User manual SBIG SGS

[. . . ] OPERATING INSTRUCTIONS FOR THE SANTA BARBARA INSTRUMENT GROUP SELF GUIDED SPECTROGRAPH (SGS) AND SPECTRA ANALYSIS SOFTWARE Alan Holmes 3/30/2001 1 SBIG Self Guided Spectrograph (SGS) Operating Instructions Alan Holmes 3/30/01 1. 0 - Overview: this document describes the operation of SBIG's self guided spectrograph and the installation and use of our analysis software. This instrument has been optimized to capture stellar spectra with high resolution, but has enough sensitivity and flexibility to allow its use on brighter galaxies and emission nebula. This unit is a scientific instrument: we expended considerable effort in making collection of the spectra easy, but you will find that a good spectrum of an object requires significant care and effort. Analysis of the data for astronomical meaning is beyond the scope of this document. [. . . ] I have investigated this problem and found that darks on separate nights indeed do not subtract well, but it is not a temperature stabilization problem. It is as if something changes in the CCD slowly over time, and does not repeat over long periods. I recommend the user plan on taking long darks to match the exposures on the same night the spectra are collected. To collect a calibration spectra, position the calibration lamp to illuminate the opal diffuser on the bottom of the spectrograph and take an exposure long enough to get at least a few hundred counts of signal in each line. You should do this either directly before or after capturing the astronomical spectra, without removing the spectrograph from the telescope or disturbing it significantly. We have tried to make a very solid unit, but measuring line position to an accuracy of fractions of a pixel requires great care. For very critical work, such as measuring doppler shifts with high accuracy, we recommend you collect the calibration data simultaneously with the astronomical data. This is done by placing the light source several feet away from the diffuser, and leaving it on during the exposure. A fluorescent lamp will not work for this technique since there is so much continuum. You will need either a mercury or neon lamp. The biggest problem is not scale changes, where the dispersion of the system in angstroms per pixel changes during an exposure, but simple shifts. For this reason neon lines, even though they are inconveniently placed in the yellow and red, are useful. After the exposure one can use a fluorescent lamp to measure the scale factor. Try to position the spectral region under observation such that one can obtain at least two calibration lines. There is no perfect calibration source commercially available. Small neon indicator lamps positioned close to the diffuser work well for the red and yellow end of the spectra, but have nothing in the green and blue. Mercury (which is used in fluorescent lights) is the best easily available lamp source for this region. If your site is significantly light polluted you will see mercury and sodium lines in long exposure data, along with an airglow line at 5577 angstroms. Sodium absorption lines in galaxies are very useful for doing red shifts, so sodium lines in the sky background are quite annoying. The analysis software enables one to remove the bulk of the light pollution signal from the data. 6. 0 Use of the Analysis Software SBIG's WINDOWS 95/98 analysis software enables one to view the collected spectra graphically, to calibrate the spectra, to print out graphical data, and to produce text files with the raw data annotated with wavelength data. These text files can be viewed with popular spreadsheet programs, which have more powerful graphing capability. [. . . ] Galaxies are hard since they are faint, extended, and mostly have a continuum spectra. Capture some spectra in low resolution mode with the wide slit and long exposures, such as 10 to 20 minutes, with multiple exposures. One usually does not need a calibration lamp on during the exposure since light pollution and natural airglow lines will be recorded, but at a low level. It is much easier to measure red shifts for Seyfert galaxies, which have an excess of H-alpha and show emission lines, than other galaxies. [. . . ]