Pub Date : 1988-12-01DOI: 10.1109/ARFTG.1988.323923
B. R. Epstein, Z. Shen, Spencer C. Chen
Much publicity has surrounded the recent introduction of nonlinear microwave CAD tools into the marketplace. However, active device modeling remains as the major limitation of these tools. Furthermore, techniques that characterize nonlinear microwave devices in a systematic, accurate, and repeatable manner still must be developed in order to assure effective and proper use of the new CAD tools. Two techniques that offer the promise of practical device characterization are discussed in this paper. The flrst technique makes use of time-domain signal sampling to "extract" the nonlinear behavior of a device in terms of device model parameters. The second technique d e s use of load-pull measurements.
{"title":"Large-Signal MESFET Parameter Extraction Techniques","authors":"B. R. Epstein, Z. Shen, Spencer C. Chen","doi":"10.1109/ARFTG.1988.323923","DOIUrl":"https://doi.org/10.1109/ARFTG.1988.323923","url":null,"abstract":"Much publicity has surrounded the recent introduction of nonlinear microwave CAD tools into the marketplace. However, active device modeling remains as the major limitation of these tools. Furthermore, techniques that characterize nonlinear microwave devices in a systematic, accurate, and repeatable manner still must be developed in order to assure effective and proper use of the new CAD tools. Two techniques that offer the promise of practical device characterization are discussed in this paper. The flrst technique makes use of time-domain signal sampling to \"extract\" the nonlinear behavior of a device in terms of device model parameters. The second technique d e s use of load-pull measurements.","PeriodicalId":235867,"journal":{"name":"32nd ARFTG Conference Digest","volume":"77 1 1","pages":"0"},"PeriodicalIF":0.0,"publicationDate":"1988-12-01","publicationTypes":"Journal Article","fieldsOfStudy":null,"isOpenAccess":false,"openAccessPdf":"","citationCount":null,"resultStr":null,"platform":"Semanticscholar","paperid":"121026958","PeriodicalName":null,"FirstCategoryId":null,"ListUrlMain":null,"RegionNum":0,"RegionCategory":"","ArticlePicture":[],"TitleCN":null,"AbstractTextCN":null,"PMCID":"","EPubDate":null,"PubModel":null,"JCR":null,"JCRName":null,"Score":null,"Total":0}
Pub Date : 1988-12-01DOI: 10.1109/ARFTG.1988.323921
M. Sipila, K. Lehtinen, V. Porra, M. Valtonen
Accurate active device models are essential for reliable circuit simulation, necessary in the design of modern integrated circuits. The trend towards higher frequencies places an increasing demand on models with good accuracy in the UHF and microwave regions. Nonlinear high frequency models are needed in the computer aided design of communication subsystems, for example power amplifiers and mixers. Such models are also imperative in the time domain simulation of fast digital circuits used in supercomputers. A good nonlinear high frequency model should ideally predict correctly both the DC and large si nal AC behavior of the device. It should also give correct small signal S-parameters
{"title":"Determination of Some Nonlinear Transistor Model Parameters by Using Periodic Time Domain Measurements","authors":"M. Sipila, K. Lehtinen, V. Porra, M. Valtonen","doi":"10.1109/ARFTG.1988.323921","DOIUrl":"https://doi.org/10.1109/ARFTG.1988.323921","url":null,"abstract":"Accurate active device models are essential for reliable circuit simulation, necessary in the design of modern integrated circuits. The trend towards higher frequencies places an increasing demand on models with good accuracy in the UHF and microwave regions. Nonlinear high frequency models are needed in the computer aided design of communication subsystems, for example power amplifiers and mixers. Such models are also imperative in the time domain simulation of fast digital circuits used in supercomputers. A good nonlinear high frequency model should ideally predict correctly both the DC and large si nal AC behavior of the device. It should also give correct small signal S-parameters","PeriodicalId":235867,"journal":{"name":"32nd ARFTG Conference Digest","volume":"1 1","pages":"0"},"PeriodicalIF":0.0,"publicationDate":"1988-12-01","publicationTypes":"Journal Article","fieldsOfStudy":null,"isOpenAccess":false,"openAccessPdf":"","citationCount":null,"resultStr":null,"platform":"Semanticscholar","paperid":"127249763","PeriodicalName":null,"FirstCategoryId":null,"ListUrlMain":null,"RegionNum":0,"RegionCategory":"","ArticlePicture":[],"TitleCN":null,"AbstractTextCN":null,"PMCID":"","EPubDate":null,"PubModel":null,"JCR":null,"JCRName":null,"Score":null,"Total":0}
Pub Date : 1988-12-01DOI: 10.1109/ARFTG.1988.323924
Chia-lun J. Hu
When a high power microwave load is changing in time due to either microwave heating effect, or an environmental weathering effect, or a manual tuning of the load at high power level, or a manual sweeping of microwave power, the changing complex impedance (both magnitude and phase) is generally very difficult to monitored in real time. This paper reports an experimental result of the real-time nonlinear measurements of a standard time varying load operated at medium high power levels. The technique used is a well developed three-probe automatic impedance tracking measurement system. The real time nonlinear measurement capacity of the system is seen to be excellent. The error of measurements when the power level is changed while the load is kept constant is estimated to be within a few percents.
