|
|
The
Mid and Near Infrared Array Camera
|
MANIAC (Mid- And
Near-Infrared Array
Camera) is an imaging camera for astronomical
observations in the mid- and near-infrared wavelength ranges.
It has been developed at the Max-Planck-Institute for extraterrestrial
Physics (MPE)
in Garching/Germany, in collaboration with the University
of New South Wales (UNSW)
in Sydney/Australia, and the University
in Jena/Germany. MANIAC is designed for
simultaneous observations in the near-infrared (NIR) from
1 to 5 µm and the mid-infrared (MIR) from 8 to 28 µm.
This allows measurement of accurate relative positions of
the mid and near-infrared images of an object. However, at
present, only the mid-infrared channel is installed. Nevertheless,
MANIAC is a fully operational instrument
at the mid-infrared wavelength regime and has been used for
numerous observations. In future phases, a cooled Fabry-Pérot
Interferometer, a bigger detector array for the mid-infrared
channel, and the near-infrared channel will be installed.
In the mid-infrared wavelength range, thermal radiation
from the atmosphere, the telescope, and the instrument reduce
the sensitivity significantly (the Planck curve for a Black
Body at room temperature peaks at about 10 µm). To avoid
a contribution to the thermal background emission from MANIAC,
the instrument has to be cooled. Furthermore, the detector
has to be cooled down below 12 K for optimum operation. Therefore,
the whole MANIAC instrument is cooled down
to 4.2 K using liquid helium. Due to high helium costs, the
present cooling system of MANIAC will be
exchanged by a closed cycle cooling system in the near future.
MANIAC is operated using a PC. All the
electronics and the PC sit in a standard 19'' rack, which
is mounted next to MANIAC on the telescope.
The PC is connected to the display and the keyboard in the
telescope control room via a 60 m long video and keyboard
extension cable.
The tanks for liquid nitrogen and liquid helium have a volume
of 20 litres. As a result, the hold time of the filled helium
tank with MANIAC in operation is about 35
h. Due to space considerations, the optical plate in MANIAC
is mounted vertical on the surface of the liquid helium tank
and the mid and near-infrared channel are set up on the opposite
side of the optical plate. Figure 1 represents a section through
MANIAC. This figure also shows the mid-infrared
channel mounted on the optical plate.
Figure 1: Mechanical
construction of the MANIAC cryostat (not
to scale)
 |
MANIAC has two filterwheels, with the stepper
motors that drive the filterwheels are mounted outside the
dewar. Figure 2 shows an overall view of the MANIAC
instrument. The box mounted to the front of MANIAC
is the analog readout electronic box. On the left hand side,
outside the dewar, the two stepper motors for the filterwheels
can be seen. Inside the dewar, the vertical optical plate
with baffles, filterwheels, and the gear of the filterwheels
are visible. Figure 3 shows a closer view of the mid-infrared
channel without the baffles. The total weight of the instrument
is roughly 100 kg.
 |
 |
| Figure 2: Overall view
of the MANIAC cryostat. |
Figure 3: View of the MIR
optics of MANIAC |
MANIAC is designed for a telescope with
a focal ratio of f/35. The focal plane of the telescope is
inside the instrument and is used as a cold field stop. The
beam enters MANIAC horizontally, as can be
seen in figure 4.
Figure 4: Schematic
of the optical layout of the MANIAC mid-infrared
channel
 |
The entrance window of MANIAC is made out
of KRS-5. All the optical components of MANIAC
are purely reflective. Therefore, the optical design of MANIAC
is intrinsically achromatic over the whole wavelength range.
The first mirror (M1; off-axis paraboloid) creates an image
of the telescope secondary mirror at the position of the second
mirror (M2; spherical), which thus acts as the aperture stop
of the system. A third mirror (M3; spherical) is used to fold
the light to the detector. One filterwheel sits in front M2
and the second filterwheel is placed just in front of the
detector. Figure 5 shows a photo of the actual optical plate
with the mirrors mounted on it.
 |
 |
| Figure 5:
Photo of the MANIAC mirror system. |
Figure 6: MANIAC
detector: Rockwell International S/N MF12 |
The detector array for the mid-infrared channel
is a 128 × 128 pixel Si:As Blocked-Impurity-Band (BIB)
-detector from Rockwell International. The detector is used
in the wavelength range between 8 µm and 23 µm.
The parameters of the detector are given in table 1.
a Signal value with 10% deviation
from ideal curve.
b Detective quantum efficiency: (real quantum
efficiency)/(excess factor)
Table 1: Parameter of the
MANIAC mid-infrared detector
| Parameter |
|
| Size |
(128 × 128) pixel |
| Pixel Size |
75 µm × 75 µm |
| Good Pixel |
99.94% |
| Number of Signal Outputs |
4 |
| Operation Temperature |
12 K |
| Active Layer Thickness |
15 µm |
| Active Layer Doping Density |
1 × 1018 cm-3 |
| Well Capacitya |
8.5 × 106 e- |
| Read Noise |
590 e- |
| Det. Quantum Efficiencyb |
0.21 |
| Dark Current |
5 e- |
|
The filters for the mid-infrared wavelength range used in
MANIAC are made out of Germanium with the
appropriate coatings for the individual band-passes. In addition
to the broad-band and narrow-band filters, a Circular Variable
Filter (CVF) is used in MANIAC. The filters
are installed in two filterwheels. The specifications of the
filters are given in Table 2.
