Search

The upgrades of the CMS and ATLAS experiments for the high luminosity phase of the Large Hadron Collider will employ precision timing detectors based on Low Gain Avalanche Detectors (LGADs). We present a suite of results combining measurements from the Fermilab Test Beam Facility, a beta source telescope, and a probe station, allowing full characterization of the HPK type 3.1 production of LGAD prototypes developed for these detectors. We demonstrate that the LGAD response to high energy test beam particles is accurately reproduced with a beta source. We further establish that probe station measurements of the gain implant accurately predict the particle response and operating parameters of each sensor, and conclude that the uniformity of the gain implant in this production is sufficient to produce full-sized sensors for the ATLAS and CMS timing detectors., 20 page, 20 figures

In this paper the results of charge and timing resolution characterization realized at Fermi National Accelerator Laboratory (Fermilab) on 3.5 × 3.5 mm2 Silicon PhotoMultipliers fabricated at STMicroelectronics Catania R&D clean room facilities are presented. The device consists of 4900 microcells and has a geometrical fill factor of 36%. Timing measurements were realized at different wavelengths by varying the overvoltage and the temperature applied to the photodetector. The results shown in this manuscript demonstrate that the device, in spite of its large area, exhibits relevant features in terms of low dark current density, fast timing and very good single photoelectron resolution. All these characteristics can be considered very appealing in view of the utilization of this technology in applications requiring detectors with high timing and energy resolution performances.

WE have previously demonstrated a stripline readout for TOFPET detectors that can achieve significant reduction in the number of readout channels while maintaining good detection performance [1-3]. The principle of this time-based multiplexing readout is depicted in Figs. 1(a) and 1(b). As shown, a stripline receives the outputs of several photodetectors (PD). When a PD produces a signal at time t e , two identical pulses are created at the position of the PD but travel in opposite directions along the stripline, arriving the ends at time, say, t 1 and t 2 . Evidently, t 1 = t e + l 1 /v and t 2 = t e + l 2 /v = t 2 + (L − l 1 )/v, where l 1 and l 2 are the path lengths from the PD to the ends, L is the total length of the stripline, and v is the pulse propagation speed on the stripline. It follows that dt = t 2 − t 1 = L/v − 2(l 1 /v) and t = (t 1 + t 2 )/2 = t e + L/2v. As L and v are fixed numbers, the differential time (DT), dt, thus encodes the PD's position on the stripline and the average time, t, the event time. The pulses are highly correlated; therefore, the size of the pulse, which is related to the event energy in PET, can be measured off either one of the pulses by using the peak or integrated amplitude. Now, the path-length between neighboring PDs on the strpiline can be small, for example, 3–6 mm. The propagation speed of the pulse, on the other hand, is approximately half the speed of light. Thus, the difference in the DT reading of two adjacent PDs can be on the order of 20–40 picoseconds (ps) and the DT measurement needs to be highly precise. To achieve this, we proposed to digitally sample the pulses at 2–5 GHz by employing an inexpensive, low power-comsumption waveform-sampling technology such as the DRS4 from the Paul Sherrer Institut (PSI) (see Fig. 1(c)) and apply digital-signal processing to extract accurate time information. Subject to the count-rate requirement, multiple striplines can be simply connected to increase the number of SiPMs per line and therefore the readout is scalable. Also, with the stripline readout the data-acquisition (DAQ) electronics can be detached from the detector (the use of up to 10-meter long cables to connect the DAQ electronics and detector have been successfully tested); this feature can be useful for developing systems that are subject to tight space constraints.

