ProtoEXIST: Advanced Prototype CZT Coded Aperture Telescopes for EXIST
ProtoEXIST1 is a pathfinder for the EXIST-HET, a coded aperture hard X-ray telescope with a 4.5 m CZT detector plane a 9070 degree field of view to be flown as the primary instrument on the EXIST mission and is intended to monitor the full sky every 3 h in an effort to locate GRBs and other high energy transients. ProtoEXIST1 consists of a 256 cm tiled CZT detector plane containing 4096 pixels composed of an 88 array of individual 1.95 cm 1.95 cm 0.5 cm CZT detector modules each with a 8 8 pixilated anode configured as a coded aperture telescope with a fully coded field of view employing passive side shielding and an active CsI anti-coincidence rear shield, recently completed its maiden flight out of Ft. Sumner, NM on the 9th of October 2009. During the duration of its 6 hour flight on-board calibration of the detector plane was carried out utilizing a single tagged 198.8 nCi Am-241 source along with the simultaneous measurement of the background spectrum and an observation of Cygnus X-1. Here we recount the events of the flight and report on the detector performance in a near space environment. We also briefly discuss ProtoEXIST2: the next stage of detector development which employs the NuSTAR ASIC enabling finer (3232) anode pixilation. When completed ProtoEXIST2 will consist of a 256 cm tiled array and be flown simultaneously with the ProtoEXIST1 telescope.
Send correspondence to Branden Allen (firstname.lastname@example.org)
CZT (CdZnTe Cadmium-Zinc-Telluride) and CdTe (Cadmium-Telluide) is a semiconductor
which has found increasing use in hard X-ray and soft -ray sensing applications
over the past decade. The wide band gap ( eV) and high Z-composition make the
detection of high energy X-rays / soft -rays (5-2000 keV) possible at room
temperature and eliminates the need for cryogenics which is traditionally required
for other commonly utilized semiconductor detector materials such as high purity germanium
(bandgap eV) or silicon ( eV).
For application in X-ray astronomy this removes a significant constraint to
operation in a space environment through appreciable mass savings
and eliminates the need for irreplaceable consumables increasing the upper bound of
the mission lifetime. Long term implementation of high energy coded aperture
telescopes utilizing CZT and CdTe detector planes have in fact been successfully
demonstrated over the past decade in the INTEGRAL-IBIS
The ProtoEXIST program was conceived as a pathfinder for the development of an instrument with such a detector plane on a much larger scale: the EXIST-HET[8, 9] (high energy telescope on EXIST). The HET is currently envisioned to utilize 4.5 m of CZT consisting of 11264 20 mm 20 mm 5 mm pixilated CZT detector units with 12 pixels total. Each detector units will have 32 32 pixels with a 600 m pixel pitch and consume 20 W per channel. In order to demonstrate the feasibility of such a detector plane using a scalable and highly redundant architecture the ProtoEXIST project has been divided into three phases:
ProtoEXIST1 (described here) is a coded-aperture telescope with a 256 cm CZT detector plane tested in a space environment with a balloon flight at an altitude of km above mean sea level. The detector plane utilizes the RADNET ASIC and is comprised of an array of individual CZT detectors with pixilated anodes. Each individual detector had 64 pixels (a array of pixels) or 4096 individual pixels over the entire detector plane. This phase culminated in a 6 hour flight out of Ft. Sumner, NM on the 9 of October 2009.
ProtoEXIST2 will demonstrate the feasibility of a nearly identical architecture utilizing the NuSTAR DB-ASIC (i.e. direct bond ASIC) [6, 10]. This enables the demonstration of a detector plane with pixels possessing a 600 m pitch (i.e. a pixel array for each detector) and fulfills the fine pixilization requirement of the EXIST-HET. The DB-ASIC currently consumes 80 W per channel exceeding the EXIST power requirement, necessitating the production of a new lower power version. Currently the design of the required detector electronics is nearing completion. At the time of this writing it is expected that testing of the first detectors of this series will begin within the next few months (i.e. October 2010).
ProtoEXIST3 will demonstrate the final unmet requirement for EXIST through a relatively minor modification of the DB-ASIC in order to produce the EX-ASIC (EXIST-ASIC) which will consume 20 W per channel. This phase of the program will conclude with a flight of the original detector module from ProtoEXIST1 and a second detector module consisting of 1/4 DB-ASIC mounted CZT detectors and 3/4 EX-ASIC mounted CZT detectors.
2 ProtoEXIST1 Payload Overview
The ProtoEXIST1 payload consists of two main parts: the gondola which supports flight operations and pointing functions, and the instrument package which contains all required subsystems for the support of the ProtoEXIST1 coded aperture telescope.
