Hybrid Pixel Detector Research Articles

Design of the AGIPD sensor for the European XFEL

For experiments at the European X-Ray Free-Electron Laser (XFEL) the Adaptive Gain Integrating Pixel Detector (AGIPD) is under development. The particular requirements include a high dynamic range of 0, 1 to more than 104 photons of 12.4 keV per pixel within an XFEL pulse of less than 100 fs duration, and a radiation tolerance for doses up to 1 GGy for 3 years of operation. AGIPD is a hybrid-pixel detector system with a total of 1024 × 1024 pixels of size 200 μm × 200 μm. AGIPD will use 16 p+n-silicon sensors of 500 μm thickness. The nominal operating voltage is 500 V, however, for special applications operation close to 1000 V should be possible. Experimental data on the X-ray-dose dependence of the oxide-charge density and of the surface-current density have been implemented in the SYNOPSYS TCAD simulation program in order to optimize the design of the pixel and guard-ring layout. The methodology of the sensor design, the optimization of the most relevant parameters, and the optimized layout are described. Finally, the simulated performance, in particular the breakdown voltage, dark current and inter-pixel capacitance as function of the X-ray dose are presented.

Synchrotron X-ray diffraction analysis for quantitative defect evaluation in GaP/Si nanolayers

In the context of III–V monolithic integration on silicon, synchrotron X-ray diffraction has been employed in this study using a bi-dimensional large area hybrid pixel detector (XPAD third generation) to characterize defects in the GaP layers. Despite a very coherent interface (low plastic relaxation) of GaP/Si, 2 types of defect are detected. Micro-twins contributions are evidenced and quantitatively evaluated from additional reflections analysis. Antiphase domains are evidenced using the Williamson-Hall-like plot method applied to transverse scans extracted directly from single XPAD images taken on specular GaP reflections.

Algorithms for minimization of charge sharing effects in a hybrid pixel detector taking into account hardware limitations in deep submicron technology

Charge sharing is the main limitation of pixel detectors used in spectroscopic applications, noting that this applies to both time and amplitude/energy spectroscopy. Even though, charge sharing was the subject of many studies, there is still no ultimate solution which could be implemented in the hardware to suppress the negative effects of charge sharing. This is mainly because of strong demand on low power dissipation and small silicon area of a single pixel. The first solution of this problem was proposed by CERN and consequently it was implemented in the Medipix III chip. However, due to pixel-to-pixel threshold dispersions and some imperfections of the simplified algorithm, the hit allocation was not functioning properly. We are presenting novel algorithms which allow proper hit allocation even at the presence of charge sharing. They can be implemented in an integrated circuit using a deep submicron technology. In performed simulations, we assumed not only diffusive charge spread occurring in the course of charge drifting towards the electrodes but also limitations in the readout electronics, i.e. signal fluctuations due to noise and mismatch (gain and offsets). The simulations show that using, for example, a silicon pixel detector in the low X-ray energy range, we have been able to perform proper hit position identification and use the information from summing inter-pixel nodes for spectroscopy measurements.

Initial results on charge and velocity discrimination for heavy ions using silicon-Timepix detectors

The Timepix ASIC is a version of the hybrid pixel detector technology developed by the Medipix2 Collaboration [1]. Within the 256 by 256 pixel matrix, the electronics for each of the 55μm individual pixels are contained in the footprint of that pixel. The Timepix has a charge-sensitive pre-amp and an associated discriminator attached to a logic unit capable of being employed in one of several different modes. For the present work, the Time-Over-Threshold mode was used to allow measurement of deposited energy in the silicon sensor layer. The general properties of a Timepix-based device with a silicon sensor have been described in [2,3], and [4].Ionization along heavy ion particle tracks in the silicon sensor results in the production of free charge carriers in the detector. The charge carrier motion under the influence of an applied bias voltage leads to charge collection at the Timepix-sensor interface in one or more pixels. Signatures within the pixel cluster patterns are currently being examined, and initial results indicate that such signatures, when coupled with stopping power information, provide enough discrimination capability to begin to resolve heavy ion charge and velocity. Here we present the salient characteristics that have been identified for heavy ion charge and velocity discrimination using Timepix Silicon detectors and discuss the application of this method for particle track characterization.

