Figure 1: A schematic of the CalUnit mounted
on the instrument support structure.
The CalUnit will be located on one port of the telescope Instrument Support Structure (Figure 1), on the opposite port from the facility adaptive optics system [5] (GAOS). The beam from the CalUnit is directed to each of the instruments via the science fold mirror. In order to comply with the requirement that the beam from the CalUnit mimic that from the telescope, the beam may be passed through the GAOS to calibrate observations taken with adaptive correction. Two flat mirrors, one deployable, one fixed, fold a 2.2' beam into the GAOS.
The basic optical layout is shown in Figure 2. Both the size of the optics (the calibration unit is ~2m from the telescope focal plane and the 7' field has a diameter of 260mm) and the broad wavelength coverage required predicates an all-reflective design for the CalUnit. The design may be considered as two halves - a pupil imaging system and an illumination system. The pupil imaging system consists two off-axis mirrors which reproduce the exit pupil of the telescope. A large off-axis angle is required to reproduce the 7' field without vignetting. The telescope vignetting will be reproduced by inserting stops in the pupil imaging system. Neutral density filters and colour-balance filters will be included, located at the aperture stop.
Figure 2: The optical layout of the CalUnit.
The flat mirror to send the beam to the GAOS is
deployed.
The calibration sources to be used for the various instruments are listed in Table 2, along with their approximate étendue. The étendue of the maximum science field (7' diameter) is 163.8mm^2 sr; for the 1' field it is 3.3mm^2 sr. One of the primary functions of the CalUnit is to diffuse these sources to fill the large fields. This has been done traditionally using an integrating sphere. Light entering an integrating sphere at one port is reflected multiple times from a near perfectly diffusing internal surface. The exit port of the sphere is a highly uniform source emitting into a hemisphere. Integrating spheres have very low transmission due both to the large number of internal reflections and the fact that the light at the output is emitted into a hemisphere, whereas only a cone of the light is accepted by the instrument. A typical integrating sphere transmission would be a few percent. Initial calculations of the flux expected from the commercial lamps indicated that the scientific requirements on exposure time would be met by a system with transmission of ~40%. The illumination/diffuser system devised is shown in Figure 3. It has some of the features of an integrating sphere, but the light is emitted into a cone matched to the 7' field rather than into a hemisphere. Light injected from the lamp is reflected from a mirror located at the aperture stop of the pupil imaging system. The beam is then reflected onto a diffuser which scatters the light in the direction of the stop. Some fraction passes through the aperture stop and into the pupil imaging system; the remainder is returned to the diffuser by a oblate spherical reflector. Aside from the reflection losses, the major loss in this system is the reduction of the spectral radiance per unit solid angle by a factor (etendue source/etendue field) which is an unavoidable consequence of the diffusion process. The diffuser will be matched to the wavelength of observations. Diffusing coatings which provide high reflectivity and non-specular reflection in the UV/optical are not efficient beyond 2.4um; a second, NIR, coating will be used for wavelengths from 1-5um. The reflectance of the diffusing coatings is expected to be 0.95. The overall throughput of the illumination system (ISys) is predicted to be ~40%.
The uniformity achievable with the ISys has been considered. Assuming that the illumination from the diffuser is perfectly uniform, the large scale variation in illumination at the aperture stop is 5% across the 7' field; for the 1' field it is ~0.5%. Aberrations in the pupil imaging system have little affect on the output from the ISys, but distortion of 1% can produce an increase in the flux at edge of the field at the output the CalUnit of 4%. Distortion in the pupil imaging mirrors affects the illumination in the opposite sense to the non-uniformitites from the ISys, affording an opportunity to optimise the design. A final uniformity of 3% over the 7' field is anticipated.
Table 2: Commercial calibration sources for
wavelength calibration and flat-fielding
Figure 3: Details of the illumination system
(ISys) showing a hollow cathode lamp illuminating the 7'
science field.