This dissertation is aimed at investigating the effect of parasitic elements on the timing performance of low light detection systems based on Silicon Photomultipliers (SiPMs). SiPMs represent a well consolidated and cost-effective technology, featuring inherent fast timing, sensitivity and high dynamic range that prove beneficial for recent applications demanding excellent timing resolution. However, in many practical cases, accuracy in time measurements can be impaired by nonidealities associated with the interconnections used to couple the photodetector to the front-end electronics. After focusing on the single-photon response of the system and going through the slope of its leading edge as a function of the input impedance and the bandwidth of the preamplifier, the objective becomes to effectively convey the design specifications in terms of circuit variables, thus providing sufficient understanding of their impact on the performance of the whole SiPM-based detection system. The proposed approach consists of two parts. Using a realistic and accurate SPICE circuit model of the SiPM based on the Corsi et al. model, the first part introduces a mathematical study for both a complete and approximate analytical characterization of the most crucial parameters involved in the determination of the timing performance of the system, with special emphasis on the effects of parasitic interconnection inductance, often underestimated. The common practice to rely on a front-end with extremely low input impedance and very large bandwidth for accurate time measurements is critically discussed, in search of theoretical evidence of the fact that, when non-idealities are envisaged, a preferred front-end architecture exists and that an optimum range of values for circuit variables can be selected to achieve the best result in terms of timing accuracy. The second part proposes the complete design of a front-end implemented in a standard 130-nm CMOS process from TSMC, aimed at achieving state-of-art timing accuracy performance when coupled to large area SiPMs indeed intended for applications such as ToF-PET. A timing resolution of 33 ps rms for single-photon events and a dynamic range from 1 to 8000 photoelectrons have been achieved and validated with post-layout simulations, thanks to innovative circuit techniques, that have been described in detail. Finally, pending the prototype fabrication, the closing section of the work has been dedicated to carry out extensive measurements on a test preamplifier, realized on a printed circuit board with discrete components and coupled to a 3x3 mm2 SiPM commercially available from Hamamatsu Ltd. The preamplifier can be configured according to the architectures proposed in the theoretical analysis, with the purpose of validating the results of that analysis. The experimental results achieved are in good agreement with the analytical expressions obtained with the developed mathematical study, thus demonstrating the effectiveness of the proposed approach in making comparative analysis of different front-end architectures.
La qui presente tesi di dottorato si ripropone di studiare l’effetto degli elementi parassiti sulla prestazione di timing dei sistemi di rivelazione di luce a bassa intensità basati sui fotomoltiplicatori al silicio (SiPMs). I SiPMs rappresentano una tecnologia ben consolidata ed efficace rispetto ai costi di produzione, caratterizzati da un timing intrinseco veloce, sensibilità e ampio range dinamico, tutti fattori favorevoli per le recenti applicazioni che esigono una eccellente risoluzione temporale. Tuttavia, in molti casi pratici, l’accuratezza nelle misure temporali può essere deteriorata dalle non-idealità associate alle interconnessioni che si usano per accoppiare il fotosensore al front-end. Focalizzando l’attenzione sulla risposta del sistema al singolo fotone e analizzando la sua pendenza iniziale in funzione dell’impedenza d’ingresso e della banda del preamplificatore, l’obiettivo diventa quello di esprimere efficacemente le specifiche di progetto in termini di variabili circuitali, fornendo così un sufficiente livello di comprensione del loro impatto sulla prestazione dell’intero sistema di rilevazione basato su SiPM. L’approccio proposto consiste in due parti. Facendo uso di un modello SPICE del SiPM realistico ed accurato basato sul modello di Corsi et al., la prima parte introduce uno studio matematico, sia completo sia approssimato, per la caratterizzazione analitica dei parametri più importanti che incidono sulla prestazione di timing del sistema, ponendo l’accento sugli effetti spesso sottostimati dell’induttanza parassita di interconnessione. La pratica comune di fare affidamento su un front-end con impedenza d’ingresso estremamente bassa e banda molto larga per misure di tempo accurate viene criticamente discussa, alla ricerca di indizi teorici del fatto che, quando si tengono in considerazione le non-idealità, esiste un’architettura di front-end conveniente e che sia possibile assegnare dei valori ottimali alle variabili circuitali per ottenere il miglior risultato in termini di accuratezza temporale finale. La seconda parte propone il progetto completo di un front-end implementato in tecnologia standard 130-nm CMOS di TSMC, che mira al raggiungimento dello stato dell’arte per la prestazione di accuratezza temporale quando viene accoppiato ai SiPM di notevoli dimensioni, peraltro destinati ad applicazioni quali ToF-PET. Grazie all’uso di tecniche circuitali innovative, che sono state descritte in dettaglio nel corso della tesi, sono state ottenute e validate, mediante lunghe simulazioni circuitali post-layout, la specifica di risoluzione temporale pari a 33 ps rms per eventi a singolo fotone e la specifica di range dinamico esteso, da 1 a 8000 fotoelettroni. Per concludere, in attesa della fabbricazione del dispositivo, la sezione finale del lavoro è stata dedicata alla misurazione delle prestazioni di timing di un preamplificatore di test, realizzato con componenti discreti su circuito stampato e accoppiato a un SiPM commerciale da 3x3 mm2 di Hamamatsu Ltd. Il preamplificatore si può configurare sulla base delle architetture proposte nell’analisi teorica, con lo scopo di validare i risultati della suddetta analisi. I risultati sperimentali rilevati sono in buon accordo con le espressioni analitiche ottenute attraverso lo sviluppo dello studio matematico, dimostrando in questo modo l’efficacia dell’approccio proposto per condurre analisi comparative di diverse architetture di front-end.
