首页> 外文学位 >Compact semi-physical modelling for HBT circuit simulations
【24h】

Compact semi-physical modelling for HBT circuit simulations

机译:紧凑的半物理建模,用于HBT电路仿真

获取原文
获取原文并翻译 | 示例

摘要

With the maturing of III-V heterojunction bipolar transistor's (HBT's) design and processing teclinology, development of a truly physical and predictive III-V HBT compact model for circuit simulation that covers geometry, bias, temperature, DC and RF characteristics becomes a major challenge. Many of the existing HBT models are based on solid physical backgroimd, but again because of the complexity of the device physics they end up with many empirical coefficients that are difficult to extract. In this thesis, aspects of InGaP/GaAs double HBTs (DHBTs) pertaining to large signal models (LSMs) are addressed. A new LSM generation methodology for III-V HBTs that is physics and process oriented, which reduces parameter extraction efforts moderately is presented. As an initial step in the LSM development methodology, an algorithm based on T-topology HBT equivalent circuit (EC) is developed using IC-CAP software to effectively generate multi-bias small-signal parameter sets. Here analytical approximations and numerical optimisation procedui-es are utilised to accurately fit measured s-parameter data, extracting the maximimi amount of information, parameter values and constraints. Based on this technique, continuously differentiable constitutive relationships are established that takes into accoimt dynamic and static temperature effects as well as high ciuTent effects such as Kirk effect and self-heating, which are important for the operation of InGaP/GaAs DHBTs. It combines the basis of the Ebers-Moll model with elements of the vertical bipolar inter-company model (VBIC) and the high current model (HICUM) as well as including specific extensions for InGaP/GaAs DHBTs. In order to assess the validity and accuracy of the LSM development methodology, an eight port symbolically defined device (SDD) model is constructed in Agilent's ADS circuit simulator. In addition, a novel teclmique for extracting directly the temperature dependence of terminal resistances using the observed kink effect is presented. Using a theoretical model for tunnelling tlirough metal-semiconductor barriers, incoiporating Fermi-Dirac statistics and Wentzel-Kramers-Brillouin (WKB) approximations, it was demonstrated that InGaAs capping layers provided stable, temperature insensitive contacts to the HBT. The LSM development methodology presented in this thesis gives a complete set of model equations that is applicable to III-V HBTs together with parameter extraction teclmiques suitable for the implementation in commercial simulators and characterisation programmes. These provide intuitive imderstanding in the development of modern III-V HBT models for circuit simulations.
机译:随着III-V异质结双极晶体管(HBT)的设计和处理技术的日趋成熟,开发一种真正的物理和可预测的III-V HBT紧凑模型以进行电路仿真,该模型涵盖几何形状,偏置,温度,DC和RF特性成为一项重大挑战。 。许多现有的HBT模型都基于可靠的物理背景,但同样由于设备物理的复杂性,它们最终具有许多难以提取的经验系数。在本文中,InGaP / GaAs双HBT(DHBT)与大信号模型(LSM)有关。提出了一种面向物理和过程的,面向III-V HBT的新型LSM生成方法,该方法可适度减少参数提取的工作量。作为LSM开发方法的第一步,使用IC-CAP软件开发了基于T拓扑HBT等效电路(EC)的算法,以有效地生成多偏置小信号参数集。在这里,利用分析逼近和数值优化程序来精确拟合测得的s参数数据,提取信息,参数值和约束的最大量。基于该技术,建立了连续可区分的本构关系,该结构关系吸收了动态和静态温度效应以及柯尼效应和自热效应等高效应,这对于InGaP / GaAs DHBT的操作非常重要。它结合了Ebers-Moll模型的基础,垂直双极公司间模型(VBIC)和高电流模型(HICUM)的元素,并包括InGaP / GaAs DHBT的特定扩展。为了评估LSM开发方法的有效性和准确性,在安捷伦的ADS电路仿真器中构建了一个八端口符号定义的设备(SDD)模型。此外,提出了一种新技术,可使用观察到的扭结效应直接提取终端电阻的温度依赖性。使用理论模型来隧穿金属半导体势垒,费米-狄拉克统计和Wentzel-Kramers-布里渊(WKB)近似,证明InGaAs覆盖层为HBT提供了稳定,对温度不敏感的接触。本文提出的LSM开发方法论提供了一整套适用于III-V HBT的模型方程式,以及适用于在商业模拟器和表征程序中实现的参数提取技术。这些为开发用于电路仿真的现代III-V HBT模型提供了直观的理解。

著录项

  • 作者

    Dharmasiri, Chalika N.;

  • 作者单位

    The University of Manchester (United Kingdom).;

  • 授予单位 The University of Manchester (United Kingdom).;
  • 学科 Electrical engineering.
  • 学位 Ph.D.
  • 年度 2006
  • 页码 294 p.
  • 总页数 294
  • 原文格式 PDF
  • 正文语种 eng
  • 中图分类
  • 关键词

相似文献

  • 外文文献
  • 中文文献
  • 专利
获取原文

客服邮箱:kefu@zhangqiaokeyan.com

京公网安备:11010802029741号 ICP备案号:京ICP备15016152号-6 六维联合信息科技 (北京) 有限公司©版权所有
  • 客服微信

  • 服务号