Search

Let us translate the zero-voltage-derivative switching (ZVDS) introduced in last month’s puzzle into plane geometry. For the power amplifier shown in Figure 1 to perform ZVDS operation, the load impedance ${Z} = {R} + {jX}$ should meet a specific condition. We can illustrate this specific condition as a continuous locus on the R – X plane. Which of the following shapes does the locus exhibit?

This paper theoretically discovers a cutoff behavior of the DC current in twin rectifiers operating in parallel. The formulation of the circuit starts from a single rectifier focusing on its flow-angle equation. The DC output voltage and current formulas are derived in terms of the RF input power and load resistance. We then consider twin rectifiers connected in parallel at the DC output port. When the twins are excited with different RF amplitudes, a specific input power ratio exists whereby the weaker rectifier goes to sleep. We name this behavior rectifier's current cutoff, and the specific cutoff power ratio is theoretically predicted as –13.3 dB. A harmonic-balance simulation and proof experiment are performed at an input frequency 6.78 MHz and power up to 100 W in total. Both simulation and experiment successfully confirm the current cutoff behavior.

To make the dream of wireless powering to running electric vehicles come true, this paper presents a viable solution to the standing wave problem inevitably arising on a long electrified roadway. We introduce a left-handed (LH) circuit to function as a standing wave mitigator. As a power loss index in the circuit, we formulate the LH circuit with Poincare length. We then consider a long transmission line, along which multiple LH circuits are periodically inserted. By extending the formula, we explore how the total Poincare length behaves when we sweep the load impedance and the number of segments. Results reveal the condition for achieving the minimum total Poincare length.

This paper presents a unique mathematical approach to impedance mapping on a complex plane. Assuming a passive load excited from an RF power source, we express the complex reflectance with respect to the standing wave ratio that involves the phase angle. Then the ratio is heuristically raised to a power, which enables us to magnify the local grid of the impedance. By pure formulation, we explore the trajectories that will appear on the plane from this local magnification. The results are visualized in seven persuasive illustrations to open up a fresh perspectives for the plane geometry from the RF engineering field.

As one of the basic structures for wireless power transfer, a symmetric dipole antenna pair is revisited in this paper. Since Friis formula is no longer applicable in the near-field region, we first employ kQ product as an alternative index of maximum power coupling efficiency. Then assuming optimum impedance matching circuits at both source and load ports simultaneously, we employ Poincare distance D as an associate index for the minimum power dissipation inside the matching circuits. Electromagnetic-field simulation is carried out to deduce the antenna's self and mutual impedances, which are converted into the above two indices. As a result, we find that the simple dipole antenna pair performs well (kQ > 1) as a wireless power coupler within the 0.2 wavelength range. Poincare distance is predicted as D < 1 for the range of 0.06 wavelength and beyond. This will be a standard reference when we explore or evaluate new wireless coupler schemes for the future.

This paper presents Q factor analysis for FET oscillators employing distributed-constant elements. We replace the inductor of a lumped constant Colpitts circuit by a shorted microstrip transmission line for high frequency applications. Involving the FET's transconductance and the transmission line's loss due to both conducting metal and dielectric substrate, we deduce the Q factor formula for the entire circuit in the steady oscillation state. We compared the computed results from the oscillator employing an uniform shorted microstrip line with that of the original LC oscillator. For obtaining even higher Q factor, we modify the shape of transmission line into nonuniform, i.e., step-, tapered-, and partially-tapered stubs. Non-uniformity causes some complexity in the impedance analysis. We exploit a piecewise uniform approximation for tapered part of the microstrip stub, and then involve the asymptotic expressions obtained from both stub's impedance and its frequency derivatives into the active Q factor formula. Applying these formulations, we calculate out the value of capacitance for tuning, the necessary FET's transconductance and achievable active Q factor, and then finally explore oscillator performances with a microstrip stub in different shapes and sizes.

