Fig 2 - uploaded by Dennis R. Schaart
Content may be subject to copyright.
Several cascaded amplifiers (A 1 , A 2 , A 3 ) increase the overall gain at the expense of increased curvature in the initial part of the pulse response.
Source publication
In a Silicon Photomultiplier (SiPM) several hundreds or thousands of Geiger Mode Avalanche Photodiodes (GM APDs) are connected in parallel so as to combine the photon counting capabilities of these so-called microcells into a proportional light sensor. SiPM-based scintillation detectors can exhibit inherent nonlinearity, which may result in a degra...
Context in source publication
Context 1
... photomultipliers (SiPMs) consist of a large number of single-photon avalanche diodes (SPADs) or Geiger-mode avalanche photodiodes (GM-APDs), here referred to as microcells, all connected in parallel. SiPMs offer a high gain and a fast response to light, which makes them interesting for fast timing applications such as scintillation detectors for time-of-flight positron emission tomography (TOF-PET). However, using SiPMs to their full potential is not trivial. For example, conventional amplifier techniques suitable for the readout of photomultiplier tubes (PMTs), photodiodes, or avalanche photodiodes (APDs), are not one-to-one applicable to SiPMs due to the very different electronic characteristics of these types of light sensor as described in the next section. To achieve the best possible timing performance with SiPMs, dedicated preamplifiers are needed that do not deteriorate the rise time and signal-to-noise ratio of the SiPM signal. Here we present a SiPM preamplifier design concept based on low noise, high speed, discrete transistors. Electronic modeling of SiPMs has been pioneered by Corsi et al. [1]. Seifert et al. [2] recently proposed an extended model and investigated two sensors from Hamamatsu in detail, viz. the 3 mm  3 mm MPPC S10362-33-25C and the 1 mm  1 mm MPPC S10362-11-25U. A main characteristic of these devices is their high terminal capacitance ( 4 300 pF for the 3 mm  3 mm device) consisting of the sum of all cell capacitances and the internal interconnect capacitance. Since this is much larger than the 3 pF–10 pF anode capacitance of a typical PMT [3], suboptimal performance may result if PMT preamplifiers are attached to SiPMs. Especially when high timing resolution is important the high capacitance of SiPMs needs to be accounted for. Another characteristic of SiPMs is the dependence of the output impedance on the number of microcells being fired, in contrast with the almost ideal current source behavior of PMTs [3]. When several photons hit a SiPM at the same time, part of the microcells may discharge, while other cells remain inactive. Each cell has its own quench resistor, diode resistance, and adherent parasitic capacitances. The total impedance of a cell is very different depending on its state (inactive or discharging). Since all cells are connected in parallel, it follows that the SiPM output impedance varies with the amount of light incident on the sensor [2]. If a SiPM is connected to an amplifier with finite input impedance, its varying output impedance may give rise to non- linear behavior. Studies have shown that the linearity of the overall response improved with decreasing input impedance of the amplifier [2]. Together with the already mentioned effect of the high SiPM capacitance on the time response, one can conclude that the lower the preamplifier input impedance, the faster the temporal response and the better the linearity of the signal will be. A common approach to amplify SiPM signals is to use a voltage amplifier in combination with a current-converting shunt resistor R , as shown in Fig. 1. However, if R is kept to a value of, for example, 10 O or smaller because of the reasons given in Section 2.1, the generated signal voltage across R will be very small, necessitating a high-gain amplifier. The small signal voltage furthermore requires this amplifier to exhibit very low noise in order not to degrade the signal - to - noise ratio. Due to gain-bandwidth limitations, high-speed amplifiers typically have a relatively low gain of about 20 dB or less, requiring several of these amplifiers to be placed in cascade as in Fig. 2 to obtain a decently measurable signal. The effect of connecting multiple amplifiers in cascade, even if they all have the same bandwidth, is an increased curvature of the initial part of the pulse response [4]. In other words, the signal slope at the onset of the amplified pulse is decreased, which may result in significant worsening of the timing