- Undergraduate
Bachelor's Degrees
Bachelor of ArtsBachelor of EngineeringDual-Degree ProgramUndergraduate AdmissionsUndergraduate Experience
- Graduate
Graduate Experience
- Research
- Entrepreneurship
- Community
- About
-
Search
All Thayer Events
PhD Thesis Defense: Wei Deng
Jan
14
Friday
2:00pm - 4:00pm ET
Videoconference
For info on how to attend this video conference, please email wei.deng.TH@dartmouth.edu.
"Noise Performance of Quanta Image Sensors"
Abstract
The Quanta image sensor (QIS), under development at Dartmouth, is a possible next-generation solid-state image sensor technology after the charge-coupled device (CCD) and the CMOS image sensor (CIS). QIS features high temporal-spatial resolution and has deep sub-electron read noise that allows photon counting. For accurate photon counting, the QIS read noise target is below 0.15 e− rms although single electron quantization becomes apparent below about 0.45 e− rms. Currently, the noise from the in-pixel source-follower (SF) transistor dominates the read noise of QIS. SF noise is composed of several different types of noise, mainly 1/f noise or flicker noise, random telegraph noise (RTN), and thermal noise. The origin of RTN is usually attributed to conduction carrier trapping and re-emission, while the theory of thermal noise is well-established. Nevertheless, the physical origin of 1/f noise has not been confirmed.
To further reduce QIS read noise and enable accurate single-photon counting, the origins of all types of noise need to be identified, especially 1/f noise. This dissertation models the 1/f noise of QIS, along with other noises such as RTN. The physical origin of 1/f noise will be investigated and a comparison between the modeling and experimental results will be discussed. Different approaches, including bandpass filtering, SF sizing, and further cooling, to achieve even lower read noise will be explored. To better understand the noise origin, a QIS prototype chip consisting of different design variations and novel structures was designed, fabricated and characterized. A multi-gate source follower is demonstrated to reduce read noise in image sensors. Finally, a high-speed QIS chip was implemented and the sensor can operate at up to 33Mcounts/s/jot with low dead time. The performance of QIS at low temperatures is also investigated for future space missions.
Thesis Committee
- Eric Fossum, PhD (Chair)
- Jifeng Liu, PhD
- Kofi Odame, PhD
- Donald Figer, PhD
Contact
For more information, contact Theresa Fuller at theresa.d.fuller@dartmouth.edu.