The near-infrared camera SWIRCAM

SWIRCAM is the near-infrared camera mounted at the focal plane of the AZT-24 telescope, at Campo Imperatore Observatory. It is designed for photometry and low-resolution spectroscopy of point-like and extended sources.

The camera was assembled by Infrared Laboratories Inc. at Tucson, Arizona in 1997 and saw its first light toward the end of 1998.

The photosensitive element is a 256 x 256 HgCdTe PICNIC array provided by Rockwell Science Corporation.

The camera optics has been designed by S. Gennari from Arcetri Observatory: tests and refinements of electronics and mechanics have been performed with the fundamental aid of people from Rome Observatory.

General scheme

The camera is a device of the EFOSC type. Travelling from the telescope to the array, the light passes through:

A Lyot stop is placed on the pupil, in order to shield the array from the light scattered by the edge and the mounting of the primary mirror and the thermal radiation emitted by other telescope components.

Optics layout

The focal plane scale is 1.04 arcsec/pixel, for a pixel size of 40 μm. The overall array, whose area is 10.242 mm2, covers a field-of-view of more than 4.4 x 4.4 arcmin2 with only a small distortion amounting to about 2.65% which can be software-corrected on the images (Li Causi, 2000, OAR Internal Report OAR/00/IR2 mar. 00).

Like in any other infrared device, the array and the optical components of SWIRCAM are hosted inside a dewar which is kept under high vacuum (residual pressure 10-6 -10-7 torr) and is cooled at liquid-nitrogen temperature (77 K, 75 K at the 2200 m a.s.l. of Campo Imperatore).
The cooling is ensured two vessels, whose overall capacity is 13 litres, which are refilled daily or after each observational session. With the good insulation achievable, the nitrogen evaporation rate is 0.26 litres/hour, so that the duration of a refilling charge is up to 50 hours (see figure). These tests, performed for the first time at Infrared Laboratories on 1997, have been repeated on September 2001 and the result is a proof of the good quality of the dewar.


SWIRCAM can perform broad-band and narrow-band photometry and low-resolution spectroscopy in the wavelength range 1.1-2.5 μm. Broad-band filters are the astronomical standards J (1.25 μm), H (1.65 μm), K (2.2 μm) and K’ (transmission curve centred at 2.2 μm but cutted at 2.32 μm). Narrow-band photometry can be performed around four selected spectral lines, namely HeI (1.083 μm), FeII (1.645 μm), H2 (2.121 μm) and Brγ (2.164 μm).

Finally, SWICAM is provided with two grisms, one for the I+J region and the other for the H+K region, both having a resolving power 270. The H+K grism is coupled with an order sorter: for the I+J grism, order sorting is naturally performed by the loss of efficiency of the HgCdTe array at wavelengths shorter than 1 μm. These grisms can be used in conjunction with three slits of increasing width (1 arcsec, 2 arcsec, 3 arcsec), placed on the first wheel.

Broad-band filters

Narrow-band filters




J (1.25 μm)

HeI (1.083 μm)


1 arcsec

H+K order sorter

H (1.65 μm)

FeII (1.645 μm)


2 arcsec


K (2.2 μm)

H2 (2.121 μm)


3 arcsec


K’ (2.2 μm)

Brγ (2.164 μm)




The spectroscopic mode has still to be completely implemented, because of technical problems which have persisted during the past years. A preliminary spectrum obtained for the star HD150205 is shown in the figure, with the corresponding spectral trace and the most prominent atmospheric absorption features.

From the identification of these spectral features, fundamental quantities like the spectral range, the central wavelength and the resolution and sampling have been computed. These preliminary results show that the I+J and H+K regions covered by SWIRCAM extend from 0.84 μm to 1.32 μm and from 1.45 μm to 2.38 μm, respectively, with central wavelengths of ~1.09 μm and ~1.92 μm. Moreover, the spectral samplings amount to ~19 Å/pixel and ~36 Å/pixel in the two cases.

All of these data are in excellent agreement with the theoretical predictions of Speziali and Vitali, 1997, OAR Internal Report OAR/97/IR6 sep. 97


The main data about the camera performances are summarized in the table below. The quantum efficiency is larger than 50%, being maximum in the H band. Dark counts are almost negligible, whereas the reported RON level is drastically reduced by a Fowler multiple correlated double sampling.

Quantum efficiency




59 %

70 %

61 %

Dark current level

0.00367 ADU / sec / pixel

Dynamic range




Readout noise

~ 30 e-


5.95 e- / ADU

Persistence ghosts


Electronic ghosts

~ 2.5 % of the source brightness


Main overheads:

Detector readout

~ 0.36 sec / DIT

Image transfer

~ 9.50 sec / group

Telescope offset

~ 3.30 sec / arcmin (low speed)

The linearity is rather good, extending from 5000 to 50000 ADUs in a dynamic range reaching its saturation at 55000 ADUs.

The flat fielding has to be performed in differential mode, being not possible estimate the bias level in single images. Flat field should be performed preferentially on the sky, at dawn or dusk, with optimal counts of 40000-50000 for the "high level" images and 500-1000 for the "low level" ones. The final flat field images show fluctuations over the image not larger than 40%.

The typical sky brightness at Campo Imperatore amounts to 15.5, 14.5 and 11.5 mag/arcsec2 in the J, H and K band, respectively. With these background luminosities, the limiting magnitudes mJ = 17.7, mH = 16.9 and mK = 16.2 can be achieved, assuming 1 min exposure time, a seeing figure of 2 arcsec and a signal-to-noise ratio 3.

For the spectroscopic mode, limiting magnitudes are expected around 14.5 for an exposure time of 1 min, a seeing figure (and a slit width) of 2 arcsec and a signal-to-noise ratio equal to 3, in the worst sky conditions (K band). If confirmed, these performances could represent a very important improvement in the potential of SWIRCAM: the spectroscopic follow-up, in addition to the photometric one, could be possible in the early- and intermediate stages of a supernova explosion. As an example of this application, the spectroscopic observations could get more light on the evolution of the gas and dust which are believed to exist around type-IIn supernovae.

All of these performances could be optimised by a better shaping of the Lyot stop, in order to shield the array also from the thermal radiation generated by the secondary mirror of the telescope. The scheme reported in figure illustrates this solution, which has already been applied (succesfully) to the Near Infrared Camera Spectrometer (NICS) at TNG.

The work is currently in progress.

Last update Dec 11, 2002

Mauro Dolci, INAF - OACT