5. K. Hirayama, E.N. Glytsis, and T.K. Gaylord, Rigorous analysis of
diffractive cylindrical lenses, J Opt Soc Am A 13 (1996), 2219 –2231.
6. Y.J. Guo and S.K. Barton, Flat printed lens and reflector antennas,
Proc IEE 9
th
Int Conf Antennas Propagat, Eindhoven, The Nether-
lands, 1995, pp. 253–256.
7. D.M. Pozar, Flat lens antenna concept using aperture-coupled micros-
trip patches, Electron Lett 32 (1996).
8. S. Chandran and J.C. Vardaxoglou, Experimental results on the focus-
ing properties of tapered frequency-selective surfaces, Microwave Opt
Technol Lett 11 (1996), 277–279.
9. D.M. Pozar and T.A. Metzler, Analysis of a reflectarray antenna using
microstrip patches of variable size, Electronic Lett 29 (1993), 657–
658.
10. J.A. Encinar, Design of two-layer printed reflectarrays using patches of
variable size, IEEE Trans Antennas Propagat 49 (2001), 1403–1410.
11. J.A. Encinar and J.A. Zornoza, Broadband design of three-layer
printed reflectarrays, IEEE Trans Antennas Propagat 51 (2003), 1662–
1664.
12. B.A. Munk, Frequency-selective surfaces: Theory and design, Wiley–
Interscience, New York, 2000.
13. C.A. Moses, A theoretical study of electromagnetic feedforward/feed-
back media and wire media, Ph.D. dissertation, University of Penn-
sylvania, 1997.
14. C.A. Moses and N. Engheta, Electromagnetic wave propagation in the
wire medium: A complex medium with long thin inclusions, special
issue on electrodynamics in complex environments, Wave Motion 34
(2001), 301–318.
15. C.A. Balanis, Advanced engineering electro-magnetics, Wiley, New
York, 1989.
16. J.A. Stratton, Electromagnetic theory, McGraw-Hill, New York, 1941.
© 2004 Wiley Periodicals, Inc.
CHARACTERISTICS OF
MICROFABRICATED RECTANGULAR
COAX IN THE Ka BAND
E. R. Brown,
1
A. L. Cohen,
2
C. A. Bang,
2
M. S. Lockard,
2
B. W. Byrne,
2
N. M. Vandelli,
2
D. S. McPherson,
2
and
G. Zhang
2
1
University of California Los Angeles
Los Angeles, CA 90095
2
MEMGen Corporation
Burbank, CA 91506
Received 13 August 2003
ABSTRACT: Miniature rectangular coaxial transmission line (with a
300 300 m outer conductor) is simulated, fabricated, and tested up
to the Ka band. It is made by the electrochemical fabrication (EFAB™)
of nickel such that the center conductor is supported primarily by /4
stubs that also establish a resonant passband. The minimum insertion
loss of a 1.67-cm-long test structure is found to be 1.74 dB at the pass-
band center of 29 GHz. This is within 0.57 dB of the lowest insertion
loss predicted by full-wave numerical simulations, indicative of the high
precision and smooth surface morphology of the EFAB process. © 2004
Wiley Periodicals, Inc. Microwave Opt Technol Lett 40: 365–368, 2004;
Published online in Wiley InterScience (www.interscience.wiley.com).
DOI 10.1002/mop.11383
Key words: rectangular coaxial transmission line; electrochemical fab-
rication; Ka band; nickel RF properties; resonant passband
INTRODUCTION AND BACKGROUND
Various 3D microfabrication techniques have improved in preci-
sion and yield to the point where they can be considered for
producing passive RF components on the deep submillimeter-size
scale. A technique familiar to the MEMS community is silicon
bulk micromachining, which has already been utilized to make
several RF components, including resonators and filters [1–3]. The
technique utilized in the present work is electrochemical fabrica-
tion (EFAB™) [4]—a process that allows for the selective depo-
sition of metals using a conformable polymeric mask that is
patterned lithographically.
A promising building block for the generation of passive RF
components by EFAB™ is rectangular coaxial transmission line
(“rectacoax”). Like its popular counterpart— circular coax, rectan-
gular coax offers several advantages over other high-frequency
transmission lines. For example, if uniformly filled with a low-loss
dielectric material, it supports a purely TEM dominant mode over
a very broad bandwidth, determined by the turn-on of the first TE
mode. If the filler material has a frequency-independent real part of
the dielectric constant and a small loss tangent, the TEM modal
dispersion will be negligible over this bandwidth. Also, because of
the peripheral distribution of surface current on the inner and outer
conductors, the microwave specific attenuation from skin-effect
absorption will be much less significant at the miniature scale than
in printed-circuit transmission lines, such as microstrip and copla-
nar waveguide. Furthermore, losses from substrate-mode excita-
tion, so pervasive in printed-circuit lines, are nonexistent in rec-
tacoax or any other TEM-coax for that matter.
The disadvantages of rectacoax pertain mostly to the practical
issues of design and fabrication. Although analytic solutions for
the fundamental properties (for example, the characteristic imped-
ance) of rectacoax do exist, they are in the form of infinite-series
or integral solutions that are not amenable to simple design rules
[5]. Therefore, fast full-wave techniques (such as transmission-line
matrix) have been developed, particularly for designing junctions
and other discontinuities [6 – 8]. And to fabricate “precision” mul-
tiport passive components such as directional and hybrid couplers,
the rectacoax must be made by expensive machine-shop tech-
niques [9]. Because of these drawbacks, rectacoax has not yet been
applied to RF integrated circuits (RFICs), particularly not with
semiconductor monolithic microwave integrated circuits
(MMICs).
Two incentives for pursuing rectacoax in RFICs are its small
allowable turn radius and its high isolation. Because the turns are
made during the fabrication, their radius can be very small and still
cause little insertion loss for the dominant TEM mode. This is in
contrast to circular coax, such as UT-08 (the smallest known
commercial coax), which has a very small center conductor
(0.05-mm diameter) and outer conductor (0.167-mm inner diam-
eter) but has a minimum turn radius of 0.8 mm [10]. This mini-
mum radius is determined more by mechanical failure (collapse of
the outer conductor) than by electromagnetic effects.
The high isolation of rectacoax means that two side-by-side
lines will have negligible crosstalk if the outer wall thickness is
just a few tens of microns or more. This is because the skin depth
in good metals is very small at microwave and mm-wave frequen-
cies (for example, 1 m in copper at 10 GHz). And it allows one
to reduce the footprint of distributed passive components by wind-
ing straight sections of rectacoax back-and-forth in a serpentine
fashion. These advantages are exploited in the design of the
present rectacoax test structure.
FABRICATION OF MINIATURE RECTACOAX
The retrocoax in this work was made by EFAB™, a fabrication
technology developed by MEMGen Corporation. It is beyond the
scope of this discussion to illustrate the EFAB process in detail,
however, it can be briefly described through the following steps: (i)
MICROWAVE AND OPTICAL TECHNOLOGY LETTERS / Vol. 40, No. 5, March 5 2004 365