A hands-on approach to avoid scientific debates and confusions about surface waves is presented. The reflection coefficient function of a TM wave incident on a flat boundary is studied as a function of the medium parameters in the most general way. Important questions are raised about some misused terminologies involving surface waves.

Waveguides with arbitrary cross-section and periodic loading are analyzed using an eigenmode projection technique. The analysis procedure is based on expanding the fields of the given structure in terms of the solenoidal and irrotational eigenmodes of a canonical waveguide with cross-section enclosing that of the original waveguide. Floquet harmonics are incorporated in the canonical eigenmodes to account for the periodicity. Results obtained using the proposed approach and other techniques are compared and show excellent agreement. The proposed approach has many advantages, namely that the problem reduces to a simple eigenvalue problem and the field distribution is directly given as weighted expansion of the canonical eigenmodes.

An analytical solution is presented for the scattering from multiple circular cylinders. The solution utilizes the spectral, plane-wave expansion of the fields in an iterative manner and accounts for all the multiple interactions between cylinders. The technique has direct applicability to the modeling of substrate-integrated waveguide (SIW) structures where many cylindrical vias are used to generate guided waves between two conducting planes. The analytical technique provides a robust method for calculating propagation properties, leakage, and coupling.

This paper presents an analytical method for the efficient full-wave modeling of bi-dimensional structures by combining the results of purely E- and H-plane analysis techniques. It is based on frequency and boundary-condition transformations, and is used to compute the fields and the Generalized Admittance Matrix of the structure. A complex 3D geometry including a bi-dimensional block is analyzed, in order to illustrate the advantages of the novel full-wave analysis technique.

We propose a methodology for developing EM-based polynomial surrogate models exploiting the multinomial theorem. Our methodology is compared against four EM surrogate modeling techniques: response surface modeling, support vector machines, generalized regression neural networks, and Kriging. Results show that the proposed polynomial surrogate modeling approach has the best performance among these techniques when using a very small amount of learning base points. The proposed methodology is illustrated by developing a surrogate model for a T-slot PIFA antenna simulated on a commercially available 3D FEM simulator.

A fast algorithm is presented for statistical anal- ysis of large microwave and high-speed circuits with multiple stochastic parameters. Using the proposed algorithm, a set of local reduced-order parameterized circuits are derived based on adaptive frequency sampling and implicit multi-moment matching projection techniques. The local models preserve the stochastic parameters as symbolic quantities. As a result, stochas- tic response of the circuit can be obtained by simulating the local reduced models instead of the original large system leading to significant reduction in the computational cost compared to traditional Monte-Carlo techniques.

In this paper, a low-cost procedure for dimension scaling and tuning of microwave filters is proposed. Our approach relies on an inverse model that outputs the filter dimensions for the required center frequency and bandwidth. The model is constructed using several reference designs obtained using surrogate-assisted optimization exploiting a response feature methodology. The initial approximation of the scaled filter design is subsequently corrected using a fast feature-based tuning procedure. The proposed methodology is demonstrated using a fifth-order Chebyshev bandpass filter for the tuning range of 10 GHz to 12 GHz (center frequency) and 6 % to 12 % (relative bandwidth).

In this paper, a novel polynomial chaos approach for the fast uncertainty analysis of multi-walled carbon nanotube interconnect networks is proposed. The key feature of this approach is the development of a decoupled sensitivity analysis methodology to intelligently identify and prune the least impactful random dimensions from the original random space. This dimension reduction strategy allows the reliable modeling of the full-dimensional uncertainty in interconnect networks at the cost of evaluating only a fraction of the full-blown PC coefficients. The validity of the proposed methodology is demonstrated using a numerical example.

Existing explicit and unconditionally stable FDTD methods rely on a global eigenvalue solution to find the unstable modes that cannot be stably simulated by the given time step. In this paper, we develop a fast explicit and unconditionally stable FDTD method requiring no eigenvalue solutions. In this method, we find the relationship between the unstable modes and the fine meshes, and use this relationship to directly identify the source of instability. We then upfront eradicate the source of instability from the numerical system before performing an explicit time marching. The resultant simulation is absolutely stable for the given time step irrespective of how large it is. If the time step is chosen based on accuracy, the accuracy of the proposed method is also guaranteed. Numerical experiments have demonstrated a significant speedup of the proposed method over the conventional FDTD as well as state-of-the-art explicit and unconditionally stable methods.

