Subject: Echelle Status, 29 Jan 1999

From: elt@astro.Princeton.EDU

Submitted: Fri, 29 Jan 1999 13:07:00 -0500 (EST)

Message number: 335 (previous: 334, next: 336 up: Index)

The attached message from Don York, co-Instrument Scientist, describes
the currently achieved performance of the new Echelle spectrograph and
some expectations for the future in some detail.  I hope that this
information will be helpful to those wishing to propose for use of the
instrument in 2Q1999.  More definitive information will be available
later, but institutional proposal deadlines are rapidly approaching
and the attached should provide a good basis for writing initial proposals.

Obviously the future can always hold unhappy surprises, but at this 
point, Echelle is still expected to be available for general community
use beginning on 1 April (2Q1999).

Ed Turner


We are now about 25% of the way through the science commissioning of the
ARC echelle spectrograph (aka CARCES-Chicago/ARC Echelle Spectrograph).
Things are going very well. Some specifics follow that may be useful to your
thinking about the use of the instrument.

Use: The echelle is to be accessed by rotating the tertiary mirror of the
3.5 meter telescope. This aspect of operations is not part of the science
commissioning, but is expected to be done by the end of March. The instrument
is commanded from the Remark interface. The instrument has been tested in a
fully remote mode including guiding (but not yet focusing). In normal use,
you can think of the instrument as having no moving parts for astronomical
targets. A mirror must be moved into the beam to do calibrations. Calibration
lamps are built into the instrument. A single flat and a calibration spectrum
will require about 6 minutes (mostly CCD readout time.) The slit should
normally be the nominal one. Others can be put in by the site staff. There are
additionally some filters useful for calibrations and guiding.

Performance: Use these values for planning purposes. They are not final.
Slit: 1.6 x 1.6 arcsec
Wavelengh coverage: 3200 to 10000 Angstroms.
Wavelength coverage with signals above 50% of peak: 4000A (range depends on
     spectral type of target.
Resolution: 7.5 km/sec (independent of wavelength) (sampled in two pixels)
Stability: better than 1/km/sec per 4 hours.
Relative wavelength precision: better than 1 km/sec
Efficiency: Assume 3% from the top of the atmosphere to the disk on a night
of 1.2 arcsec seeing. This measured number includes slit losses, atmospheric
transparency, reflectivity of telescope optics, pointing above 45 degrees
to the zenith, spectrograph optics, CCD efficiency. We cannot yet allocate
the losses between all the factors, but the spectrograph seems to be operating
at the efficiency planned some years ago.
Read noise: 7 electrons
Relative wavelength response: For an unreddened B star, the efficiency is
very low below 3600A and above 9500A, so longer exposures will be needed
for features in those regions.
Signal to noise: >200/1, when allowed by photon noise. The real number is
    better, but we do not have adequate reductions to specify the limit.

Scattered light: The orders are only 2 pixels wide and are adequately separated.
    We do not have good numbers on interorder light or intraorder scattering.
    Neither is an obvious problem and the latter is readily calibrated by use
    of the telluric lines, once the scattering is understood as a function of

Note: All of the telluric lines fall in each and every spectrum, within the
    quoted range. Thus, the wavelength scale, the scattered light and the
    instrumental profile are readily available in each exposure.

Guider: There is a slit viewing camera which is not operating much below
    ambient. Guiding is done by hand (i.e., human hand on computer). It is
    hard to see the fuzz falling on the slit plate below 15th mag at the 
    present time, but we are working to cool the camera more (which is the
    limit of interest.) One can do better with >1 minute integrations if

Dark current: ~0.001 electron per pixel per second.

I hope to put spectra on an anonymous FTP site in a few days for anyone to look
at for planning purposes. However, I am observing this weekend so cannot commit
to this for sure. I will do what I can.

Neither the science CCD or the guider CCD have any significant defects that
affect usage.


APO APO APO APO APO  Apache Point Observatory 3.5m  APO APO APO
APO  This is message 335 in the apo35-general archive. You can find
APO  the archive on
APO  To join/leave the list, send mail to
APO  To post a message, mail it to