Biased and (Possibly) Helpful Information for Using the Goodman Spectrograph on SOAR


Not all grating lists you might come across are up to date.This one is: list. We typically use the 400, 930, 1200Red, and 2100 l/mm gratings. (The 1800 l/mm grating is a nice, newer grating, but I haven't used it). The 2100 l/mm grating is the highest-resolution grating that can be used at H-alpha. The 2400 l/mm grating can be used up to a central wavelength of only ~ 550 nm, and was specially fabricated to cover the Mgb region. The 930 l/mm grating is fantastic in the UV/blue.


Given the 0.15" pixels of Goodman, the rarity of better than delivered 0.6" seeing at SOAR, and the dispersion of the gratings, we nearly always bin 2x2. An exception would be if you are using the highest resolution gratings with the 0.45" slit: in this case you will not want to bin (at least in the spectral direction).

Slits and Seeing

Due to the typical delivered image quality and that fact that we often work at intermediate airmasses, we nearly always use the 1.2" slit to ensure light losses are minimal. Getting a delivered image quality of 0.7" or better is rare.

Region of Interest (ROI)

Unless we are imaging a large field, we take all observations---both acquisition and science---in the Spectroscopic 2x2 ROI for fast readout and data transfer. There is a custom ROI for acquisition that may offer slight gains.

Readout Modes

You are recommended to use 344kHz/ATTN 3 for science data and 750kHz/ATTN 0 for acquistion, and these almost always work well unless, for instance, you don't want that high of a gain.


Focus in increments of 200 units. You might find that the focus offsets between different gratings are consistent night-to-night within a run, but in my experience they are typically not consistent between runs, presumably because the gratings are removed and remounted.

Wavelength Calibration

We use the FeAr lamp (not selectable from the GUI) for all wavelength calibration, with a baseline exposure time of 120s for a 1.2" slit for the high-resolution gratings. Typical rms of wavelength solutions for the high-resolution gratings (2100;2400 l/mm) is < 0.01 A (~ 0.5 km/s); your systematics will always be worse than this.

Flexure and Precise Radial Velocities

Despite flexure compensation, there is significant residual flexure in the instrument at a level that matters for obtaining precise radial velocities. We obtain arcs every 30 min at a minimum and additionally try to always use a wavelength range that includes reliably bright sky lines (of which the bluest is the 5577.34 A line) or telluric absorption features. If you are working further in the blue than this and care about accurate velocities, obtain even more arcs.

If you want to make sure to include 5577.34 in your setup, central lambdas that ensure this include 2400 l/mm @ 541 nm and 1200 l/mm @ 610nm.

Note that the grating angle does not repeat perfectly, i.e., if you move the grating (e.g., to recenter an object in the slit), you need to take another arc.

Time Stamps

The FITS keyword OPENTIME gives a GPS-synced shutter opening time. Add (EXPTIME/2) to this value and you have a UTC at observation midpoint that's quite good.

Slewing and Pointing

The slew speed and absolute and offset pointing accuracy of SOAR are excellent. You will find that SOAR is better than most other telescopes in these aspects.