Rudi P. Paganelli

Accurate prediction of PHEMT intermodulation distortion using the nonlinear discrete convolution model

A general-purpose, technology-independent behavioral model is adopted for the intermodulation performance prediction of PHEMT devices. The model can be easily identified since its nonlinear functions are directly related to conventional DC and small-signal differential parameter measurements. Experimental results which confirm the model accuracy at high operating frequencies are provided in the paper.

Dynamic switching conversion for piezoelectric energy harvesting systems

The current advances in ultra-low power design let foresee great opportunities in energy harvesting platforms for self-powered systems. This paper presents a switching conversion scheme based on active control for harvesting energy with a higher efficiency than traditional approaches. The approach has been validated for piezoelectric energy harvesters with mixed-signal circuital simulations of non-linear equivalent electromechanical systems and a prototype has been developed. The proposed switching converter may increase harvested power of about 30% with respect to solutions based on traditional rectifiers and is expected to achieve up to 755 muW in a train passenger car running at about 100 km/h with a 32 times 6 mm2 cantilever with a 35 g mass attached at the free end.

Equivalent-voltage approach for modeling low-frequency dispersive effects in microwave FETs

In this paper, a simple and efficient approach for the modeling of low-frequency dispersive phenomena in FETs is proposed. The method is based on the definition of a virtual, nondispersive associated device controlled by equivalent port voltages and it is justified on the basis of a physically-consistent, charge-controlled description of the device. Dispersive effects in FETs are accounted for by means of an intuitive circuit solution in the framework of any existing nonlinear dynamic model. The new equivalent-voltage model is identified on the basis of conventional measurements carried out under static and small signal dynamic operating conditions. Nonlinear experimental tests confirm the validity of the proposed approach.

An energy autonomous switching converter for harvesting power from multiple piezoelectric transducers

This paper introduces a power conversion technique based on active control for harvesting power from multiple piezoelectric transducers with higher efficiency than conventional passive approaches. The proposed scheme allows each source to operate efficiently, especially in case of irregular vibrations and heterogeneous transducers. A prototype system was developed and validated in realistic conditions: three 0.5×12.7×31.8 mm3 piezoelectric cantilevers with 18g tip masses were excited with the vibrations of a train passenger car. The converter harvested up to 142µW@0.22g-RMS and up to 90µW@0.11g-RMS. An implementation based on discrete components including passive startup and basic power management consumes a fraction of the extra harvested power as little as 5.5µW per source during operation. The structure is modular and the number of sources can be easily increased without affecting power consumption.

Actively Controlled Power Conversion Techniques for Piezoelectric Energy Harvesting Applications

The current advances in ultra-low power design let foresee great opportunities in energy harvesting platforms for self-powered systems. This paper presents two conversion schemes based on active control for harvesting energy with a higher efficiency than traditional passive approaches. A prototype has been developed and the approaches have been validated for piezoelectric energy harvesters with both measurements in realistic conditions (i.e. irregular vibrations) and mixed-signal circuital simulations. The proposed converters may increase harvested power of at least 25% and up to three times with respect to a passive rectifier. The harvested power is about 40 μW in presence of weak vibrations (aRMS = 1.18 m/s2) obtained in a train passenger car in motion with a 28 × 6 × 0.5 mm3 cantilever made of PZT-A4E with a 20 g mass attached at the free end.

Hybrid High Power Amplifiers for L-Band Space Application

This paper describes the design and implementation of 2 hybrid high power amplifiers at L band for a space application. Indeed, the amplifiers represent prototype test vehicles for a larger hybrid amplifier to be used as the final power stage in the transmitting chain of a T/R module of an L-band SAR antenna for earth observation. The amplifiers described in this paper exploit a discrete bar of a commercial 0.35 mum pHEMT process as active device. The first amplifier, featuring a single discrete device, delivers 12 Watts with 56.5% PAE and 12.3 dB gain at 2 dB compression. The second amplifier exploits 4 discrete pHEMT bars using input/output microstrip splitter/combiner and lumped wire bonding/ceramic MIM capacitor output matching networks. It delivers 42 watts with 50% PAE and 13 dB gain at 2.5 dB compression. Since it is a space application, these performances have been achieved with the required de-rating on break-down voltages, current densities and operating channel temperature. The latter has been evaluated by means of a 3-D device model implemented in the framework of a finite differences numerical thermal simulator. In spite of the constraints due to space de-rating rules, the obtained output power densities are 1 W/mm and 0.875 W/mm for the single-discrete and the 4-discrete amplifiers, respectively, which represent a value very close to the state of the art for pHEMT processes.

A simplified approach for quasi-linear power amplifier distortion evaluation

The paper presents a simplified approach for the evaluation of mild distortion in highly linear power amplifiers (PA) for microwave communications. In particular, it is shown how intermodulation distortion (IMD) can be accurately predicted on the basis of a single-tone power-swept Harmonic Balance analysis instead of using two-tone or multi-tone analyses leading to time-consuming and computationally expensive iterative PA design procedures. Moreover, simple equations provided in the paper show that common design specifications given in terms of maximum acceptable IMD are conveniently converted into constraints on a suitable non linearity index, involving both AM/AM and AM/PM amplifier characteristics. Experimental validation is provided in the paper on the basis of a 50¿-loaded GaAs 600¿m-PHEMT based power amplifier simulated with Agilent ADS.

A Method Based on Scattering Parameters for Model Identification of Piezoactuators with Applications in Colloidal Suspension Monitoring

Liquid colloids are mixtures where one substance is dispersed evenly throughout another. Piezoactuators can be reliably used to sense physical properties of colloids. However, to better understand and forecast the behavior of the system, a suitable and reliable model should be used to fit experimental data. In this work we will show how the scattering parameters, introduced for the first time in this field, can greatly improve model identification accuracy and give deeper insight into the resonant response of the system, especially in the case of viscous systems, whenever resonance peaks tend to shift or to vanish. Any change on the density and/or viscosity of the interacting fluid causes a wellpredictable modification on each possible frequency response computable from measured data: the best choice in terms of sensitivity and orthogonal correlation has been investigated by means of parametric simulations with variable loads. Experimental results demonstrate that the conceived sensor is able to reliably discriminate different bentonite clay suspensions.

Critical issues in highly-linear power amplifier design

A recently proposed technique for the design of power amplifiers is here improved and specifically tailored for the application in the field of highly-linear PAs, such as those required by modern high-capacity digital radio links. The proposed design procedure is aimed at providing minimum intermodulation distortion in the presence of hard constraints of minimum guaranteed output power and gain. The proposed algorithm can be efficiently exploited both in conjunction with a highly accurate nonlinear device model, obtaining a numerically efficient design procedure completely CAD-based, or alternatively, as a performance-driven controlling strategy for experimental source-load pull setups. The description of the algorithm also gives the opportunity of revising critical issues concerning both design and electron device modeling aspects related to high-linearity PAs.