resolution, especially since the best timing with SiPM-based scintillation detectors is typically achieved using very low threshold values for time pick-off [5–7]. To illustrate this, the normalized simulated time responses of 4 cascaded amplifiers, all having an arbitrarily chosen bandwidth of 1 GHz, are shown in Fig. 3. Trace 0 is the initial input step voltage, while traces 1 to 4 are the output signals of amplifiers 1 to 4, respectively. The signal slope at 5% of the maximum amplitude is about 10 times lower for the 4th amplifier output than for the 1st amplifier output. The simulated bandwidth, however, is only about 3 times lower, which means that for amplifier systems with multiple stages in cascade, considering bandwidth only can give a false impression of speed. Thus, the transient response should also be examined. In principle, a transimpedance amplifier might be expected to provide a suitable solution for both the high capacitance of the detector and the required low input impedance. Nowadays there are numerous high speed operational amplifiers with high bandwidth ( 4 1 GHz), using which a transimpedance amplifier can easily be made. An example of such a circuit is shown in Fig. 4. Since the current-converting resistor R f can be in the order of 100 O –500 O , the transimpedance gain of this amplifier can be high compared to the shunt resistor method. Most suitable [8] for high speed applications is the so-called current-feedback operational amplifier (CFA), which provides a true current input at the inverting terminal. The input impedance R int of the inverting input is inherently low, in the order of 8 O –30 O for commercially available devices. Using feedback, the effective input impedance R eff of the entire circuit is reduced to much lower values following the equation: Z 0 b þ 1 where Z 0 is the open-loop transimpedance gain and b the ratio of R int and feedback resistor R f . Z 0 can be 10 6 at low frequencies. Since the optimum value for the feedback resistor is in the order of 500 k O –1 k O for a CFA, b typically has a value of about 0.01 to 0.02, giving a Z 0 b in the order of 10 4 , thus resulting in a very low value of R eff . Initial tests using an AD8000 CFA from Analog Devices and a R f of 470 O have shown that the rise times obtained with such a circuit in combination with 3 mm  3 mm SiPMs illuminated with a picosecond laser are longer than with a voltage amplifier with a small 5 O shunt resistor. The pulse shape also showed ringing, which can be explained by the reduced open-loop gain Z 0 at high frequencies. For example the AD8000 CFA has a Z 0 of $ 200 at 1 GHz, giving a Z 0 b of only about 2–3. The result is a non-zero impedance at the virtual ground input (indicated as ‘‘0’’ in Fig. 4) for these frequencies. Since the high sensor capacitance is also connected to the virtual ground the result is an increased rise time. Furthermore, the sensor capacitance together with the input resistance forms a reactive impedance, resulting in peaking of the gain at high frequencies and corresponding ringing of the output signal. For smaller sensors with a smaller capacitance (e.g. the MPPC S10362-11-25U), this amplifier has been shown to give good results [9]. A common-base (also called grounded-base) amplifier without feedback is a good candidate to overcome the problems described in the preceding sections. Fig. 5 shows the basic scheme of a common-base amplifier. The emitter of the transistor forms the input of the amplifier. The base of the transistor has a DC biasing voltage but is decoupled to ground by means of a capacitor, in order to decrease the input impedance for signal frequencies. By design, the input impedance is very small without using any feedback. The absence of feedback insures low input impedance even at very high frequencies, resulting in a negligible effect of the detector capacitance on the overall time response and minimizing non-linearity at the same time. The input signal current i flows from the emitter to the collector of the npn transistor and generates a voltage difference over the collector resistor R . Since the collector capacitance is very small, this resistor can be relatively large, enabling high transimpedance gain without compromising speed. As compared to the shunt resistor method described in Section 2.2, the common base amplifier converts the signal current to a signal voltage that is at least an order of magnitude higher ( R 1⁄4 100 O ). Furthermore, in the case of the shunt resistor method, the required fast time response dictates the choice of the shunt resistor since the response is a ...