The main ProtoEXIST1 gondola (pictured in figure 1) was refurbished using
the existing EXITE  gondola frame which flew multiple
times during the late 1980s and early 1990s. The pointing system as well as
all supporting electronic subsystems were replaced using the HERO
(High Energy Replicated Optics) gondola support electronics and pointing system design
as a model. As in HERO the coarse azimuthal pointing system is based on a
differential global positioning system (DGPS) unit with a backup magnetometer in
place in the event of DGPS failure. A daytime star camera system (also based on
a design from HERO) and inertial
guidance system are utilized for fine pointing control. The star camera has a
field of view with a small pixel size () which
together with a long ( m) baffle enables the resolution of stars over the
daytime sky background. In addition two transmitters are mounted to the gondola, one for the
transmission of reduced data to the ground for storage and analysis, and the other
for the monitoring of the gondola and payload status. A single receiver is present
for command and control and is linked to the standard command interface package (CIP)
provided for scientific balloon flights by the Columbia scientific balloon facility (CSBF)
The remainder of the payload is contained within or mounted to a single 1.252 m diameter cylindrical aluminum pressure vessel with two removable end caps for payload access, and two removal self-sealing connector flanges which provide the electrical interface between the vessel and the gondola. For ProtoEXIST1 a single telescope consisting of:
the CZT detector plane (§3)
a lower (internal to the pressure vessel) side shield (graded Z-shield) (§4.1)
a upper side shield and mask support structure (§4.1)
the active CsI rear shield (§4.2)
a single 198.9 nCi tagged Am calibration source (§4.2)
the coded aperture mask (§4.4)
is mounted in one of four available telescope slots, each of which is capable of accommodating a detector plane mounted within a EMI/RFI shielded enclosure with an approximate maximum size of 20 cm 20 cm. The shielded enclosure has a thin (6 mil) Al X-ray window for the observation of external sources. The internal structure of the pressure vessel (represented by the 3D model in figure 2) contains two main component mounting plates which are oriented perpendicular to the axis of the main vessel cylinder and are separated from one another by a distance of 288 mm. The mounting plate located nearest to the bottom cap serves as the optical bench for the detector planes and as the mount point for the active rear shield (see §4.2). The top surface also contains mount points for the internal half of the passive side shielding (see §4.1) which surrounds the detector plane and runs up to the bottom surface of the top cap. The flat pressure vessel top cap is located approximately 400 cm above the top surface of the detector plane and is nominally 25.4 mm thick. The inner surface of the top cap has a honeycomb structure with a pocket depth of 14.48 mm to minimize distortion of the top cap due to internal pressure during flight while reducing the overall weight by about 30%. 4 ports are cut into the top cap onto which an airtight low mass X-ray window is mounted through which the detector plane is exposed to the overhead sky. The transparent region of the X-ray window assembly measures 28 cm 28 cm and is composed of a external 0.61 mm thick Kevlar layer and a single internal 0.12 mm thick mylar layer attached to an aluminum mounting frame. The Kevlar provides mechanical strength against the internal vessel pressure while the mylar provides the airtight seal. The thin, light design enables transmission of X-rays down to approximately 5 keV. Mounting points are located on the top surface of the pressure vessel top cap for the superstructure which supports coded aperture mask which in the final integrated payload is located 513 mm above the upper surface of the top cap.
Additional space within the vessel adjacent to the detector plane mounting ports serve as mount points for the science package electrical subsystems. There are eight additional ports in all: 4 centrally located between detector planes and 4 located in the area between the mounting plate and the outer vessel skin. In the current (ProtoEXIST1) configuration a flight computer, a transmitter interface card, 3 Ethernet controlled heater control relays, the active shield and tagged Am calibration source control box, 2 primary pressure vessel heating elements, a power control, distribution, and monitoring box, and an Ethernet switch which serves as the hub of the science package internal communication system are mounted along with the prototype detector plane inside the vessel (see figure 2). The flight computer is the primary command/control and monitoring device for the science package and oversees all internal operations with the exception of the primary power control. The individual payload components communicate via an internal Ethernet network which serves as the primary communications link between the various subsystems. The flight computer is connected with a serial port (RS-422) to the receiver on the gondola which serves as the command uplink for the science package. A command interpreter running continuously on the flight computer monitors the uplink for incoming commands and, when a command is received, determines the required action to be taken. The command is then (in most cases) forwarded across the internal network to the relevant subsystem. The flight computer is also linked to a transmitter interface card using a RS-422 interface and also acts as the primary downlink. The downlink is used to monitor the status of the instrument package and its associated subsystems as well as transmit reduced data packets from the detector plane to the ground. Each subsystem within the pressure vessel, including the detector plane, transmits status information over the network back to the flight computer where the downlink monitor, another persistent process running on the flight computer, monitors the incoming data, collates it and transmits all information back to the ground station. In order to prevent deadlocks the power monitor and distribution subsystem receives commands directly from the CIP interface, however status information is still sent via the flight computer to the downlink.
