High Performance CMOS Range Imaging Device Technology and Systems Considerations

1 Introduction

2 State of the art range imaging
2.1 Triangulation
2.2 Interferometry
2.3 Time-of-flight
2.3.1 Direct time-of-flight
2.3.2 Continuous wave method
2.3.3 Pulsed wave method
2.4 Comparison of optical range imaging methods

3 Temporal noise
3.1 Introduction to noise analysis
3.1.1 Basic probabilistic concepts for the analysis of uncertainties
3.1.2 Stochastic processes
3.1.3 Propagation of noise in linear time-invariant circuits
3.2 Noise analysis in non-linear and time-variant systems
3.2.1 Transformation of probability density functions
3.2.2 Employing z-transform for noise analysis
3.2.3 LPTV methods
3.2.4 Propagation of noise in non-linear time-variant systems
3.2.5 Noise in the time domain
3.2.6 A sequential method using a switching time-frequency domain
3.3 Fundamental noise processes in electronic devices
3.3.1 Thermal noise
3.3.2 Shot noise and photon noise
3.3.3 Remarks on thermal noise
3.3.4 Generation-recombination noise
3.3.5 Random telegraph signal noise – burst noise
3.3.6 Flicker noise
3.4 Noise processes under time-varying bias
3.5 Impedance field method

4 Noise performance of devices available in the 0.35μm CMOS process
4.1 Transistor noise basics
4.1.1 Bipolar transistor noise model
4.1.2 Field-effect transistor noise modeling
4.2 Noise performance of standard MOS Field-Effect Transistors
4.3 Noise performance of available bipolar devices

5 Noise in active pixel sensors
5.1 Photodetector principle
5.2 Photodetector noise and reduction techniques
5.2.1 Dark noise
5.2.2 Photon noise
5.2.3 Reset noise
5.2.4 Thermal, flicker and RTS noise
5.3 Correlated double sampling
5.4 Novel JFET readout structure for CMOS APS

6 On the design of PM-ToF range imagers
6.1 Basic concept and constraints
6.2 Physical limitations due to photon induced shot noise
6.3 Design objectives and considerations
6.3.1 Design objectives
6.3.2 Photodetector selection
6.3.3 Sensor system architecture
6.3.4 Fabrication technology
6.4 Detector design and evaluation
6.4.1 Readout circuitry
6.4.2 ToF-LDPD design
6.4.3 Evaluation of the first generation LDPD based PM-ToF imager
6.5 Speed considerations for ldpd based TOF image sensors
6.5.1 Design Considerations for charge transfer speed improvement
6.5.2 Evaluation of the second generation LDPD based PM-ToF imager
6.6 Matching considerations
6.6.1 Alternative ToF-LDPD concept
6.6.2 Evaluation of the third generation LDPD based PM-ToF imager
6.7 Impact of finite charge transfer speed and parasitic light sensitivity on PM-TOF
6.7.1 Concept of the generalized MSI ToF model
6.7.2 Verification
6.7.3 Fitting and comparison of the ToF-LDPD designs
6.7.4 Impact on precision

7 Conclusions

Appendix A Derivation of the autocorrelation formula of shot noise

Appendix B Measurement setups

B.1 Noise measurement setup

B.2 Setup to measure according to the emulated TOF principle

Appendix C Photon transfer method

Nomenclature

Abbreviations

Bibliography

Index

Biography

Andreas Süss received his BSc from the University of Applied Sciences Düsseldorf in 2008 and a PhD degree from the University of Duisburg-Essen in 2014. From 2007 until 2014 he was affiliated to the Fraunhofer Institute IMS where he was working mainly on high-speed, low-noise imagers for e.g. ToF applications. From 2014 until 2015 he had a scholarship from the KU Leuven and worked as a postdoctoral researcher on global shutter imaging at the MICAS department in collaboration with IMEC, Leuven. As of 2015 he is hired as an R&D engineer in the IMEC imaging division, where he is currently responsible for the pixel development for global shutter and high-speed applications. His research interests include modeling, temporal noise, optimization, compressed sensing and depth imaging.