{"title":"Nonlinear Measurements of a Microwave Time-Varying Load","authors":"Chia-lun J. Hu","doi":"10.1109/ARFTG.1988.323924","DOIUrl":"https://doi.org/10.1109/ARFTG.1988.323924","url":null,"abstract":"When a high power microwave load is changing in time due to either microwave heating effect, or an environmental weathering effect, or a manual tuning of the load at high power level, or a manual sweeping of microwave power, the changing complex impedance (both magnitude and phase) is generally very difficult to monitored in real time. This paper reports an experimental result of the real-time nonlinear measurements of a standard time varying load operated at medium high power levels. The technique used is a well developed three-probe automatic impedance tracking measurement system. The real time nonlinear measurement capacity of the system is seen to be excellent. The error of measurements when the power level is changed while the load is kept constant is estimated to be within a few percents.","PeriodicalId":235867,"journal":{"name":"32nd ARFTG Conference Digest","volume":"56 1","pages":"0"},"PeriodicalIF":0.0,"publicationDate":"1988-12-01","publicationTypes":"Journal Article","fieldsOfStudy":null,"isOpenAccess":false,"openAccessPdf":"","citationCount":null,"resultStr":null,"platform":"Semanticscholar","paperid":"130954154","PeriodicalName":null,"FirstCategoryId":null,"ListUrlMain":null,"RegionNum":0,"RegionCategory":"","ArticlePicture":[],"TitleCN":null,"AbstractTextCN":null,"PMCID":"","EPubDate":null,"PubModel":null,"JCR":null,"JCRName":null,"Score":null,"Total":0}
Pub Date : 1988-12-01DOI: 10.1109/ARFTG.1988.323916
C. Gingras, R. Bosisio
The six-port junction has already proven to be very useful in small-signal measurements of multi-port networks. The aim of this paper is to demonstrate that it can also be used to determine equivalent non-linear circuit models of selected components, through measurements only at the fundamental frequency. The examples taken are packaged PIN and varactor diodes. Small-signal measurements are presented together with the plan to obtain optimized linear and non-linear models.
{"title":"Ion-Linear Measurements Using Six-Port Network Analyzer","authors":"C. Gingras, R. Bosisio","doi":"10.1109/ARFTG.1988.323916","DOIUrl":"https://doi.org/10.1109/ARFTG.1988.323916","url":null,"abstract":"The six-port junction has already proven to be very useful in small-signal measurements of multi-port networks. The aim of this paper is to demonstrate that it can also be used to determine equivalent non-linear circuit models of selected components, through measurements only at the fundamental frequency. The examples taken are packaged PIN and varactor diodes. Small-signal measurements are presented together with the plan to obtain optimized linear and non-linear models.","PeriodicalId":235867,"journal":{"name":"32nd ARFTG Conference Digest","volume":"49 1","pages":"0"},"PeriodicalIF":0.0,"publicationDate":"1988-12-01","publicationTypes":"Journal Article","fieldsOfStudy":null,"isOpenAccess":false,"openAccessPdf":"","citationCount":null,"resultStr":null,"platform":"Semanticscholar","paperid":"115901712","PeriodicalName":null,"FirstCategoryId":null,"ListUrlMain":null,"RegionNum":0,"RegionCategory":"","ArticlePicture":[],"TitleCN":null,"AbstractTextCN":null,"PMCID":"","EPubDate":null,"PubModel":null,"JCR":null,"JCRName":null,"Score":null,"Total":0}
Pub Date : 1988-12-01DOI: 10.1109/ARFTG.1988.323918
J. M. Cadwallader
Noise-figure measurement techniques at millimeter-wave frequencies are discussed and example results are shown at V-band (50 to 75 GHz) using an automated instrumentation set-up to cover the full WR-15 waveguide bandwidth. The approach shown involves extending available microwave noise-figure measurement instruments into the millimeter-wave region by the use of broadband millimeter-wave down-converters and frequency agile local oscillator sources. A developmental approach to full waveguide bandwidth noise-figure measurements in W-band (75 to 110 GHz) is also described which employes a double-conversion technique with fixed local oscillator at W-band followed by a broadband swept LO and down-converter at the microwave IF.