The entire science package is physically supported by the mounting plates which are coupled to 8 hard points embedded in the pressure vessel main body skin which in turn are attached directly to the octagonal support structure (visible surrounding the vessel in figure 1). This serves simultaneously as the main load bearing structure for the internal mechanical structure of the vessel and as the elevation axis for the coded-aperture telescope. The octagonal support structure mounts directly to the ProtoEXIST1 gondola via two elevation flanges which rotate on bearings affixed to the gondola superstructure. The elevation is monitored using a shaft angle encoder (SAE) located on the flange.
3 The ProtoEXIST1 Detector Plane
The ProtoEXIST1 detector plane utilizes a modular architecture for redundancy and scalability. This design (as outlined section 1) is to be utilized for the construction of a much larger array i.e.EXIST. The ProtoEXIST1 detector plane consists of three primary components: a single flight control board (FCB), 8 detector crystal arrays (DCAs), 64 detector crystal units (DCUs)[14, 15, 16].
As shown in figure 3 each DCU consists of a
19.5 mm 19.5 mm 5 mm CZT crystal with a 8 8
pixilated anode bonded to a single interposer board (IPB) with single RadNet ASIC
bonded to the side opposite the CZT. The CZT crystals are procured from Redlen technologies
The DCA is composed of two printed circuit boards (PCB) with matching dimensions. Located on the upper surface of the upper board are sockets to which individual DCUs are mounted to form a approximately 4 cm 8 cm (32 cm) detector module. The upper board also contains analog-to-digital converters (ADC) for the digitization of the ASIC sample readout as well as digital-to-analog converters (DAC) for the input of settings into the ASIC (i.e. thresholds, etc.). The bottom surface of the board supports two connectors which make the connection to the lower DCA board. The lower DCA board contains a single FPGA which handles the processing of data as well as command and control of each of the 8 DCUs mounted to the DCA. In order to reduce the number of required lines in the board the DCA FPGA handles 4 independent data channels each of which consists of the readout from 2 DCUs (a single DCU pair). Each DCU pair is effectively treated as a single detector except that threshold for each DCU may still be set independently. The bottom surface of the lower DCA board contains three surface mount connectors which interface to the flight control board (FCB).
The FCB supports a array of DCAs (a total of 64 individual DCUs) which constitutes
the complete 16 cm 16 cm (256 cm) detector plane. Two EMCO HVPSs are mounted to
the underside the FCB, each with 4 leads (one per DCA) mounted on opposite sides of the detector plane.
The HV leads and the cathodes of the individual DCUs are coupled with conductive Al tape
(see figure 3). A thin flexible RFI/EMI shield consisting of a copper ground
plane sandwiched between two layers of kapton tape surrounds the individual DCU to
eliminate cross-talk between neighboring DCUs as well as reduce in the influence of any external
sources of noise (c.f. Allen et. al. in preparation).
For handling of the data as well as command and control functions the FCB contains a single
FPGA. All 32 (4 from each DCA) independent data channels are routed into the FCB-FPGA and
combined into a single data stream. The FPGA also serves as
the terminal input point for timing, anti-coincidence, and calibration source signals.
The FCB has 8 TTL inputs to accommodate discriminator output from the CsI rear shield,
the Am tagged calibration source, and the 1 PPS time synchronization signal from
the gondola GPS clock. In the ProtoEXIST1 flight configuration four of these inputs
were utilized: two were required for each PMT on the CsI rear shield, and a single input was
required for both the calibration source the 1 PPS GPS input. All of these signals are routed
into the FCB FPGA where tagging and timing information may be applied to the event data as
it is received. To ensure that the correct events are assigned the correct timing/tagging
flags each of the 32 data channels from the detector pairs supports a fast trigger line,
which connects directly to the FPGA. The fast trigger signals in combination with the
external discriminator inputs are compared with an internally set 5 s
anti-coincidence window to identify background and calibration source events in the
detector. After application of timing and tagging information the single data stream
is sent from the FCB-FPGA to a Coldfire MCF5282 microcontroller
4 Coded Aperture Telescope
The ProtoEXIST1 telescope occupies a single bay (out of a total of 4) in the pressure vessel and consists of four primary components: the graded Z-shield (§4.1) which restricts the detector plane’s field of view to the coded aperture mask and partially shields the it from atmospheric X-ray albedo, the CsI active rear shield (§4.2) which enables the tagging atmospheric X-ray albedo events as well as the identification of cosmic ray background. The coded aperture mask (§4.4), a perfect binary uniform redundant array (URA) which is mounted 900 mm away from the cathode surface of the detector plane and enables the reconstruction of hard X-ray source positions, and the pixilated 16 cm 16 cm position sensitive detector plane (§3) described above.