An extraction and calibration technique is proposed that allows electronically steerable parasitic array radiator (ESPAR) antennas to achieve high efficiency and precision in performance characterization. The proposed technique is based on a reactive-near-field measurement or direct acquisition of RF current. We present a simple formulation that requires minimal current measurements to successfully extract the mutual impedance. Currents are predicted for an arbitrary combination of bias voltages that are used to control the reactance of varactor diodes integrated with the antenna's parasitic elements. These predicted currents are compared by conducting an experiment with a seven-element ESPAR. For the bias voltage combinations that have reasonably good symmetry, the differences in relative gain, beam direction and half power beam width (HPBW) actually fall within 0.5 dB, 4deg and 9deg, respectively. Directivity patterns obtained by calculating the predicted currents are verified with those obtained by calculating the directly measured currents. Furthermore, the calculated patterns are verified through comparison with far-field measurements in a conventional anechoic chamber. The differences in beam direction and HPBW are also found to be acceptable. Using the proposed technique, the required number of measurements is reduced from 85,766,121 to just 149 for a seven-element array

We examine that an inductively loaded dipole becomes electrical invisible showing that the influence on a nearby antenna's input impedance and on its radiation pattern (scattering) are minimized. We also verify that the direction of the current reverses on the electrically invisible dipole so that the integral of the dipole current becomes zero. By making an inductive load with a distributed constant line, we fabricate an electrically invisible dipole without using chip devices and evaluate its performance. Furthermore, we propose electrically invisible conductor lines formed by connecting these electrically invisible dipoles in series using open stubs as connectors, and analytically examine their invisible state

A technique to efficiently characterize the azimuth radiation patterns of monopole arrays based on evanescent reactive-near-field measurement in a compact anechoic box is presented. We describe a desktop-size measurement setup for determining the azimuth far-field pattern of an electronically steerable parasitic array radiator (ESPAR) at 2.5 GHz, and present a thorough statistical evaluation of the ESPAR antenna characterization technique. In addition, the sensitivity to probe position and the errors caused by possible fluctuations in the probe position are investigated. Furthermore, comparisons among patterns obtained by the proposed method, simulations and conventional far-field measurements are carried out, and a good agreement among them is obtained. We also provide a statistical evaluation of the proposed desktop-size measurement setup using forty ESPAR antennas and conclude that the proposed setup can be an inexpensive candidate to replace the conventional measuring system

In this paper, we show that the current distribution along the parasitic dipole elements loaded by variable reactors varies greatly depending on the variable reactor value. Thus, when the directionality of an Electronically Steerable Passive Array Radiator (ESPAR) antenna constructed from parasitic dipoles is calculated, we must consider the changes in the current distribution along these elements. Therefore, we propose a method where the current weight is the value of the vector effective length of each element multiplied by the current value at the port. We also show that when the dipole length is less than approximately the half wavelength, the vector effective length is calculated for the most part by the ratio of the port voltage to the port current. From this relationship, even if the current distribution along an element is not determined, the vector effective length of a parasitic element can be calculated from the value of the variable reactor loading the element. Thus, the current distribution along the elements can be easily considered when calculating the directionality of the ESPAR antenna. We present a method that calculates the directionality of the feed dipole array antenna having mutual coupling among the elements that considers the current distribution along the element where not only the port current but also the port voltage are set as the weights. © 2005 Wiley Periodicals, Inc. Electron Comm Jpn Pt 1, 89(2): 22–35, 2006; Published online in Wiley InterScience (www.interscience.wiley.com). DOI 10.1002/ecja.20232

Nonlinear distortions in an anti-series varactor pair (ASVP) are analyzed by a perturbation method. To the authors' knowledge, this paper presents the first derivation of an analytical expression that explicitly shows intermodulation and harmonic distortions of the ASVP. In addition to canceling the expected even-order distortion, the third-order distortion can be suppressed simultaneously when a certain condition is satisfied. We also find that the second- and third-order distortions can be greatly suppressed without dependence on dc bias voltage if the varactors in the ASVP have an ideal abrupt p-n junction. These theoretical predictions are verified by measuring the second- and third-order harmonic distortions of an ASVP. The experimental results show that the second-order harmonic distortion is suppressed by approximately 20 dB. The third-order harmonic distortion is suppressed to the same extent in the theoretically predicted dc bias voltage range.

