Training on quantum detection, single-photon imaging, SiPMs, SPADs

22 May 2013, 08:00 → 24 May 2013, 21:00 Europe/Zurich

TU-Delft

The aims of the training are understanding sigital SiPMs (theory, design, applications), practicing d-SiPM design, understanding readout issues and time-to-digital conversion (theory, design and integration in d-SiPM) and practicing FPGA based TDC.

The 3 day course is conceived as a mix of theory and practice.

• check the list of talks
• check the timetable

• register
• info about the booking procedure
• see the social events details

Wednesday 22 May

09:00 → 12:15 Silicon Photomultipliers and Applications (1st Module)

Advanced manufacturing technologies for modern image sensors have advanced to the point where uncooled single-photon image can be considered a consumer product. Single-photon imaging brings its own challenges, and has developed a number of niche applications, but the nature of building detectors at these extreme performance levels means performance developments that can filter down through more mainstream sensor technologies and applications.

This lecture has two segments: the first is devoted to avalanche theory, SPAD fundamentals, and silicon photomultipliers (SiPMs), both analog and digital; the second is focusing on timing estimation and processing in multi-channel digital SiPMs, including a discussion of the expected and measured statistical behavior of the device and its application in many fields of science and medical imaging.

---

Edoardo Charbon (SM’10) received the Diploma from ETH Zurich in 1988, the M.S. degree from UCSD in 1991, and the Ph.D. degree from UC-Berkeley in 1995, all in Electrical Engineering and EECS. In 2000, he joined Canesta Inc. as its Chief Architect, leading the development of wireless 3-D CMOS image sensors. Canesta was sold to Microsoft in 2010. Since November 2002, he has been a member of the Faculty of EPFL in Lausanne, Switzerland, working in the field of CMOS sensors, biophotonics, and ultra low-power wireless embedded systems. In Fall 2008 he has joined the Faculty of TU Delft, as full professor in VLSI design, succeeding Patrick Dewilde. His current research includes medical and space based image sensing, single-photon technology, and picosecond electronics.

Convener: Edoardo Charbon (TU Delft)

09:00 Basics: avalanching theory, SPAD fundamentals, analog vs. digital SiPMs, readout architectures

Advanced manufacturing technologies for modern image sensors have advanced to the point where uncooled single-photon image can be considered a consumer product. Single-photon imaging brings its own challenges, and has developed a number of niche applications, but the nature of building detectors at these extreme performance levels means performance developments that can filter down through more mainstream sensor technologies and applications. This lecture has two segments: the first is devoted to avalanche theory, SPAD fundamentals, and silicon photomultipliers (SiPMs), both analog and digital; the second is focusing on timing estimation and processing in multi-channel digital SiPMs, including a discussion of the expected and measured statistical behavior of the device and its application in many fields of science and medical imaging.

Speaker: Edoardo Charbon (TU Delft)

Slides pres052213_v4.pdf

10:30 Coffee break

10:45 Timing evaluation and processing, multi-channel digital SiPMs: a statistical approach, system issues, testing and metrology, other applications

Advanced manufacturing technologies for modern image sensors have advanced to the point where uncooled single-photon image can be considered a consumer product. Single-photon imaging brings its own challenges, and has developed a number of niche applications, but the nature of building detectors at these extreme performance levels means performance developments that can filter down through more mainstream sensor technologies and applications. This lecture has two segments: the first is devoted to avalanche theory, SPAD fundamentals, and silicon photomultipliers (SiPMs), both analog and digital; the second is focusing on timing estimation and processing in multi-channel digital SiPMs, including a discussion of the expected and measured statistical behavior of the device and its application in many fields of science and medical imaging.

Speaker: Edoardo Charbon (TU Delft)

Slides pres052213_v4.pdf

12:30 → 14:00 Lunch break

14:00 → 17:30 Single-Photon Technology in Medicine (2nd Module)

Convener: York Haemisch (P)

14:00 Single-photon technology for HEP

The lecture will introduce the concept of disruptive technologies using the example of the Digital Photon Counter (DPC, dSiPM) developed at Philips since 2004. The major characteristics of disruptive technologies will be worked out and examples given. As the development of the technology at Philips was triggered by its potential application in medical imaging, in particular in Positron-Emission-Tomography (PET), the benefits of using DPC in PET will be explained. The concepts of analog and digital SiPM will be compared and the advantages of the early digitization concept will be highlighted in particular in view of industrial applications. One of the most important prerequisites for this is scalability, so special focus and attention will be given to this aspect by introducing the Philips digital systems concept. The question of how to bring such a technology to market and how to define the product will also be discussed. First application examples will be shown and a brief outlook on future developments will be provided.