4.1 Graded Z-Shield
As previously alluded to the graded Z-shield is a passive shield installed in order to restrict the detector plane field of view to the coded aperture mask and to minimize the effects of the atmospheric albedo. The shield for a single telescope is divided into a 26 cm 26 cm 39.7 cm lower section, which sits inside the pressure vessel and surrounds the detector plane on 4 sides, and a 40 cm 40 cm 50 cm upper section which is affixed to the top cap of the science package pressure vessel. Both sections of the side shield are composed of laminated sheets of Pb-Sn-Cu (ordered from the outermost to the innermost layer) attached to an aluminum support frame. The design of the side shielding is based on the SWIFT-BAT graded Z-shield, however SWIFT-BAT employed a Z-shield composed of Pb-Sn-Ta-Cu. Due to the prohibitive cost of Ta at the time of construction this layer was dropped from the ProtoEXIST1 design and the thickness of the Sn and Cu layers were increased in order to compensate for its removal.
The inner layers of the graded Z-shield serve to attenuate the Pb fluorescence lines and, similarly, the innermost Cu layer attenuates the remaining Sn fluorescence lines, the Cu fluorescence lines ( keV) lie below the minimum detector threshold. Due to the discovery of a minor mechanical problem with the internal vessel structure a rebuild of the lower section of the side shield was required. As a result a portion of the Pb layer at the top of the lower section approximately 30 cm distant from the detector plane surface was left unshielded due the lack of additional Sn and Cu in the field. As a result the Pb-K and -K florescence lines were observed emanating from the exposed section during the flight (see §5).
4.2 Active Rear Shielding and Am Calibration Source
The shield electronics box is the subsystem/module that is responsible for the readout of raw PMT signals from the active rear shield and the Am calibration source. For this purpose the shield electronics box contains HV outputs for the biasing of PMTs, and computer controlled discriminator thresholds and outputs. The shield system is controlled by the flight computer over a serial port. A connection to the internal Ethernet network is available and used to send raw PMT pulse rates for both the CsI rear shield and the Am source. The discriminator outputs for each PMT are linked directly to the detector plane for anti-coincidence timing. The active rear shield itself is composed of a single 26 cm 26 cm 2 cm CsI crystal coupled to two PMTs across two wave shifter bars mounted on opposite sides of the scintillator. The PMTs are monitored individually and have separate discriminator output lines. The calibration source consists of a small aluminium cylinder that is filled with a Am doped scintillator. A clear window to which a single PMT is optically coupled is located at the top end of the cylinder which enables the detection 5.47 MeV -particles which are emitted in coincidence with the 59.6 keV X-rays used for detector calibration. The calibration source PMT has an independent discriminator input/output line on the shield electronics box as well that enables the identification or tagging of calibration X-ray events in the detector plane. The calibration source tagging efficiency observed in the hanger and during the flight is approximately 75%.
4.3 Detector Integration and Performance
The integration of the ProtoEXIST1 proceeded in two steps. Newly bonded DCUs were received at a rate of about 8-10 per week during peak production and integrated onto a single DCA for preliminary testing. Due to the production rate and time constraints most DCUs were not tested individually but tested together in a fully populated DCA. Previous tests have not revealed a significant difference in performance for healthy detectors operating in an array provided that the EMI/RFI shielding completely surrounds each DCU and is properly grounded. The test array would then undergo a short pulser test at 0 V for the detection of dead pixels, a pulser test at 600 V for the detection of hot pixels and any other possible anomalies, and finally undergo a Co calibration test. Detectors with more than 3 dead pixels or other anomalies were replaced and the DCA retested as needed. Once a good candidate DCA was completed it was set aside for integration into the flight detector.
The flight detector was then integrated 1 DCA at a time. After the installation a single DCA the 0 V pulser, 600 V pulser, and Co calibration tests were repeated on the partially completed array to detect any changes with the addition of a single DCA. This was repeated until the installation of all 8 fully populated DCAs was complete. During the detector integration and after arrival in the field at Ft. Sumner some detectors developed anomalies and required replacement.
During each integration step, with the exception of a few anomalous detectors which went bad during the course of pre-flight testing, no substantial changes in detector performance were observed between the operation individual isolated DCUs, DCUs integrated into a single array on a DCA, and DCAs integrated into the array of DCAs ( DCUs). In each case the detectors maintained an average resolution of 3.5 keV (FWHM) at 59.6 keV. A more complete account will be available in Allen et. al..
