HgCdTe Detector Array
This section discusses the characteristics of the HgCdTe array in OSIRIS
and in particular the following topics:
Overview
The infrared array currently in OSIRIS is a 1024x1024 HgCdTe HAWAII array
supplied by CTIO. The array is a hybrid of a silicon multiplexor and an
array of infrared sensitive (HgCdTe) detectors. The two pieces are pressed
together, with indium bumps on each piece making the electrical connection.
The multiplexor is an array of discrete read-out transistors and is unlike
an optical CCD. In particular, the array can be read non-destructively
and each pixel can be read separately and in any order. Although these
differences do not matter for typical observing programs, future upgrades
could permit additional observational flexibility. For example, designated
sub-arrays could be read-out at high frequency, then shifted and added
later to create diffraction limited images.
The band gap energy of this HgCdTe is approximately 0.5 eV, smaller
than Si, so this detector is sensitive to radiation with wavelengths as
long as 2500 nm. The relatively small band gap energy also causes high
dark current at relatively high temperatures and thus in the low count
rate regime characteristic of astronomy it must be cooled to liquid nitrogen
temperatures (~77 Kelvin).
This HgCdTe array has 18.5 micron pixels. The scale (arcsec/pixel) depends
on the telescope and choice of camera lens. The various possibilities are
summarized in the section on the Imaging
Mode.
[ Top of Page | Contents
]
Dark Current and Read Noise
The dark current in the array is low. We currently measure about 0.15
e-/sec, some of which may be due to background radiation in the
dewar. If you measure dark current significantly higher than 1 DN/sec,
please check that the instrument is still boiling off its liquid nitrogen.
The read noise of the array is 15 e- and
the gain is
3.0 electrons per DN. The effective
read noise of the detector can be reduced by reading each pixel's signal
non-destructively many times; factors of 2-3 improvement are possible.
Note, however, that a read noise of 19 e- is comparable to the shot noise
for signals of about 100 DN, which happens quickly in most imaging and
low-resolution spectroscopic observations. If your observations are read-noise
limited, please contact Darren DePoy (
depoy@astronomy.ohio-state.edu) for more information.
[ Top of Page | Contents
]
Linearity and Saturation
Here
is a plot showing the basic linearity behavior of the array. The plots
were generated using large image sections over the upper and lower halves
of the array. The array becomes significantly non-linear before the full
well capacity of the detector is reached. Further, an accurate non-linearity
correction depends on the signal rate as well as the total signal collected.
However, non-linearity is 1.0%, for 11,000
DN
and 2.0% for 17,000
DN
at a signal rate of 800 DN/sec (a typical
rate at K looking at the dome white spot with the f/7 camera). Non-linearity
coefficients are given on this plot.
The array becomes seriously non-linear (>5%) at about 28,000
DN.
The linearity correction is better fit by a 3rd order polynomial than
a 2nd order one. The latter has been used previously with CIRIM and is
implemented in all previous versions of the IRAF task "irlincor."
This task has been updated (as of 20 March 2000) to allow for 3rd order
fits. OSIRIS users can corrrect their data using the updated irlincor task
(you must upgrade your IRAF installation for the revised "ctio" package)
or with the "linfits" program. This FORTRAN program is available by anonymous
ftp. It may be used stand alone, or as part of the basic osiris IRAF data
reduction script, osiris.cl (irlincor can also be called from osiris.cl).
Both linfits and osiris.cl are included in the CTIO IR reduction package
cirred.
[ Top of Page | Contents
]
Bad Pixels
The OSIRIS array is generally very cosmetically clean. There are scattered
hot and cold pixels on the array, though these only comprise less than
1% of the illuminated area of the array. Dithering the telescope during
the course of an observing sequence is an effective way of eliminating
the effects of bad pixels.
[ Top of Page | Contents
]
After-Image (Residual Charge)
The detector array in OSIRIS exhibits a residual signal from bright sources.
The magnitude of the residual image depends on the brightness of and total
signal recorded from the source. For typical observations the residual
will be 0.5-2% of the originally detected signal. Reading the array several
times reduces the magnitude of the residual image to <<1% of the
original signal.
We suggest that if a 1% residual image will seriously compromise your
results, you dither the telescope and, perhaps, read
the array several times (set the exposure time to 0 seconds, do
a go 3, and throw those images away) between each science exposure.
[ Top of Page | Contents
]
Updated: 1999 December 21