Similar publications
The scintillator bismuth germanate (BGO) has attractive properties for positron emission tomography (PET) systems such as high stopping power, high photo-fraction, and relatively low cost. However, its moderate scintillation light yield and slow rise and decay time compared to lutetium (yttrium) oxyorthosilicate (L(Y)SO) results in a degradation of...
In this work we describe the results of performance tests and measures of SiPM of several sizes (1 x 1, 3 x 3, 5 x 5) delivered from MEPHI. The SiPMs have been studied both in steady and pulsed stimuli. Aging and temperature behavior are also discussed. Another test has been performed in order to obtain an electrical model of the SiPM to be used in...
In this paper, we propose a fully time-based multiplexing and readout method that uses the principle of the global positioning system. Time-based multiplexing allows simplifying the multiplexing circuits where the only innate traces that connect the signal pins of the silicon photomultiplier (SiPM) channels to the readout channels are used as the m...
Fully 3-D time-of-flight (TOF) PET scanners offer the potential of previously unachievable image quality in clinical PET imaging. Timing calibration is critical to achieve the best possible timing resolution, to maximize the true counts and minimize the random counts. In this paper, we consider a data-driven timing calibration based on TOF data con...
The design of combined positron emission tomography/magnetic resonance (PET/MR) systems presents a number of challenges to engineers, as it forces the PET system to acquire data in a space constrained environment that is sensitive to electro-magnetic interference and contains high static, radio frequency and gradient fields. In this work we validat...
Citations
... The same electronic collimation described in Section D was used; with the PMT placed on a linear stage to adjust the irradiation depth along the 20 mm long array. The 22 Na source was placed centrally between both detectors with a source to detector distance of 6 cm. The 22 Na source size was estimated to be ~0.5 mm FWHM by microCT. ...
... If a charge sensitive preamplifier (CR-113) is used the rise time increases from 100 ns to over 500 ns. We are exploring other preamplifier concepts, which have been investigated by other groups [21], [22], to improve the PS-SSPM's cathode channel timing response. Though the timing properties of PS-SSPMs are only slightly worse than PS-APDs and should be acceptable for most small animal imaging applications. ...
We present an analysis of the signal properties of a position-sensitive solid-state photomultiplier (PS-SSPM) that has an integrated resistive network for position sensing. Attractive features of PS-SSPMs are their large area and ability to resolve small scintillator crystals. However, the large area leads to a high detector capacitance, and in order to achieve high spatial resolution a large network resistor value is required. These inevitably create a low-pass filter that drastically slows what would be a fast micro-cell discharge pulse. Significant changes in the signal shape of the PS-SSPM cathode output as a function of position are observed, which result in a position-dependent time delay when using traditional time pick-off methods such as leading edge discrimination and constant fraction discrimination. The timing resolution and time delay, as a function of position, were characterized for two different PS-SSPM designs, a continuous 10 mm ×10 mm PS-SSPM and a tiled 2 ×2 array of 5 mm ×5 mm PS-SSPMs. After time delay correction, the block timing resolution, measured with a 6 ×6 array of 1.3 ×1.3 ×20 mm3 LSO crystals, was 8.6 ns and 8.5 ns, with the 10 mm PS-SSPM and 5 mm PS-SSPM respectively. The effect of crystal size on timing resolution was also studied, and contrary to expectation, a small improvement was measured when reducing the crystal size from 1.3 mm to 0.5 mm. Digital timing methods were studied and showed great promise for allowing accurate timing by implementation of a leading edge time pick-off. Position-dependent changes in signal shape on the anode side also are present, which complicates peak height data acquisition methods used for positioning. We studied the effect of trigger position on signal amplitude, flood histogram quality, and depth-of-interaction resolution in a dual-ended readout detector configuration. We conclude that detector timing and positioning can be significantly - mproved by implementation of digital timing methods and by accounting for changes in the shape of the signals from PS-SSPMs.