4.4 Coded Aperture Mask and Imaging Performance
The ProtoEXIST1 telescope utilizes a element prefect binary uniform
redundant array (URA) (c.f. references[19, 20])
placed 900 mm from the surface of the detector plane for image reconstruction.
4.2 mm square pixels with a pitch of 4.7 mm are utilized for the mask pattern,
the support grid (i.e. the material for support of the mask structure running between pixels)
has a thickness of 0.5 mm. The mask is composed of 12 identical laminated tungsten sheets
into which the URA pattern was chemically etched
The final assembled telescope has a fully coded field of view and an angular resolution of .
Prior to the construction of the mask a series of imaging simulations were carried out for optimization of the mask pattern. 10 separate masks were simulated utilizing the current URA pattern with different scale sizes between 4.5 mm and 5.5 mm at equal 0.1 mm steps within this range. In each case the 0.5 mm support grid was retained and the mask pixel size was reduced to achieve the change in pixel pitch. The detector plane was modeled to represent the real detector plane as closely as possible with 600 m gaps between each DCU. Due to the structure of the detector plane and mask additional induced coding noise was predicted to occur in detector images generated with the planned 5.0 mm mask. From these simulations it was determined that a mask with a 4.7 mm pitch minimized this effect and was ultimately adopted for use in the ProtoEXIST1 telescope. An imaging verification test was carried out with the fully integrated telescope prior to flight utilizing a single 3 mCi Co source placed at an approximate distance of 3.0 m from the surface of the mask. The performance of the test image was consistent with that of the simulated images after taking the magnification factor due to the placement of the source at a finite distance into account (Allen et. al. in preparation).
5 Flight Performance
ProtoEXIST1 began its maiden flight at 14:40 UT on the 9 of October 2009. The gondola reached float altitude (nominally 39 km) approximately 1.5 hours after liftoff (see figure a and table 2). The observation plan was initiated with an attempted observation of the quasar 3C 273, however during the first half of the flight due to instability in the elevation axis a stable pointing count not be achieved, presumably due to a stiff wire on the elevation axis. After the end of the 3C 273 observation run testing of the CsI rear shield thresholds was carried out over a period of approximately 30 min. At 15:00 an observation of the black hole binary Cyg X-1 was initiated and carried out over the course of one hour. A short system power control test and detector restart was initiated after which time the Cyg X-1 observing run was re-initiated until flight termination. Cyg X-1 was successfully observed and the payload was recovered with minimal damage to the gondola. The science package was recovered unharmed.
In spite of the short duration of the flight and the additional difficulties encountered with the pointing system the primary objectives: the detection of a celestial source (Cyg X-1) with the ProtoEXIST1 coded aperture telescope, the successful operation of the detector plane and measurement of backgrounds in a near space environment, and verification that all the payload subsystems functioned as designed in flight were all accomplished.
5.1 Pointing System Performance
During the flight 3 major difficulties with the pointing system were encountered which complicated flight operations and the analysis of the flight data: the failure of the differential global positioning system (DGPS) shortly after liftoff, the failure of the daytime star camera to acquire an aspect solution during the flight, and sporadic instability in the elevation axis.
During the first half of the flight instability of the elevation axis was a major issue (see figure b) and prevented the stable observation of 3C 273. Fortunately stability improved as the flight progressed and a long observation of Cyg X-1 was possible. The suspected cause is a combination of stiff wiring at low temperature in combination with a relatively weak elevation axis motor. In anticipation of future flights the replacement of the elevation axis motor is being considered.
The most serious issue encountered was the loss of the DGPS module during the ascent
of the payload (see figure a). As the DGPS is the primary method by which
an accurate azimuthal position for the coarse pointing system is determined its loss forced
the aspect system to fall back on the magnetometer. Unfortunately magnetometers have large
intrinsic systematics associated with their orientation relative to the position of other
structures on the payload as well as their position on the earth; a post-flight
analysis of the magnetometer data revealed substantial systematic offsets of up to
. On ascent attitude control was offline and the gondola was allowed to
rotate freely, which enabled the collection of simultaneous measurements
of the azimuth by both the magnetometer and the DGPS over the entire 360. These
measurements were later used to calibrate the magnetometer, which removed the most egregious
systematic errors in the azimuth measurement. After the application of an additional
correction taking into account the change of the magnetic declination over the course
of the flight
utilizing GEOMAG IGRF10
The next major difficulty was the failure of the daytime star camera to acquire a pointing solution during the flight. This was primarily the result of bad columns in the star camera CCD, a large offset between the true pointing position and the measure position due to the loss of the DGPS, and to a lesser extent insufficient shielding from the atmospheric background light at certain positions on the sky. The offset of the measured azimuthal value using the magnetometer likely prevented the determination of an aspect solution during the majority of the flight due to the finite star catalog search radius (approximately ) about the coarse aspect system’s reported position. To correct for these problems the CCD has been returned to the manufacturer for repair, an evaluation of the refurbished CCD is planned within the next year. In addition to this alternative star field search algorithms are being considered for the star camera in order to increase the size of the search radius. In spite of these difficulties usable star camera images archived during the flight proved useful in reconstruction of the telescope aspect information. Using cleaned archived star camera images the extraction of star positions in the image using SExtractor then an aspect solution using astrometry.net-0.25 was possible. Images were cleaned by removal of bad columns from the image and replacement of the bad column pixel values with the average taken from their 6 nearest neighbors, a rough flat fielding and application of a 2D Gaussian filter with pixels was applied to the entire image (see figure 6). After the aspect information for all star camera images was determined an aspect solution was generated then compared to the corrected magnetometer / elevation aspect data. Using this it was possible to independently confirm the validity of the corrected magnetometer aspect solution (see figure 5).
