Design methods to mitigate the effects of variation in analogue and mixed-signal circuits
Design methods to mitigate the effects of variation in analogue and mixed-signal circuits
The continued scaling of CMOS process features enables ever-faster and denser circuits, which comes at the cost of increased device parameter variation. The performance of analogue and mixed-signal circuits in particular degrades in such a high variation environment, which poses an extraordinary challenge in the design and fabrication of such circuits.
This thesis develops a set of tools and methodologies for a post-fabrication calibration system called the Configurable Analogue Transistor (CAT). The principle of the CAT technique is to replace certain transistors in a circuit with calibration devices, which allow adjustment of circuit performance after fabrication to compensate the effects of device parameter variation. Building on initial research on the CAT, this thesis develops a methodology to identify the most suitable calibration devices in their circuit and determine their optimal sizes. Furthermore, the applicability of CAT is extended beyond parameter variation to also include direct compensation of temperature.
A complementary technique to post-fabrication calibration is robust design, where a circuit is designed to be inherently robust against variation in device parameters.
In this thesis, a novel closed-loop pick-off circuit for force-balanced MEMS accelerometers is presented. It is comparable in performance to other state-of-the-art techniques, but provides vastly improved robustness against parameter variation and a more intuitive design process.
Rudolf, Robert
83a73233-ff6f-4d49-919c-c59f4b31c196
November 2014
Rudolf, Robert
83a73233-ff6f-4d49-919c-c59f4b31c196
Wilson, Peter
8a65c092-c197-4f43-b8fc-e12977783cb3
Rudolf, Robert
(2014)
Design methods to mitigate the effects of variation in analogue and mixed-signal circuits.
University of Southampton, Physical Sciences and Engineering, Doctoral Thesis, 227pp.
Record type:
Thesis
(Doctoral)
Abstract
The continued scaling of CMOS process features enables ever-faster and denser circuits, which comes at the cost of increased device parameter variation. The performance of analogue and mixed-signal circuits in particular degrades in such a high variation environment, which poses an extraordinary challenge in the design and fabrication of such circuits.
This thesis develops a set of tools and methodologies for a post-fabrication calibration system called the Configurable Analogue Transistor (CAT). The principle of the CAT technique is to replace certain transistors in a circuit with calibration devices, which allow adjustment of circuit performance after fabrication to compensate the effects of device parameter variation. Building on initial research on the CAT, this thesis develops a methodology to identify the most suitable calibration devices in their circuit and determine their optimal sizes. Furthermore, the applicability of CAT is extended beyond parameter variation to also include direct compensation of temperature.
A complementary technique to post-fabrication calibration is robust design, where a circuit is designed to be inherently robust against variation in device parameters.
In this thesis, a novel closed-loop pick-off circuit for force-balanced MEMS accelerometers is presented. It is comparable in performance to other state-of-the-art techniques, but provides vastly improved robustness against parameter variation and a more intuitive design process.
Text
__soton.ac.uk_ude_personalfiles_users_jo1d13_mydesktop_Rudolf.pdf
- Other
More information
Published date: November 2014
Organisations:
University of Southampton, EEE
Identifiers
Local EPrints ID: 374300
URI: http://eprints.soton.ac.uk/id/eprint/374300
PURE UUID: ff62a2cf-7475-4e5d-a750-7dbd4ac3c9fd
Catalogue record
Date deposited: 16 Feb 2015 14:28
Last modified: 14 Mar 2024 19:05
Export record
Contributors
Author:
Robert Rudolf
Thesis advisor:
Peter Wilson
Download statistics
Downloads from ePrints over the past year. Other digital versions may also be available to download e.g. from the publisher's website.
View more statistics