{"title":"Millimeter-Wave Noise-Figure Measurement Techniques & Results","authors":"J. M. Cadwallader","doi":"10.1109/ARFTG.1988.323918","DOIUrl":"https://doi.org/10.1109/ARFTG.1988.323918","url":null,"abstract":"Noise-figure measurement techniques at millimeter-wave frequencies are discussed and example results are shown at V-band (50 to 75 GHz) using an automated instrumentation set-up to cover the full WR-15 waveguide bandwidth. The approach shown involves extending available microwave noise-figure measurement instruments into the millimeter-wave region by the use of broadband millimeter-wave down-converters and frequency agile local oscillator sources. A developmental approach to full waveguide bandwidth noise-figure measurements in W-band (75 to 110 GHz) is also described which employes a double-conversion technique with fixed local oscillator at W-band followed by a broadband swept LO and down-converter at the microwave IF.","PeriodicalId":235867,"journal":{"name":"32nd ARFTG Conference Digest","volume":"35 1","pages":"0"},"PeriodicalIF":0.0,"publicationDate":"1988-12-01","publicationTypes":"Journal Article","fieldsOfStudy":null,"isOpenAccess":false,"openAccessPdf":"","citationCount":null,"resultStr":null,"platform":"Semanticscholar","paperid":"131408834","PeriodicalName":null,"FirstCategoryId":null,"ListUrlMain":null,"RegionNum":0,"RegionCategory":"","ArticlePicture":[],"TitleCN":null,"AbstractTextCN":null,"PMCID":"","EPubDate":null,"PubModel":null,"JCR":null,"JCRName":null,"Score":null,"Total":0}
Pub Date : 1988-12-01DOI: 10.1109/ARFTG.1988.323914
R. Gilmore, J. Gerber, M. Eron
This paper illustrates the methodology for designing a nonlinear circuit using state-of-the-art CAD tools applied to several examples of FET circuits. The parameters needed for characterization of the (nonlinear) FET are extracted using RoMPE, a parameter extraction program based on gradient optimizers which have been enhanced by adjoint sensitivity calculations to provide exact gradients. Measured DC and small-signal S-parameter data are simultaneously fitted to a modified Materka FET model. Simultaneous fitting of the data to both DC and AC parameters is essential to preserve the implicit dependence of the small-signal model parameters upon the bias. Microwave Harmonica, a harmonic-balance based nonlinear circuit simulator, is then used with the model parameters to design a single-gate FET mixer circuit. The use of an optimizer to improve the conversion gain and spectral purity of the mixer, and its use in the design of oscillator circuits is also demonstrated.
