The Coude Echelle Spectrographs
1.1 Introduction
Sections:
UCLES & UHRF | The
user manual | Development
of the spectrographs | Echelle spectrographs
| Description of UCLES &
UHRF | Coude stability | Further
reading
Next: UCLES Hardware
|
Contents | Home
UCLES & UHRF
The University College London Echelle Spectrograph (UCLES) is located in
the east coudé room on the 4th floor of the AAT building. UCLES
can record spectra of astronomical sources as faint as V = 16-17 with a
resolving power (lambda/(delta lambda)) of 50,000-80,000 depending upon
the detector used. The spectrograph is operated entirely under computer
control from within the ADAM software environment.
UHRF is a cross-dispersed echelle which provides resolving powers of
300000, 600000 and 1000000, intended primarily for interstellar line studies.
UHRF shares pre-slit optics with UCLES, and is controlled from the same
software.
The User Manual
The UCL Coude Spectrographs Manual describes the components and operation
of UCLES and UHRF on the AAT. This version differs from its predecessor
in removing all mention of the IPCS detector and adding more recent CCD
information. It has also been streamlined to become more web-friendly.
Those working with archival data may wish to refer to AAO
UM 25.2, by Sean Ryan and Adrian Fish which was produced in 1995 following
the unification of UCLES and UHRF under a common control system. It draws
heavily on the two earlier manuals ``The UCL Echelle Spectrograph" by R.
D. Robinson, F. Diego, A. C. Fish, W. F. Lupton, M. Pettini, & D. D.
Walker, and ``Ultra High Resolution Facility Operating Manual" by J. Spyromilio.
Development of the Spectrographs
The UCL Coudé Echelle Spectrographs form the high resolution optical
spectroscopic facility at the AAT. The Anglo-Australian Telescope (AAT)
was originally designed to include a classical photographic coudé
spectrograph. The mechanical support structure for a horizontal layout
was installed with the telescope, but the spectrograph didn't materialize
due to financial pressures. Around the early 1980s, AAT users became increasingly
aware of the need for a high resolving power facility and in 1983 the Advisory
Committee for Instrumentation at the AAT (ACIAAT) gave a high priority
to the construction of an echelle spectrograph.
The f/36 coudé focus was preferred over Cassegrain for several
reasons:
-
Mechanical and thermal stability is guaranteed in this environment and
simple mechanical structures and mechanisms can be employed.
-
The large space available allows a large collimated beam size which can
partially overcome the throughput disadvantage due to light losses in the
five mirror coudé train.
-
Optical components can be remotely interchangeable and duplicated with
coatings optimised for different wavelength bands.
-
The coudé environment lends itself to an open, optical bench layout
which can accommodate future and unexpected enhancements.
-
The coudé system can be permanently operational, permitting fast
changeover from other foci during the night due to instrument failure,
changing observing conditions, flexible scheduling etc.
As a result of ACIAAT's recommendations, UCLES was built by David Walker
and his team at the Department of Physics and Astronomy, University College
London under contract to the AAO.
In early 1990, the Ultra High Resolution Facility (UHRF) was added to
the coudé spectrograph. UHRF was built by the Optical Sciences Laboratory
of University College London, funded by a grant from SERC and the AAO.
Echelle Spectrographs
-
An echelle operates at high blaze angle (~63 deg) and order numbers (50
- 150).
-
The blaze angle is the angle between the surface of the grating and the
reflective groove faces.
-
The high blaze angle permits a comparatively wide spectrograph entrance
slit and therefore better light throughput at a given resolution.
-
The high order numbers result in many orders overlapping, but these can
be separated by cross-dispersing, thus giving large spectral coverage while
maintaining high spectral resolution.
Wavelengths which overlap in the undispersed echelle spectrum are related
by
n1W1 = n2W2 = ... = niWi
where ni are the order numbers. The wavelengths Wi
thus occupy similar positions in adjacent orders of the cross-dispersed
echelle. A given wavelength will appear in several orders, but each echelle
order has its own intensity profile (blaze profile) resembling a sinc function,
so the intensity at the same wavelength will differ in each order. In practice,
a wavelength is concentrated in only one or two orders.
The Free Spectral Range (FSR)
The FSR is delineated in each order by two wavelengths which appear at
equal distances from the peak of the order, and peaks of adjacent orders.
Observations restricted to the FSR would provide complete wavelength coverage
without any duplication, whereas if the detector is narrower than the FSR,
there will be gaps in the spectral coverage. The length of the FSR on the
detector increases proportional to the central wavelength of the order,
giving the familiar trapezoidal format of an echelle spectrum. Note that
the spectrum does not terminate at the end of the free spectral range -
it continues at lower intensity away from the blaze peak.
Dispersion and Resolution
The dispersion in Å mm-1 is proportional to wavelength,
but the resolving power (lambda/delta(lambda)) and the velocity resolution
is constant. Also, the dispersion at a given wavelength depends only on
the angle at which the grating is used, not on its groove-spacing. Thus,
the two echelle gratings available in UCLES (with 31.6 and 79 lines mm-1)
give the SAME dispersion because their blaze angles are the same. The primary
difference is in FSR and order separation.
Description of UCLES & UHRF
UCLES
-
UCLES gives images with FWHM < 20 um over the 38.5 x 18.8 mm field of
its 70 cm focal length camera. The resolving power is therefore dictated
by the slit width and pixel size. As one example, a detector resolution
element of 48 um corresponds to R = 50000, and projects to 0.9" (or 1.2"
with the focal modifier) at the slit. Existing CCDs yield maximum resolving
powers of ~50000 (Tek) and ~80000 (MITLL2A/MITLL3)
-
There is a choice of two gratings: 31.6 and 79 lines mm-1. Both
produce the same dispersion. The free spectral range (FSR) and order
separation are 2.5 times larger for the 79 lines mm-1 grating
than for the 31 lines mm-1 grating.
