We designed and built in our workshop a new spectrograph for our 60cm F20 Cassegrain telescope. New design take advantage of experience
from the use of previous longslit spectrograph that is working since october 2004 in monitorning of novae
and symbiotic stars within ANS program .
First mechanical parts were machined on Feb 2006 and test at the telescope started on July 2008.
Picture 1: Two projects are starting at home. Carlo and first parts of the echelle-longslit spectrograph. In both cases I had acquired some
experiece before. Carlo is following Giulio and Lucia, and the echelle is following at least 3 generation of longslit grating spectrographs.
When we purchased gratings from Optometrics, we saw that echelle gratings were available too at an affordable price and decided to make
echelle-mode an optional optical beam in the spectrograph. Following drawings show how we obtained this, using an additional mirror and a
transmission grating (300 l/mm) as cross-disperser. Echelle grating is a 79 l/mm R2 grating (25x50mm) and due to the 14░ working angle is a little overfilled
by the 20 mm diameter optical beam resulting in a 10% loss of light.
Picture 2: Both echelle and longslit mode are available in our spectrograph. Switching from echelle to normal grating can be done in less than
5 minutes. Both dispersive system use the same parallel F/20 beam from the spherical collimator f=400 mirror.
Picture 3: Some mechanical parts ready for assembly. On the left size, an hollow Cathode Lamp Fe(Ar) used for reference.
Picture 4: Time is running and Carlo is now very interested in testing mechanical parts...
Picture 5: Chassis is ready for painting black. On the background the tools I used. nothing more than a drill and a lathe.
Picture 6: Slit of the spectrograph is cutted into two stainless steel knife edged lips that are mirrorized. A decker is used to limit the slit height for
echelle mode (in echelle mode slit is usualli setted to 2x16arcsec)
Picture 7: An alignement setup was prepared that simulate the telescope entrace axis with a laser beam
Picture 8: Aligmement of optical parts is going on... Light from ref lamps come throught a 1mm core optical fiber that is focalized as F20 beam
onto the slit.
Picture 9: Gratings are interchangeable. Available dispersions are 4.3 A/pix (300 l/mm), 2.1 A/pix (600 l/mm), 0.75 A/pix (1200 l/mm) and 0.2 A/pix (approx) with echelle.
Picture 10: Lamps for wavelength calibration are contained in a box attached to the tube of the telescope. Available lamps are at the moment a mixture of
Ar (a lot of Ar and Ar+ lines) and Ne lamp. Intensity of Ar lines redward 6995 A is reduced with filters. A HCL Fe(Ar) lamp is also available for high resolution.
We realized also the power supply for this lamp. Light from the lamps is collimated with mirrors and lenses and then an aspheric condenser focuses the light
into a 1mm core fiber that goes to the spectrograph.
Picture 12: A fascinating image of all spectral lamps swtched on. HCL Fe(Ne) lamp is the bigger one. The long blue is Ar and the smalla one in Ne.
Unfortunately Fe(Ne) lamp is rich of line only in the regions 4000-6000 and over 7300 A. A Th(Ar) lamp is recognized to be the best solution for
wavelength calibration but the cost is at the moment too high for our budget.
Picture 11: Ar+Ne ref lamp recorded in echelle mode from 4100 (left size) to 7700 A (right size). Number of lines per order is adequate only in the
blue-green-orange only untill 6000 A. We hope that HCL could improve number of lines for calibration in the range 6000-7800.
Picture 11: Spectrograph on July 2008 is finally mounted at Cassegrain focus of the 60 cm telescope atop Campo dei Fiori mountain (1226 m) and
test can start.
Picture 12: Twilight sky spectrum with solar lines. Measured spectral resolution with 2 arcsec slit is 17'000.
Picture 13: spectrum of Radial Velocity standard HR 8308. Tests on a set of radial velocity standards show that radial velocities measured with our
echelle spectrograph are accurate to 500 m/sec if calibration lamp frames are recorded before and after the science exposure of 900 sec.
Picture 14: Raw spectrum of CI Cyg at beginning of outburst is the first science target of our new spectrograph in echelle mode. 6x900 sec exposure on 16th Sept 2008.
Picture 15: Orders extracted from previous 2D image containing Halpha, Hbeta and He I 5876. Amount of information on a single echelle spectrum is
huge. S/N is not bad for a V=9.6 star.
Picture 16: Histeresis pattern of flexures in the spectrograph (echelle mode) as a funcion of Hour Angle. Note that this measurement was done
in 1x1 CCD binning while scientific operation is with 2x2 binning of CCD.
Picture 17: Data on dispersion, central wavelenght and spectral coverage of the orders that are recorded on the CCD in our echelle spectrograph.
Improvement in spectral coverage and overlap of orders will be obtained with a new 100mm F2 camera optics for the echelle that can be exchanged
with the 135 mm F2.8 that is working now.
Picture 18: With the long-slit mode we have available 2 gratings 600 and 1200 l/mm. Dispersion are 2.1 A/pixel and 0.7 A/pixel. Spectral coverage
is particularly suited for novae spectral recording because we can secure in a single frame low dispersion spectra from 3800 to 8500 A thus including
important Ne and O lines at both end of the spectrum. Resolving power is slightly better than R=1000 with 3" slit. With 1200 l/mm grating the spectral range
can be selected with a micrometric screw but usually it encompass from 5500 to 7000 A including He I, interstellar Na and Halfa.
In this picture is illustrated a low dispersion spectrum of Symbiotic star CI_Cyg. At V=9.6 it is an easy target. A 1h exposure gives very high S/N in the
range 4200-8000 A and still acceptable down to 3800 A. The insert shows a zoomed region with faint and sharp Fe lines.