SOPHIE échelle spectrograph
Catalog Generator *
Article SPIE *
SOPHIE publications *
SOPHIE Archive *
SOPHIE is a
cross-dispersed échelle spectrograph permanently located in a
temperature-controlled chamber in the second floor of the 1.93-m telescope
light with this instrument was achieved on July 31, 2006, but the first
wavelength solution was obtained only on 24 August 2006. The spectrograph
is fed from the Cassegrain focus through either one of two separate optical
fiber sets, yielding two different spectral resolutions (HE and HR modes).
presently cover the wavelength range 3872-6943 Å.
The instrument is entirely
computer-controlled and a standard data reduction pipeline automatically
processes the data upon CCD readout. For late-type stars (F,G,K,M) this includes
radial velocities by numerical cross-correlation techniques
which can yield very accurate velocities (down to 2-3 m/s using simultaneous
Th calibration), depending on the signal-to-noise ratio.
Apertures & modes Each fiber set consists
of two circular entrance
apertures (3 arc-sec wide) separated by 1.86 arc-min
in the focal plane (nominally in the
E-W direction). They feed light into the spectrograph through optical
fibers of 100-micron diameter. Fiber A is normally used for the target while
fiber B can be used to obtain the sky spectrum or for a
simultaneous Thorium calibration lamp exposure. Fiber B can also be masked.
Both fibers A and B can be simultaneously illuminated by one of the
calibration lamps, or by the sky.
The HR fiber pair includes optical scramblers designed to render the
measured radial velocities largely insensitive to the exact position of
the star in the entrance aperture. These fibers also incorporate
a 40 micron exit slit to achieve high spectral resolution (HR mode,
R=75000). The other fiber pair is used when a higher throughput is desired
(HE mode, R=40000), particularly in the case of faint objects.
These average resolving powers near 5500 Å
have been derived from the FWHM of extracted Th-Ar spectra in both modes.
in throughput between the HE and the HR modes is a factor of 2.5 (1 magnitude).
Switching between the two modes involves moving the fiber heads in the adapter
which takes about 3 minutes.
Acquisition and guiding
The adapter (or bonnette) carrying the optical
fiber heads is mounted at the
Cassegrain focus of the 1.93-m telescope. This is the same
adapter used by Elodie which has been modified for Sophie.
It includes a new autoguider CCD camera, the calibration lamps and atmospheric refraction
correctors. The field of view of the autoguider is 4.6 arc-min.
The telescope pointing system features digital encoders in both declination and
hour angle with an accuracy of 5 arc-sec. A pointing model corrects for
mechanical flexure and allows for accurate setting of the telescope. The
coordinates are displayed in the observing room and in the dome.
There are limits
on the pointing of the telescope due to its asymmetrical english-type
The spectrograph is enclosed in a thermally insulated chamber whose temperature
is accurately regulated and humidity is under control.
This chamber is located in an air-conditioned room where are located the instrument
and telescope electronics. This room is separated from the observing area,
which also air-conditioned, by a special insulated wall. This three-level temperature control
ensures a high degree of thermal stability for the instrument. The spectrograph itself
is mounted on shock absorbers and is supported by the telescope pier structure. Its
dispersive elements are in a constant-volume vessel filled with Nitrogen so as
to insure a constant pressure and thus a constant index of refraction.
SOPHIE has now its own thermal insulation cover, so the final thermal
configuration has been achieved. The temperature within the constant-volume
vessel is now stable to ±0.01°.
illustration shows the principal optical elements of the SOPHIE
The 52.65 grooves/mm
R2 échelle grating manufactured by Richardson (RGL), blazed at
65°, has dimensions of 20.4x40.8 cm and is mounted in a fixed
configuration. The spectrum, as projected onto the e2V 44-82
CCD detector (4096x2048 pixels), yields 41 spectral orders, of which
39 are curently extracted, between
3872 and 6943 Å. See the list of
orders for information on the details of the spectral format.
Cross-dispersion is done with a OHARA PBL25Y glass prism of angle 31° and
dimensions 28x22 cm.
color image of the echelle spectrum projected onto a translucent screen,
obtained with a digital camera during the instrument integration process,
shows the approximate spectral area sampled by the CCD chip.