Speaker: York Haemisch (P)

Slides Delft_spring_school_lecture_Haemisch_pdf.pdf

15:30 Coffee break

15:45 The transition form HEP to Medical Imaging

The lecture will introduce the concept of disruptive technologies using the example of the Digital Photon Counter (DPC, dSiPM) developed at Philips since 2004. The major characteristics of disruptive technologies will be worked out and examples given. As the development of the technology at Philips was triggered by its potential application in medical imaging, in particular in Positron-Emission-Tomography (PET), the benefits of using DPC in PET will be explained. The concepts of analog and digital SiPM will be compared and the advantages of the early digitization concept will be highlighted in particular in view of industrial applications. One of the most important prerequisites for this is scalability, so special focus and attention will be given to this aspect by introducing the Philips digital systems concept. The question of how to bring such a technology to market and how to define the product will also be discussed. First application examples will be shown and a brief outlook on future developments will be provided.

Speaker: York Haemisch (P)

Slides Delft_spring_school_lecture_Haemisch_pdf.pdf

18:30 → 22:30 Delft channels boat trip and dinner

Thursday 23 May

09:00 → 12:15 SiPM Materials and Modelling (3rd Module)

Conveners: Mr Angelo Gulinatti (POLIMI), Prof. Lis K. Nanver (TU Delft)

09:00 New materials and fabrication for APDs, SPADs and SiPMs

The “new materials” that will be discussed in this lecture are depositions of pure boron and pure gallium, given the names PureB, PureGa and PureGaB. It may seem out of place to speak of boron as a novice in silicon technology since it has always been, and still is, the most commonly used p-dopant. Nevertheless, it is now recognized that boron, deposited by chemical-vapor deposition (CVD) in its pure form, exhibits both electrical and processing properties that make it a useful and unique supplement to the long list of materials already playing a role in silicon device integration. In particular, extremely shallow junctions for p+n diodes can be fabricated with PureB as the anode. To underline the unique behaviour of these junctions the lecture begins with an introduction to the device physics of truly ultrashallow junctions followed by the basics of the junction formation process. At the moment, the most prominent application is for PureB Si photodiodes that are sensitive right up to the front-entrance window that can be made with PureB down to 2 nanometers thin. Therefore these photodiodes are particularly interesting for the detection of very low-penetration-depth radiation such as DUV/VUV/EUV light and charged particles such as low-energy electrons. Their performance has been shown to surpass that of other existing technologies on points such as internal/external quantum efficiency, dark current, degradation of responsivity. At the same time they readily lend themselves to detector integration schemes that allow low parasitic resistance and capacitance as well as the on-chip combination with other electronic elements. Focus will be placed on the special requirements for integration as SPADs. Similarly outstanding results have been achieved with a pure-gallium deposition capped with a PureB deposition (PureGaB), also when applied to Ge devices. Infrared Ge-on-Si photodiodes with photon-counting capabilities at room temperature have been demonstrated in this PureGaB technology. All in all, the PureB/PureGaB diodes perform better than other ultrashallow junction technologies. The reasons for this will be highlighted while we walk through the applications to detectors/imagers that now span the whole electromagnetic spectrum going from terahertz – IR – visual – DUV/VUV/EUV – X-rays through to electron detectors.

Speaker: Prof. Lis K. Nanver (TU Delft)