In addition to the filters on the filterwheels a cooled
(4.2 K) Fabry-Pérot interferometer for the N-band will
be installed soon.
The MANIAC Fabry-Pérot Interferometer
(FPI) is currently under construction at the UNSW.
The FPI which will operate at 4.2 K can be used over the whole
N-band range with a resolution of about 3000 and it can be adjusted
to the wavelength of interest during observation. The etalons
for the FPI are from Queensgate Instruments. They are made out
of ZnSe and have a diameter of 70 mm. The flatness of the etalons
is about /18 at
= 630 nm. One
side of each etalon is coated with an anti-reflection coating.
On the other side of each etalon the inner diameter of 50 mm
is coated with a high-reflection coating with a reflectivity
of 95.5%. This results in a finesse of about 61. Each etalon
has five gold pads, placed around the periphery of the high-reflectivity
side. These gold pads act as capacitors and are used to monitor
the tilt and the distance between both etalons. Three piezoelectrical
elements (PZT) are used to correct for a tilt between the etalons.
A fourth PZT is used to adjust the separation of the etalons.
Since the FPI will only be used for observations of line-emission
it can be moved out of the beam for broad-band observations.
The principle of the MANIAC read-out electronics
is shown in figure 7. The analog box, which is mounted outside
the dewar, reads the four detector outputs and amplifies them.
It also generates the detector biases and the clock signals
for the array. The digital box reads the preamplified detector
outputs from the analog box, digitizes, and co-adds them at
up to 400 frames per second. It also also communicates with
the PC over two standard serial IO boards. Two read-out modes
are available for the detector: the single read mode (SR)
and the uncorrelated double sampling mode (UDS). In UDS, the
detector is reset twice at the beginning of each integration.
Since the UDS mode reduces the 1/f noise, this mode is normaly
used for the observations. However, the read-out in UDS mode
is slower than in SR mode. To avoid saturation of the detector
in N and Q band, the faster SR mode has therefore to be used
for broadband observations.
Figure 7: Schematic
of the MANIAC readout electronics.
 |
The sensitivity of MANIAC in the N-band
at the 2.2 m telescope is typically 3 - 5 mJy per square arcsec
for a 1 detection within 1 h (on source) for an
extended source (Lehmann, T., Böker, T., & Krabbe,
A. 1997, ESO Msngr, 88, 9). For a point source the sensitivity
is roughly 2 mJy for 1
within 1 h (on source) in N-band.
Because of the high background radiation of the atmosphere
and the telescope, the chop-nod-technique has to be used for
the observations. Chopping means to look alternately at the
object of interest and the empty sky. This is usualy done
by wobbling the secondary mirror of the telescope at frequencies
of 1 to 10 Hz. By chopping between the object and the empty
sky and subtracting the sky-image from the object-image, the
background radiation of the atmosphere is eliminated. However,
because of the chopping, the beam hits slightly different
parts of the main mirror of the telescope. Due to temperature
gradients (and imperfections), the main mirror of telescopes
is never perfectly uniform and after the chopping there is
some ''structure'' left in the image. By nodding with the
telescope one can get rid of this ''structure''. Nodding is
carried out by moving the whole telescope so that the sky-beam
will be the object-beam and the former object-beam will be
a sky-beam on the opposite side (with respect of the object)
of the former sky-beam. Nodding is done on the order of a
few minutes.
|
So far, observations with MANIAC have
been carried out at the ESO/MPG 2.2 m telescope at La Silla/Chile. The
field of view of MANIAC at this telescope
is 44.2'' × 44.2'' and the pixel size is 0.345''.
MANIAC is diffraction limited at the
2.2 m telescope. The diameter of the diffraction disk
is roughly 1.5''. |

Figure 8: MANIAC
mounted on the ESO/MPG 2.2 m telescope at La Silla.
|
For future observations, we intend to use
MANIACs at the Anglo-Australian Telescope
( AAT)
which is operated by the Anglo-Australian Obervatory ( AAO).
This is a 4m-class telescope and will provide a field of
view of about 25.6'' × 25.6'' with a pixel size of
0.2''.
This web-page is based on the PhD thesis
of Torsten Böker
and the MANIAC instrument paper:
Böker, T., Krabbe, A., Storey, J.W.V., & Lehmann,
T. 1997, PASP, 109, 827. I would like to thank Torsten
Böker for the permission to use his MANIAC
pictures.
|