Several types of absorbing boundary conditions and simple matched absorbers, employed for mesh truncation in time and frequency domain numerical techniques (FDTD, FEM) work well for normally incident waves. However, their performance at other angles of incidence is often deemed unacceptable, motivating their replacement by perfectly matched layers. We propose a medium that can collimate incident waves, producing almost normally incident plane waves that can be readily absorbed even by a low order absorbing boundary condition such as Mur's. This perfectly matched layer collimator (PMLC) is a lossless, uniaxially anisotropic medium; its numerical implementation is much simpler than that of the conventional PML. Numerical results demonstrate the effectiveness of this medium as a means of rendering the performance of a first order absorbing boundary condition comparable to that of a PML.

A novel surface integral equation(IE) formulation for magneto-quasi-static analysis of current flow in two-dimensional conductors situated in lossy layered substrates is proposed. Traditional approach is based on the volume IE formulated with respect to the unknown current distribution in the conductors’ cross-section. In this work, we transform the volume IE to a surface IE via representation of the volumetric current density as a superposition of the waves emanating from the conductors’ boundaries. The resulting surface IE features a single unknown surface current density unlike the traditional surface IEs which require both electric and magnetic surface current densities. Also, the resultant single-source surface IE formulation involves only the electric field Green’s functions instead of both electric and magnetic field Green’s functions as required by the traditional surface IE formulations. The latter property makes the proposed IE formulation particularly suitable for analysis of current flow in conductors embedded in lossy layered substrates.

This paper presents a novel technique for the numerical modeling of three-dimensional waveguide components. The component is preliminary segmented into boxed building blocks, and then the generalized admittance matrix of each building block is determined by the Boundary Integral-Resonant Mode Expansion (BI-RME) method. The kernel of the integral equation is the Green’s function of the box, which is efficiently calculated by using the Ewald technique. Finally, the admittance matrices of the building blocks are cascaded to retrieve the frequency response of the whole component. The analysis of a dual-band orthomode transducer is reported as an example, demonstrating the effectiveness of the proposed algorithm.

The development of transient gate resistance thermometry (T-GRT) is reported. It is a technique used to measure the transient self-heating of a FET’s gate metal. Demonstrations of T-GRT are presented at the wafer level on a GaAs pHEMT and an AlGaN/GaN-on-SiC HEMT. Dynamic self heating is monitored from hundreds of nanoseconds to hundreds of milliseconds. Preliminary finite-element simulations across a range of power dissipation levels agree with T-GRT to within 8% at, and beyond, 1 s after the applied drain pulse. Characterization of dynamic self heating and its application to pulsed applications such as radar are discussed.

A fast and simple method for the direct characterization of nonlinear charge functions of devices is presented. The input and output transistor ports are simultaneously excited through single-tone sources at different frequencies and calibrated large-signal waveforms are measured by means of a NVNA-based setup. Proper choice of the two frequencies guarantees an almost complete coverage of the voltages domain in a single and very fast measurement and allows the extraction of the charge functions by direct integration of currents in the frequency domain, since, contrary to other methods, the measured waveforms are both iso-thermal and iso-dynamic (i.e. at fixed charge trapping status). The method is validated by characterizing the gate charge function of a 5W 8x125μm GaN FET and implementing a simple table-based model of the transistor input port. Very good results are achieved by comparison with large-signal measurements under conditions different than the ones used for the characterization.

This paper presents a non-linear electro-thermal AlGaN/GaN model for CAD application with a new additive thermal-trap model to take into account the dynamic behavior of trap states and their associated temperature variation. The thermal-trap model is extracted through low-frequency small-signal CW S-parameter measurements and large-signal pulsed-RF measurements at different temperatures. This thermal-trap model allows accurately predicting the physical temperature activation of traps and also the thermal signature of traps. It is also demonstrated that extrapolation of trap model parameters by stretched multi-exponential function of drain current transient measurements during pulsed-RF excitations allows deeply improving the envelope simulations.