{"title":"Design Methodologies for Nonlinear Circuit Simulation and Optimization","authors":"R. Gilmore, J. Gerber, M. Eron","doi":"10.1109/ARFTG.1988.323914","DOIUrl":"https://doi.org/10.1109/ARFTG.1988.323914","url":null,"abstract":"This paper illustrates the methodology for designing a nonlinear circuit using state-of-the-art CAD tools applied to several examples of FET circuits. The parameters needed for characterization of the (nonlinear) FET are extracted using RoMPE, a parameter extraction program based on gradient optimizers which have been enhanced by adjoint sensitivity calculations to provide exact gradients. Measured DC and small-signal S-parameter data are simultaneously fitted to a modified Materka FET model. Simultaneous fitting of the data to both DC and AC parameters is essential to preserve the implicit dependence of the small-signal model parameters upon the bias. Microwave Harmonica, a harmonic-balance based nonlinear circuit simulator, is then used with the model parameters to design a single-gate FET mixer circuit. The use of an optimizer to improve the conversion gain and spectral purity of the mixer, and its use in the design of oscillator circuits is also demonstrated.","PeriodicalId":235867,"journal":{"name":"32nd ARFTG Conference Digest","volume":"13 1","pages":"0"},"PeriodicalIF":0.0,"publicationDate":"1988-12-01","publicationTypes":"Journal Article","fieldsOfStudy":null,"isOpenAccess":false,"openAccessPdf":"","citationCount":null,"resultStr":null,"platform":"Semanticscholar","paperid":"131140115","PeriodicalName":null,"FirstCategoryId":null,"ListUrlMain":null,"RegionNum":0,"RegionCategory":"","ArticlePicture":[],"TitleCN":null,"AbstractTextCN":null,"PMCID":"","EPubDate":null,"PubModel":null,"JCR":null,"JCRName":null,"Score":null,"Total":0}
Pub Date : 1988-12-01DOI: 10.1109/ARFTG.1988.323925
Paul E. Jeroma, G. Martin
A method is presented for changing the positions of the reference planes for automatic network analyzer (ANA) measurements using a single set of calibration standards and the THRU-REFLECT-LINE (TRL) calibration procedure. The reference planes can be set from either the THRU or REFLECT calibration standard. However, if the THRU is used, the reference planes can also be set in various positions by changing the defined delay of the THRU and calibrating. The reference planes can be moved symmetrically from the center of the THRU toward the edges of the THRU with only one set of calibration standards. A simple experiment is presented to show the accuracy of setting the reference planes using this method. Test results show that reference planes can be set to within a 5 micron window on a 150 micron transmission line.
{"title":"Moving Reference Planes for On-Wafer Measurements Using The TRL Calibration Technique","authors":"Paul E. Jeroma, G. Martin","doi":"10.1109/ARFTG.1988.323925","DOIUrl":"https://doi.org/10.1109/ARFTG.1988.323925","url":null,"abstract":"A method is presented for changing the positions of the reference planes for automatic network analyzer (ANA) measurements using a single set of calibration standards and the THRU-REFLECT-LINE (TRL) calibration procedure. The reference planes can be set from either the THRU or REFLECT calibration standard. However, if the THRU is used, the reference planes can also be set in various positions by changing the defined delay of the THRU and calibrating. The reference planes can be moved symmetrically from the center of the THRU toward the edges of the THRU with only one set of calibration standards. A simple experiment is presented to show the accuracy of setting the reference planes using this method. Test results show that reference planes can be set to within a 5 micron window on a 150 micron transmission line.","PeriodicalId":235867,"journal":{"name":"32nd ARFTG Conference Digest","volume":"3 1","pages":"0"},"PeriodicalIF":0.0,"publicationDate":"1988-12-01","publicationTypes":"Journal Article","fieldsOfStudy":null,"isOpenAccess":false,"openAccessPdf":"","citationCount":null,"resultStr":null,"platform":"Semanticscholar","paperid":"114414029","PeriodicalName":null,"FirstCategoryId":null,"ListUrlMain":null,"RegionNum":0,"RegionCategory":"","ArticlePicture":[],"TitleCN":null,"AbstractTextCN":null,"PMCID":"","EPubDate":null,"PubModel":null,"JCR":null,"JCRName":null,"Score":null,"Total":0}
Pub Date : 1988-12-01DOI: 10.1109/ARFTG.1988.323922
L. Lerner, C. McGuire
This paper presents an automated methodology for the measurement and non-linear modeling of GaAs MESFETS. The FET model extraction is based on the optimization of bias-dependent representations of DC-IV and linear equivalent circuit element values to small signal measurements of DC-IV and s-parameter data acquired over a wide range of operating bias, Vgs and Vds. The data acquisition is fully automated through the use of ANACAT(TM) and is collected and managed in an open ASCII format. The single set of measurement data serves as the input to the extraction of several popular bias-dependent FET models using Xtract(TM), a fully interactive FET model extraction tool developed by EEsof. The Xtract model file output may be used directly by Touchstone(TM), Libra(TM), or Microwave Spice(TM) to simulate the large-signal performance of the device in addition to serving as a master linear model for replicating linear small signal s-parameters at any operating point within the device's useful range. As an example, the paper presents the model extraction for a half-micron TriQuint GaAs MESFET. The example discusses the fully automated FET measurement and the subsequent extraction of three bias-dependent models. Both linear and non-linear model simulations are compared with actual measurements of the same device.
