the grating path length to put the lines at a sharp focus.  This is done using a screw adjustment on the spherical mirror that collimates the light onto the grating.  The last step is to adjust the offset screw of the micrometer so that the reading of the micrometer corresponds to the wavelength in the high resolution mode.  The last three steps were not difficult.  The alignment process was done on a table in a darkened room in less than an afternoon.

Setting up the Spectrograph-  The camera is removed from the telescope and the spectrograph coupling is attached to the camera face.  The coupling positions the spectrograph parallel to the camera but the camera must be precisely rotated so that the spectral lines are vertical on the imaging CCD.  This is done by taking a series of images with the calibration source illuminated while slightly varying the rotational alignment.  Since the coupling plate initially sets the angle quite close 3 to 4 iterations is all that is necessary to complete the setup.  The changeover from using the filter wheel with the camera to using the spectrograph can be done in about 15 minutes.  The spectrograph/camera unit is then attached to the JMI focuser on the telescope.  Since the smaller tracking CCD must be used to find and position the star or nebula on the slit for imaging, the telescope must be able to point to the object of interest with a high degree of precision.  Several steps are necessary to do this .  First, the telescope must be precisely polar aligned.  In our case this means better than one arc minute in both azimuth and elevation.  Second, the telescope must be accurately balanced.  Third, when the telescope power is shut down after the prior use the telescope must be centered on the intersection of the meridian and the Celestial Equator.  This corresponds to declination equal to  zero and right ascension equal to the local sidereal time (See the shutdown section on the operating page for details). When the telescope is powered up this procedure will insure that the tracking CCD is automatically aligned since the spectrometer's optics places it at the center of the telescope's field of view.  After turning on the power, the telescope is slewed to a bright nearby star, centered on the tracking CCD using CCDOPS and synced using the Meade hand controller.  The Sky program is then linked to the telescope and the system is ready for use.  It is useful but not necessary at this point to do a short tracking run using T Point to optimize the pointing accuracy of the telescope.  The slit is illuminated and a pixel that corresponds to the middle of the slit approximately 2/3 from the top is identified using the tracking mode of CCDOPS.  The bright star is centered on this pixel using the "guide to crosshair" command of CCDOPS.  This is the only effective means of positioning the star on a desired pixel.  A short exposure of the imaging CCD (about 10 seconds) is then made to obtain a test spectra.  The intensity of the spectra is measured at a selected  point and the value recorded.  The star is then shifted left and right one pixel at a time until the five pixels closest to the slit are tested.  The pixel with the brightest spectra corresponds to the optimal position for the star.  This optimal pixel location remains constant for night to night as long as the camera to spectrograph coupling is not disturbed.  The focus is then adjusted in a similar manner to optimize the intensity of the spectra.  This point should correspond to the sharpest star image if the calibration was precise.  If not then a slight offset of the star focus to obtain the best spectra focus is required.  With the star centered approximately 2/3 from the top of the slit the high resolution spectra will be in the upper half of the frame and the low resolution spectra will be in the lower half of the frame.  The spectral intensity is not a strong function of the vertical position on the slit.  The spectrograph is now ready for use. 

Using the Spectrograph  The telescope is slewed to the desired object and the Meade controller is used to position the object near the optimum pixel.  Using CCDOPS the self guide track command is used with the "guide to crosshair" option to precisely center the object on the optimal pixel.  The object is then guided for the desired time using 1xN resolution and vertical binning of 4 for the imaging CCD. Following this exposure, the calibration source is turned on and a calibration spectra is immediately recorded.  The standard exposure time for the neon bulb calibration source is 30 seconds both for low and high resolution.  It is critical that this be done without disturbing the setup.  If desired the calibration source can be switched on for thirty seconds during the guided exposure to superimpose the calibration spectra on the object spectra.  In our case the stellar spectra and calibration spectra are normally obtained sequentially.  A trip to the telescope is necessary only to switch between the low and high resolution gratings or to change the micrometer setting.  When the desired spectra have been recorded the telescope is slewed to a star near the intersection of the meridian and the celestial equator.  The star is centered in the tracking CCD and synced using the Meade hand controller.  Then using the Meade controller the telescope is slewed to a declination of zero and a right ascension equal to the local sidereal time.  power to the telescope is then switched off.  This assures that the telescope will be aligned when the power is switched back on.

Processing the Spectra  The spectral images are processed first using CCDOPS  to remove the dark current and crop the spectra images to a size of 10 pixels high by 765 pixels long.   A cropped low resolution spectra is shown below.

The cropped spectra is then processed using a SBIG program called Spectra.  Spectra does several things:

1. It presents the spectra graphically
2. Enables the precise calibration of the spectra
3. Precisely measures the wavelength of a selected peak or null
4. Generates a table of intensity vs. wavelength in "txt" format that can be exported to other programs for analysis or presentation purposes
5. It can filter the data to improve the signal to noise ratio
6. Expanded spectra can be created (the image below is the expanded version of the spectral image shown above).

An example of the graphical presentation of spectra is shown below.

The red line is a selected portion of the spectra and the green lines are used to identify the regions of the spectra for analysis such as line identification.

The program allows the user to identify lines on the calibration spectra and then generates a wavelength value for each pixel.  The wavelength values are then applied to the object spectra.  The program can then be used to calculate the wavelength of features in the object spectra.  The calculation is done to two decimal places.  A sample section of a table of wavelength vs. intensity that corresponds to the peaks in the image above is shown below.