Practical linear and quadratic gradient errors suppression techniques in string DACs

Gradient errors in device arrays cause mismatch between device parameters, which in turn degrade linearity performance of data converters realized with these arrays. A practical "outputs averaging" technique for string DACs is presented to release complex routing problems in gradient reduc...

Full description

Saved in:
Bibliographic Details
Published in2017 IEEE 60th International Midwest Symposium on Circuits and Systems (MWSCAS) pp. 1176 - 1179
Main Authors Nanqi Liu, Lash, Chris, Todsen, Jim, Degang Chen
Format Conference Proceeding
LanguageEnglish
Published IEEE 01.08.2017
Subjects
Arrays
Layout
Linearity
Performance evaluation
Resistance
Resistors
Routing
Online AccessGet full text

Cover

More Information
Summary:Gradient errors in device arrays cause mismatch between device parameters, which in turn degrade linearity performance of data converters realized with these arrays. A practical "outputs averaging" technique for string DACs is presented to release complex routing problems in gradient reduction patterns. An N-bit string is divided into multiple substrings and the substrings' outputs are averaged to reduce gradient errors induced nonlinearity. A novel layout methodology, called "periodic INL shift", is further developed to suppress quadratic gradient errors. Based on these two techniques, a practical structure with linear gradient cancellation and quadratic gradient reduction is proposed. MATLAB simulations are performed in an 8-bit string DAC and significant improvements on linearity performance are obtained.
ISSN:1558-3899
DOI:10.1109/MWSCAS.2017.8053138