In this paper we present a new small signal multiport modelling approach for III-V High Electron Mobility Transistors (HEMT) that is capable for internal transistor analysation and optimization as well as scaleable in gate width and finger-number. The new model decomposes the planar transistor structure into single multiport elements that are separately described by electrical equivalent circuits and connected to each other over discrete ports. With this new modelling topology we only need to extract a couple of multiport elements to predict the correct behavior for a high amount of different planar transistor structures. This point gives the circuit designer a wide range of possibilities to analyze and optimize a given transistor structure according to special needs, like low-noise, input-output matching or cryogenic behavior on a computer based level.

In this paper we extract a small-signal model of a mm-wave deep-recess N-polar GaN MISHEMT exhibiting record 94GHz power density. We show that certain existing methods for extrinsic parasitic extraction cannot be easily employed because of the device design but that an existing cold-bias method provides accurate extraction. The small-signal model with pad layout parasitics is then validated with the gain measured at low input powers by a 94 GHz loadpull system. The factors impacting the measured gain are analyzed to show their origins and relative impact, giving guidance and predictions for future improvement.

The paper proposes an optimization procedure for the design of a common mode filter based on removable electromagnetic bandgap structures. Our algorithm exploits auxiliary response surface approximation models. As demonstrated, it allows us to quickly refine the preliminary analytical design and precisely achieve the required filtering frequency and to maximize the filter bandwidth at the level of a full-wave EM simulation model of the structure of interest.

This paper proposes a novel parametric modeling technique for electromagnetic (EM) based multiphysics analysis of microwave passive components. Multiphysics parameters usually affect the EM performance by indirectly influencing the geometrical variables, such as thermal effect causing an expansion in the geometrical parameters, and stress inducing the physical deformation. In the proposed technique, the input classification and correlating mapping is introduced to transform the multiphysics input parameters into geometrical input parameters. Further, the combined neural network and transfer function technique (neuro-TF) is used to model the EM responses w.r.t. the transformed geometrical variables. The model obtained using the proposed technique can achieve good accuracy with low complexity of neural networks, and further can be used in the high-level design. One tunable evanescent mode cavity filter example is used to demonstrate the validity of this technique.

This paper presents a powerful and general parallel artificial neural network training technique with parallel computing on cluster systems. Large numbers of training samples are distributed to multiple computers on a cluster system to achieve a high speed-up for training. The method is evaluated with respect to two examples of an advanced dynamic nonlinear simulation model for GaN transistors where the training set is measured large-signal waveform data from an NVNA. For advanced models in a high dimensional space with large training sample size, the proposed approach is demonstrated to reduce the total training time by a factor of 35.

A novel methodology for miniaturization-oriented design of a class of wideband branch-line couplers is proposed. The initial design is chosen from a family of optimized circuits that feature a simplified two-section topology. Compact size of the coupler is attained by using quasi-periodic slow-wave structures instead of conventional lines. Our approach explicitly aims at circuit size reduction by adjusting the number of elements within the recurrent slow-wave structure and its designable parameters to reach the smallest coupler layout possible. This is achieved at a low computational cost by exploiting a surrogate-based optimization (SBO) process with the underlying model of the recurrent slow-wave structure composed of multiple response surface approximations (RSAs). The SBO scheme incorporates adaptively adjusted design specifications to converge in just two iterations. A rapid fine-tuning procedure is applied to account for T-junction effects omitted during the design process. Our methodology is illustrated through an example supported by experimental verification.

We demonstrate for the first time the use of the Circuit Envelope method with the solution of the fundamental semiconductor device equations. The Harmonic Balance method is the starting point for this work and used to illustrate how the spectrum of conduction and displacement current density distributions can be used to infer the sources of device non-linear conductance and capacitance. The class-A simulation of a 0.5um by 1mm gate periphery MESFET is used to demonstrate the simulation of spectral regrowth using a 16 QAM modulated signal with 1GHz carrier frequency and 2MHz bandwidth. The discrete Fourier transform of the complex time varying envelope waveforms of current density distribution, in band and adjacent to the spectral mask, are illustrated as a potential approach for troubleshoot adjacent channel performance issues.

