Light detection and ranging (LiDAR) exploits the “echo” light reflected back by an object illuminated by a light emitter. Nowadays, such technique is emerging in respect to others, due to the fine resolution and the possibility to reconstruct a 3D map of the scene. The 3D ranging spans from long-range automotive applications towards shorter scenes, such as in augmented reality and in smartphone cameras The main goal of this Ph.D. dissertation was to present the design and study of two different SPAD-sensors, for scanning single-spot LiDAR application with solar background rejection. Both sensors have been commissioned by a market leader customer and must fit into a specific scanning LiDAR system. The main idea behind the first microchip was the possibility to move toward a single-shot acquisition, using multiple SPADs within the single-pixel. The photon coincidence is exploited to track the incoming light and follow the detected peaks. Only the TOF related to the highest detected peak is provided, so it is possible to read out only one TOF per frame. The second chip implements a more standard approach, known as Time-Correlated Single Photon Counting (TCSPC), exploited in many SPAD based setups. More TOFs can be computed, thanks to the multi-hit timing electronics, and a novel background rejection technique has been implemented: the background is observed for a defined interval, and a threshold is set accordingly. Every TOF computed after each threshold crossing, is associated to the number of triggered SPADs (i.e., number of photons), so a weighted histogram can be built off-chip. These new methods will be validated in the customer’s final setup. The two microchips have been conceived, designed, and fabricated in a 160 nm BCD technology node. Such technology features SPADs with excellent performances, and is a mature node used since many years in the SPADlab at POLIMI. The first SPAD chip has been produced in 2021 and some preliminary validation and characterization have been performed after fabrication. A powerful 905 nm laser has been assembled to execute some preliminary long-range measurements. The second chip has been entirely designed in 2021 and the tape out should have been in April 2022. Unfortunately, some manufacturing errors during fabrication required to start a second batch, and the tape out will be at the end of 2022. The last section describes a parallel activity concerning the design of a SPAD array for quantum communication, already started during the master thesis.
La tecnica LiDAR sfrutta “l’eco” di luce riflessa da un oggetto illuminato da un emettitore. Ai giorni nostri, questa tecnica è in continua evoluzione, grazie alla sua risoluzione e la possibilità di ricreare una mappa 3D della scena. Il 3D ranging può avere un lungo fondo scala, utile soprattutto per applicazioni automotive, realtà aumentata e smartphone. L’obiettivo principale di questo dottorato è presentare lo studio e la progettazione di due diversi sensori SPAD, per Lidar a scansione singolo punto con grande luce ambientale. Entrambi i sensori sono stati commissionati da un leader nel market, e devono essere integrati in un sistema specifico per LiDAR. L’idea principale del primo chip è quella di poter arrivare al regime di singolo colpo, usando più SPAD in parallelo nello stesso pixel. La tecnica di coincidenza a più fotoni è usata per seguire i picchi di corrente in rivelati. Il tempo di volo relativo al picco più alto è utilizzato per determinarne la distanza dell’oggetto, leggendo quindi un solo tempo di volo per frame. Il secondo chip implementa un approccio più standard: la TCSPC, usata in molteplici applicazioni che sfruttano il tempo di volo dei fotoni. Con questa architettura è possibile calcolare molteplici tempi di volo, grazie a un TDC multi-hit e una nuova tecnica di reiezione della luce di fondo: il background è osservato per un certo intervallo di tempo, determinandone la soglia. Ogni volta che il segnale in ingresso supera questa soglia, il tempo di volo viene calcolato insieme al numero di SPAD che è scattato, dando la possibilità di creare un istogramma pesato off-chip. I due chip sono stati progettati e fabbricati in BCD 160 nm. Il primo chip è stat prodotto nel 2021 e alcune validazioni preliminari sono state effettuate, mentre il secondo chip, a causa di alcuni errori nel processo di fabbricazione non è stato rilasciato nel 2022. Nell’ultima parte di questa tesi è descritto lo sviluppo di un’attitivà parallela riguardante lo sviluppo di un array SPAD per quantum communication, realizzato durante la tesi magistrale.
SPAD arrays and SIPM for time-of-flight LiDAR and quantum communication
Incoronato, Alfonso
2022/2023
Abstract
Light detection and ranging (LiDAR) exploits the “echo” light reflected back by an object illuminated by a light emitter. Nowadays, such technique is emerging in respect to others, due to the fine resolution and the possibility to reconstruct a 3D map of the scene. The 3D ranging spans from long-range automotive applications towards shorter scenes, such as in augmented reality and in smartphone cameras The main goal of this Ph.D. dissertation was to present the design and study of two different SPAD-sensors, for scanning single-spot LiDAR application with solar background rejection. Both sensors have been commissioned by a market leader customer and must fit into a specific scanning LiDAR system. The main idea behind the first microchip was the possibility to move toward a single-shot acquisition, using multiple SPADs within the single-pixel. The photon coincidence is exploited to track the incoming light and follow the detected peaks. Only the TOF related to the highest detected peak is provided, so it is possible to read out only one TOF per frame. The second chip implements a more standard approach, known as Time-Correlated Single Photon Counting (TCSPC), exploited in many SPAD based setups. More TOFs can be computed, thanks to the multi-hit timing electronics, and a novel background rejection technique has been implemented: the background is observed for a defined interval, and a threshold is set accordingly. Every TOF computed after each threshold crossing, is associated to the number of triggered SPADs (i.e., number of photons), so a weighted histogram can be built off-chip. These new methods will be validated in the customer’s final setup. The two microchips have been conceived, designed, and fabricated in a 160 nm BCD technology node. Such technology features SPADs with excellent performances, and is a mature node used since many years in the SPADlab at POLIMI. The first SPAD chip has been produced in 2021 and some preliminary validation and characterization have been performed after fabrication. A powerful 905 nm laser has been assembled to execute some preliminary long-range measurements. The second chip has been entirely designed in 2021 and the tape out should have been in April 2022. Unfortunately, some manufacturing errors during fabrication required to start a second batch, and the tape out will be at the end of 2022. The last section describes a parallel activity concerning the design of a SPAD array for quantum communication, already started during the master thesis.| File | Dimensione | Formato | |
|---|---|---|---|
|
2023_04_PhD_Incoronato.pdf
non accessibile
Dimensione
7.06 MB
Formato
Adobe PDF
|
7.06 MB | Adobe PDF | Visualizza/Apri |
I documenti in POLITesi sono protetti da copyright e tutti i diritti sono riservati, salvo diversa indicazione.
https://hdl.handle.net/10589/198638