We have developed a filtering method to reduce RF interference in PET data for simultaneously acquired PET-MR. Data was acquired from a SiPM-based PET detector coupled to LYSO scintillators using a transmission-line readout scheme and high-speed waveform sampling. The detector was placed within a 9.4 T small animal MR system with either copper or no shielding, and PET data from a low-activity 22Na source and LYSO background were acquired while MR images were also acquired. MR images showed no image quality degradation. 4% and 7% of PET waveforms demonstrated RF interference with and without shielding. After applying a notch filter, the energy spectrum of RF-affected events was comparable to RF-unaffected events due to improved measurement of the waveform maximum amplitude. Differential time measurements of events on the transmission-line were also improved. This study demonstrates the utility of waveform sampling readout which allows straightforward post-processing in PET data acquisition for simultaneous PET-MR.

We have investigated the strip-line based SiPM signal readout method for TOF PET application; output of multiple SiPMs are connected on a single transmission-line, and signal waveforms from SiPMs are acquired at both ends of the strip-line by using high speed waveform sampler. This approach enables to decode the position of interaction along the transmission-line by using arrival time difference, and thus could reduce the number of readout electronics channels efficiently while keeping the fast time response of SiPMs. The scalability is one of the features of the strip-line readout method; the length of strip-line and the number of SiPMs connected on the strip can be easily extendable without increasing electronics channels. Previously we have built a prototype board with four strip-lines each with eight SiPMs (5.2mm pitch) installed on it, and demonstrated that the strip-line readout method is applicable for TOF PET. To check the scalability of the readout method, we modified the prototype board to connect two separate strip-lines into one single strip-line. From the laser test bench, the electronic time resolution along the strip is measured to be ∼36ps fwhm (equivalent to 1.7mm fwhm in position) and found that the time resolution is not degraded by doubling SiPMs on a strip.

We have developed a new time calibration method for DRS4 waveform sampler. The method is based on the correlation between differential amplitude and time interval of adjacent switched-capacitor cells responding to a simple saw-tooth shape pulses. In the experiment, saw-tooth pulse with 40ns period from a Tektronix AWG7104 is fed to DRS4 evaluation board for calibrating all 1024 cells of the board. The electronic time resolution of the DRS4 board with the new time calibration is measured to be ∼5.6ps FWHM using two identical simulataneous Gaussian pulses. The time resolution dependencies on the delay and the sampling speed are measured also. The new method is applicable for other switched-capacitor array technology-based waveform samplers for precise time calibration.

We are developing a human-brain imaging TOF PET detector with high sensitivity and lower radioactive dose to patients. The proposed detector design consists of 36 detector modules, and each module has 4×8 LYSO scintillators coupled to SiPMs individually. For signal readout for detector modules, the transmission-line readout with high speed waveform sampling is adopted to fully exploit the fast time response of SiPM with reduced electronics channels. A detector simulation has been developed using Geant4 package to check the performances of the proposed detector. Previously we demonstrated that our signal readout approach is feasible with SiPM. Based on the experience with the proto-type readout board, second generation readout board has been developed to implement the transmission-line signal readout for the proposed detector. In the paper, we present the initial results obtained from the simulation studies and experimental tests using the newly developed strip-line board.

We are investigating a new way of SiPM signal readout for Time-of-flight Positron Emission Tomography detector. Our new method adopts a transmission-line connected to multiple SiPMs in a row and high speed waveform sampling technology. The event information are retrieved from the digitized waveform measured at the two ends of the transmission-line; the interaction position along the strip direction is inferred from the arrival time difference at the two ends. With this approach, the number of readout channels can be efficiently reduced for larger detection area coverage by exploiting the fast time characteristic of SiPM while preserving the advantages of its compact size. We have built prototype transmission-line boards which holds up to eight SiPMs (3.5x3.5 mm2 and 5.0 mm pitch) on the strip. The waveforms of signals from the strip were recorded using DRS4 evaluation board at 5 GHz sampling rate. From the tests using pulsed laser and light emitting diode, the position along the strip could be determined in ~1 mm FWHM from the measured time difference on the strip. The responses to 511 keV photon were measured also in a coincidence setup; two LYSO scintillators (3×3×10 mm3) were coupled to the SiPMs on each transmission-line board, and a 22Na source was placed in the middle of the two LYSO scintillators with the distance of 12 cm. Preliminary results show that the position resolution of ~3 mm along the strip-line is achieved with LYSO+SiPM signal in addition to ~15% energy resolution at 511 keV and ~560 ps FWHM of coincidence time resolution. The results indicate that the investigated approach is a promising way for SiPM signal readout for TOF PET.