We propose a novel numerical approach for the microwave circuit variability analysis through efficient physics-based simulation of devices operated in nonlinear conditions. The proposed technique allows for a direct link between nonlinear circuit performances and technological parameter variations, and is validated against a class A and a deep class AB power amplifier example.

Substrate network modeling of RF Power LDMOS devices are important for accurate modeling at higher frequencies. Substrate losses can account for a considerable amount of the losses in the device and directly affects the efficiency, which is one of the most critical performance criteria of a power amplifier. In this paper, an improved substrate network model for RF Power LDMOS devices is presented that can more accurately predict these losses and be of help in designing improved device structures. Nonlinear resistors representing a depleting drain to bulk junction as a function of drain to source bias are included in a physical way. It is shown that the model gives better agreement with s-parameters for varying LDD lengths and as a function of drain to source bias.

This paper investigates the influence of TSV noise coupling on nearby devices based on an extended 3D TSV circuit model. This model not only takes into account the complex RF field distributions in bulk Si, but also incorporates the anomalous TSV capacitance inversion behavior, which has been found to occur due to the presence of fixed charges in the backside passivation layer after wafer thinning. The extended 3D TSV circuit model is validated by the excellent agreement between the simulation results and experimental data, and demonstrates that the inversion behavior of TSV capacitance increases the noise coupling to adjacent devices mainly in the low frequency range. Furthermore, we show that noise mitigation techniques can be easily implemented in this 3D circuit model to predict the extent of noise coupling alleviation. Index Terms — TSV, 3D circuit model, inversion layer, inversion behavior of TSV capacitance, noise coupliing

Nonlinear charge-trapping observed in the electrical characteristics of GaN FETs can introduce distortion in GaN-based power amplifiers (PA), especially in supply-modulated (envelope tracking) transmitters. A measurement approach is developed for large signal characterization of GaN-based PAs operated with dynamic bias supplies for efficiency enhancement. A new pre-pulsing technique is introduced which forces the active device to operate in trapping and thermal states close to those found in the actual application. The characteristics obtained with this technique are shown to give an accurate description of the PA performance. The measured data are used for the direct computation of pre-distortion functions for the linearization of a 10-GHz Envelope Tracking (ET) 12-W GaN MMIC PA for amplitude-modulated pulsed radar transmitters. The demonstrated measurement method can be also exploited for the identification of PA behavioral models, which take into account trapping effects.

We provide a technique to generate and characterize a precision wideband millimeter-wave linear chirp signal. This linear chirp signal has 2 GHz of bandwidth, a center frequency of 30 GHz and repeats every 0.5 microseconds. The signal can be used as a reference waveform for characterizing the phase dispersion in wideband receivers. The generation of the signal is based on frequency multiplication of a linear chirp signal that is generated by an arbitrary waveform generator (AWG). The signal is characterized by an equivalent time sampling oscilloscope that is synchronized to the 12 GHz sampling clock of the AWG. The oscilloscope measurements are corrected for timing jitter, mismatch and impulse response.

This paper proposes a novel phase calibration approach to enable multi-harmonic modulated signal measurements using existing nonlinear network analyzers receivers configured in wideband mode. The envelopes around the harmonics are captured sequentially so that the dynamic range of the measurement system is not compromised. A reference signal generated by a comb generator is used to ensure phase coherency between the harmonics, avoiding the need to use a small frequency grid that typical solutions must rely on. To validate its ability for accurate measurements of multiharmonic modulated signals, the proposed NVNA-based method is tested against a state-of-the-art oscilloscope. The proposed approach is applied first on an 8-tone signal that covers 14 MHz then on a 5 MHz 1C-WCDMA signal. Both signals have three harmonics with the fundamental at 1 GHz. The phase accuracy of the measured signal is within ±1.5 and ±10 degree for the first and test second signals, respectively.

A high-resolution microwave non-contact current probe is described. Novel techniques are used in order to reduce the intrusive E-field pickup, which has restricted the performance and usefulness of such probes in the past. These include a probe structure that has built-in E-field screening and a high performance broadband differential amplifier. A rigorous testing procedure is described which asserts that the probe is truly responding almost entirely to the H-field. The probe operates up to 9GHz and has been used to profile the current distribution between the multiple cells of a power transistor at 3GHz, the first time such a direct measurement of this kind has been reported.

