J.R. Richardson

Improved calibration and measurement of the scattering parameters of microwave integrated circuits

A novel procedure for the calibration of microwave integrated circuit test fixtures, based on a generalization of the through-reflect-line (TRL) algorithm, is presented. Its advantages compared with previous methods, namely bandwidth of validity and standards availability, are discussed. The approach is verified through the characterization of a particular microstrip verification standard using both the generalized TRL and precision 7-mm calibration techniques. Comparison of the results obtained from these schemes indicates that both the effective directivity and the source/load match are better than 30 dB. >

A large-signal physical MESFET model for computer-aided design and its applications

A quasi-static, large-signal MESFET circuit model is presented. It is based on a comprehensive quasi-two-dimensional, semiclassical, physical device simulation, and its unique formulation and efficiency make it suitable for the computer-aided design of nonlinear MESFET subsystems. Using this approach the semiconductor equations are reduced to a consistent one-dimensional approximation requiring substantially less computing resources than a full two-dimensional simulation. CPU time is typically reduced by a factor of 1000. A single/two-tone harmonic balance analysis procedure which uses the describing frequency concept is also developed and combined with the MESFET model. Numerical load-pull contours as well as intermodulation distortion contours have been simulated; their comparison with measured results validates the approach taken. >

On-wafer measurement uncertainty for 3-terminal active millimetre-wave devices

An experimentally derived frequency-dependent residual error budget to 60 GHz in well-calibrated state-of-the-art measurement systems for millimeter-wave devices is presented. The budget identifies errors resulting from the imperfections in the calibration standards, both on- and off-wafer, probe-wafer interface, and instrumentation. Root-mean-square errors are assumed uncorrelated and are added to arrive at a worst-case scenario. The impact in terms of level of significance and sensitivity of measurement uncertainty on small-signal modeling is shown. The effects of measurement uncertainty on extracted device parameters can be minimized by selecting a suitable frequency range and appropriate calibration structures.<<ETX>>

Efficient computation of the steady-state response of periodic nonlinear microwave circuits using a convolution-based sample-balance technique

The authors describe an efficient and robust approach to the computation of the steady-state response of periodic nonlinear microwave circuits. The problem of solving a set of differential equations is converted into that of solving a system of nonlinear algebraic equations using a technique called convolution-based sample balance. Although exact in all cases for which harmonic-balance techniques are exact, this technique does not require the use of discrete Fourier transforms, and calculating the Jacobian is straightforward. For the solution of the resulting system of nonlinear equations, an efficient and yet robust algorithm has been developed. In the examples given, savings in computational effort of over 85% are reported when this algorithm is compared with Newtons method. >

An Automated Measurement Technique for Measuring Amplifier Load-Pull and Verifying Large-Signal Device Models

A physical model has been used to characterise the large-signal behaviour of a GaAs MESFET terminated with a wide range of load impedances and a new, automated, load-pull system developed to make measurements in order to verify the performance of the model. The heart of the measurement system is a motorised slide-screw tuner which operates over 2-18 GHz with reflection coefficients greater than 0.93 above 3 GHz.

Characterization of thermal effects on microwave transistor performance using an efficient physical model

An efficient physical model capable of predicting the effects of self-heating as well as ambient temperature on the DC and microwave characteristics of MESFETs is outlined. Excellent agreement has been obtained with measured characteristics for a range of gate length devices for both DC and microwave characteristics. The model provides great insight into device behavior and is capable of relating physical structure to device performance while still being efficient enough to run on a personal computer.<<ETX>>

Direct use of a MESFET physical model, in nonlinear CAD

The authors present a CAD (computer-aided design) tool which combines a novel quasi-two-dimensional MESFET physical model with an efficient nonlinear circuit analysis technique. Device technological parameters (dimensions, doping profile, etc.) can be directly related to electrical performance with this tool, which can therefore be used in nonlinear circuit yield study and optimization. An excellent agreement between simulated and measured results was obtained.<<ETX>>

Millimetre-wave device modelling differences in microstrip and coplanar waveguide

The measurement of an active device produces results which depend on the transmission medium in which the device is embedded. These differences, insofar as a particular device is concerned, are related solely to its extrinsic elements, the intrinsic device being the same in all cases. Results are presented for on-wafer measurement and 2-D simulation of a specially designed active device and its associated transmission structure in both microstrip and coplanar waveguide. The measured and simulated results are obtained for pseudomorphic HEMTs (high electron mobility transistors) with a 0.2- mu m gate length arranged in a 6- mu m*15- mu m configuration with typical extrinsic f/sub t/s of 70 GHz. The dual approach considered here has the advantage of enabling a model for the intrinsic device to be validated and the extrinsic model elements for both structures to be obtained.<<ETX>>

The design of microwave subsystems based on a device physical model

A numerical simulation to characterize both the DC and RF behaviour of GaAs MESFETs is described. It is based on coupling the four semiclassical semiconductor equations with analytical expressions for the channel and solving these using a forward-difference scheme. In the approach taken, the semiconductor equations are reduced to a one-dimensional approximation using the expressions for the channel, requiring substantially less computing resources than a full two-dimensional simulation. CPU time is typically reduced by a factor of 1000. DC and RF results for a 0.5- mu m MESFET are presented.<<ETX>>

The effects of process variations on microstrip wafer calibration

The coplanar-to-microstrip transition part of on-wafer microstrip line-reflect-line (LRL) calibration structures has been modeled and measured to determine calibration repeatability due to variations in this structure. On-wafer small-signal device characterization is dependent on the quality and repeatability of the calibration, which depends on the calibration structures. When these are fabricated on the same wafer as the devices to be characterized they are also affected by process variations, the most important of which, when using LRL calibration structures in microstrip, is the variation in the thickness of the wafers. The LRL calibration solution is modified to include the process variation in the microstrip calibration structures in terms of propagation constant and effective line length. Both probe contact and probe position repeatability are identified in the experimental results as they are inevitably measured with process variations.<<ETX>>