High accuracy injection circuit for the calibration of a large pixel sensor matrix

Semiconductor pixel detectors, for particle tracking and vertexing in high energy physics experiments as well as for X-ray imaging, in particular for synchrotron light sources and XFELs, require a large area sensor matrix. This work will discuss the design and the characterization of a high-linearity, low dispersion injection circuit to be used for pixel-level calibration of detector readout electronics in a large pixel sensor matrix. The circuit provides a useful tool for the characterization of the readout electronics of the pixel cell unit for both monolithic active pixel sensors and hybrid pixel detectors. In the latter case, the circuit allows for precise analogue test of the readout channel already at the chip level, when no sensor is connected. Moreover, it provides a simple means for calibration of readout electronics once the detector has been connected to the chip. Two injection techniques can be provided by the circuit: one for a charge sensitive amplification and the other for a transresistance readout channel. The aim of the paper is to describe the architecture and the design guidelines of the calibration circuit, which has been implemented in a 130nm CMOS technology. Moreover, experimental results of the proposed injection circuit will be presented in terms of linearity and dispersion.

The data acquisition system of the SuperB-SVT beam test

Prototypes of a new hybrid pixel detector and a high resistivity detector with short strips, developed by the VIPIX Collaboration and aimed at equipping the layer-0 of the SuperB vertex detector, have been tested in September 2011 with a 120GeV pion beam at the SPS H6 beam line at CERN. They are placed at the center of a reference telescope consisting of six planes of silicon detector with a double-sided strip readout. Both the telescope and the detectors under test (DUT) are equipped with a custom-design, data-push digital readout. The main elements of the trigger and data acquisition system are two VME boards (EDRO) organized in a master-slave configuration and responsible for programming the front-end chips of both the telescope and the DUT. The master board distributes a global synchronization clock and the triggers to all devices, including two 3×3 analog-maps matrices placed behind the DUT and supplied with an independent readout. Both EDROs act as event-fragment builders. These are sent out to a remote PC for event building, buffering and storage.

Detective quantum efficiency model of single-X-ray-photon counting hybrid pixel detectors

A Detective Quantum Efficiency (DQE) model of single-X-ray-Photon Counting Hybrid Pixel Detectors (PC-HPDs) is presented. It applies to PC-HPDs based on semiconductor sensors such as silicon and CdTe pixel sensors. Charge-sharing effects are introduced in the expressions of imaging performance parameters such as large-area gain factor, presampling modulation transfer function and digital noise power spectrum, using the concept of threshold-dependent effective fill-factor. A simple X-ray induced charge distribution approximation is used to derive a practical formula for the threshold-dependent large-area gain factor, i.e. the integral X-ray spectrum which can be indirectly measured with a PC-HPD. This detector model was applied to standard synchrotron X-ray PC-HPDs: MEDIPIX3, PILATUS and XPAD detectors.

The detector simulation toolkit HORUS

In recent years, X-ray detectors used and developed at synchrotron sources and Free Electron Lasers (FELs) have become increasing powerful and versatile. However, as the capabilities of modern X-ray cameras grew so did their complexity and therefore their response functions are far from trivial. Since understanding the detecting system and its behavior is vital for any physical experiment, the need for dedicated powerful simulation tools arose.The HPAD Output Response fUnction Simulator (HORUS) was originally developed to analyze the performance implications of certain design choices for the Adaptive Gain Integrating Pixel Detector (AGIPD) and over the years grew to a more universal detector simulation toolkit covering the relevant physics in the energy range from below 1 keV to a few hundred keV. HORUS has already been used to study possible improvements of the AGIPD for X-ray Photon Correlation Spectroscopy (XPCS) at the European XFEL and its performance at low beam energies. It is currently being used to study the optimum detector layout for Coherent Diffration Imaging (CDI) at the European XFEL. Simulations of the charge summing mode of the Medipix3 chip have been essential for the improvements of the charge summing mode in the Medipix3 RX chip. HORUS is universal enough to support arbitrary hybrid pixel detector systems (within limitations). To date, the following detector systems are predefined within HORUS: The AGIPD, the Large Pixel Detector (LPD), the Cornell-Stanford Pixel Array Detector (CSPAD), the Mixed-Mode (MMPAD) and KEKPAD, and the Medipix2, Medipix3 and Medipix3 RX chips.