A composite right/left handed metamaterial with almost dispersion-free phase-shifting nonreciprocity over a wideband is proposed. The difference of refractive indices between forward and backward propagation directions is nearly constant with respect to the operational frequency. To achieve this condition, periodic insertion of asymmetric capacitive stubs for nonreciprocity and symmetric inductive stubs for negative permittivity into a normally magnetized ferrite-based CRLH microstrip line is proposed and designed. Numerical simulation and measurement results clearly show phase-shifting nonreciprocity approximately proportional to the operational frequency in the range over 5-7 GHz.

The blazing behavior of non-planar sawtooth gratings are mimicked with a planar metasurface. We show that with an appropriate design, an equivalent planar "blazed metasurface" can reflect most of the incident power back in the path of incidence, and with reduced power reflected in the specular direction. We then utilize such metasurface to create high rejection stopband in a waveguide, by lining the sidewall of the guide with this metasurface. The results are justified by theory, as well as simulations and measurements. The blazed metasurface may find various applications in microwaves up to optical frequencies, e.g. for stopband filters, Fabry-Perot resonator mirrors, or replacing corner reflectors.

In this paper, an artificial gradient-index lens based on a single layer fishnet metamaterial is presented. By introducing a variation of the geometric features in the unit cell, a phase variation between -180 and +180 degress can be achieved with only one fishnet layer, i.e. a single substrate. The relation between the geometric dimensions and dispersion properties is presented. Furthermore, the beam-scanning capability of the gradient-index fishnet lens is demonstrated at 27.5 GHz and its performance is evaluated by nearfield measurements making the presented single layer fishnet a good candidate for artificial lenses with low weight and fabrication costs.

The objective of this work is to investigate a hybrid deterministic and stochastic formulation for a quantitative statistical analysis of in-situ IC and electronics in complex and wave-chaotic enclosures. The key technical ingredients include: (ii) a stochastic dyadic Green’s function method for wave interaction with wave-chaotic media, which quantitatively describes the universal statistical property of chaotic systems through random matrix theory; (ii) a hybrid deterministic and stochastic formulation based on the optimized multi-trace integral equation domain decomposition method, which enables a statistical prediction of in-situ IC and electronics in complex wave-chaotic environments. The capability and benefits of the computational algorithms will be exploited, illustrated and validated through a variety of 3D product-level IC and electronic systems.

The filtering of unwanted modes that can propagate in oversized rectangular waveguides is addressed, focusing on longitudinal corrugations partially filled with an absorber. Lengthwise slots, in the middle of top and bottom waveguide walls, can indeed extract the power of some modes from the main waveguide with negligible insertion losses for the working dominant mode. Once some power is coupled through the slot into the corrugation, it travels toward the absorbing material, where it is damped. This behavior is studied comparing the performance of a single continuous slot, where various modes can propagate, with the one achieved by several shorter single-mode apertures. The excitation and absorption of modes in the corrugations is estimated by means of analytical expressions under a small coupling approximation. Mono-modal corrugations achieve better performances with respect to their overmoded alternative.

Multipactor effect can lead to RF components deterioration which could be fatal to RF systems in space communication payloads or in experimental fusion devices. To avoid such risk, oversized margins are used. Multipactor simulations are used to get voltage threshold predictions. Since the power breakdown depends on the Total Electron Emission Yield (TEEY) curve a sensitivity study has been made to determine which parameters of the TEEY properties are critical. An evaluation of multipactor threshold sensitivity to TEEY curve variations is realized and two critical parameters are found for parallel-plate geometry: first cross-over energy and the curve definition for incident electron energy between the first cross-over and the maximum curve energies. Six TEEY models and their accuracy to predict voltage threshold are compared. Electron emission experimental measurements have to be accurate on the first cross-over energy and TEEY model must respect this value to obtain coherent multipactor voltage threshold.

Horizontal Carbon-nanotube-based monopole antennas are studied for wireless on-chip and chip to chip interconnects. 3D EM simulation including dedicated CNTs model is first described. Then, applying a design approach already validated, we design horizontal CNTs monopole antenna. Technological processes are described. Theoretical results are presented and commented.