Two lamps are mounted in the Cassegrain
adapter : a tungsten lamp for relative flux calibration ("flat field")
and a Thorium lamp for wavelegth calibration. Exposure times for both
lamps are set automatically through the STS. Light from the Th lamp is
now filtered by a specially designed filter
to cut-off light from strong near-infrared Argon lines which were polluting the Th spectrum.
The present level of scattered light in the interorder space is about
0.6 % of the signal in the orders for stellar spectra, as measured at the center of order 26
(corresponding to the center of the V band). This scattered
component is due to infrared light from orders redward of 7000 Å which are not
directly recorded by the CCD but which illuminate the optical elements and create a faint diffuse
light which is superposed onto the observed spectrum. This was seen when the
Thorium filter was first installed (see above).
The CCD is a thinned, back-illuminated,
anti-reflection coated e2V 44-82 chip with 4096x2048 pixels of
15 microns. The chip can be read out using two different settings:
||RON||Read-out time||Saturation||Fast||2.85 e-/ADU||6.0 e- RMS||18.8 s||62000 ADU = 177000 e-
||Slow||0.68 e-/ADU||2.1 e- RMS||175 s
||65535 ADU = 44560 e- (25% well depth)
'Slow' read-out speed is recommened for S/N < 30. The bias level is measured over the full image when doing a
calibration sequence, but accurate values are measured for
every exposure using the 53-pixel overscan zones on each side of the CCD
frame. WARNING: Make sure to do a bias exposure prior to any observing
when changing CCD read-out speed (both for Fast to Slow and for Slow to Fast).
Calibration of the CCD gain and read-out noise can now be obtained using the new
LED lamps installed in the spectrograph. Preliminary results confirm the
above values for the Fast and Slow read-out speeds.
The dewar maintains the CCD at -99° for a maximum duration of 48 h, but is
automatically refilled every 24 h.
A plot shows the expected
S/N ratio (per pixel) at 5500 Å as a function of the magnitude of a star
(for standard exposure times of 5 minutes and 1 hour) in the two different
modes: High-Efficiency (HE) and High-Resolution (HR). Observations obtained
after the aluminization of the telescope optics in early September 2006
confirm these calculations.
An exposure meter is presently available which gives the real-time plot of
level. This system employs a H8259-01 Hamamatsu photomultiplier which
measures part of the light ultimately lost due to obturation by the grating. A
calibration of the count rate as a function of
magnitude has been obtained
in July and August 2007 for the HR and HE modes (plotted as black circles
and red triangles, respectively). This calibration is valid for obj_A
and obj_AB modes, but only valid for V<7 for thosimult mode.
The gain in efficiency with SOPHIE in the V band has been determined to be
3 magnitudes in HE mode relative to ELODIE (and consequently
2 magnitudes in HR mode) by comparing S/N values (per unit wavelength)
for the same star taken with the two spectrographs. Spectra of relatively faint
sources are now well within reach, as shown during the science demonstration
phase where a S/N (per pixel) of 27 (in V) was reached in 90 min for a V=14.5 object
in HE mode. For a comparison between spectra of the same star taken with
SOPHIE and ELODIE, see the case of a couple of interstellar lines present in
the line of sight to Zeta Persei.
The stability of radial velocity measurements in HR mode
is presently of the order of 3-4 m/s over several months. The
short term stability is much better, as revealed by a recent observing
run on Procyon where acousting mode oscillations of 50cm/s amplitude were
detected. The relative velocity drift between fibers A and B has been
measured to be 1.3 m/s, compatible with the error on the velocity
Three (actually four) different systems are used when observing with SOPHIE.
Please read the new version of the Software User's Guide (in french).
- All spectrograph functions are controlled via dedicated software on a
Windows workstation called 'pcsophie' which displays the complete status of
the spectrograph and of the exposure in progress. All exposures are started from
this computer. It talks to the CCD controller via a Linux workstation
- Observing sequences are prepared on a second Linux workstation called 'sophiests',
or STS for short. All information needed for a given exposure (calibration
or science frame) is normally entered through the STS. Parameters for the science
exposures come from a user catalog which must be prepared in advance.