5.2 Telescope System Performance
The telescope and the associated subsystems performed without incident during the flight. The science payload heating system and control mechanism maintained a steady internal temperature of C for the entire duration of the flight. The flight computer and associated command and control system did not experience a single failure enabling the receipt of data from liftoff through flight termination. The internal power control and monitoring mechanism proved adequate; toward the end of the flight a system test was conduced to confirm the stable operation of the independent CIP interfaced power control mechanism as well as to test the complete remote reinitialization of the detector plane and other subsystems. The restart procedure, save a minor glitch in the data readouts, was a complete success. Within 15 min. after the restoration of power the telescope and all subsystems were running within nominal parameters and celestial observations were resumed.
In order to verify the operation of the coded-aperture telescope an observation of Cyg X-1 and 3C 273 was carried out. As outlined in §5.1 the observation of 3C 273 proved untenable due to issues encountered with the pointing system, however sufficient pointing stability was achieved during the observation of Cyg X-1. After application of corrected aspect data Cyg X-1 was detected at 7.2 with offset after 10 minutes of stable observation at the end of the flight (c.f. Hong et. al. in preparation ).
During the flight the background spectra and detector behavior were also monitored for any deviations in behavior due to charge saturation in the high radiation environment. In spite of the large increase in the background event rate the detector plane operated normally without any degradation in performance. The spectral response of the detector was monitored in flight using the tagged Am source and, as was observed on the ground the and during the flight. A slight increase in the average energy resolution was observed, 4.0keV in flight vs. 3.5 keV in the hanger (FWHM) at 59.6 keV averaged over all active pixels for the duration of the flight. The detector gain remained constant over the duration of the flight as did the background at cm s keV above 120 keV; in line with expectations. Additionally one of the key advantages of the tiled detector configuration proposed for EXIST, namely the reduction of background rates through shielding of the individual detector edges with a neighboring detector, was clearly demonstrated as well. For pixels on the edge of the detector plane, 252 in all, the measured backgrounds were 30% higher than observed in pixels located in the central (i.e. non-edge) pixels of the DCUs. For edges shielded by an adjacent detector, 1540 pixels total, a higher background rate was observed. Extrapolating from these rates had the ProtoEXIST1 detector not been closely tiled fully exposing all DCU edges a 10% increase in the background rate would be expected (see figure 9).
6 Future Telescope Development
Development for EXIST is currently continuing with ProtoEXIST2. ProtoEXIST2 will utilize the DB-ASIC for the readout of 19.5 mm 19.5 mm 5.0 mm CZT detectors with a array of anode pixels with a 600 m pitch. Using a slightly modified version of the ProtoEXIST1 modularization scheme (see §3); a sub-array of these detectors (1/4 of a detector plane) will be fabricated and demonstrated. It is anticipated that testing will begin on the first detector modules for ProtoEXIST2 within the next few months with the completion of the first sub-array occurring before April of 2011. In ProtoEXIST3 the remaining 3/4 of the detector plane will be built using a lower power version derived from the DB-ASIC which consumes about a factor of four less power, the EXIST Final or EX-ASIC. With the completion of the detector plane a test flight with both the original ProtoEXIST1 and new ProtoEXIST2/3 detector planes will be be carried out in the fall of 2012. The ProtoEXIST2/3 telescope will feature an extended 2 m focal length, be capable of imaging sources with an angular resolution of and have a fully coded field of view. Parallel to this effort a further modification of the EX-ASIC design for the replacement of wirebond pads, which prevent close tiling over the entire detector plane, with microvias to form the final EXIST ASIC (EXF-ASIC).