{"title":"An Automated Methodology for the Measurement and Non-Linear Model Parameter Extraction of GaAs MESFETs","authors":"L. Lerner, C. McGuire","doi":"10.1109/ARFTG.1988.323922","DOIUrl":"https://doi.org/10.1109/ARFTG.1988.323922","url":null,"abstract":"This paper presents an automated methodology for the measurement and non-linear modeling of GaAs MESFETS. The FET model extraction is based on the optimization of bias-dependent representations of DC-IV and linear equivalent circuit element values to small signal measurements of DC-IV and s-parameter data acquired over a wide range of operating bias, Vgs and Vds. The data acquisition is fully automated through the use of ANACAT(TM) and is collected and managed in an open ASCII format. The single set of measurement data serves as the input to the extraction of several popular bias-dependent FET models using Xtract(TM), a fully interactive FET model extraction tool developed by EEsof. The Xtract model file output may be used directly by Touchstone(TM), Libra(TM), or Microwave Spice(TM) to simulate the large-signal performance of the device in addition to serving as a master linear model for replicating linear small signal s-parameters at any operating point within the device's useful range. As an example, the paper presents the model extraction for a half-micron TriQuint GaAs MESFET. The example discusses the fully automated FET measurement and the subsequent extraction of three bias-dependent models. Both linear and non-linear model simulations are compared with actual measurements of the same device.","PeriodicalId":235867,"journal":{"name":"32nd ARFTG Conference Digest","volume":"4 1","pages":"0"},"PeriodicalIF":0.0,"publicationDate":"1988-12-01","publicationTypes":"Journal Article","fieldsOfStudy":null,"isOpenAccess":false,"openAccessPdf":"","citationCount":null,"resultStr":null,"platform":"Semanticscholar","paperid":"126090498","PeriodicalName":null,"FirstCategoryId":null,"ListUrlMain":null,"RegionNum":0,"RegionCategory":"","ArticlePicture":[],"TitleCN":null,"AbstractTextCN":null,"PMCID":"","EPubDate":null,"PubModel":null,"JCR":null,"JCRName":null,"Score":null,"Total":0}
Pub Date : 1988-12-01DOI: 10.1109/ARFTG.1988.323920
C. Beccari, A. Ferrero, U. Pisani
We present a technique which resorts to the time domain capabilities of a vector network analyzer and to the network synthesia tools, in order to perform an in-fixture calibration of the S-parameter measurement system directly to the ports of the device under test. The effects of the customer's non ideal fixtures can be removed without requiring the insertion of standard components or particular loads, which can affect the calibration efectiveness. The inaccuracies due to the precision of the actual loads and to the connection repeatability are also avoided. Some simulation reeults demonstrate the very good capability of the technique. Experimental tests were also carried out on an actual microstrip transistor fixture, showing a very satisfactoty launcher modeling and de-embedding.
