In an unusual approach, LRS-J (see Fig. 3) places the detector in a sandwich-like assembly between the field flattener and copper cooling plate.
Matching the instrumental pupil (142 mm) to the 18 jim pixels of the 1RG at a plate scale of 2 jrad/pix required an unusually fast f /1 camera. The final optical element in this assembly is a PK50 field flattener (a meniscus shape with outer convex ri=50.5969 mm and inner concave r2=-461.0109 mm). The PK50 material was chosen to sharpen 1RG's long wavelength cutoff. At f /1 the overall size of the FF is approximately that of the 1RG package. Coincidentally, this nearly matched the HET central obscuration and, on axis, the detector assembly does not obscure the beam by using the oversized 1RG package in a Schmidt camera configuration. However, we were not afforded much room beyond the package to make the assembly. We were also restricted to a very short (25 mm) overall length by the separation between the detector's active layer and cavity face of the J-band blocking filter.
To constrain the separation between FF and detector, we make use of the 1RG's mask features, 4 1-72 blind tapped holes located next to the detector island. Using precision turned vented pins threaded to fit these tapped holes, we locate a spacer with a hole/slot combination, on the front surface of the 1RG package. The spacer, of stress relieved SST 416, has a spherical surface precision machined to match the concave final FF surface. Its CTE was chosen to match the PK50 rather than that of the molybdenum 1RG package so that it would not move, during thermal cycling, relative to the FF. Regardless, the total relative movement during a thermal cycle, between detector and spacer, is 28 jm.
The SST spacer and field flattener are together clamped into a die-form EDM machined aluminum retaining shell with a 25 jm thick SST shim. This is only necessary for alignment and assembly purposes as the final assembly is preloaded together. The shell itself is 0.5 mm larger than the elements it contains to allow a small amount of tip/tilt alignment tolerance between the axis of the FF and that of the detector.
The FF is constrained in three DOF by the geometry of the spherical retaining shell, and in the remaining three, xy translation and z rotation by a frictional force equal to many hundreds of times its weight. Likewise the SST spacer is constrained to the FF in the same manner. The H1RG is constrained to the spacer in all six DOF with the aforementioned locating pins, and the copper cold block to it in z by spring preload.
Three McMaster-Carr 9435K11 springs are compressed approximately 1.5 mm yielding a force of 125 N keeping the assembly together, allowing thermal conduction between its members, and to allow for CTE mismatches between the Al gimbal assembly and the various low CTE components which make up the detector assembly.
Was this article helpful?