In this paper, we introduce generation of multiplexed signals on the millimeter-wave bands for fiber-radio systems where an optical millimeter-wave generator is based on a two-mode locked Fabry-Pe/spl acute/rot (FP) slave laser, whose injection current is directly modulated by a signal source. We qualitatively consider the distortion of the millimeter-wave signals from the FP slave laser. The distortion components on the millimeter-wave bands are induced from the simultaneous modulation of the locked modes and the nonlinear modulation response of the FP laser. Two-tone modulation of the locked FP laser is examined to evaluate the dynamic range of the millimeter-wave signals against the second- and third-order distortion components. We also perform fiber transmission of three 156-Mb/s-BPSK signals on the 60-GHz band to demonstrate fiber-radio down-link systems. The total capacity of the down-link system is discussed. In addition, two methods for multicarrier generation on the millimeter-wave bands are proposed. Multicarrier generators supported by these methods can be used as local signals for up-link millimeter-wave signals. The first method is based on multitone modulation of the FP slave laser. We attempt the down-conversion of a 52-Mb/s ASK signal on the 60-GHz band by using the millimeter-wave local signals. The second method depends on the distorted modulation of the FP slave laser by using a single continuous wave signal where the DC-bias level of the FP laser's injection current is partly under the threshold value. We confirm that five carriers on the 60-GHz band are effectively generated by using the second method. Furthermore, the influences of the chromatic dispersion effects on the millimeter-wave local signals are investigated for both methods.

The radiation pattern of the ESPAR antenna is controlled by variable reactances loaded to its passive elements. In this paper, the effects of the number of elements and the variable range of the reactance value on the beam and null forming capability are studied by simulation using the method of moments. It is found that a beam with a directivity of more than 8 dBi and a null of less than −40 dBi can be scanned independently in all directions when the number of elements is more than 5 and the reactance value range is ±50 Ω. The reactance range of ±50 Ω is needed for simultaneous forming of a single beam of more than 5 dBi and a single null of less than −30 dBi, while ± 100 Ω is needed for simultaneous forming of a single beam of more than 5 dBi and multiple nulls of less than −15 dBi. The direction of the beam and null can be close as the number of elements is increased. The input impedance of the feed element is also studied and the standard return loss of a seven-element antenna is found to be 15.8 dBi. © 2003 Wiley Periodicals, Inc. Electron Comm Jpn Pt 1, 86(12): 1–11, 2003; Published online in Wiley InterScience (www.interscience.wiley.com). DOI 10.1002/ecja.10141

This paper concerns an optical millimeter-wave signal generator for fiber-based millimeter-wave systems. The millimeter-wave signal generator is based on two-mode injection locking of a Fabry-Perot (F-P) laser. The millimeter-wave signal can be induced by self-heterodyne detection of the locked two modes in a high-speed photodetector (PD). The locking characteristics of the F-P slave laser and the tunability for the millimeter-wave carrier frequency are demonstrated. When the F-P laser is directly modulated by data signals, the locked two modes are modulated simultaneously. The data signals can then be up-converted to the millimeter-wave band at the PD output. By this direct modulation method, the effect of fiber chromatic dispersion on the millimeter-wave signal components at the PD output can be moderate according to mixing of amplitude and phase modulation on the locked two modes. In virtue of the wide response of the F-P laser, relatively high-speed data (2.5 Gbit/s nonreturn to zero (NRZ)-ASK or 622 Mbit/s NRZ-BPSK) on the millimeter-wave band (52 or 60 GHz) can be transmitted on a 32-km single-mode fiber without bit error. A wide tunable range (56-63 GHz) for the central frequency of the millimeter-wave signals and a wide optical bandwidth (1530-60 nm) of the F-P slave laser are also confirmed by the bit-error measurements of the transmitted data signals.