Slides Nanver_PicSEC_2013.pdf

10:30 Coffee break

10:45 Modeling of SPADs and SiPMs

Single Photon Avalanche Diodes (SPADs) and Silicon PhotoMultipliers (SiPMs) have since long ceased to be a scientific curiosity and have become a valuable tool for many applications both in research and industrial fields. Such a transition, on one hand, has been favored by the remarkable improvement in SPADs and SiPMs performance obtained in the last years; on the other hand, the widespread use of these devices in many cutting edge applications push the research community for a continuous improvement in their performance and complexity. A detector development based on a “trial and error” approach is in most cases ineffective and impractical both because of the remarkable time and costs needed for each iteration of the fabrication process and because of the difficulties in identifying the suitable modifications to the device structure. Actually, an effective detector’s development requires a detailed understanding of the physical phenomena that regulate its behavior and the availability of physical models that allows the device designer to evaluate the effect of the main parameters (doping and electric field profiles, defects type and distribution, etc.) on detector performance. In this lecture we will discuss the physical modeling of SPADs and SiPMs as a whole. In fact, although the latters are significant different from the formers in terms of applications, architectures and sometimes even in terms of the corresponding electronics circuits, the core of the detectors is the same and therefore a unified treatment is possible. In this optics we will present the main figures of merit common to both the detectors (Breakdown Voltage, Photon Detection Efficiency, Dark Count Rate, Afterpulsing Probability, Optical CrossTalk and Timing Jitter) and for each of them we will analyze the physical phenomena that influence its behavior. Then we will discuss the modeling of such phenomena highlighting both relevant parameters and possible limitations.

Speaker: Mr Angelo Gulinatti (POLIMI)

Slides PicoSEC_training2013_Gulinatti.pdf

12:30 → 14:00 Lunch break

14:00 → 17:30 SiPM Simulation and Design Lab (Training Session)

14:00 SiPM Simulation and Design Lab (Training Session)

Silicon photomultipliers (SiPMs) are an alternative to photomultiplier tubes because of their robustness to magnetic fields, compactness, and low bias voltage. To take advantage of the merits of SiPMs, the deep understandings of SiPM characteristics are required. In this workshop, we will model and simulate the SiPM using Cadence spectre simulator to investigate the SiPM characteristics as well as building the schematic, drawing layout and verifying the design in Cadence design environment. We will also simulate the SiPM behavior assuming that a gamma photon hits a crystal scintillator and produce various number of photons to be detected by small cells in the SiPM.

Speakers: Mr C. Veerappan, Esteban Venialgo (TU Delft), Shingo Mandai (TU Delft)

Friday 24 May

09:00 → 12:15 SiPM Architecture and time-processing (4th Module)

Conveners: Dr Francesco Regazzoni (TU Delft), Dr Lucio Pancheri (FBK)

09:00 CMOS SiPM design and signal compression

This lecture reviews several design solutions adopted in both analog and digital SiPMs to improve their characteristics, with particular focus on timing and spatial resolution. A brief overview of SiPM technology developed at Fondazione Bruno Kessler (FBK) is given. The problem of optimum timing pickoff is analyzed and the solutions adopted at FBK are described. The main limitations to optimum timing resolution are addressed, indicating the directions of future SiPM design improvements. Position encoding approach, which was proposed to reduce the complexity of high-spatial resolution detector modules for preclinical PET systems, is also reviewed. The architecture of a CMOS digital SiPM developed inside the EU project SPADnet is then presented. In this design, signal compression is used to reduce the complexity and area occupation of focal plane processing electronics. Each pixel, composed by 720 SPADs, provides the total counts and timestamp of detected gamma events. SiPM total counts, used to discriminate gamma event detection, are sampled at up to 100MS/s by an adder tree overlaid on top of the pixel array. Experimental results validating the approach are presented and critically discussed.

Speaker: Dr Lucio Pancheri (FBK)

Slides Pancheri_PicoSec2013.pdf

10:30 Coffee break

10:45 Time-to-digital converters

Time-to-digital converters (TDC) are devices which allow, with a certain precision, to digitally represent the time occurred. Their usage spawn over a large number of applications where a precise time time stamp is required. Due to technological progresses, reconfigurable hardware devices (such as FPGAs) have become an attractive platform for implementing low cost and high performances TDCs. This module presents the design and the implementation of a TDC on FPGA. More in details, the module will cover the basics of TDCs and of hardware design, with particular emphasis on design for re-configurable devices, their hardware description language and the tools used for programming them.

Speaker: Dr Francesco Regazzoni (TU Delft)

12:30 → 14:00 Lunch break

14:00 → 17:00 TDC Design and Test Lab (Training Session)

14:00 TDC Design and Test Lab (Training Session)

The previous module will be completed by an hands-on laboratory session in which the students will design a small TDC and implement it on the provided FPGA platform. The Xilinx design environment will be utilized for this task, and the resulting configuration file will be tested on an FPGA board.

Speakers: Mr C. Veerappan, Dr Francesco Regazzoni