... This ability to decode a large number of crystals with only 16 SiPM pixel elements is a strength of this multiplexing approach. Our simple resistor multiplexing readout approach for this SiPM detector results in timing resolution that is considerably worse than that achieved by buffering the output of each SiPM pixel using methods such as those proposed by Seifert [19], [20]. However, the simple resistor multiplexor design only requires four op-amps, thus minimizing power consumption and reducing the number of active components on the detector readout circuit. ...
A silicon photomultiplier (SiPM)-based positron emission tomography (PET) detector was developed using a resistor network charge division multiplexing circuit for detector readout. The detector consists of a lutetium-yttrium oxy-orthosilicate (LYSO) scintillation crystal array, an SiPM array detector (SPMArray 4, SensL Inc., Cork, Ireland) and the resistor multiplexing network implemented in a through-hole package to facilitate changing of resistor values. For purposes of optimizing the readout circuit, the LYSO array used was a 4 4 crystal array with crystal size mm on a pitch of 3.37 mm, matched to the SiPM pixel size. Flood image, energy resolution, photopeak amplitude, timing resolution, and signal time-pickoff measurements were performed using standard NIM electronics. The resistor network values were optimized through an iterative process. The performance of the detector was evaluated over a range of temperatures from 23°C to 60°C by heating the detector. The ability of the detector to resolve crystals smaller than the SiPM pixel pitch was evaluated using a dual-layer LYSO array with crystals of 1.67-mm pitch. The optimal resistor network values were found to be 100 Ω along the rows connecting the SiPM pixels and 56 Ω for the columns. For these resistor value settings, the average energy resolution for the central four crystals in the array at 23.5°C was 13.3% ± 0.3% and degraded to 16.3% ± 0.3% at 60°C. The photopeak amplitude decreased by 2%/°C, and the timing resolution degraded from 3.43 ± 0.22 ns to 4.64 ± 0.25 ns for a 350-750-keV energy window over this temperature range. The signal time-pick-off point shifted earlier by 2.7 ns as the temperature increased, an effect likely due to changes in the signal shape with temperature. The detector was able to resolve all 113 crystals in the dual-layer LYSO array. These results demonstrate that the resistor multiplexing reado- t circuit functions well for reading out SiPM array based detectors, which use scintillator crystal arrays much smaller than the SiPM pixel pitch. The reduced number of output signals achieved through this signal multiplexing greatly reduces the number of signal cables required. In addition, the ability of this detector to function over a wide range of temperatures offers significant flexibility in defining the system operating temperature set point.
... A variation of the direct voltage readout scheme is the indirect scheme illustrated in Figure 4.4(b). This scheme is also quite straight forword and widely used [94]. The input impedance is further decreased to R 0 /A by the amplifier feedback scheme. ...
Silicon Photomultipliers (SiPMs) are novel kind of solid state photon detectors with ex-
tremely high photon detection resolution. They are composed of hundreds or thousands
of avalanche photon diode pixels connected in parallel. These avalanche photon diodes are
operated in Geiger Mode. SiPMs have the same magnitude of multiplication gain compared
to the conventional photomultipliers (PMTs). Moreover, they have a lot of advantages
such as compactness, relatively low bias voltage and magnetic field immunity etc. Special
readout electronics are required to preserve the high performance of the detector. KLauS
and STiC are two CMOS ASIC chips designed in particular for SiPMs. KLauS is used for
SiPM charge readout applications. Since SiPMs have a much larger detector capacitance
compared to other solid state photon detectors such as PIN diodes and APDs, a few special
techniques are used inside the chip to make sure a descent signal to noise ratio for pixel
charge signal can be obtained. STiC is a chip dedicated to SiPM time-of-flight applications.
High bandwidth and low jitter design schemes are mandatory for such applications where
time jitter less than tens of picosends is required. Design schemes and error analysis as well
as measurement results are presented in the thesis.
