A 20-bit ±40-mV Range Read-Out IC With 50-nV Offset and 0.04% Gain Error for Bridge Transducers

Rong Wu, Youngcheol Chae, Johan H. Huijsing, Kofi A.A. Makinwa

Research output: Contribution to journalArticleScientificpeer-review

55 Citations (Scopus)
39 Downloads (Pure)

Abstract

This paper presents a 20-b read-out IC with ±40-mV full-scale range that is intended for use with bridge transducers. It consists of a current-feedback instrumentation amplifier (CFIA) followed by a switched-capacitor incremental ΔΣ ADC. The CFIA's offset and 1/ f noise are mitigated by chopping, while its gain accuracy and gain drift are improved by applying dynamic element matching to its input and feedback transconductors. Their mismatch is reduced by a digitally assisted correction loop, which further reduces the CFIA's gain drift. Finally, bulk-biasing and impedance-balancing techniques are used to reduce the common-mode dependency of these transconductors, which would otherwise limit the achievable gain accuracy. The combination of these techniques enables the read-out IC to achieve 140-dB CMRR, a worst-case gain error of 0.04% over a 0-2.5 V common-mode range, a maximum gain drift of 0.7 ppm/°C and an INL of 5 ppm. After applying nested-chopping, the read-out IC achieves 50-nV offset, 6-nV/°C offset drift, a thermal noise floor of 16.2 nV/√Hz and a 0.1-mHz 1/ f noise corner. Implemented in a 0.7-μm CMOS technology, the prototype read-out IC consumes 270 μA from a 5-V supply.
Original languageEnglish
Pages (from-to)2152-2163
Number of pages12
JournalIEEE Journal of Solid State Circuits
Volume47
Issue number9
DOIs
Publication statusPublished - 2012

Keywords

  • $1/ f $noise
  • bridge transducer
  • chopping
  • current feedback instrumentation amplifier (CFIA)
  • dynamic element matching
  • gain accuracy
  • gain drift
  • incremental delta-sigma ADC
  • linearity
  • offset
  • readout-IC

Fingerprint Dive into the research topics of 'A 20-bit ±40-mV Range Read-Out IC With 50-nV Offset and 0.04% Gain Error for Bridge Transducers'. Together they form a unique fingerprint.

Cite this