The design of high performance analog interfaces is becoming progressively harder as silicon semiconductor technologies (like CMOS) keep scaling and circuits can only operate with smaller supply voltages. At the same time, demands on circuit performance and power efficiency are increasing. This program targets to research and demonstrate a switched-mode signal-processing paradigm. It allows close-to rail-to-rail signal swings and high efficiency for the output stages of operational amplifiers while being very linear. It further facilitates the easy translation of signals between switched mode, and analog or digital representations. Switched-mode circuits are expected to enable performance gains in a versatile set of analog functions including amplifiers, filters and oversampling analog-to-digital converters.

The proposed signal representation and processing paradigm uniquely exploits the strengths of nanoscale devices, in particular their high switching speed, while avoiding the challenges associated with their limited output impedance or voltage tolerance and resulting swing reductions. It further pursues new avenues for circuit and architecture explorations by representing analog information in a non-classical way, in particular by moving the representation from the voltage to time domain in a linear manner. Additionally, switched-domain representation offers seamless conversion from one domain to the other, a challenge that has hindered progress for other time-domain based representations.

Broader Impact

Analog circuits perform the critical function of interfacing the physical world to the digital world of signal processing and storage. The performance of digital systems in nanoscale technologies can only be reaped with analog interface with equally high performance in fully integrated systems-on-a-chip. Such devices are a key enabling technology for advances in many areas with broad societal and scientific impact from technology to medicine, from communications to scientific discoveries. The majority of applications have migrated to digital-centric architectures, owing to the advantages of cost or robustness and the availability of significantly larger digital signal-processing capability in nanoscale semiconductor technologies. The paradigm and circuits researched here show promise for impact across a variety of applications in multiple disciplines.

The research includes theoretical, design and experimental efforts for the design of several proof-of-principle demonstrators. Graduate and undergraduate students will gain expertise in theoretical, fundamental, practical and experimental aspects of the design of high-performance, analog interface circuits in nanoscale silicon technologies. The results of this research will be disseminated through publications and seminars and will be incorporated in graduate level courses.

Project Start
Project End
Budget Start
2013-09-01
Budget End
2018-08-31
Support Year
Fiscal Year
2013
Total Cost
$368,000
Indirect Cost
Name
Columbia University
Department
Type
DUNS #
City
New York
State
NY
Country
United States
Zip Code
10027