-
The 31 echelle gives more complete spectral coverage but at the expense
of good sky coverage. (The permissible slit length can be increased by
37% by using the focal modifiers) With this echelle, it is possible to
cover the full FSR of every order using the MITLL CCDs (or all orders blueward
of 6000 Å with the Tek), but with orders separated by only a few
arcsec.
-
With the 79 lines mm-1 grating, even the MITLL CCDs do not cover
the full FSR redward of ~5400 Å, but the order separation is greater,
allowing a ~9" long slit in the red and greater in the blue.
Links:
Tables for the 31 grating
(Appendix A1) and 79 gratings
(Appendix A2) giving the wavelength coverage, free spectral range,
dispersion and order separation for each grating.
Echellograms for 31
and 79 gratings indicating
the relative sizes of detectors.
Fact sheet summarising the basic properties
of UCLES.
Cookbook for UCLES
with some useful hints.
UHRF
UHRF has three resolving powers of nominally 300000 (0.3M), 600000 (0.6M),
and 1000000 (1.0M - actually 940000). The lower resolving powers are obtained
by inserting focal reducing lenses into the UHRF camera optical path, which
also change the wavelength coverage. Note that the detector must also be
moved to a different mounting point for each resolution, so changing over
during the night is not practical.
UHRF has a single echelle with a suite of cross dispersing gratings
suitable for different wavelength ranges. The very small wavelength coverage
of the detector (typically 2 - 15 Å) dictates most setup parameters.
Usually only a single order of spectrum is observed, widened in the spatial
direction due to the necessity of using an image slicer to obtain adequate
throughput with the small slit width. A few orders can be observed simultaneously
at the lowest resolving power (300000), but extra orders rarely contain
a wavelength of interest.
Note that since observations are normally made with the detector binned
by 4x or 8x to minimise the readout noise contribution, the MITLL3 detector
(which does not offer binning) is not suitable for use with UHRF.
Links:
Tables (Appendix A3) giving the wavelength
coverage, free spectral range, dispersion and order separation.
Fact sheet summarising the basic properties
of UHRF.
Cookbook for UHRF with some useful hints.
Coude Stability
Important Note to Observers:
All spectrograph mechanisms are under software control, and
user access to the coudé room is neither necessary nor desirable.
Pressure changes affect the refractive index of air in the room, and these
will be evident in the data. The user is not permitted to enter
the main spectrograph room unless prior authorization from the support
staff has been obtained, and then clean room clothing must be worn. Do
not switch on the fluorescent lights in the coudé room during a
run, or even 12 hours before the observations are due to start, as the
walls fluoresce and the dark count will be seriously increased. Use only
tungsten lamps while in the room.
The coudé location of the spectrograph provides great stability.
For UCLES, random shifts over one hour are <0.01 pixels (peak to peak),
and slow drifts are consistent with refractive index changes due to pressure
variations. Even when grating settings are changed many times during a
night, it is possible to return to a previous setting with an RMS accuracy
of 4 um (0.2 pixels).
Repeat exposures with UHRF at R = 1.0M using a stabilized laser have
shown the instrument to be stable to +/-0.2 pixels over 30 minutes. To
maintain resolution on a star, limit integration times to around 30 minutes
to avoid smearing by changes in the heliocentric correction of the star.
Although the coudé optical bench floats on airbags to vibrationally
isolate it from the hammerhead, which is itself separate from the building,
arcs should not be taken while the telescope and/or dome are slewing. It
is advisable to bracket stellar exposures by arc calibrations.
The CCD dewars can be filled from a 25 litre storage dewar by activating
a switch in the coude ante-room, so it is no longer necessary to disturb
the coude room to fill the CCDs. The 25 litre dewar must be replaced every
three days.
Further Reading
For additional information on these spectrographs, readers may wish to
consult the following papers:
-
Barlow, M. J., Crawford, I. A., Diego, F., Dryburgh, M., Fish, A. C., Howarth,
I. D., Spyromilio, J., & Walker, D. D. 1995, MNRAS,
272, 333-345
First results from the UHRF: ultra-high-resolution observations of
atomic interstellar lines towards Zeta Ophiuchi"
-
Diego, F. de A., 1988, Ph.D. Thesis, University College London.
``U.C.L. Echelle Spectrograph (UCLES) for the Coudé Focus of
the Anglo-Australian Telescope: Optical Design and Performance"
-
Diego, F., et al. 1995, MNRAS,
272, 323-332
``The Ultra-High-Resolution Facility at the Anglo-Australian Telescope"
-
Diego, F. & Walker, D. D. 1985, MNRAS,
217, 347
``On the possibility of increasing the throughput of astronomical spectrographs
by overfilling the dispersing element"
-
Walker, D. D. & Diego, F. 1985, MNRAS,
217, 355-365
``Design philosophy of the forthcoming échelle spectrographs
for the AAT and LPO"
Sections:
UCLES & UHRF | The
user manual | Development
of the spectrographs | Echelle spectrographs
| Description of UCLES &
UHRF | Coude stability | Further
reading
Next: UCLES Hardware
|
Contents | Home
Ray Stathakis (ras@aaoepp.aao.gov.au)
Last Update: 3 October 2000