These parameters are automatically transfered to 'pcsophie' where
the exposure will be started manually once the telescope has been
pointed to the target and the autoguider has been enabled. An estimate of the
required exposure time can be requested from the STS, but is computed for an
observation at the zenith.
- The reduction of the data is done automatically on a third Linux worstation
called 'sophiedrs', or DRS for short. For a science frame, this reduction
includes bias subtraction, optimal order extraction,
cosmic-ray removal, flat-fielding, wavelength calibration,
cross-correlation with a suitable numerical mask and merging of the spectral
orders. An off-line facility is available on the DRS for detailed
examination of the reduced data and for additional cross-correlations.
The SOPHIE spectrograph features an entirely automatic data-reduction pipeline,
adapted from the HARPS software designed by Geneva Observatory. This software
performs localization of the orders on the frame, optimal order
extraction, cosmic-ray rejection and wavelength calibration. Also
computed for each order are the blaze function and a "one-dimensional"
flat field, both derived from a Tungsten lamp exposure. The latter is
derived for each order from the average of the rows perpendicular to
the dispersion. The limited extent of the orders and the high quality of the
CCD cosmetic response makes any further second-order flat-field correction
unnecessary. The extracted spectrum is saved as an 'e2ds' FITS file, which
contains the S/N values for each order and the wavelength solution
The orders are then reconnected after correction
for the blaze function, yielding a 's1d' FITS file with a 0.01 Å
wavelength step, which includes the barycentric correction.
All identification, calibration and reduction
information is written into the file headers.
See the data products page for more
See also several examples of e2ds and s1d
spectra for the hot star HD 34078 observed in HR mode.
The data reduction
pipeline on the DRS machine, which is controlled by the template specified
through the STS, proceeds entirely automatically including the
radial velocity cross-correlation and merging of the orders.
Several numerical cross-correlation masks are
available : F0, G2, K0, K5 and M4 and the closest one to the spectral type
of the star is selected by the DRS. If none corresponds, by default
the G2 mask is used.
The on-line velocities are corrected to the barycenter of the solar system
using the coordinates given in the STS catalog.
On-line estimates of the line bisector, error on the velocity and v*sin(i)
[if the B-V color is known] are also available.
An off-line facility is available on the DRS machine for detailed
examination of the results and further cross-correlations with any of the
See the data products page for more
Each new observer is given a
complete introduction by local staff to recommended observing procedures
and tutored on software operation and data reduction as needed for
succesful observing. This introduction normally takes place in the
afternoon (14h local time) preceeding the first night of the run.
Beginning of the night
In order to prepare for an observing night,
observers are advised to check in the afternoon (14-16h) that the user
'sophie' is logged on to the following machines :
If this is not so, log in
with the (same) username/password available in the document you will find in the
observing room. Follow carefully the detailed instructions given in this
document which should lead you up to the start of an exposure and its full
reduction. This will ensure you that the system is working correctly.
In case of any problem, contact software support
(nominally C.Carol, phone 6475).
End of the night
In addition to filling out the night report paper form, the observer can
report on any telescope or spectrograph technical problems using a new
web report form. It is no
longer necessary that you log off from the different machines at the
end of the night.
At the end of your run, a special software tool, called DAU,
is available to copy the raw and the reduced data onto a DVD or a USB pocket
disk. All data from SOPHIE are automatically archived the next day onto an
exernal system, a Mac OS X Server called 'atlas', which is equipped with a
large-capacity RAID disk array (2x2.3 Tb) and which hosts the
The observer has exclusive use of his/her data for one year,
after which it will be made available to the community. Longer protection periods
can be granted for special projects by the relevant Time Assignment Committee.
A night assistant is on duty at the 1.93-m telescope at all times
(except for meal time around midnight) from sunset until start of
nautical twilight and is responsible for
telescope safety. He will open, close and turn the dome,
point the telescope, start and set up the auto-guider. Observing is the
responsibility of the astronomer. The night assistant
is normally able to deal with most problems in the course of the night,
but in very difficult cases he might call up resident staff for help.
Phone numbers of on-duty personnel are listed in the observing room.
Should minor problems arise
Detailed instructions on how to overcome possible minor problems ("bugs")
with the software and hardware are available in the observing room. Check with
the SOPHIE support astronomer or with the night assistant.