... In this first proof-of-principle setup, the SiPMs are applied as gamma photon counters. The signals of the 12 SiPMs are first preamplified by a 16-channel read out board (the design of which is described and characterized in Seifert et al (2008)) and subsequently summed. The summed signal is shaped and amplified by an Ortec 572 spectroscopy amplifier. ...
High spatial resolution γ-imaging can be achieved with scintillator readout by low-noise, fast, electron-multiplying charge-coupled devices (EMCCDs). Previously we have shown that false-positive events due to EMCCD noise can be rejected by using the sum signal from silicon photomultipliers (SiPMs) mounted on the sides of the scintillator. Here we launch a next generation hybrid CCD-SiPM camera that utilizes the individual SiPM signals and maximum likelihood estimation (MLE) pre-localization of events to discriminate between true and false events in CCD frames. In addition, SiPM signals are utilized for improved energy discrimination. The performance of this hybrid detector was tested for a continuous CsI:Tl crystal at 140 keV. With a pre-localization accuracy of 1.06 mm (full-width-at-half-maximum) attained with MLE the signal-to-background ratio (SBR) was improved by a factor of 5.9, 4.0 or 2.2 compared to the EMCCD-only readout, at the cost of rejecting, respectively, 47%, 9% or 4% of the events. Combining the pre-localization and SiPM energy estimation improved the energy resolution from 50% to (19 ± 3)% while maintaining the spatial resolution at 180 µm.
... This would slightly change the single photon signal, but this effect is mostly seen with higher photon flux, and can be reduced by having a low impedance input amplifier. Using a transimpedance pre-amplifier this can be kept close to zero [11]. A histogram was generated (Fig. 3) where the arrival time places each photon in one or two bins. ...
Monte Carlo simulations of a time-of-flight Positron Emission Tomography front end readout system (TOF-PET) are performed to investigate obtainable timing resolution at different ADC sample times using LYSO scintillators. Results from simulations with different combinations of ADC sample times and analog filter time constants are presented. To determine the coincidence resolving time (CRT), a trigger system based on a digital constant fraction zero crossing discriminator (CFD) is investigated. Such a trigger system would be suitable for running real time in a field-programmable gate array (FPGA). It is seen that below a certain sample time, the time resolution is limited by the scintillator and MPPC combination, and not the readout system.
... The SiPM array was operated at the manufacturer-specified bias voltage of 29.3 V, exceeding the nominal breakdown voltage by 2.0 V and corresponding to a gain of . The SiPM signals were preamplified using a 16-channel readout board described by Seifert et al. [20]. The preamplified SiPM pulses were shaped and their pulse heights digitized using the multichannel data acquisition system described by Maas et al. [6]. ...
... Previous works reported in the literature have demonstrated that this scheme for DOI PET detectors produce good results in terms of spatial and energy resolution [5]; however, the replacement of the PS-PMT with a SiPM matrix introduces some uncertainties that required further study even though SiPM have demonstrated good imaging capabilities when used with pixelated crystal scintillators [6], and good timing performance [7][8][9]. ...
PET detector modules that implement depth of interaction information reduce uncertainties about the actual line-of-response after positron annihilation. This, in turn, has an effect in the spatial resolution and uniformity across the field of view of the reconstructed image. It also has an effect on the system sensitivity since it enables the use of thicker crystal with respect to a non-DOI system without deteriorating significantly the spatial resolution. In this work we evaluate a DOI detector design based on a scintillator phoswich of two types of pixelated crystals with different time decay constants. A matrix of silicon phomultipliers (SiPM) was coupled to this crystal array without any light-guide, and its outputs were connected to an anger logic type of charge divider which outputs were buffered using a transimpedance amplifier. These position signals were digitized using charge-to-digital converters, and the time decay constant of the crystal of interaction was measured using a delayed charge method. Previous works reported in the literature have demonstrated that this scheme for DOI PET detectors produce very good results in terms of spatial and energy resolution; however, the replacement of PMT with a SiPM matrix introduces some uncertainties that required further study. SiPM have demonstrated their good imaging capabilities when used with pixelated crystal scintillators, and their good timing performance allow us to predict that the DOI resolution was not going to be a problem with this technology. In this work we present quantitative results that assess the goodness of these detectors for its use on small-animal PET scanners.
