The intrinsic characteristics of LDMOS transistor amplifier can be extracted by removing the simulated characteristics of the input side and output side of test fixtures from the overall measured characteristics. In addition, impedance-matching networks are designed with these extracted S-parameters of the LDMOS transistor amplifier. The transistor used in this work is an MRF21030 LD-MOS 30W transistor provided by Freescale Semiconductor, Inc., and biased such that I DS_DQ is equal to 200 mA. Figures 4(a) and 4(b) show the measured and simulated S-parameters of the impedance matched transistor amplifier in the frequency range from 2.0 to 2.3 GHz. The maximum deviations of the measured ͉S 11 ͉, ͉S 21 ͉, and ͉S 22 ͉ from the simulated ͉S 11 ͉, ͉S 21 ͉, and ͉S 22 ͉ are 0.66, 0.17, and 0.13 dB, respectively. The simulated characteristics of impedance matching blocks can accurately predict the measured ones, as shown in Figures 4(a) and 4(b). CONCLUSIONIn this paper, we have presented a method of extracting 3D field simulation parameters of RF-band transistor amplifier test fixtures and demonstrated that a circuit model of the test fixtures can be constructed via pure 3D field simulation using the extracted material constants. The developed circuit model predicts measurement results within 0.5 dB.ABSTRACT: A simple square-slot antenna capable of providing a very wide impedance bandwidth of larger than 9 GHz is presented. The ultra-wideband operation for the proposed antenna is realized by using a novel feeding mechanism of loading an offset rectangular stub at the end of the microstrip feed line. With the proper offset distance of rectangular stub chosen, the proposed antenna can operate in the 2.78 -12.02-GHz frequency range and covers the UWB operating bandwidth of 3.1-10.6 GHz. The antenna radiation patterns at 3, 5.3, 7.9, and 10 GHz are also presented.ABSTRACT: An H-plane hybrid ring 3-dB power divider constructed with substrate integrated rectangular waveguide (SIRW) is proposed. Theoretical simulations are carried out at the X-and Ka-bands. The results show that the isolation between two output ports is high up to 35 dB and good 3-dB power division within the band of interest is observed also, that is, less than 0.3-dB derivation in the band of interest. In addition, the simulation results at the X-band shows a bandwidth of 10% at Ϫ15 dB. A prototype in the Ka-band is fabricated and the measured data at the Ka-band shows a bandwidth of 7% at Ϫ15 dB, in good agreement with simulation.

Since the imaginary part of the propagation constant and its second derivative are involved, without knowing the unperturbed results, one cannot use Eq. (14) to calculate wavelength shift. However, the equations are useful for qualitative analysis of SPR sensor. At the resonance wavelength, the attenuation reaches to a maximum (␤ i reaches to a minimum), so the second derivative Ѩ 2 ␤ i /Ѩ 2 is positive. Equation (14) predicts that when the sensed medium index change ⌬ Ͼ 0 (⌬ Ͻ 0), the resonance wavelength will shift to longer (shorter) wavelength direction. The conclusion agrees with experimental results from all kinds of fiber/waveguide SPR sensor. The shape of attenuation dip in transmitted spectra reflects ␤ i and Ѩ 2 ␤ i /Ѩ 2 . Usually, the change of this shape is small in relatively broad wavelength range. This means that the index sensitivity roughly remains at a constant value. Although the factor I r is increasing as the sensed medium thickness t increases, due to the decay of the modified Bessel function, only the dielectric perturbation within the penetration depth makes real contribution. In the case of uniform index perturbation, the wavelength shift is proportional to ⌬. To obtain a good sensitivity, one should match the penetration depth of the electric field with the thickness of the sensed dielectric film. If the dielectric perturbation is not uniform, the wavelength shift must be determined by integrating Eq. (13a). In this case, it is impossible to distinguish dielectric perturbation effect from sensed medium thickness change by resonance wavelength shift. The effect of the thickness of the metallic thin film is partially included in the factor I r . The surface plasma mode cannot be developed if the thickness d is too small [10]. On the other hand, when the metallic film is too thick, the field ⌽(r,) almost vanishes in region r Ͼ a ϩ d, thereby I r Ϸ 0. This implies that there is an optimal metallic film thickness for SPR sensing. APPENDIXFrom Eq. (5a) we haveNote that the derivative is taken at resonance wavelength, we omit the terms that contain Ѩ␤ i /Ѩ in Eq. (A1). Substituting the relation (A2)into Eq. (A1), we obtain Eq. (9). Let ␤ r ϭ k 0 n r . In fiber SPR sensor, usually k 0 a ϾϾ 1, the mode index n r approaches a constant. ThereforeEquation (10a) and (10b) can be rewritten asClearly, it can be seen that C r Ͻ 0 and C m Ͻ 0. . In this paper, an alternative square slot antenna with an offset rectangular stub and an U-like slot is proposed for achieving the band-notched UWB operation. The proposed square slot antenna is simple in structure and easy to construct. The offset rectangular stub is first employed for achieving UWB operation. Then, a quarter-wavelength U-like slot is inserted into the rectangular stub and a notched frequency band at 5 GHz can thus be created. Details of the antenna design are described, and experimental results of the proposed antenna are presented and discussed. BAND-NOTCHED ULTRA-WIDEBAND SQUARE SLOT ANTENNA ANTENNA DESIGNThe geometry of the proposed band-no...

No abstract