(c) Maurice Gavin - 2000
   Simple spectroscopes [this page]   New spectrograph    Amateur 'players'    Terminology  Books
These pages are intended to provide some practical guidance for newcomers to the exciting field of astronomical spectroscopy. A spectroscope can provide data on the composition, temperature, rotation, relative velocity [via the Doppler Effect] and ionization [atmospheric electron stripping] of a star and indirectly its luminosity, mass, age and distance.
  • Doppler Effect : we're all familiar with a passing express train - the drop in pitch as it recedes caused by fewer soundwave reaching us.  The same effect would be observed for a lamp on the front of the train.
  • train moving at say 100km/h [0.028km/s] or
  • a distant star typically moves x1000 faster [28km/s].
  • quasar 3C273 receding 1.6 million times faster [48,000km/s] than the train
  • a quasar Doppler shift is relatively easy to measure.
  • a quasar recessional velocity or redshift is a sizeable fraction of lightspeed [300,000km/s] and the effects of Einstein's relativity comes into play.
  • A spectroscope is used to compare lines [see below] from the moving light source with those from a fixed reference source applied at the instrument.
    • Now let's start with some simple spectroscopes that will quickly get you involved.....

    Simple spectroscopes: try some of these....

    [1] - solar spectroscope: the data recorded onto a CD-ROM [~600 lines or 'pits' per millimetre] acts as a reflective diffraction grating breaking up sunlight into a rainbow of colours or spectrum. With the eye placed a centimetre or so from the CD-ROM surface several orders of spectra can be seen.  By tilting the CD-ROM further so sunlight just skims its surface, dark absorption lines may be seen crossing the spectrum. No other optics are needed !  These lines are caused by elements within the sun's atmosphere like hydrogen [H], helium [He], magnesium [Mg], sodium [Na], iron [Fe] and so on.  Although ideal to demonstrate to your friends as a party piece, the CD-ROM surface is unfortunately inadequate for practical spectroscopy.    NOTE: don't allow direct sunlight reflected off the CD-ROM to enter the eye - only the spectrum !


    This solar spectrum was captured by holding a fixed-focus digital camera 10mm from a 1200 lines per millimetre reflection grating [see items 4/5 below] with sunlight skimming the grating surface.   The narrow width of the spectrum represents the sun's unmagnified disk of 30'arc seen in the sky. The grating disperses it sideways into a spectrum where some dark absorption lines can just be seen in blue and green.

    Practical solar spectrograph: In a darkened room focus a camera on a chrome cylinder [eyepiece] over 1m away and via a plain mirror play sunlight [or fluorescent light] onto the cylinder.  The cylinder acts as a reflective slit spraying a narrow sample of light onto the camera.   A transmission grating or prism placed before the camera will disperse the light into a spectrum for recording by the camera.     Typically a grating  will give a higher resolution spectrum [like this sample and the page heading] than a prism - showing many dark absorption lines originating in the sun's atmosphere or bright emission lines from a fluorescent lamp.  Cheap acetate replica gratings in 9" x 9" sheets or 35mm slide-mounts [ideally suited for experiment and a solar spectroscope] plus high quality reflective gratings and prisms are available from EdmundScientific.

    Swapping a small finder telescope [or half binocular i.e. monocular] for the camera converts the instrument into a visual spectroscope.

    [2]  - a simple stellar spectroscope can be made by supporting a right-angled prism in a lenshood and allowing starlight to pass through one 45o face of the prism, as shown adjacent, into a regular film or CCD camera with the lens focused at infinity.  This arrangement is called on objective prism spectroscope.  The camera, supported on a fixed tripod, is aimed at a bright star [or planet] and the shutter opened for a minute or so.  The diurnal motion of the stars across the sky cause the narrow spectrum from the star to be trailed out into a rectangle with occasionally dark absorption lines superimposed.    If the prism/camera combo is piggy-backed on an equatorially driven telescope, many fainter stars can be recorded simultaneously with the shutter opened for say 5 minutes.   Experiment with camera lenses of different focal length - a 'longer' lens will record a longer more detailed spectrum but it will be somewhat fainter. NOTE: this arrangement only works on point sources like stars and for the best results the blank [frosted] sides of the prism should be parallel to the lines of RA in the sky.  These are vertical when looking due south.

    Discarded for free by a local binocular repair company, this slightly chipped prism [above right] is more than adequate for experimental use in a spectroscope.


     [3] - stellar spectrum via diffraction: simple devises like a tennis/badminton racket or a fine kitchen sieve can be placed before a telescope to form a crude spectrum - acting like a transmission diffraction grating.   These gratings are really too course to resolve a good single spectrum  [many overlapping spectra are created] but it's enlightening to test the principle.  A proper grating will have between 100 lines [or grooves] per millimetre [100 l/mm] and 1200 l/mm supported on an optically flat surface. The more lines per grating the higher the resolving power or detail in the spectrum.
    More advanced spectroscopes...
    [4] - classic focal plane stellar spectrograph using camera lenses and reflective grating: this shows a basic layout of a potentially powerful instrument for coupling to a focal plane of an equatorially driven telescope at the slit position.   The slit can be omitted for some types of work [i.e. extended objects like planetary nebulae or some emission line stars] which make aiming the telescope far less critical - the slit would typically subtend only a couple or arc-seconds wide on the sky and the selected star must be located on the slit for the duration of the longish exposure!  The spectrograph must be contained in a light-tight box and preferably rotated about the slit axis so that the spectrum is vertical [i.e. parallel to the lines of RA in the sky] thus minimize blurring of the spectrum through telescope tracking errors.  Any broadening [at right-angles to dispersion] can be recovered later by pixel binning. Full details on the construction of such a spectrograph, describing the resolving power, optical arrangement and use of gratings can be found on Christian Buil's excellent website.
    [5] - stellar spectrograph with autocollimator: first shown by the author at the BAA Exhibition Meeting in London in 1998, this compact and light-weight design is based around the Starlight Xpress range of CCD cameras.  It uses a tiny plane pick-off mirror [or right-angle prism] to send starlight from the telescope onto the grating via a single camera lens which serves the dual role of collimator and camera lens where both the telescope focal plane and CCD chip are at the common infinity focus of the lens. The spectrograph is light-tight and needs no enclosure except for the grating.

    As the spectrum occupies a few rows of pixels across the CCD, vignetting from the pick-off mirror is negligible.  Increasing the lens focal length increases the resolution of the spectrograph.  The optional slit is imaged at the focal plane full size [ratio 1:1] and thus typically has a width equal to the pixel size. Care needed to mask the slit image [reflected in the lens elements] from reaching the CCD detector and degrading results.