High speed wave-form sampling is becoming a more reliable method for TOF PET detector signal readout; event information including precise time are extracted from the digitized waveform in an unified signal processing chain. In this readout approach, the processing of larger data due to higher sampling rate would be the challenging issue. We conducted a study on the optimal waveform sampling speed using DRS4 evaluation board. The experimental setup consisted of two Hamamastu R9800 PMTs coupled with LSO crystals (6.3×6.3×25mm3). 22Na was used for positron source, and the two PMTs outputs of coincidence event were digitized varying DRS4 sampling rate in 0.7 - 5 giga-samples per second (GS/s). Initial results show that the coincidence time resolution of ~250 ps was measured in 2 - 5 GS/s, and becomes larger at the lower sampling rate.

Silicon Photomultipliers (SiPMs) are fabricated in two different configurations: P-on-N and N-on-P junctions. More particularly P-on-N SiPMs are more suitable for application in Positron Emission Tomography (PET), due to their higher sensitivity in blue wavelength range where common PET scintillators have their emission spectrum. In this paper we will report on the electro-optical performances of the first P-on-N SiPMs manufactured at STMicroelectronics-Catania. The results obtained on these devices are compared with those measured on the standard N-on-P technology. We will show that the performances of the new devices are extremely promising in terms of higher photon detection efficiency in blue wavelength range and lower cross talk effects.

Silicon Photomultipliers (SiPMs) have known a fast development in recent years, due to their excellent single photon detection capability and very fast timing response. In this paper we present the results of the electro-optical characterization performed on the first STMicroelectronics P on N SiPMs prototypes properly designed for their possible application in Positron Emission Tomography (PET). We will show that the performances of the new devices are extremely promising in terms of high photon detection efficiency and fast timing response in blue wavelength range.

A new setup has been made at Fermilab for detector timing measurements at picosecond level. The core timing resolution of the amplifiers, discriminators and TAC/ADC combination is approximately 2 picosecond. We have made a study of a single photoelectron time resolution (SPTR) measured for signals coming from silicon photomultipliers (SiPm) made by different manufacturers. The obtained SPTR is of the order of 180 picosecond with SiPms illuminated by red (635 nm) PiLas laser light. IRST SiPms show better SPTR when illuminated by the blue laser light (405 nm). Most of the data were taken with 1 Volt of the overvoltage. The SiPms time resolution is inversely proportional to the square root of the number of photoelectrons. A time-of-flight (TOF) system with few tens of picosecond time resolution, based SiPms with quartz Cherenkov radiators looks practically achievable. A simple model is proposed to explain the difference in SPTR of the IRST SiPms when illuminating by the blue and red light. The explanation of the origin of the tail in the MPPC SiPm’s single photoelectron time spectra is presented. Finally, requirements for the SiPms temperature and bias voltage stability to maintain few picosecond time resolution are discussed.

The CMS Hadron Outer Calorimeter (HO) is undergoing an upgrade to replace the existing hybrid photodetectors (HPDs) with silicon photo-multipliers (SIPMs). The chosen device is the Hamamatsu 3 × 3 mm 2 50 μm pitch MPPC. The system has been developed to be a “drop-in” replacement of the HPDs. A complete control system of bias voltage generation, leakage current monitoring, temperature monitoring, and temperature control using solid state Peltier coolers has been developed and tested. 108 channels of the system have been installed into CMS and operated for more than 2 years. The complete system of about 2200 channels is in production and will be installed in the next LHC long shut-down scheduled for 2013.