The AGIPD detector for the European XFEL

The Adaptive Gain Integrating Pixel Detector (AGIPD) is a hybrid pixel detector with bump-bonded silicon sensor, which is being developed for the European X-ray Free Electron Laser (XFEL). The XFEL provides photons in an energy range between 0.2 and 24.8 keV [1]. XFELs unique time structure comprises bunch trains at a repetition rate of 10 Hz, where each train consists of 2700 bunches with a temporal separation of 220 ns. Each bunch itself consists of ∼ 1012 photons impinging in less than 100 fs. The challenges posed by this time structure will be addressed using a charge integrating device with dynamic gain switching and the possibility to store the analog signals on at least 200 storage cells. Up to now, three different prototype chips have been built, in order to investigate the different blocks of the final AGIPD chip. This paper will report on the results of the characterization of the first two prototype chips, namely AGIPD0.1 and AGIPD0.2. While the investigations on AGIPD0.1 focus on the readout chain (dynamic gain switching, linearity, etc.), the tests of AGIPD0.2 concentrate on the measurements involving the bump-bonded silicon sensor, such as the possibility of resolving single photons and pixel-to-pixel variations as well as the performance of the analog storage pipeline.

Electrical measurements of a multi-mode hybrid pixel detector ASIC for radiation detection

We present the first electrical measurements of an application-specific integrated circuit (ASIC) to be used in a hybrid pixel detector intended for dosimetry and radiation detection. The dosimeter has three programmable modes of operation: photon counting mode, energy integration mode, and dosimetry mode. The ASIC comprises a matrix of 16 by 16 (256 total) square pixels of 220 μm pitch, providing 12.4 mm2 of segmented active area. Each pixel can be configured to operate in one of the three radiation measurement modes, with programmable-depth counters and shift registers to tailor the data word size and optimise the readout frame-rate in a given mode. The individual energies of impinging photons are determined through programmable analogue energy threshold discrimination, time over threshold measurement, or a combination thereof. Furthermore, the dosimetry mode contains 16 digital energy thresholds and automatically sorts data into 16 corresponding energy bin registers. The chip's output is therefore pre-processed charge spectra of the radiation field. This paper discusses results from measurements taken using programmable test-pulses to inject controlled stimuli into the pixel circuits.

Synchrotron X-ray diffraction experiments with a prototype hybrid pixel detector

A prototype X-ray pixel area detector (XPAD3.1) has been used for X-ray diffraction experiments with synchrotron radiation. The characteristics of this detector are very attractive in terms of fast readout time, high dynamic range and high signal-to-noise ratio. The prototype XPAD3.1 enabled various diffraction experiments to be performed at different energies, sample-to-detector distances and detector angles with respect to the direct beam, yet it was necessary to perform corrections on the diffraction images according to the type of experiment. This paper is focused on calibration and correction procedures to obtain high-quality scientific results specifically developed in the context of three different experiments, namely mechanical characterization of nanostructured multilayers, elastic–plastic deformation of duplex steel and growth of carbon nanotubes.

Optimal fine ϕ-slicing for single-photon-counting pixel detectors

The data-collection parameters used in a macromolecular diffraction experiment have a strong impact on data quality. A careful choice of parameters leads to better data and can make the difference between success and failure in phasing attempts, and will also result in a more accurate atomic model. The selection of parameters has to account for the application of the data in various phasing methods or high-resolution refinement. Furthermore, experimental factors such as crystal characteristics, available experiment time and the properties of the X-ray source and detector have to be considered. For many years, CCD detectors have been the prevalent type of detectors used in macromolecular crystallography. Recently, hybrid pixel X-ray detectors that operate in single-photon-counting mode have become available. These detectors have fundamentally different characteristics compared with CCD detectors and different data-collection strategies should be applied. Fine φ-slicing is a strategy that is particularly well suited to hybrid pixel detectors because of the fast readout time and the absence of readout noise. A large number of data sets were systematically collected from crystals of four different proteins in order to investigate the benefit of fine φ-slicing on data quality with a noise-free detector. The results show that fine φ-slicing can substantially improve scaling statistics and anomalous signal provided that the rotation angle is comparable to half the crystal mosaicity.

