Previous Article
Next Article
Table of Contents

CCD News (1Sep93)
(from CTIO, NOAO Newsletter No. 35, 1 September 1993)

ArCon Implementation

As predicted in Newsletter No. 34, the Tek 1024 #2 CCD was converted to
operation with a new ArCon CCD controller (ArCon 3.2) in early May. It
was subsequently successfully used (12-22 May) at the 4-m telescope,
with the R-C spectrograph and Folded Schmidt Camera. On 14 July ArCon
3.2 was installed at the 0.9-m, for a continuous six-week period. At
the same time operation of the filter wheels was changed to a STD bus
controller. Both the STD bus computer and the Telescope Control System
computer are fully integrated with the ArCon.

An astronomer operates an ArCon CCD controller from a Sun SparcStation
10 computer, via an IRAF-based user interface. This user interface is
very similar to the KPNO "ICE" interface, and is easy to operate. A
second Sun computer is available in the control room; if desired this
computer can be used for IRAF reductions instead of the IRAF reduction
window on the Data Acquisition computer. Both Instrument and Software
Manuals are available from the CTIO FTP Archive (see article in this
Newsletter); be cautioned that over the next few months the Software
Manual in particular will be frequently updated as enhancements and
modifications are made.

Initial feedback from observers has been very positive. Users who have
learned to dread the appearance of "MUX TIMEOUT" messages when using
the old VEB controllers will be pleased to learn that the ArCon
hardware appears to be ultra- reliable. We summarize some of the
advantages of the ArCon CCD controllers over the old VEB controllers,
in no particular order:

o Quad channel readout for CCDs with multi-amplifiers.
o Unsigned 16 bit data (0-65535) allows greater dynamic range.
o Dc coupled preamp means no memory of image in overscan.
o Real time display with saturated pixels in red.
o Direct data interface to Sun/IRAF.
o Faster readout in single channel mode for most CCDs.
o Freedom from fixed pattern noise at fast read rates.
o Greatly increased hardware reliability.
o Similar observing environment at CTIO and KPNO.

The conversion of our CCDs to ArCon operation is unfortunately a rather
slow process, due to the need to schedule the CCDs we have, implement
new CCDs, and also to convert instruments to operate in the new (Sun,
ArCon, IRAF) environment. In addition, we are taking the opportunity to
simultaneously update the motor controllers on some instruments, using
STD Bus computers. At present we have the following ArCons; all these
CCDs have four working amplifiers:

o ArCon 2.1, Thomson #1 CCD, used at the Schmidt telescope.
o ArCon 3.1, Thomson #2 CCD, used as Lab. development system.
o ArCon 3.2, Tektronix 1024 #2 CCD.
o ArCon 3.3, Tektronix 2048 #3 CCD (scheduled for August 1993).

Beyond this, our plans are as follows. The completion dates are
approximate, and depend on the available manpower:

o ArCon 3.4, Nicmos III IR array (mid 1994)
o ArCon 3.5, Tektronix 1024 #1 CCD (late 1993)
o ArCon 3.6, Tektronix 2048 #2 (STIS) CCD (late 1993)
o ArCon 3.7, Loral 3K CCD (early 1994)

For the next few months, predicting whether or not a given observer
will get an ArCon-based CCD is not easy. It is likely that all future
0.9-m observers, excepting those requesting the TI CCD, will be
scheduled with a Tek 1024 or 2048, operated with an ArCon controller.
Direct imaging with ArCon CCDs will start at the 1.5-m as soon as a STD
bus motor controller is built for that telescope, presently scheduled
for November. Direct imaging at the 4-m prime focus, and all 4-m and
1.5-m Echelle long camera runs should use the ArCon-based Tektronix
2048 in the future. The remaining CCDs, used for 4-m R-C spectroscopy,
1.5-m spectroscopy, and with the Rutgers Fabry-Perot, are scheduled to
be converted to ArCon operation during 1994.

4-m Prime Focus Imaging with the Tek 2048

For the period 11-23 June the Tektronix 2048 CCD was installed for the
first time at the 4-m prime focus, with the Large Format PFCCD camera.
The blue triplet corrector was used, while the instrument control
employed a hybrid arrangement (old VEB controller for the CCD, new STD
bus instrument controller operated stand-alone for the motors). After a
few hiccups due to poor connections between the prime focus cage and
the motor-controller terminal, all worked reasonably reliably.
Considerable field curvature was apparent; the resulting defocus could
be minimized (as an example, 1.0 arcsec FWHM center, 1.1 arcsec FWHM
corners) by focussing approximately mid-way center-to-edge of the CCD.
DAOPHOT II with a quadratically-varying PSF did a good job of fitting
the profiles.

Ray tracing by Brooke Gregory indicates that the great majority of the
field curvature is due to the triplet corrector, and the curvature of
the CCD has only a minor effect. This is good news, since it implies
that the new ADC corrector should produce images with no defocus over
the whole CCD. Future PFCCD runs are scheduled to use the ADC corrector
and ArCon CCD controller, and results from this combination will be
reported in the next Newsletter.

Alistair Walker, Steve Heathcote

Previous Article
Next Article
Table of Contents