Analog-to-digital converter (ADC) is a very fundamental and key part to nearly all kinds of electronics. The applications cover a wide range requiring different resolution to different sampling rate, including UWB systems, radar detection, wide band radio receivers, optical communication links, CCD imaging, ultrasonic medical imaging, digital receivers, base stations, digital video (for example, HDTV), xDSL, cable modems, and fast Ethernet. Among them, lower resolution very high speed ADC is a critical part for building UWB system, disk drive read channels and optical communication.
This thesis consists of two parts. The first part focuses on the design of a high speed low resolution flash ADC in 90nm technology. Capacitive interpolation technique was used in this flash ADC in order to reduce the hardware requirement and input capacitance. No sample-and-hold (S/H) circuit is needed since the distributed capacitors (including capacitors in the very front end and the interpolated capacitors) serve to sample and hold the signals. Offset cancellation and averaging techniques are also implemented to reduce the offsets and the non-linearity. The ADC design achieves a sampling speed of 2.3GSps with 4 bits resolution in 90nm CMOS technology.
The second part describes a new comprehensive ADC design methodology for capacitive interpolated flash ADCs, aiming to provide a quantitative, yet handy design guideline for circuit designers to conduct practical ADC designs. This new ADC design methodology provides a quantitative and comprehensive mapping between ADC chip level performance specs and various design parameters at different levels, such as, interpolation factor, number of stages, pre-amplifier bandwidth, loading effects, transistor sizes, technology parameters and etc. It serves to allow IC designers to conduct quick and quantitative flash ADC designs for well-balanced overall chip performance in practices. A dynamic power consumption analysis technique for capacitive interpolated flash ADCs is also discussed.