A study of the effect of parasitic elements on the timing performance of SiPM readout electronics / Calo', Pietro Antonio Paolo. - ELETTRONICO. - (2020). [10.60576/poliba/iris/calo-pietro-antonio-paolo_phd2020]
A study of the effect of parasitic elements on the timing performance of SiPM readout electronics
Calo', Pietro Antonio Paolo
2020-01-01
Abstract
This dissertation is aimed at investigating the effect of parasitic elements on the timing performance of low light detection systems based on Silicon Photomultipliers (SiPMs). SiPMs represent a well consolidated and cost-effective technology, featuring inherent fast timing, sensitivity and high dynamic range that prove beneficial for recent applications demanding excellent timing resolution. However, in many practical cases, accuracy in time measurements can be impaired by nonidealities associated with the interconnections used to couple the photodetector to the front-end electronics. After focusing on the single-photon response of the system and going through the slope of its leading edge as a function of the input impedance and the bandwidth of the preamplifier, the objective becomes to effectively convey the design specifications in terms of circuit variables, thus providing sufficient understanding of their impact on the performance of the whole SiPM-based detection system. The proposed approach consists of two parts. Using a realistic and accurate SPICE circuit model of the SiPM based on the Corsi et al. model, the first part introduces a mathematical study for both a complete and approximate analytical characterization of the most crucial parameters involved in the determination of the timing performance of the system, with special emphasis on the effects of parasitic interconnection inductance, often underestimated. The common practice to rely on a front-end with extremely low input impedance and very large bandwidth for accurate time measurements is critically discussed, in search of theoretical evidence of the fact that, when non-idealities are envisaged, a preferred front-end architecture exists and that an optimum range of values for circuit variables can be selected to achieve the best result in terms of timing accuracy. The second part proposes the complete design of a front-end implemented in a standard 130-nm CMOS process from TSMC, aimed at achieving state-of-art timing accuracy performance when coupled to large area SiPMs indeed intended for applications such as ToF-PET. A timing resolution of 33 ps rms for single-photon events and a dynamic range from 1 to 8000 photoelectrons have been achieved and validated with post-layout simulations, thanks to innovative circuit techniques, that have been described in detail. Finally, pending the prototype fabrication, the closing section of the work has been dedicated to carry out extensive measurements on a test preamplifier, realized on a printed circuit board with discrete components and coupled to a 3x3 mm2 SiPM commercially available from Hamamatsu Ltd. The preamplifier can be configured according to the architectures proposed in the theoretical analysis, with the purpose of validating the results of that analysis. The experimental results achieved are in good agreement with the analytical expressions obtained with the developed mathematical study, thus demonstrating the effectiveness of the proposed approach in making comparative analysis of different front-end architectures.File | Dimensione | Formato | |
---|---|---|---|
32 ciclo-CALÒ Pietro Antonio Paolo.pdf
accesso aperto
Descrizione: Tesi di dottorato completa di frontespizio
Tipologia:
Tesi di dottorato
Licenza:
Creative commons
Dimensione
11.68 MB
Formato
Adobe PDF
|
11.68 MB | Adobe PDF | Visualizza/Apri |
I documenti in IRIS sono protetti da copyright e tutti i diritti sono riservati, salvo diversa indicazione.