L.J. du Toit

Advances in the design of Jaumann absorbers

The authors consider electric screens with the material spacers assumed homogeneous, nonmagnetic, lossless, commensurate, and isotropic. Normal incidence is assumed. Such an N-layered structure may be modeled as a transmission line network. The unknowns are the N spacer characteristic impedances and the N sheet surface resistivities. All are normalized by the free-space impedance. The maximally flat solution of Fante and McCormack (1988) is extended to any number of layers, with proven solvability up to 20 layers. An equiripple solution, which, for a given number of layers and specified maximum reflection coefficient, exhibits a wider bandwidth than the corresponding maximally flat solution, is described. The sheet resistivity spread is also much narrower, thus increasing realizability.<<ETX>>

A design process for Jaumann absorbers

Results of an investigation into the design of Jaumann absorbers are reported. Jaumann absorbers are effective in minimizing plane wave reflection from a conducting metal surface and exhibit reasonably wide frequency and incidence angle bandwidths. The spacers between the resistive sheets are usually assumed to be lossless and a quarter-wavelength long at the center frequency. It is shown that the numerical minimization technique can result in a design which improves upon the purely analytical methods. Some representative results are shown.<<ETX>>

Linear patch array pattern degradation due to corporate feed radiation

A specific linear array of square patches is considered. The array, measured feed radiation, and sources of feed radiation are examined. It is shown that radiation from a coplanar microstrip feed network can degrade the patterns of a linear patch array to an extent that is unacceptable in many applications. The problem arises when the microstrip substrate parameters are chosen to maximize radiation at the patch edges, thus undermining the ability of the feed network to guide waves. One solution is to use a relatively thin, high-relative permittivity substrate. This has the disadvantage of reduced bandwidth. Another approach is to sacrifice the low-cost coplanar feed by placing the feed on the opposite side of the ground plane from the radiating elements, and using coaxial pins to drive the patches. In this way, the radiation and guided wave functions are separated and the substrate parameters can be separately optimized for each. The feed can also be screened.<<ETX>>

Optimal equiripple solution to the Jaumann absorber problem

The general Chebyshev approximation method is used to find the optimal solution to the Jaumman absorber problem. The Chebyshev-like (CL) solution with all reflection zeros at real frequencies is used as a starting point in the iteration. The optimal solution, with the reflection zeros moved slightly off the imaginary S-axis, is in the immediate vicinity of the CL-solution, and convergence properties are good.<<ETX>>

Using a small array to optimize dipole match in the presence of mutual coupling

Tuning the center dipole in a small array, such that it will ultimately exhibit a good active input impedance when used in a large array, is discussed. The presentation addresses theory, timing techniques, measured results, and comments on general applicability.Tuning the center dipole in a small array, such that it will ultimately exhibit a good active input impedance when used in a large array, is discussed. The presentation addresses theory, timing techniques, measured results, and comments on general applicability.

Equivalent circuits for an electrically thick square patch directly fed by a coaxial probe or by a microstrip line

The authors present lumped-element equivalent circuits for two specific square patches on an electrically thick (0.05 lambda /sub 0/) substrate. The equivalent circuits were derived from experimental data and could be useful in the verification of theoretical techniques. For the particular geometry studied the probe fed patch requires sidelengths approximately 6% shorter than the microstrip fed patch, for the same resonant frequency. The microstrip-fed patch has a sharper resonance and approximately half the bandwidth of the probe-fed patch. Presumably the probe introduces more losses than the microstrip, thus damping the patch resonance.<<ETX>>