Of Wavelengths and Calibration

The eShel spectrograph from Shelyak Instruments in France has arrived, and with it a lot of excitement about all the possibilities of projects for the future! The unit came into the country and stopped for a brief stint at an airport customs office before making its way here to Exeter. The box was about 45 pounds in weight and had a ton of individually wrapped goodies inside, all in excellent condition. Packaging was perfect and survived the trip from France without any issues.

Unpacking the unit is best done with a checklist and table nearby in a clean area. I have a very clean basement, dry too. The job was carried out there. Some serious time was spent learning the proper id of all the connecting pieces, some of which are not all that obvious at first glance. The majority were easy to identify: CCD camera and power supply, the optical head which attaches to the telescope, the fiber optic and power cables, the spectrograph and its lens… all easy enough. How to connect them? That is the fun part!


Here is the system completely hooked up and ready to go… with everything but a telescope for the little optical head. If you get one of these: note the use of the correct fiber cables for the starlight going into the spectrograph and the feed for the calibration lamps. The calibration lamp cable is larger than the 50 micrometer diameter cable for the starlight. Also note that the AC power goes into the calibration light box first, then is jumpered to the power supply for the thorium lamp which feeds a 10mA supply back to the calibration box. It is a little confusing, but the insides of the calibration box need power to handle logical switching directions fed into it via a com port. This brings up some interesting points: you need a PC with com ports.

The supplied software is an extension of the AudeLA package. The whole systems runs on top of that. It will manage your CCD imager (power and temperature), a guiding video camera, telescope guide control, the automatic on/off of the calibration lamps (yes, two: one is thorium for wavelength calibration, the other a white LED for flat fields! Cool!), the flip mirror in the guide head which allows either object spectra or imaging the calibration lamps, and the processing of all the final images automatically. It is pretty amazing. The setup is a bit tricky, but not bad. Make note: the software initially expects you to have already taken and formed an instrumental response FITS file… but you can’t do that until you have taken spectra of some standard star or blackbody radiator with a known temperature like a tungsten lamp. There is a checkbox to turn that feature off until you are ready. None of your object spectra will be calibrated automatically until you take care of this.


The spectrograph uses a Canon EOS lens to focus unto a QSI CCD imager (blue above). Focus is achieved by focusing the lens as you would any camera. No, it does not autofocus! The beauty of this design is that the spectrograph can reside 20m away from the telescope in a vibration free temperature controlled environment which makes for very sharp spectra and excellent calibration. Finding exoplanets using radial velocity measurements will be pretty easy with this!

The system is an Echelle spectrograph using an Echelle grating to make a merged spectrum first then is spread out using a secondary prism to break the orders of spectra apart for imaging. The result is an interesting and complicated field of many curved spectral orders. All this must be sorted out! Here is the spectrum of the sun (blue clouds actually) for example.


How to make sense of that!? Well, the software can take the calibration lamp spectrum of thorium and use it to identify the field curvature, the angle of the curves, and the wavelengths across the field: automatically. It then provides both individual spectra for each order and for the whole spectra combined.



Above are the spectra of the sun near H-alpha and then the combined spectra of the orders from about 500nm to 670nm. Note that the true blackbody curve of the sun is not evident here, because the unit has not been calibrated (YET) for instrumental response. We’re almost there! Once that is done, the wavelengths AND the flux values will be calibrated and the curves shall be accurate.

About johnb

- Director of Grainger Observatory, Phillips Exeter Academy. - Variable-star-crazed astronomer, but have done research in other areas. - Drummer, archer, pilot, chef, friend, pet owner, husband, father, Train-nut.
This entry was posted in Astronomy, Education, Tech Talk and tagged , . Bookmark the permalink.