{"title":"A non-quasi-static modular model for HBTs","authors":"A. Morris, R. Trew, C. T. Kelley, G. J. Hayes","doi":"10.1109/CORNEL.1993.303108","DOIUrl":null,"url":null,"abstract":"Heterojunction bipolar transistors (HBTs) show promise as a high speed and high power density device for many circuit applications. However the quasi-static models found in standard circuit simulation tools can not treat fast transients in HBTs properly. This leads to inaccurate simulations at high frequency and of strongly non-linear operation. To properly account for the charge in transit through the device, non-quasi-static (NQS) models must be used. This work presents a model for the bipolar transistor formed from regional modules. Each module is a NQS solution to a specific region of the transistor and uses material and geometry inputs. These modules are solved for physical consistency during non-linear circuit simulation. The modularization allows appropriate approximations for each region to yield analytic solutions. The input parameters for the model reflect the physical structure of the device as much as possible to provide intuitive results and verifiability. This allows direct device optimization since all parameters are either uncorrelated or their correlations can be derived from process parameters. Thus the device can be optimized in its circuit environment. The model provides for many effects which previously required numerical simulation for accurate results. These include forward and reverse Early, Webster/Rittner, and Kirk/quasisaturation effects. By following the modular modeling scheme, these effects are simply the result of varying boundary conditions on each of the regional solutions. The modular model provides much of the physical insight of numerical models but with computational requirements on the same order as conventional circuit models.<<ETX>>","PeriodicalId":129440,"journal":{"name":"Proceedings of IEEE/Cornell Conference on Advanced Concepts in High Speed Semiconductor Devices and Circuits","volume":"16 1","pages":"0"},"PeriodicalIF":0.0000,"publicationDate":"1993-08-02","publicationTypes":"Journal Article","fieldsOfStudy":null,"isOpenAccess":false,"openAccessPdf":"","citationCount":"0","resultStr":null,"platform":"Semanticscholar","paperid":null,"PeriodicalName":"Proceedings of IEEE/Cornell Conference on Advanced Concepts in High Speed Semiconductor Devices and Circuits","FirstCategoryId":"1085","ListUrlMain":"https://doi.org/10.1109/CORNEL.1993.303108","RegionNum":0,"RegionCategory":null,"ArticlePicture":[],"TitleCN":null,"AbstractTextCN":null,"PMCID":null,"EPubDate":"","PubModel":"","JCR":"","JCRName":"","Score":null,"Total":0}
引用次数: 0
Abstract
Heterojunction bipolar transistors (HBTs) show promise as a high speed and high power density device for many circuit applications. However the quasi-static models found in standard circuit simulation tools can not treat fast transients in HBTs properly. This leads to inaccurate simulations at high frequency and of strongly non-linear operation. To properly account for the charge in transit through the device, non-quasi-static (NQS) models must be used. This work presents a model for the bipolar transistor formed from regional modules. Each module is a NQS solution to a specific region of the transistor and uses material and geometry inputs. These modules are solved for physical consistency during non-linear circuit simulation. The modularization allows appropriate approximations for each region to yield analytic solutions. The input parameters for the model reflect the physical structure of the device as much as possible to provide intuitive results and verifiability. This allows direct device optimization since all parameters are either uncorrelated or their correlations can be derived from process parameters. Thus the device can be optimized in its circuit environment. The model provides for many effects which previously required numerical simulation for accurate results. These include forward and reverse Early, Webster/Rittner, and Kirk/quasisaturation effects. By following the modular modeling scheme, these effects are simply the result of varying boundary conditions on each of the regional solutions. The modular model provides much of the physical insight of numerical models but with computational requirements on the same order as conventional circuit models.<>