7 Conclusions and Future Plans
The successful integration of the ProtoEXIST1 detector plane and its operation at high altitude in a near space environment have demonstrated the feasibility of operating a large area tiled detector plane similar to the EXIST-HET utilizing individual pixiliated CZT detectors (DCUs) as the core building block. Measured spectra revealed that the high background rates and charge saturation due to incident cosmic rays had little impact () on the detector energy resolution and no discernible impact on the detector gains. The elevated event rates also did not pose any difficulties for the detector plane electronics or the data acquisition system. It has also been demonstrated that the interference between the individual DCUs on the ground and in flight does not pose a significant challenge to the construction of large tiled arrays. In short the operation of the telescope and science package was a complete success on the ground and in flight. Additionally a key advantage of the tiled array, the mutual shielding of detector edges by neighboring DCUs, was also demonstrated in flight showing that a reduction in backgrounds of at least 10% is possible over a non-tiled detector plane configuration.
In the near future repair of the landing damage on the gondola will be carried out as will improvements to the pointing system to improve reliability. Once the detector integration for ProtoEXIST2/3 is completed a flight together with the original ProtoEXIST1 detector plane is planned for characterization and comparison of both systems in flight.
- Launch Date: 17 October 2002 (Baikonur)
- Launch Date: 20 November 2004 (Cape Canaveral)
- Low temperature solder bonding carried out at Delphon/QuikPak (http://www.delphon.com)
- TLPS bonding carried out at Creative Electron (http://www.creativeelectron.com)
- Photo-etched at Tech-Etch, Plymouth, MA http://www.tech-etch.com
- Ubertini, P., Lebrun, F., Di Cocco, G., Bazzano, A., Bird, A. J., Broenstad, K., Goldwurm, A., La Rosa, G., Labanti, C., Laurent, P., Mirabel, I. F., Quadrini, E. M., Ramsey, B., Reglero, V., Sabau, L., Sacco, B., Staubert, R., Vigroux, L., Weisskopf, M. C., and Zdziarski, A. A., “IBIS: The Imager on-board INTEGRAL,” Astronomy & Astrophysics 411, L131–L139 (Nov. 2003).
- Lebrun, F., Leray, J. P., Lavocat, P., Crétolle, J., Arquès, M., Blondel, C., Bonnin, C., Bouère, A., Cara, C., Chaleil, T., Daly, F., Desages, F., Dzitko, H., Horeau, B., Laurent, P., Limousin, O., Mathy, F., Mauguen, V., Meignier, F., Molinié, F., Poindron, E., Rouger, M., Sauvageon, A., and Tourrette, T., “ISGRI: The INTEGRAL Soft Gamma-Ray Imager,” Astronomy & Astrophysics 411, L141–L148 (Nov. 2003).
- Barthelmy, S. D., Barbier, L. M., Cummings, J. R., Fenimore, E. E., Gehrels, N., Hullinger, D., Krimm, H. A., Markwardt, C. B., Palmer, D. M., Parsons, A., Sato, G., Suzuki, M., Takahashi, T., Tashiro, M., and Tueller, J., “The Burst Alert Telescope (BAT) on the SWIFT Midex Mission,” Space Science Reviews 120, 143–164 (Oct. 2005).
- Barrett, H. H., Eskin, J. D., and Barber, H. B., “Charge Transport in Arrays of Semiconductor Gamma-Ray Detectors,” Physical Review Letters 75, 156–159 (July 1995).
- Doty, F. P., Barber, H. B., Augustine, F. L., Butler, J. F., Apotovsky, B. A., Young, E. T., and Hamilton, W., “Pixellated CdZnTe detector arrays,” Nuclear Instruments and Methods in Physics Research A 353, 356–360 (Dec. 1994).
- Harrison, F. A., Christensen, F. E., Craig, W., Hailey, C., Baumgartner, W., Chen, C. M. H., Chonko, J., Cook, W. R., Koglin, J., Madsen, K., Pivavoroff, M., Boggs, S., and Smith, D., “Development of the HEFT and NuSTAR focusing telescopes,” Experimental Astronomy 20, 131–137 (Dec. 2005).
- Baumgartner, W. H., Tueller, J., Krimm, H., Barthelmy, S. D., Berendse, F., Ryan, L., Birsa, F. B., Okajima, T., Kunieda, H., Ogasaka, Y., Tawara, Y., and Tamura, K., “InFOCuS hard x-ray telescope: pixellated CZT detector/shield performance and flight results,” in [Society of Photo-Optical Instrumentation Engineers (SPIE) Conference Series ], J. E. Truemper & H. D. Tananbaum, ed., Society of Photo-Optical Instrumentation Engineers (SPIE) Conference Series 4851, 945–956 (Mar. 2003).