... The SiPM array was operated at the manufacturer-specified bias voltage of 29.3 V, exceeding the nominal breakdown voltage by 2.0 V and corresponding to a gain of . The SiPM signals were preamplified using a 16-channel readout board described by Seifert et al. [20]. The preamplified SiPM pulses were shaped and their pulse heights digitized using the multichannel data acquisition system described by Maas et al. [6]. ...
Several improvements of the k-nearest neighbor (k-NN) method for the determination of the entry point (x, y) of a gamma photon in a monolithic scintillator PET detector have been investigated with the aim to obtain better spatial resolution and/or to enable faster detector calibration by reducing the amount of required reference data and by allowing for calibrating with a line source. These methods were tested on a dataset measured with a SiPM-array-based monolithic LYSO detector. It appears that ∼10% to ∼25% better spatial resolution can be obtained compared to the standard approach. Moreover, some of the improved methods using two orders of magnitude less reference data, yield essentially the same spatial resolution as the standard method, which reduces the time needed for calibration as well as entry point computation. Finally, line source calibration is shown to be possible with some of the methods, yielding better results than the standard method and allowing much faster and easier collection of the reference data.
... To reduce an influence of the MPPC 33-050C capacitance of 320 pF and preserve parameters of an input scintillation signal a low input resistance current preamplifier was build following [10]. The preamplifier posses two outputs: slowaccepted by spectroscopy amplifiers and fast -reflecting the direct SiPM output and used for timing. ...
Silicon photomultipliers (SiPM) are relatively new photodetectors studied presently by many groups in the world. Experiments requiring good time resolution are one of the areas of interest for application of SiPMs. Experiments with high photon statistics and fast picosecond lasers showed that SiPMs time resolution can be extremely good, much below 100 ps. Unfortunately, in case of a full detector with a scintillation crystal like LSO, a high capacitance of this type of devices deteriorates the rise time of the output pulse. In consequence, the timing resolution measured with the slow pulses from SiPM could be worse than results obtained with photomultipliers. In this work a detailed studies of the time resolution with scintillation detectors based on a 3×3 mm2 Multi Pixel Photon Counters (MPPCs) from Hamamatsu and LSO or LFS-3 crystals are presented. Most of the measurements were done using a detector with a pixel size of 50 μm (S10362–33–050C). The results of coincidence experiments with an 22Na gamma source are analyzed in terms of number of photoelectrons, time jitter, excess noise factor and output pulse characteristics. The aim of the work is to present timing capabilities of scintillation detectors based on MPPC in comparison with detectors based on classic timing photomultipliers.
... A 1 mm x 1 mm SiPM is preferable in this respect, due to 9 times lower capacitance. In the tested MPPC, it could be largely improved by a drastic reduction of the load resistor, below 10 ohms [17]. Both tested MPPCs give comparable time resolution, which follows measured numbers of fired pixels. ...
The performance of multi pixel photon counters (MPPC) of 3 mm × 3 mm size, with 14400 and 3600 pixels, were studied by means of the signal from a laser light pulser and using the 3 mm × 3 mm × 20 mm LSO pixel scintillator. Special attention was paid to measure number of fired pixels, generated by the light of pulser and that of the LSO crystal, using a direct method of a comparison of the light peak position in the pulse height spectrum with that of the single photoelectron. The tests of the LSO crystal showed 1550 ± 80 fired pixels per MeV in the MPPC with 14400 pixels assuring a good linearity of the response up to about 1 MeV energy of gamma rays absorbed in the LSO crystal. Energy resolution of 14.8% for 662 keV gamma rays from <sup>137</sup>Cs source and a time resolution of about 850 ps for 511 keV annihilation quanta were limited by a rather low number of the fired pixels compared to the number of photoelectrons in photomultipliers.