{"title":"In-Fixture Calibration of an S-Parameter Measuring System by Means of Time Domain Reflectometry","authors":"C. Beccari, A. Ferrero, U. Pisani","doi":"10.1109/ARFTG.1988.323920","DOIUrl":"https://doi.org/10.1109/ARFTG.1988.323920","url":null,"abstract":"We present a technique which resorts to the time domain capabilities of a vector network analyzer and to the network synthesia tools, in order to perform an in-fixture calibration of the S-parameter measurement system directly to the ports of the device under test. The effects of the customer's non ideal fixtures can be removed without requiring the insertion of standard components or particular loads, which can affect the calibration efectiveness. The inaccuracies due to the precision of the actual loads and to the connection repeatability are also avoided. Some simulation reeults demonstrate the very good capability of the technique. Experimental tests were also carried out on an actual microstrip transistor fixture, showing a very satisfactoty launcher modeling and de-embedding.","PeriodicalId":235867,"journal":{"name":"32nd ARFTG Conference Digest","volume":"53 1","pages":"0"},"PeriodicalIF":0.0,"publicationDate":"1988-12-01","publicationTypes":"Journal Article","fieldsOfStudy":null,"isOpenAccess":false,"openAccessPdf":"","citationCount":null,"resultStr":null,"platform":"Semanticscholar","paperid":"132524348","PeriodicalName":null,"FirstCategoryId":null,"ListUrlMain":null,"RegionNum":0,"RegionCategory":"","ArticlePicture":[],"TitleCN":null,"AbstractTextCN":null,"PMCID":"","EPubDate":null,"PubModel":null,"JCR":null,"JCRName":null,"Score":null,"Total":0}
Pub Date : 1988-12-01DOI: 10.1109/ARFTG.1988.323913
D. Root, B. Hughes
Principles of nonlinear active device modeling for circuit simulation in SPICE and harmonic balance programs are presented, motivated by an examination of three common problems of standard nonlinear GaAs MESFET models. The first problem is that simulations done in large signal analysis don't fit the measured imaginary parts of the Y-parameters versus bias. On the other hand, good fits to the measurements can be obtained in linear (small signal) analysis. The solution is to use nonlinear Voltage Controlled Charge Source (VCQS) elements in place of two-terminal capacitors in the nonlinear model. A VCQS is a reactive analogue of the familiar voltage controlled current source. This improves the accuracy of the large signal reactive FET model. Linearizing the resulting equations produces an accurate small signal model, consistent with the large signal simulations in the appropriate limit. Linearizing a VCQS element produces another unfamiliar quantity, the transcapacitance. A transcapacitance is a reactive analogue of a transconductance. The second problem is that large signal FET models do not simulate time delays of the FET. The transcapacitance associated with a VCQS element between the source and the drain provides a reactive output current proportional to the time derivative of the gate voltage. This current approximates the effect of the delay.
{"title":"Principles of Nonlinear Active Device Modeling for Circuit Simulation","authors":"D. Root, B. Hughes","doi":"10.1109/ARFTG.1988.323913","DOIUrl":"https://doi.org/10.1109/ARFTG.1988.323913","url":null,"abstract":"Principles of nonlinear active device modeling for circuit simulation in SPICE and harmonic balance programs are presented, motivated by an examination of three common problems of standard nonlinear GaAs MESFET models. The first problem is that simulations done in large signal analysis don't fit the measured imaginary parts of the Y-parameters versus bias. On the other hand, good fits to the measurements can be obtained in linear (small signal) analysis. The solution is to use nonlinear Voltage Controlled Charge Source (VCQS) elements in place of two-terminal capacitors in the nonlinear model. A VCQS is a reactive analogue of the familiar voltage controlled current source. This improves the accuracy of the large signal reactive FET model. Linearizing the resulting equations produces an accurate small signal model, consistent with the large signal simulations in the appropriate limit. Linearizing a VCQS element produces another unfamiliar quantity, the transcapacitance. A transcapacitance is a reactive analogue of a transconductance. The second problem is that large signal FET models do not simulate time delays of the FET. The transcapacitance associated with a VCQS element between the source and the drain provides a reactive output current proportional to the time derivative of the gate voltage. This current approximates the effect of the delay.","PeriodicalId":235867,"journal":{"name":"32nd ARFTG Conference Digest","volume":"1 1","pages":"0"},"PeriodicalIF":0.0,"publicationDate":"1988-12-01","publicationTypes":"Journal Article","fieldsOfStudy":null,"isOpenAccess":false,"openAccessPdf":"","citationCount":null,"resultStr":null,"platform":"Semanticscholar","paperid":"128637334","PeriodicalName":null,"FirstCategoryId":null,"ListUrlMain":null,"RegionNum":0,"RegionCategory":"","ArticlePicture":[],"TitleCN":null,"AbstractTextCN":null,"PMCID":"","EPubDate":null,"PubModel":null,"JCR":null,"JCRName":null,"Score":null,"Total":0}
京公网安备 11010802042870号
京ICP备2023020795号-1
扫码关注我们
求助内容：
应助结果提醒方式：