- Hong, J., Grindlay, J. E., Allen, B., Barthelmy, S. D., Skinner, G. K., and Gehrels, N., “The high energy telescope on EXIST,” in [Society of Photo-Optical Instrumentation Engineers (SPIE) Conference Series ], Society of Photo-Optical Instrumentation Engineers (SPIE) Conference Series 7435 (Aug. 2009).
- Hong, J. et. al.., “The proposed high-energy telescope (het) for exist,” These Proceedings 7732.
- Rana, V. R., Cook, III, W. R., Harrison, F. A., Mao, P. H., and Miyasaka, H., “Development of focal plane detectors for the Nuclear Spectroscopic Telescope Array (NuSTAR) mission,” in [Society of Photo-Optical Instrumentation Engineers (SPIE) Conference Series ], Society of Photo-Optical Instrumentation Engineers (SPIE) Conference Series 7435 (Aug. 2009).
- Grindlay, J. E., Garcia, M. R., Burg, R. I., and Murray, S. S., “The energetic X-ray imaging telescope experiment (EXITE),” IEEE Transactions on Nuclear Science 33, 750–754 (Feb. 1986).
- Ramsey, B. D., Alexander, C. D., Apple, J. A., Benson, C. M., Dietz, K. L., Elsner, R. F., Engelhaupt, D. E., Ghosh, K. K., Kolodziejczak, J. J., O’Dell, S. L., Speegle, C. O., Swartz, D. A., and Weisskopf, M. C., “First Images from HERO, a Hard X-Ray Focusing Telescope,” Astrophysical Journal 568, 432–435 (Mar. 2002).
- Dietz, K. L., Ramsey, B. D., Alexander, C. D., Apple, J. A., Ghosh, K., and Swift, W. R., “Daytime aspect camera for balloon altitudes,” Optical Engineering 41, 2641–2651 (Oct. 2002).
- Hong, J., Allen, B., Grindlay, J., Chammas, N., Barthelemy, S., Baker, R., Gehrels, N., Nelson, K. E., Labov, S., Collins, J., Cook, W. R., McLean, R., and Harrison, F., “Building large area CZT imaging detectors for a wide-field hard X-ray telescope–ProtoEXIST1,” Nuclear Instruments and Methods in Physics Research A 605, 364–373 (July 2009).
- Hong, J., Grindlay, J. E., Chammas, N., Allen, B., Copete, A., Said, B., Burke, M., Howell, J., Gauron, T., Baker, R. G., Barthelmy, S. D., Sheikh, S., Gehrels, N., Cook, W. R., Burnham, J. A., Harrison, F. A., Collins, J., Labov, S., Garson, III, A., and Krawczynski, H., “Wide-field hard x-ray survey telescope: ProtoEXIST1,” in [Society of Photo-Optical Instrumentation Engineers (SPIE) Conference Series ], Presented at the Society of Photo-Optical Instrumentation Engineers (SPIE) Conference 6706 (Sept. 2007).
- Hong, J., Grindlay, J. E., Chammas, N., Copete, A., Baker, R. G., Barthelmy, S. D., Gehrels, N., Cook, III, W. R., Burnham, J. A., Harrison, F. A., Collins, J., and Craig, W. W., “CZT imaging detectors for ProtoEXIST,” in [Society of Photo-Optical Instrumentation Engineers (SPIE) Conference Series ], Presented at the Society of Photo-Optical Instrumentation Engineers (SPIE) Conference 6319 (Aug. 2006).
- Allen, B. et. al.., “The ProtoEXIST1 Coded-Aperture Hard X-Ray Telescope,” In Preparation (2010).
- Robinson, D. W., “Building a new kind of graded-Z shield for Swift’s burst alert telescope,” in [Society of Photo-Optical Instrumentation Engineers (SPIE) Conference Series ], J. E. Truemper & H. D. Tananbaum, ed., Presented at the Society of Photo-Optical Instrumentation Engineers (SPIE) Conference 4851, 1374–1381 (Mar. 2003).
- Fenimore, E. E. and Cannon, T. M., “Coded aperture imaging with uniformly redundant arrays,” Applied Optics 17, 337–347 (Feb. 1978).
- Busboom, A., Elders-Boll, H., and Schotten, H. D., “Uniformly Redundant Arrays,” Experimental Astronomy 8, 97–123 (1998).
- Bertin, E. and Arnouts, S., “SExtractor: Software for source extraction.,” Astronomy & Astrophysics Suppliment Series 117, 393–404 (June 1996).
- Lang, D., Hogg, D. W., Mierle, K., Blanton, M., and Roweis, S., “Astrometry.net: Blind Astrometric Calibration of Arbitrary Astronomical Images,” Astronomical Journal 139, 1782–1800 (May 2010).
- Hong, J. et. al.., “Flight performance of an advanced czt imaging detector in a wide-field hard x-ray telescope- protoexist1,” In Preparation (2010).