Influence of electromagnetic interference on the analog part of hybrid Pixel detectors

The analog signal from the sensor of hybrid semiconductor pixel detectors is prone to electro-magnetic interference. The study and diagnosis of induced and common electro-magnetic coupling between the analog part and digital part of these devices is required. The influence of electro-magnetic interference was tested on the setup with a pixel detector Timepix or Medipix and a FITPix read-out interface. Measurements were carried out of external as well as internal interference. We evaluated the influence of both sources of electro-magnetic interference to the noise recorded by pixels. We measured the local spatial intensity distribution and frequency spectrum of the electro-magnetic field originating inside the readout chip during its own operation. In context of this test we exposed the detector chip to a locally generated artificial electro-magnetic field evaluating its sensitivity to induced interference. Consequently, the whole setup of the detector and read-out interface was exposed to a distant source of electro-magnetic radiation, during which we tested efficiency of the electro-magnetic shielding of various arrangements. Further, tests measured the coupling over power supply lines. In particular, the noise generated by the operation of the detector itself was determined. In addition, the detector sensitivity to deliberately induced noise was evaluated. By means of these tests weak points of the setup sensitive to the intrusion of electro-magnetic interference are revealed. When locations of susceptible places are identified proper methods can be applied to increase immunity of the detector setup against the electro-magnetic interference. Experiences gained are planned to be used in development of the EMI shielded version of the FITPIX interface shielded to electro-magnetic interference.

Characterisation of the NA62 GigaTracker End of ColumnDemonstrator Hybrid Pixel Detector

The architecture and characterisation of the NA62GigaTracker End of Column Demonstrator Hybrid Pixel Detector (HPD) arepresented. This detector must perform time stamping to 200 ps (RMS)or better, provide 300 μm pitch position information and operatewith a dead time of 1% or less for 800 MHz−1 GHz beamrate. The demonstrator HPD Assembly comprises a readout chip with atest column of 45 pixels, alongside other test structures, bump bondedto a p-in-n detector 200 μm in thickness. Validation of theperformance of the HPD and the time-over-threshold timewalkcompensation mechanism with both beam particles and a high precisionlaser system was performed and is presented. Confirmation of betterthan the required time stamping precision has been demonstrated andsubsequent work on the design of the full-scale ASIC, dubbed TDCPix,is underway. An overview of the TDCPix architecure is given.

TSV processing of Medipix3 wafers by CEA-LETI: a progress report

The next generation of hybrid pixel detectors in particle physics experiments require reduced material budget, increased interconnection density and, ideally, they should be tileable to cover large areas seamlessly. These criteria cannot be fulfilled with present day interconnection techniques. As a result the particle physics community has recently put in a lot of effort to investigate and evaluate a variety of novel interconnection technologies.This paper focuses on describing a recently launched Through Silicon Via process development project with CEA-LETI. The project aims to use Medipix3 wafers and an existing ``via last'' TSV process made available by CEA-LETI to demonstrate the feasibility of TSVs on functional detector chips. The status of the project, TSV design and future plans are presented.

Charge Signal Processors in a 130 nm CMOS Technology for the Sparse Readout of Small Pitch Monolithic and Hybrid Pixel Sensors

Hybrid pixel detectors are nowadays a robust and mature technology for particle detection as well as for medical and X-ray imaging, but the demands for improved performance for the next generation high energy physics experiments ask for the development of novel devices. Monolithic CMOS pixels have the potential to provide high granularity thin detectors as the sensor and the readout electronics are integrated in the same substrate. However, they suffer from a few limitations closely related to their working principle. 3D vertical integration has the potential of providing a performance breakthrough toward the design of fast, radiation tolerant and ultra thin CMOS radiation sensors. This work will discuss the design of analog circuits for processing the signals from small pitch monolithic and hybrid pixel detectors designed and fabricated in a planar 130 nm CMOS technology and in a 130 nm CMOS technology with vertical integration capabilities. Various solutions complying with different S/N ratio and detector capacitance constraints will be described in this paper by means of circuit simulations and experimental results.