A nanomachined ferromagnetic and non-ferromagnetic superlattice approach is used for tunable radio frequency (RF) conductors, where the current flow profile through the volume of the conductor is magnetically manipulated and so is the RF resistance. The realized superlattice conductor consists of an alternating 20 layers of Cu/Ni and Cu/NiFe thin films with a thickness of 150 nm/25 nm. The multi-layer conductor has shown to have a reverse RF resistance performance as a function of frequency compared to its solid Cu counterpart with the same thickness, we call this artificial conductor as the “metaconductor”. In 10 MHz – 20 GHz frequency range, the metaconductors show an increased ohmic resistance in less than 10 GHz, but a reduced ohmic resistance in greater than 10 GHz range as “low loss conductors” in Ku and K bands. The RF frequency showing maximum resistance is magnetically tunable by 700 % (1 GHz ~ 7 GHz).

We report on the design and characterization of a 200 GHz resistive subharmonic mixer based on a single, multichannel CVD graphene field effect transistor (G-FET). The device has gate length 0.5 µm and width 2x40 µm. The integrated mixer circuit is implemented in coplanar waveguide (CPW) technology and realized on a 100 m thick high resistive silicon substrate. The measured mixer conversion loss (CL) is 34 plus minus 3 dB across 190-210 GHz band with 10 dBm local oscillator (LO) pumping power and the overall minimum CL gives 31.5 dB at 190 GHz.

Few-layer phosphorene MOSFETs with 0.3-µm-long gate and 15-nm-thick Al2O3 gate insulator was found to exhibit a forward-current cutoff frequency of 2 GHz and a maximum oscillation frequency of 8 GHz after de-embedding for the parasitic capacitance associated mainly with the relatively large probe pads. The gate lag and drain lag of the transistor was found to be on the order of 1 μs or less, which is consistent with the lack of hysteresis, carrier freeze-out or persistent photoconductivity in DC characteristics. These results confirm that the phosphorene MOSFET can be a viable microwave transistor for both small-signal and large-signal applications.

Carbon nanotube thin films deposited on sapphire substrates have been used to realize a microwave power sensor that operates at and above room temperature. The detector includes a power-sensitive resistor that has been incorporated into a voltage divider circuit. Using lock-in detection, experiments were performed with 915 MHz test signals that showed detection down to -45 dBm. A sensitivity of 0.36mV/mW was achieved with the device held at a temperature of 15°C. Additional experiments (which included static and pulsed current versus voltage measurements) indicate that the primary physical mechanism responsible for power detection near room temperature is Joule heating.

In this paper, novel microwave gas sensors based on graphene-loaded substrate integrated waveguide (SIW) cavity resonators are presented. Two SIW-based cavity resonators, a ring-slot resonator and a complementary split ring resonator (CSRR), are fabricated and coated with chemical vapor deposited (CVD)-grown graphene. The fabricated graphene contains a layer of polymethyl methacrylate (PMMA) on its top. The graphene sheets exhibit high sensitivity to various kinds of polar and non-polar gases. When polar gas contacts the graphene sheet, it will donate or receive electrons, thereby changing its conductance. The SIW cavities thus perform a resonant frequency shift from the perturbation of electron exchange. In the experiment, a frequency shift of 59 MHz and 167 MHz for the SIW ring-slot resonator and CSRR, respectively, can be observed after pure ammonia gas is injected into a closed chamber filled with air at standard atmospheric pressure and temperature.

An in-depth investigation of oscillation modes in free-running oscillators loaded with multi-resonance networks is presented. It focuses on the mechanisms leading to the coexistence of stable oscillation modes, which may give rise to uncertainty in the physical behaviour. The multiple periodic and quasi-periodic solutions are detected and related to the stability properties of the dc solution and each of the periodic modes. Two different types of Hopf-bifurcation loci enable a global understanding of the onset of qualitatively different steady-state solutions. The investigation is initially carried out with an analytical formulation and then extended to harmonic-balance simulations. The results have been experimentally validated through their application to a HEMT-based cross-coupled oscillator, at 0.65 and 2.4 GHz.

