New Spectrograph

In November of 2012 I purchased a new spectrograph for the observatory from a fellow amateur and, as it turns out, long-time ProTools victim, Ron DiIulio in Texas.  It’s a LHiRes III (for Littrow High Resolution), manufactured in France by Shelyak Instruments.  Ron had purchased the instrument early in 2011 but just did not have the time to use it.  I found his ad on Astromart where he’d listed it for sale back in January.  It was quite a nice package deal as it included every grating and every slit width that Shelyak provides and so is quite versatile, offering resolution as high as 0.11 Angstroms.  Including considerations of slit width, camera pixel size, and sampling theory it should allow me to achieve a systematic maximum resolution of around 0.4 Angstroms.

The new LHiRes III attached to the tail end of the #1 Celestron 14-inch telescope.

The new LHiRes III attached to the tail end of the #1 Celestron 14-inch telescope.

Over the past month or so I’ve had the opportunity to learn a bit about the instrument and to do some calibrations and finally to baptize the new instrument with first light, that having occurred while many of my friends were either in or just off the coast of Australia for the total solar eclipse that happened on November 13th.

Calibrating the Grating Micrometer

The first thing I calibrated on the spectrograph was the micrometer that adjusts the grating angle, thus bringing different parts of the spectrum into the field of view of the acquisition camera.  Each grating will need to be similarly calibrated, but for the time being I only measured the 2400 line/mm grating – the grating which provides the highest resolution for the instrument.  I took a series of spectra of both the internal Neon calibration lamp as well as a Mercury gas discharge tube and measured the micrometer reading required to place various emission lines at the center of the ccd chip.  The figure, below, shows a plot of the results.

The wavelength range for the 2400 line/mm grating and ST8xme camera is about 175 Angstroms, at a scale of about 0.11 Angstroms per pixel.  In practice, however, I’ll likely be using a slit width that is about four times the detector’s native pixel size.  So I will usually be binning the data 2×2, resulting in a system resolution in the area of 0.4 Angstroms. Also, just looking at the plot and noting the locations of the Neon and Mercury lines it’s pretty clear that I’ll need to build or acquire an improved calibration lamp setup as wavelength calibration will be difficult or impossible anywhere blueward of about 5600 angstroms. I see that some people are using Thorium-Argon lamps, which makes sense given the wealth of bright lines across the entire spectrum, but at well over $2000 it is something that will have to wait!

Focusing the LHiRes

The next item up was to focus the instrument.  In the Littrow design the same lens is used both to collimate the light coming from the spectrograph’s slit onto the dispersing element (in this case a reflective grating) and to focus the light returning from the grating onto the acquisition camera’s CCD.  In fact, the distance from slit to collimator must be the exact same as the distance from the collimator to the imaging device.  The process of focusing the spectrograph is achieved in two steps; first the collimator is adjusted until the light coming from the slit is collimated (the rays made parallel) onto the grating, then the camera is adjusted to best focus by moving it either closer or farther from the collimator.

Focusing the Collimator

When the collimator is at the correct distance from the slit the light rays heading towards the grating are made parallel.  This is the same condition for light rays originating from an object which is at infinite distance,  So the strategy for setting the collimator position is to remove the spectrograph’s grating and in its place insert a telephoto camera lens and camera.  It is then easy to simply take exposures of light coming from the slit and adjust the collimator position until the slit is focused in the camera.  I used my Nikon D5100 DSLR along with a 300mm telephoto lens which I’d previously adjusted to focus at infinity.  The collimator lens in the LHiRes is a simple two-element achromat housed in a helical focusing device with a 1mm thread pitch; that is, a full turn of the lens in it’s focusing holder will move lens 1mm either towards or away from the slit.  The images, below, show the focusing setup and an image of the focused slit.

Focusing the Camera

Once the collimator lens is locked into position it is time to assure that the acquisition camera is at the proper distance from the collimator.  Shelyak Instruments makes a variety of camera adapters for this purpose and the spectrograph includes, as standard equipment, an adapter for the Santa Barbara Instrument Group ST-series cameras.  That’s plenty convenient since that’s precisely the model of camera I use.

In a perfect world that would be the end of it; simply attach the camera and away you go.  There can, however, be small manufacturing inconsistencies in both the focal length of the collimator optics and in the positioning of various components within the spectrograph itself, typically on the order of small parts of a millimeter.  The ST cameras, also, have undergone several revisions and changes.  Taken together these considerations mean that once the camera is attached it may be necessary to tweak the collimator to get the best focus onto the camera’s CCD chip.   Tests conducted by several people, most notably by French astronomer Christian Buil (one of the true pioneers in the field – his website is a treasure trove of information), have shown that measurable image degradation occurs if the collimator has to be adjusted more than about 250 microns (one quarter millimeter) from it’s optimal location relative to the slit.

Once I finally got the correct adapter attached I was relieved to find that the collimator required only very little adjustment, about 60 microns, to bring images of Neon emission lines from the internal wavelength calibration lamp into focus.  Shown below is one of the first spectra taken with the LHiRes from its new home.  It shows the absorption lines of H-alpha (Hydrogen, on the left side) and Helium (to the right).



  1. Tõnis Eenmäe says:

    I’m not familiar with Lhires III and dimensions of it’s calibration lamp. Maybe you’ll find useful a much cheaper ThAr lamp, e.g.

    • SleepIsWrong says:

      Hello again Tõnis:

      Since posting that article I’ve upgraded the calibration lamp and mechanism to include a Neon+Argon lamp, giving me many more lines for calibration. I’ve looked into the ThAr lamps and considered trying a flip mirror to inject the lamp’s light into the spectrograph, but since adding the new NeAr lamp the number of calibration lines available has been sufficient. The LHiRes is a pretty small device and the lamp you pointed to would definitely not fit inside the enclosure. Also, while the lamp is almost $600, the power supply looks to be around $2100. The same company that makes the LHiRes (Shelyak Instruments) also offers a fiber-fed echelle spectrograph which uses a ThAr lamp for wavelength calibration.


Leave a Comment