11. A DIGITALLY ASSISTED DYNAMICALLY RECONFIGURABLE ANALOG FRONT-END WITH DATA-DEPENDENT POWER REDUCTION FOR PHYSIOLOGICAL SENSING APPLICATIONS

Department: Electrical & Computer Engineering
Research Institute Affiliation: Agile Center for Wearable Sensors (CWS)
Faculty Advisor(s): Drew A. Hall

Primary Student
Name: Somok Mondal
Email: somondal@ucsd.edu
Phone: 858-361-4217
Grad Year: 2018

Abstract
A typical bio-signal acquision system consisting of a low-noise bio-potential amplifier, an ADC and a DSP back-end is traditionally designed corresponding to the worst-case operation scenarios. Hence, such a design, with the various performance parameters being fixed and over-designed, generally consumes a higher power than needed. However, a front-end as the one presented in this work which could be adapted or reconfigured based on an appropriate feedback from the DSP can ovecome such a drawback. Implemented in 65nm CMOS, the analog front-end (AFE) has a number of digitally programmable modes with the input referred noise ranging between 1.90μV_rms in 250Hz to 3.89μV_rms in 420Hz and drawing 307nW to 769nW power from a 0.8V supply. The designed AFE with agile, on-the-fly, noise-power reconfiguration is used to achieve signal-activity dependent power savings of 2.5 for an ECG signal while still acquiring pristine signal-features of interest.

Industry Application Area(s)
Electronics/Photonics | Life Sciences/Medical Devices & Instruments | Semiconductor

« Back to Posters or Search Results


Contact:   researchexpo@soe.ucsd.edu   (858) 534-6068