A Grism Spectrometer

John A. Blackwell
johnb@regulusastro.com

Introduction:

First, an introduction. The reason I have written this is simply because I am getting many emails concerning how I have made a grism spectrograph and how it is used. I highly recommend that you also read Christian Buil's marvelous website located at: http://astrosurf.org/buil/ I make no attempt to recreate his work here, but endeavor to show you how I made my grism and how it works for me.

What is a Grism?

A Grism is part diffraction grating and part wedge prism: Grating + Prism = Grism. It is really that easy, too.

My grating is a Rainbow Optics transmission grating available from Astrovid. It is a grating in between two pieces of glass in a standard 1.25" filter housing. It has 200 lines per mm and is blazed for the first order to maximize efficiency. Efficiency seems to vary from one to another. Christian Buil's site states that it is 67% efficient in the first order.

The wedge prism is from Edmund Scientific. It is part number H43651 for a 3.87 degree wedge prism that fits within the ring of the 1.25" filter holder of the grating. You can get it with or without coatings.

The raw materials to make a grism These are the two primary components to the grism: a wedge prism (top) and a transmission grating in a filter cell (bottom).

Building Your Own:

Attaching the wedge prism (now "WP") to the grating is tricky, important and tedious. Do it once and do it well. The goal is to align the WP such that the first order spectrum is brought back to the center of the optical axis. This reduces coma and allows the spectrum to come to a focus across its length on your CCD chip. The zero order star image will, of course, not be in focus when the spectrum is in focus. This is not an issue.

To align the two parts prior to gluing, I did the following: I made a PowerPoint slide that was black. On it, I made four white dots and a couple of white lines. The dots act like stars. The lines have to be parallel to each other, one above the other. Holding the grating up to this slide as it appears on a CRT will show the spectrum of your CRT! Note that you will see two spectra: one to each side of the bright dot. The brighter of the two is the first order spectra, and is the one you want to bring back to the center of the optical axis. So, for the sake of this paper, rotate the grating until the 1st order spectrum is to the right of the dot as you look at your CRT. Make sure the spectrum is aligned with the parallel lines you made on the slide. Now, grab the wedge prism. Be careful not to mark it with finger prints. Hold it in front of the grating. Rotate just the WP until the spectrum is brought back towards the zero order image. Hold it in place and mark the edges of the WP and the grating with a marker to establish alignment.

Aligning the wedge prism and the grating

To attach one to the other I used hot melt glue. It works. It is not elegant. Apply it to the edge of the WP and to the cell of the grating. It can be removed easily and does not hurt anything. Three small dabs will do the job. It is important to have the WP on the telescope side of the grating. If you are placing your grism into a CFW-8 filter wheel, then the wedge prism must be glued onto the threaded side of the grating. NOTE this assembly will NOT fit into an SBIG CFW-10 filter wheel. The grism is just too thick.

Place this newly made grism assembly into your filter wheel or thread it into the nose of your CCD. I like to keep the grating aligned such that the spectrum runs north to south. That way any periodic error will not harm the spectrum image but will just spread it out.

The finished product
This is the finished product. I have used two longer strips of hot melt glue for this one, and it works just fine. Be careful to avoid getting the stringy glue across the optics. Yuk. It does peel off, but can leave a residual blemish depending on the glue's ingredients.

Using the Grism:

Since this is a transmission grating, you can see stars to guide on! It works well, but remember to focus on the spectrum and not the stars. This is best done by finding a star with prominent lines and focusing on them.

Armed with this system, you have the ability to take a full-field of spectra of multiple objects. You are limited only by the size of your CCD chip and the telescope you use. For my ST-7 on an FSQ-106n (106mm f/5) refractor, I get about 17 Angstroms per pixel. This is good enough to show the hydrogen Balmer series and a whole host of other features in point source spectra. Here are two examples:

This first spectra is of Gamma Gem, a type A0iv star which is used as a spectral standard. You can see the obvious H-Balmer absorption series in this.

The spectra of Gamma Gem A0iv

This second image is of the variable star Theta CrB, a hot class B dwarf with a very fast rotational period of about 10 hours! It has an excited gas shell causing emission lines and is labeled a Be (B-Emission) star. Note that I have left the 0-order image of the star itself for you to see here. Having the 0-order image helps in several ways: Identification of fainter objects is made easier. If you know the typical angstrom/pixel value of your system, wavelength calibration is much easier when studying unknown objects.

The spectra of Theta CrB

Additional Resources:

Please do check out these great sites. I have not covered the math, theory, and incredible detail that Christian Buil has on his site. Also, I know you are worried about the calibration process. You must get a copy of the superb freeware by Valerie Desnoux called VSpec.

A Low Cost Spectrograph, by Christian Buil
Visual Spec Home Page
Spectra that I have taken

Contents	What's Up	Feature	Papers	Photos
	Links	Observations	Spectra	Sketches