Triggerless Micro Vertex Detector with low material budget in the PANDA experiment

Besides the typical features characterizing the silicon detectors in the High Energy Physics such as good spatial resolution, energy loss measurement, limited material budget and radiation hardness, the Micro Vertex Detector (MVD) for the PANDA experiment (at the FAIR Facility in Darmstadt connected to the GSI Laboratory) is requested to cope with special issues. The expected study of many physics channels at high antiproton-proton annihilation rates (2·107 ann/s) is asking for a triggerless readout with a continuous data transmission and a good time resolution. The wide particle momentum distribution starting from some hundreds of MeV/c asks both a very limited material budget and a high dynamic range. The MVD is composed of: hybrid pixel (100μm ×100μm) detectors with thin silicon epitaxial sensor and readout developed by 130nm CMOS technology and double sided silicon strip detectors with supporting mechanics based on carbon fibres. Carbon foam layers will improve the heat transfer towards the cooling system based on water and working in vacuum-operated mode. Last developments in the MVD design and results concerning the prototyping phase are reported.

Optimal fine ϕ slicing for single photon counting pixel detectors

The data-collection parameters used in a macromolecular diffraction experiment have a strong impact on data quality. A careful choice of parameters leads to better data and can make the difference between success and failure in phasing attempts, and will also result in a more accurate atomic model. The selection of parameters has to account for the application of the data in various phasing methods or high-resolution refinement. Furthermore, experimental factors such as crystal characteristics, available experiment time and the properties of the X-ray source and detector have to be considered. For many years, CCD detectors have been the prevalent type of detectors used in macromolecular crystallography. Recently, hybrid pixel X-ray detectors that operate in single-photon-counting mode have become available. These detectors have fundamentally different characteristics compared with CCD detectors and different data-collection strategies should be applied. Fine φ-slicing is a strategy that is particularly well suited to hybrid pixel detectors because of the fast readout time and the absence of readout noise. A large number of data sets were systematically collected from crystals of four different proteins in order to investigate the benefit of fine φ-­slicing on data quality with a noise-free detector. The results show that fine φ-slicing can substantially improve scaling statistics and anomalous signal provided that the rotation angle is comparable to half the crystal mosaicity.

A pixel detector asic for dosimetry using time-over-threshold energy measurements

In this work we present the design of a chip which provides the readout of a highly segmented diode array, in which signals induced by individual X-ray photons are processed discretely. There are several benefits to this approach, including the ability to achieve a high signal to noise ratio due to the inherently low sensor capacitance, and the suppression of background noise ( e.g . dark current) using an analogue threshold. The segmentation also ensures a linear behaviour even at very high dose rates. A time over threshold (ToT 1 1 ToT: Time over threshold. ASIC: Application-specific integrated circuit. APS: Active pixel sensor. DAC: Digital to analogue convertor. CSA: Charge sensitive preamplifier. ENC: Equivalent noise charge. ) energy measurement technique provides an immediate digital value corresponding to the energy deposited onto the diode by each individual photon. Deadtime-free operation is achieved by reading out a subset of the detector segments at a time while the rest of the detector continues to process signals. This paper describes the application-specific integrated circuit (ASIC) chip which was designed to provide pre-processing of photo-induced signals in the detector and readout of the processed digital data. ► We have designed the ASIC of a hybrid pixel detector for photon dosimetry. ► The pulse processing circuits measure the time over threshold of discrete photons. ► Each pixel automatically sorts ToT measurements into 16 digital energy bins. ► The binned data is used to reconstruct the charge spectrum and dose in real time.

Application of Hybrid Pixel Detectors for Searches of Rare Decays

The new generation of hybrid pixel detectors like the Timepix detector provides access to the track information of the energy deposition in the used sensor, which allows better background discrimination in experiments for searches of rare decays. Due to the hybrid design several combinations of an ASIC with a sensor are possible. Assemblies are available with an attached Silicon sensor or CdTe sensor respectively. The detector measures the energy deposition using the time over threshold method. In this contribution we present simulation results of the detector response of the Timepix detector in applications for searches of rare decays. One application would be the search for the neutrinoless double beta decay of 116 Cd using Timepix detectors with enriched CdTe as sensor material. In addition to the simulation results we present first experimental background measurements using a Timepix detector with Silicon sensor in the underground laboratory in Dresden. Using cluster analysis methods it is possible to categorise the single events.