Large signal characterization of microwave GaAs MESFETs

Large Signal modelling of GaAs MESFETs has often been based on the device material and electrical parameters. This approach, while helping in elucidating the physics of the device, does not help the device user very much. There is the problem of modelling and computation complexity, and of simulatio...

Full description

Bibliographic Details
Main Author: Amaeshi, Lawrence Lemchukwu Nnanyelugo
Published: University of Surrey 1988
Subjects:
Online Access:https://ethos.bl.uk/OrderDetails.do?uin=uk.bl.ethos.232883
id ndltd-bl.uk-oai-ethos.bl.uk-232883
record_format oai_dc
spelling ndltd-bl.uk-oai-ethos.bl.uk-2328832018-09-11T03:20:01ZLarge signal characterization of microwave GaAs MESFETsAmaeshi, Lawrence Lemchukwu Nnanyelugo1988Large Signal modelling of GaAs MESFETs has often been based on the device material and electrical parameters. This approach, while helping in elucidating the physics of the device, does not help the device user very much. There is the problem of modelling and computation complexity, and of simulation time. Furthermore predictions based on such models may not be consistent with practical reality. A real-time large Rf signal characterization of the device, will help in the understanding of the device behaviour under Large Signal drive, and would also yield valuable results/information, useful to the device designer and user, especially in Large Signal applications. A large Rf signal characterization of GaAs MESFETs, employing Large Signal S-parameter (LSSP), and waveform distortion analysis techniques, is carried out. LSSP design is a natural extension of the SSSP approach where the LSSPs are known. And the LSSP design approach is simplified if the LSSPs are determined easily. Waveform distortions affect the device performance. A LSSP measurement system, (also applicable to SSSP measurements), including an uncomplicated, direct deembedding technique is developed. A direct technique of measuring the current and voltage waveforms of the microwave signals, at the device terminals, is also developed. Measurements of the LSSPs show that only the input parameters: S21 and S11 vary with the Rf. The results are explained against reported trends of variation. The non-linear elements are identified, and a subsequent Large Signal Model (LSM) of the DUTs developed and verified. It is demonstrated that LSMs cannot be generalised. However a systematic approach of determining the LSM of a given device is given. An improved model of the transconductance, Gm, in terms of the S-parameters, and a method to determine the LSSP from small signal parameters are developed and verified. The optimum incident Rf to determine the LSSPs at a given bias is given. The flow of forward conduction,IF is known to damage, by burn-out, the DUT. A limiting resistor was included in the gate external circuit to limit this effect, when large enough Rf was employed. The interaction of the IF with this circuit is investigated, and the self-limiting actions explained. The flow of Is. is found to degrade the output performance and device power added efficiency also. The waveform distortions are investigated, the main causes - the nonlinear elements, and the manner in which they affect the distortions are explored. The non-linearity in the Gm is shown to be the main cause of the output waveform distortion, especially before the onset of forward conduction by the gate Schottky diode. While the forward conduction If and the non-linearity in the depletion capacitance are responsible for the input waveform distortion, hence are secondary causes of output waveform distortions. In particular, the flow of If, due to large Vgs > 0, causes the saturation of the drain voltage, hence the output power. But the waveforms, were in particular insensitive to the output conductance. The results reaffirm the LSM developed. Finally future work is discussed.530.41Semiconductor electronicsUniversity of Surreyhttps://ethos.bl.uk/OrderDetails.do?uin=uk.bl.ethos.232883http://epubs.surrey.ac.uk/843717/Electronic Thesis or Dissertation
collection NDLTD
sources NDLTD
topic 530.41
Semiconductor electronics
spellingShingle 530.41
Semiconductor electronics
Amaeshi, Lawrence Lemchukwu Nnanyelugo
Large signal characterization of microwave GaAs MESFETs
description Large Signal modelling of GaAs MESFETs has often been based on the device material and electrical parameters. This approach, while helping in elucidating the physics of the device, does not help the device user very much. There is the problem of modelling and computation complexity, and of simulation time. Furthermore predictions based on such models may not be consistent with practical reality. A real-time large Rf signal characterization of the device, will help in the understanding of the device behaviour under Large Signal drive, and would also yield valuable results/information, useful to the device designer and user, especially in Large Signal applications. A large Rf signal characterization of GaAs MESFETs, employing Large Signal S-parameter (LSSP), and waveform distortion analysis techniques, is carried out. LSSP design is a natural extension of the SSSP approach where the LSSPs are known. And the LSSP design approach is simplified if the LSSPs are determined easily. Waveform distortions affect the device performance. A LSSP measurement system, (also applicable to SSSP measurements), including an uncomplicated, direct deembedding technique is developed. A direct technique of measuring the current and voltage waveforms of the microwave signals, at the device terminals, is also developed. Measurements of the LSSPs show that only the input parameters: S21 and S11 vary with the Rf. The results are explained against reported trends of variation. The non-linear elements are identified, and a subsequent Large Signal Model (LSM) of the DUTs developed and verified. It is demonstrated that LSMs cannot be generalised. However a systematic approach of determining the LSM of a given device is given. An improved model of the transconductance, Gm, in terms of the S-parameters, and a method to determine the LSSP from small signal parameters are developed and verified. The optimum incident Rf to determine the LSSPs at a given bias is given. The flow of forward conduction,IF is known to damage, by burn-out, the DUT. A limiting resistor was included in the gate external circuit to limit this effect, when large enough Rf was employed. The interaction of the IF with this circuit is investigated, and the self-limiting actions explained. The flow of Is. is found to degrade the output performance and device power added efficiency also. The waveform distortions are investigated, the main causes - the nonlinear elements, and the manner in which they affect the distortions are explored. The non-linearity in the Gm is shown to be the main cause of the output waveform distortion, especially before the onset of forward conduction by the gate Schottky diode. While the forward conduction If and the non-linearity in the depletion capacitance are responsible for the input waveform distortion, hence are secondary causes of output waveform distortions. In particular, the flow of If, due to large Vgs > 0, causes the saturation of the drain voltage, hence the output power. But the waveforms, were in particular insensitive to the output conductance. The results reaffirm the LSM developed. Finally future work is discussed.
author Amaeshi, Lawrence Lemchukwu Nnanyelugo
author_facet Amaeshi, Lawrence Lemchukwu Nnanyelugo
author_sort Amaeshi, Lawrence Lemchukwu Nnanyelugo
title Large signal characterization of microwave GaAs MESFETs
title_short Large signal characterization of microwave GaAs MESFETs
title_full Large signal characterization of microwave GaAs MESFETs
title_fullStr Large signal characterization of microwave GaAs MESFETs
title_full_unstemmed Large signal characterization of microwave GaAs MESFETs
title_sort large signal characterization of microwave gaas mesfets
publisher University of Surrey
publishDate 1988
url https://ethos.bl.uk/OrderDetails.do?uin=uk.bl.ethos.232883
work_keys_str_mv AT amaeshilawrencelemchukwunnanyelugo largesignalcharacterizationofmicrowavegaasmesfets
_version_ 1718732099212541952