Geometry of a coded sphere telescope. a) cutaway view illustrating inward-facing detector elements. b) cross-sectional view. c) detailed view of a detector/mask element 

Contexts in source publication

Context 1

... at hard X-ray energies (10 250 keV) faces the basic difficulty that reflective and refractive optics cannot be used at such short wavelengths. However, coded mask telescopes provide a method for imaging the sky in this waveband: the images are reconstructed from the geometric shadows cast by opaque mask elements when detecting elements are illuminated by X-ray sources. This technique is now well established in practice (e.g. Skinner 1984), and it has opened up a new window on the universe. Observations of cosmic hard X-ray sources provide us with direct information on some of the universe’s most en- ergetic phenomena. The short timescale variability seen in this emission underlines the compactness, and hence the intensity, of these sources. Long-term monitoring in this energy band would provide information on source variability over a wide range of timescales, from which we would be able to infer much about the structure of the emitting region. Unfortunately, most coded mask telescopes to-date have had a relatively small field of view, and so continuous monitoring of many sources over long periods has not been possible. Even wide-field instruments, such as those on the SAX satellite (Scarsi 1993), cover only a small frac- tion of the sky at any given time. Much new science would be possible if we were able to monitor sources all over the sky simultaneously. Even more extreme examples of source variability are provided by transients and bursters. These sources can rise from undetectable flux levels to become the brightest sources in the sky with durations from milliseconds in the case of bursters up to years for the more long-lived transients. The BATSE experiment on CGRO has revolu- tionized the study of these sources by providing data for many hundreds of bursts (Fishman et al. 1994). However, the geometric techniques used for deriving the location of the burst from this instrument are fundamentally limited by systematic effects to an accuracy of ∼ 4 degrees, which is rather too large for quiescent counterparts (if they exist) to be identified in other wavebands. The origin of gamma ray bursts remains a mystery. In this paper, we present a proposal for a new type of hard X-ray instrument, the coded sphere telescope. This design combines coded mask techniques with the all-sky geometric source location techniques of BATSE, and so it can be used to monitor known sources anywhere in the sky, and to detect and localize transients to within a small error box. Section 2 describes the basic geometry of the telescope, and defines the design parameters for a practical implementation of the concept. In Sect. 3, we show one method by which this design can be used to monitor many sources simultaneously, and present the analysis of simulated observations to show the sensitivity of the method. Section 4 describes the use of the telescope for detecting transients and bursts, and presents simulations which show the accuracy with which these sources can be located. 2 . 1 . Basic geometry The geometry of the proposed telescope design is illus- trated in Fig. 1. It consists of a transparent spherical structure on the surface of which are mounted circular pixel elements. These elements contain disk detectors which face inward toward the centre of the sphere, with their outside faces shielded against incident photons. Thus, the elements on one side of the telescope act as masks to the detectors on the other side of the sphere. The detection plane is very incomplete, but, as we shall see, it is still quite practical to form images with such an instrument. Planar coded mask systems generally have their mask elements arranged in a uniformly redundant array (URA). This arrangement has the desirable property that artifact- free images can be reconstructed by simple correlation techniques (Gunson & Polychronopulos 1976; Fenimore & Cannon 1978). If the detector plane is not completely filled, many of the advantages of a URA mask disappear, and correlation techniques produce images with significant artifacts (Byard & Ramsden 1994). For the proposed spherical geometry with a significantly incomplete detection “plane”, it is not immediately apparent whether analogues of the URA can be constructed. However, for an all-sky monitor, there are more powerful techniques for deriving the fluxes of sources than simple correlation (see Sect. 3), and so the absence of a URA mask pattern is not terribly restrictive. We therefore place the mask/detector pixels at random locations on the surface of the sphere by analogy with the random pin-hole camera (Dicke 1968). If the sphere has a radius R t , and the N d detector elements have a radius R d , then the filling factor of the sphere occupied by detectors ...

Context 2

... can be seen from Fig. 1, the basic structure of the coded sphere is very simple. A set of identical mask/detector pixels must be mounted at random locations on the surface of a spherical supporting structure. This structure could either be a solid shell of material which is transparent in the energy range of interest, or it could form a skeletal framework like a geodesic sphere. The choice of materials for the mask/detector pixels depends on the operational energy range for the proposed telescope. For a moderately sized instrument, we must use small mask/detector elements to achieve reasonable spatial resolution (see Eq. (3)), and so we are constrained to use detectors which can be implemented as small pixels. Detectors which might be used in a coded sphere telescope include: i) Silicon pin photodiodes , which can be used as rugged low-cost X-ray detectors. Standard devices with de- pletion depths of up to 500 μ m offer useful detection efficiency up to ∼ 30 keV. A low energy threshold of ∼ 5 keV is attainable with standard electronics at room temperature; cooling the detector and associated electronics to ∼ − 30 C reduces the detection threshold to ∼ 1 keV. At these low energies, a simple passive lead or tungsten shield at the rear of each detector would be adequate to provide both background shielding and the necessary mask function. Care would have to be taken to provide a supporting structure which mini- mizes the amount of extra attenuating material along the path of incident photons ii) Cadmium telluride detectors would allow the telescope to operate at higher energies. This high-density room- temperature semiconductor would provide good detection efficiency over the range ∼ 15 − 200 keV with a detector thickness of ∼ 2 mm. In this case, a more sophis- ticated lead-tin-copper graded shield would be needed to avoid fluorescence photons from the shield falling within the operational energy band of the detectors. iii) Scintillator-photodiode combinations would extend the sensitive range of the telescope to even higher energies. However, it would not be possible to keep such detectors thin and maintain useful detection efficiency. Fur- ther, passive shielding is no longer practical at these higher energies and active veto systems would be required. As a result, the size and mass of detectors operating at higher energies would undoubtedly increase ...