← 返回论文列表 📄 下载原文 PDF  ISSCC 2026 · 36.11
ISSCC 2026Session 36 · NEURAL AND BIOMEDICAL INTERFACESAI / ML

A 0.62μW/sensor 82fps Time-to-Digital Impedance Measurement IC with Unified Excitation/Readout Front-End for Large-Scale Piezo-Resistive Sensor Array

⚡ 本页包含 AI 生成的分析内容,仅供参考

📋 论文概要

本文提出一种用于大规模压阻传感器阵列的快速阻抗测量IC,采用统一差分时域解调架构,通过激励电路直接读出阻抗,并利用预饱和自适应偏置技术提高能效。该芯片在125kHz下扫描253个传感器仅需12.2ms(82fps),功耗158μW,实现了0.5%误差和71.1dB SNR。

💡 主要创新点

核心指标
158μW总功耗,82fps扫描速率,0.5%误差,71.1dB SNR
重要性
发表年份
ISSCC 2026

🏷 关键词

阻抗测量时域解调压阻传感器阵列低功耗自适应偏置

📄 原文摘要

Abstract This paper presents a fast impedance-measurement IC for large-scale piezo-resistive sensor arrays. It features a unified differential time-to-digital demodulation architecture that reads out impedance directly through the excitation circuit. A pre-saturation adaptive bias technique further improves power efficiency. The chip scans 253 sensors in 12.2ms (82fps) at 125kHz, consuming 158μW (7.5nJ/sensor). With loads from 20$ to 500k$, it achieves 0.5% error and up to 71.1dB SNR. Frailty increases sharply with age and is the major cause of falls, leading to severe injury and fatality. While fall risks can be effectively predicted using gait analysis, it is often limited in laboratory settings. Piezo-resistive crossbar sensor arrays (Fig. 36.11.1 (top)) can be used towards lab-grade, yet wearable insoles. However, several challenges remain: 1) high sensor density (>250 per insole); 2) high temporal resolution required (50 to 100 frames

👥 作者与机构

Jiayang Li1,2, Qingyu Zhang2, Sohmyung Ha2,3, Andreas Demosthenous2, Dai Jiang2, Yu Wu2

University of Bristol, Bristol, United Kingdom, 2University College London, London, United Kingdom, 3New York University Abu Dhabi, Abu Dhabi, United Arab Emirates

分类:AI / ML · 年份:ISSCC 2026