2.3 The LWS Optics

Radiation from an astronomical source entered the ISO telescope, which was a 60cm diameter () Ritchey-Chrétien system with an overall focal ratio of (, where is the effective telescope focal length). The telescope plate scale is given by and is 22.9 /mm; i.e. an object of 22.9 on the sky had a linear size of 1mm in the focal plane of the telescope. The radiation was divided into four beams, one for each ISO instrument, by a pyramidal mirror near the Cassegrain focus. Figure 2.1 shows the path radiation took through the LWS, superimposed on a photograph of the instrument taken before the integration of the Fabry-Pérot subsystem. After entering the LWS it was reflected from Mirror 1 onto Mirror 2, which was coincident with the telescope focus. The size of Mirror 2 thus precisely defined the Field of View (FOV) on the sky for the LWS. At these extreme infrared wavelengths the spatial resolution of the ISO telescope was completely determined by diffraction, with an achievable spatial resolution of 83 ( ) at the longest wavelength (). Although this lowers with decreasing wavelength, the LWS was designed to detect this longest wavelength with reasonable efficiency yet have a nearly constant FOV on the sky. A study of the intensity received from a point source as a function of the focal plane aperture size by Duncan 1983, [14] showed that the diameter of the central Airy diffraction disc ( ) can be reduced to about 65% before significant losses occur. Note that the energy within the central spot of the Airy disc is 84%, whilst Figure 5 of Duncan 1983, [14] shows that this only reduces to 72% with an iris set to 0.65 of the Airy disc diameter. For the longest LWS wavelength this gave a diameter of mm ( ), which was the criterion used to set the size of Mirror 2. The maximum geometrical FOV for the LWS was given by 106 .

Figure 2.1: Photograph of the LWS with an overlay of the optical path (note the FP wheel is not shown) by S. Adams (QMW).

In practice the beam size was determined by the aperture size at the detector feed horn entrance which also determined the spectral resolution achievable. This will be addressed again in the next section which describes the grating. This unavoidable linkage between the spectral and spatial performance of the LWS makes it a particularly difficult instrument to understand and hence its output data products are also difficult to interpret. To further confuse these issues, internal stops and diffraction losses within the instrument also modify the FOV. The beam size on the sky, while in-flight, was 80 (see Section 5.9). This was measured by scanning a point source across the FOV in-flight.

Unfortunately Mirror 2 was stepped, as shown in Figure 2.2, with a much larger diameter annular surface parallel to the mirror surface. The intention was that this surface should have been black, to absorb all of the straylight from around the target source. The blackening method chosen was to anodise this annular ring with nickel dye to a thickness of 55m, which proved not to be efficient enough at long wavelengths (see Ungar et al. 1990, [44] for a study of black paints for ISO). The step, , between the front and annular planes was 1.5mm, so in effect, at the longer wavelengths, radiation from sources extended with respect to the beam (or non-centred point sources) would partly reflect off the annular surface and interfere with that from the desired path from the front of Mirror 2. This interference is seen as fringing in the spectra as described in Section 6.2.

Figure 2.2: The unintentionally stepped Mirror 2. Dashed lines show the undesired optical path with consequent phase delay 2 h.

Figure 2.3: The LWS (photographed by S. Adams, QMW)

Following reflection from Mirror 2, the radiation then met Mirrors 3 and 4, which made a powered reverse Cassegrain that collimated the beam. The collimator focal length (500mm) was necessarily large to reduce the angular spread of the parallel beam, which limited the attainable resolution of the Fabry-Pérot (the Fabry-Pérot Spectrometers are discussed in Section 2.5). Because the focal length was large, mirrors (5, 6 and 7) were used to fold the beam to adhere with the space constraints on board ISO. Between Mirror 5 and Mirror 6 the radiation passed through the interchange wheel. This wheel was driven by a pinnon engaging gear-teeth on the rim of the wheel: a new type of cryogenic motor was especially developed to drive the wheel. The wheel could be set to four positions: a hole for grating-only medium resolution mode, the long wavelength Fabry-Pérot or the short wavelength Fabry-Pérot for high resolution modes and a blanking plate for dark current measurements (this was never used in operation). The Fabry-Pérot mechanism can be seen in Figure 2.3, which shows the complete instrument. Mirror 7 then directed the parallel radiation onto the diffraction grating, which had an off-centre rotationally symmetric Schmidt profile to correct for the spherical aberration. The basic action of the grating was to disperse different wavelengths from the source spatially. The grating was rotated to cover a range of wavelengths.

The dispersed radiation from the grating was focused by the large spherical condensing Mirror 8 (see Figure 2.1, focal length 97.5mm) onto the cylindrical surface of the ten element detector array which was aligned along the dispersion direction.