Spitzer Documentation & Tools
MIPS Instrument Handbook

2.3  Detectors

2.3.1        Focal Plane Detector Array Design

Si:As (24 m) Array

The MIPS 24 m array was developed and furnished to MIPS through the Spitzer InfraRed Spectrograph (IRS) program. It is a Boeing 128x128 pixel blocked impurity band (BIB) Si:As device identical to those in the two short wavelength spectrometers of IRS except that the MIPS array has an anti-reflection coating optimized for the 24 m region. The device has four output amplifiers, each of which reads every fourth pixel in a row. That is, the readout is in columns, with an output amplifier for each of columns 1, 2, 3, and 4, repeated for columns 5, 6, 7, and 8. The array is described in more detail by van Cleve (1995; Proc. SPIE Vol. 2553, p. 502) and in the IRS Instrument Manual, as well as Houck et al. (2004, ApJS, 154, 18). A summary of the characteristics of the flight array is given in Table 2.4.

 

Figure 2.7: 70 m array 4x32 submodule.

 

Ge:Ga (70 and 160 m) Arrays

The 70 m and 160 m arrays both utilize traditional gallium-doped germanium extrinsic photoconductor detector technology. Because many of the performance characteristics of the two arrays are very similar, the performance of both is discussed below, after the physical layout of the two arrays is described. Differences between the performance of the two arrays are pointed out as needed.

70 m Array Design

The 70 m array is described by Young (1998; Proc. SPIE Vol. 3354, p. 57-65). Figure 2.7 shows a single 4x32 pixel subarraymodule from the 70 m array. Each row of 32 pixels is fabricated from a single 24 mm x 0.5 mm x 2 mm piece of Ge:Ga. The 24 mm length is divided into 32 pixels by photolithography, yielding a dimension of 0.75 mm x 0.5 mm for the illuminated face of each pixel. Absorption of photons occurs over the 2 mm depth of the pixel and is enhanced by a mirror at the back of the pixel that directs transmitted photons back through the sensitive volume. Individual rows are separated by spaces of 0.25 mm. Wedge-shaped germanium optical concentrators are placed in front of each row of pixels to expand the smaller dimension of the pixels to give essentially complete filling of the focal plane and square pixels. Each module contains four readouts, one for each block of 4x8 pixels, mounted on a multilayer ceramic electronics board. A molybdenum frame secures the assembly and the cabling. The vertical dimension of the frame exactly matches that of the 4x32 module, allowing 8 of the modules to be stacked to form the full 32x32 focal plane array (FPA).

 

Figure 2.8 is a schematic of the full 32x32 70 m FPA assembly. The 8 modules are stacked and attached to a baseplate. An additional enclosure houses electrical connections and connectors for the camera wiring harness.

 

 

Figure 2.8: Assembled 70 m focal plane.

Figure 2.9: 1x5 submodule of the 160 m array showing discrete pixel elements, integrating cavities, and readout.

160 m Array Design

The 160 m array is described by Schnurr (1998; Proc. SPIE Vol. 3354, p. 322-331). It also has a modular design. Figure 2.9 shows a 5-pixel submodule of the 160 m array. Each pixel of the array is a discrete block of Ge:Ga 0.8 mm x 0.8 mm x 1 mm in dimension. The 5 pixels are mounted on an alumina electronics board on 3 mm centers, resulting in a very low intrinsic filling factor for this array (see below). Also attached to the board are a single readout for the 5 pixels and an array of 5 integrating cavities.

 

Figure 2.10: Stressing rig for the 160 m array viewed edge-on.

The pixel material normally exhibits photoconductive behavior only for wavelengths shortward of 120 m. To extend that behavior to longer wavelengths, the pixels are mechanically stressed to a level of roughly 50% of the material strength. A schematic diagram of the mechanical assembly used to apply and maintain that stress is given in Figure 2.10. The stressing rig contains 2 of the submodules shown in Figure 2.9, oriented back-to-back, and separated by the central metal anvil of the rig. Once the submodules are aligned within the rig, mechanical pressure is applied to the pixels individually, literally by the turn of a screw. A pressure plate between the tip of the adjustment screws and the detector material ensures the even application of force and eliminates the possibility of lateral motion of any of the parts. The amount of force applied is monitored by measuring the conductivity of the pixels as the screws are tightened. Once the appropriate stress level is attained, the screws are mechanically secured to prevent slippage.

 

Figure 2.11: Fully-assembled 160 m focal plane array showing 4 stressing rigs, each containing two 1x5 submodules, and showing the photon concentrator cones.

Figure 2.11 illustrates the final configuration of the 160 m FPA. Four of the stress rigs illustrated in Figure 2.10, each containing 10 pixel elements, are mounted to the curved upper surface of a baseplate. Affixed to the top of each stress rig is a 2x5 array of gold plated, light-concentrating, conical apertures. These concentrating cones expand the effective area of each pixel in the array to ~16x~18, and are machined such that each row of 20 has a filling factor approaching 100% in spite of the considerable inter-pixel space required to accommodate the stress rigs and individual integrating cavities. There is a ~16 wide dead strip down the center of the array which is automatically filled in during observations using multiple exposures and scan mirrormotions.