CMOS subthreshold-mode I/V converter for analogue neural network applications

被引：1

作者：

Coue, D

Wilson, G

机构：

[1] Sch. Electron. Commun. Elec. Eng., University of Plymouth, Plymouth, Devon PL4 8AA, Drake Circus

来源：

ELECTRONICS LETTERS | 1996年 / 32卷 / 11期

关键词：

neural networks; CMOS integrated circuits;

D O I：

10.1049/el:19960639

中图分类号：

TM [电工技术]; TN [电子技术、通信技术];

学科分类号：

0808 ; 0809 ;

摘要：

An improved I/V converter for analogue neural network applications is presented, using devices operating in the subthreshold mode of conduction. The proposed scheme employs diode connected transistors biased by a network incorporating feedback. Analysis and simulation suggest that the modified structure offers a substantial reduction in the sensitivity to supply voltage variations. as the proposed scheme uses only NMOS transistors, it can be implemented in a standard single well process.

引用

页码：990 / 991

页数：2