The first MMIC local oscillator for 16-QAM digital microwave systems is presented. The key advance is achieving the very low phase noise required by such systems. Low phase noise is realized with a low phase noise VCO (voltage-controlled oscillator) that is installed into a new PLL (phase-locked loop). Although the developed local oscillator is 90% smaller than the existing Dielectric Resonator-based local oscillators, it achieves comparable receiver performance. >

In order to realize a compact low-profile antenna that can steer beam/null continuously in full azimuthal directions, this paper proposes a new configuration of beam-steering antenna with mutual coupling between inverted-F antenna (IFA) elements. The antenna array consists of two active and two passive IFA elements loading variable reactors, thereby realizing a low-profile antenna array within a small space. It is expected that the antenna possesses a high ability to form beam and/or null, since it behaves as both Yagi-Uda array and phased array. Furthermore, we derive here a general expression of equivalent weight [1] for multiport parasitic IFA arrays. Finally, the beam steering of the proposed antenna is demonstrated by the numerical simulation.

We proposed a spatial fading simulator using CECA loaded with phase shifter for performance evaluation of antenna arrays. The statistic investigation of received signal such as LCR and AFD are achieved for verifying Rayleigh fading. SCC of received signal is investigated for AS. Based on the statistical characteristics obtained from experiment, it is sure that CECA loaded with phase shifter works well as a spatial fading simulator.

This paper describes the experimental study of a parasitic microstrip array antenna operated in circular polarization (CP). This antenna applies a microstrip antenna with dual-orthogonal feed as an antenna element for excitation of the CP wave from both fed and parasitic elements. A new switched reactance circuit using an SP3T switch taking three reactance states is connected to two ports on each parasitic element. The radiation pattern is controlled by switching the reactance state. The measured radiation pattern and the axial ratio of the two-element array (one fed and one parasite) indicated that the main beam was switched depending on the reactance state (capacitive, short, and inductive) with an axial ratio below 3 dB. The capability of 360° beam-switching in the azimuth direction was verified by measurement of the five-element two-dimensional array (one fed and four parasites). By setting proper reactance values, the main beam was tilted in the elevation plane at the azimuth of every 45° interval, and the axial ratio was acceptable. The measured results agreed with the calculated ones. From these experimental observations we concluded that the antenna works as a switched-beam antenna operated in CP wave. (5 pages)

This paper describes the simulated design for a fed circuit having a radial structure for use in a newly proposed millimeter-wave switched sector-beam antenna. The goal is to realize an ultra-high-speed gigabit-rate wireless LAN. The radial structure is designed to achieve the purpose that the input power from a unique RF port at the structure's center that can be sent to any one of selected sectors with maximum power under the required bandwidth. The simulated results of the proposed radial structure show that the bandwidth of S21, with an insertion loss of -1.0 B in the frequency range of 59-66 GHz is 7.0 GHz for the case of four, five, and six sectors, and that the bandwidths are 2.3 GHz and 1.7 GHz for the case of seven and eight sectors, respectively

We have successfully demonstrated high-quality, 40-GHz, 4.5-ps, 1550-nm, passively-mode-locked pulse generation, with our original SOA-based pulse-generator scheme.

The generation of small sub-pulses in the delayed-interference signal-wavelength converter (DISC), which has been studied for use in future 160-Gb/s optical time division multiplexing–wavelength division multiplexing (OTDM–WDM) communication systems, was recently predicted as a potential problem. In this work, we have experimentally verified the generation of such pulses and its mechanism. In the experiments we used 3.8-ps-long 1.56-µm input pulses with repetition frequencies from 12.5 to 25.0 GHz and a cross-correlation monitoring system with a time resolution of approximately 2 ps.

A theoretical analysis based on the multiple scale method is presented for the behavior of the leaky wave in a corrugated ferrite image guide. the attenuation of the guided mode due to leakage is given analytically as a complex perturbation of the propagation constant. As an application of this structure, a millimeter wave leaky wave antenna has been considered and its radiation efficiency and directivity have been studied. Next, experiments are carried out for a leaky wave antenna using a poly-crystalline YIG slab. the radiation pattern has been measured in the 40-GHz band and compared with theoretical values. A change in dc magnetic bias by 1 Wb/m2 shifts the main beam to be scanned by 17.5 degrees. the experimental results agree well with theoretical data.

The delay and loss characteristics of magnetostatic forward volume waves in a YIG slab have been studied experimentally where an inhomogeneous bias field is applied normal to the slab surface by using artificially designed magnetic pole pieces. The nondispersive group delay characteristics with a low propagation loss are observed.