In this paper, a new technique is presented for the analysis of the transient dynamics of microwave oscillators. The technique makes use of a nonlinear admittance function that can be identified in commercial Harmonic Balance software. This function is included in a time-frequency domain equation governing the transient dynamics. The equation provides the growth rate function of the first harmonic amplitude, which allows an exhaustive analysis of the transient speed from the neighborhood of the dc solution to the oscillation establishment, with no need for a numerical integration, as in time domain or envelope-transient methods. The technique has been applied to predict the length of the transient towards the oscillating state of a FET oscillator at 5 GHz

Determining the origin and nature of the possible instabilities is key for an effective elimination of the unstable dynamics in multistage power amplifier design. In this work, a novel technique is proposed to provide a quantitative metrics that serves to locate and categorize the sensitive sections of the amplifier at which the unstable dynamics can be controlled and eliminated. The technique is based on an automatic Multiple-Input Multiple-Output (MIMO) frequency identification performed at different observation ports, followed by a residue analysis of critical poles. A three-stage amplifier exhibiting two common types of instabilities has been used to illustrate the complete approach.

In this paper, a channelized sideband distortion model is proposed to suppress the unwanted emission of Q-band millimeter wave (mmWave) transmitters in wideband contiguous carrier aggregation scenarios. By employing this model, the compensating bandwidth or center frequency can be agilely controlled. Experiments were conducted on a 40 GHz mmWave power amplifier to validate this idea. The satisfactory performance proved that the proposed model is a promising candidate for future 5G applications.

Transient thermal response provides rich information about the thermal behavior of advanced transistors and devices. We present the transient thermal modeling and analysis based on an instantaneous thermal imaging technology with transistor level spatial resolution for a high power GaN HEMT. The time responses show the clear separations of sources in temperature rises. At the transistor level or die level, failure or thermal irregularities can be determined in a short time scale with much smaller impact to the neighbor dies if characterized on a wafer. The differences in contact conditions on a vacuum chuck for wafer or die attach in a package can be separated from intrinsic thermal response of the die in different time scale.

Kelvin measurements of series inductance and resistance of an on-chip inductor at frequencies from 5 to 10 GHz is demonstrated using a CMOS process. The measurements require 3 DC voltage meters, 3 DC current sources, an AC signal source, and one high frequency probe. The resistance is lower and within 0.5 Ω of that from a calibrated measurement using a vector network analyzer (VNA), while the inductance is 10% (50 pH) lower. Due to the reduction of the variability of contacts, the range of series resistance variation measured using the proposed technique is 2 to 5 times lower than that using a VNA. 3-D electromagnetic simulations suggest that the inductance measured using the proposed Kelvin technique is more accurate due to the elimination of de-embedding errors.

An accurate phasor-sum method is proposed for the built-in detection of relative phase and amplitude between neighboring RF chains in phased array systems. The phasor sum of two RF signals provides a distinct amplitude-phase trigonometric relationship, from which the relative phase can be determined solely by the signal amplitudes. However, the amplitude distortion, attributed from the impedance mismatch or nonlinearity of detection circuits and devices, introduces considerable phase error. In our proposed method, the amplitude distortion error is first eliminated and then the phase error is significantly reduced. The measurement results of a 3-GHz four-chain RF array demonstrate that the root-mean-square phase error is less than 3.2° over the entire 0° to 180° range and the power error is less than 0.3 dB.

This paper presents an on-wafer calibration algorithm, with fixed probe position constraint, for scattering parameter (S-parameter) measurements without impedance-standard substrate (ISS). The fixed probe position feature is suitable for production tests of radio-frequency integrated circuits (RFICs), where the probe positions designed for the dimension of the device-under-test (DUT) are applied to the associated calibration standards as well, and the probes need not to be moved in X-Y directions during calibrations and measurements. Three on-chip standards are adapted including a transmission line (TL), a series resistor with offset line segment, and a shunt resistor with offset line segment, where the characteristics of the on-chip standards are solved using the self-calibration technique. To show the robustness of the proposed calibration algorithm, simulation studies with noise effects are conducted. Further experimental results are shown with verifications of an independent multiline thru-reflect-line (TRL) calibration data.