See discussions, stats, and author profiles for this publication at: https://www.researchgate.

net/publication/263473560

Accurate impedance calculations for transmission lines with adjustable inner

conductor

Article in Microwave and Optical Technology Letters · September 2014

DOI: 10.1002/mop.28497

CITATIONS READS

2 208

3 authors, including:

Elena Vinogradova Yury Aleksandrovich Tuchkin

Macquarie University Gebze Technical University
100 PUBLICATIONS 443 CITATIONS 135 PUBLICATIONS 616 CITATIONS

SEE PROFILE SEE PROFILE

All content following this page was uploaded by Elena Vinogradova on 07 August 2020.

The user has requested enhancement of the downloaded file.

 for WLAN/WiMAX applications, Microwave Opt Technol Lett 54 cross-sections. The developed algorithm allows fast and accurate com-
(2012), 1049–1052. putations of impedance values for various positions of the inner conduc-
10. J.H. Yoon and G.S. Kil, Compact monopole antenna design for tor in the infinitely long transmission line. V
C 2014 Wiley Periodicals,

WLAN/WiMAX triple-band operations, Microwave Opt Technol Inc. Microwave Opt Technol Lett 56:2066–2070, 2014; View this article
Lett 54 (2012), 465–473. online at wileyonlinelibrary.com. DOI 10.1002/mop.28497
11. X. Sun, G. Zeng, H.C. Yang, and Y. Li, A compact quad-band
CPW-fed slot antenna for M-WiMAX/WiMAX applications, IEEE Key words: two-dimensional electrostatics; characteristic impedance;
Antennas Wireless Propag Lett 11 (2012), 395–398. method of regularization
12. A. Mehdipour, A.R. Sebak, C.W. Trueman, and T.A. Denidni, Com-
pact multiband planar antenna for 2.4/3.5/5.2/5.8 GHz wireless appli-
cations, IEEE Antennas Wireless Propag Lett 11 (2012), 144–147. 1. INTRODUCTION
13. R.L. Li, X.L. Quan, Y.H. Cui, and M.M. Tentzeris, Directional triple The long-standing interest in the conformal transmission lines is
band planar antenna for WLAN/WiMAX access points, Electron explained by their wide usage as transmission links of radio fre-
Lett 48 (2012), 305–306. quency energy, as impedance-transforming sections, or as com-
14. Y. Xu, Y.C. Jiao, and Y.C. Luan, Compact CPW-fed printed monop- ponents in electron tubes. Most of the studies focus on the
ole antenna with triple band characteristics for WLAN/WiMAX coaxial circular and square transmission lines. Less attention is
applications, Electron Lett 48, (2012), 1519–1520.
paid to other geometries, partly because corresponding transmis-
15. J.H. Yoon, Y.C. Rhee, and Y.K. Jang, Compact monopole antenna
sion line types are less common, and partly because noncanoni-
design for WLAN/WiMAX triple-band operations, Microwave Opt
Technol Lett 54 (2012), 1838–1846. cal shapes (where contours does not coincide with the
16. J.H. Yoon, Y.C. Rhee, and W.S., Kim, A rectangular ring open- coordinate lines) essentially complicates the problem. Very few
ended monopole antenna with two symmetric strips for WLAN and articles considering a possibility of the rotational displacement
WiMAX Applications, Int J Antennas Propag Article ID 109450 of the conductors mostly focus on a short discrete set of dis-
(2013), 1–9. placement angle values.
17. H.Q. Zhai, Z.H. Ma, Y. Han, and C.H. Liang, A compact printed Because of the conformal transmission line topology, the
antenna for triple band WLAN/WiMAX applications, IEEE Anten- conformal mapping method (CMM) is intrinsically well-suited
nas Wireless Propag Lett 12 (2013), 65–68. for its analysis (see, e.g., [1–3]). The case of cylindrical con-
18. H. Chen, X. Yang, Y.Z. Yin, J.J. Wu, and Y.M. Cai, Triband rectan-
formal (coaxial) transmission line is the most popular in prac-
gular loaded monopole antenna with inverted L slot for WLAN/
tical applications and well-studied [4–6]. Among others, the
WiMAX application, Electron Lett 49 (2013), 1261–1262.
19. H. Chen, X. Yang, Y.Z. Yin, S.T. Fan, and J.J. Wu, Triband planar “square in a square” transmission line studies are discussed in
monopole antenna with compact radiator for WLAN/WiMAX appli- [1]. This type of the conformal transmission lines is in a
cations, IEEE Antennas Wireless Propag Lett 12 (2013), 1440–1443. focus of our investigation undertaken in this article. Bow-
20. J. Pei, A. Wang, and W. Leng, A novel arc shaped printed antenna man’s computations [7] of the capacitance per unit length of
for WLAN applications, Appl Mech Mater 130–134 (2012), 4006– the square coaxial line have been widely used by others for
4010. validating different, mainly numerical, techniques [8–10]. Of
21. J.H. Lu and C.H. Yeh, Planar broadband arc-shaped monopole special interest for our studies is the problem of the rotating
antenna for UWB system, IEEE Trans Antennas Propag 60 (2012), coaxial squares considered in [11]. Any disturbance of geome-
3091–3095.
try in conformal coaxial lines essentially complicates the use
22. M. Chongcheawchamnan, K. Meelarpkit, S. Julrat, C.
of the CMM. This becomes especially apparent in the case of
Phongchareonpanich, and M. Krairiksh, Extending bandwidth of the
CPW-fed monopole antenna using circular arc structure, Microwave offset conductors or in the case of rotating conductors.
Opt Technol Lett 54 (2012), 1412–1415. Because of this reason, other methods are used for calculation
23. C. Wang, Z.H. Yan, and S. Li, Compact ultra wideband CPW-fed of the characteristic impedance of such perturbed conformal
fan shaped antenna with dual band notched using arc shaped slots, coaxial lines. The quasianalytical multipole theory method is
Microwave Opt Technol Lett 55 (2013), 368–371. presented in [12] for analysis of a rectangular transmission
24. Ansoft High Frequency Structure Simulator (HFSS) Version 10.0, line. In [13], the parameters of the square coaxial lines with
Ansoft Corporation, Pittsburgh, PA, 2005. a circular and a square inner conductor are investigated using
the finite element method, method of moments, and curves
C 2014 Wiley Periodicals, Inc.
V fitting techniques. The position of the inner conductor is arbi-
trary, but its orientation is assumed to be parallel to the top
and bottom walls of the outer conductor. The partial charge-
ACCURATE IMPEDANCE CALCULATIONS
simulation method is presented in [13] and [14] for calculat-
FOR TRANSMISSION LINES WITH ing the capacitances of rectangular shielded lines with offset
ADJUSTABLE INNER CONDUCTOR inner conductors.
Galyna Safonova1,2,3, Elena Vinogradova1, and Yury Tuchkin4 The various positions of a square inner conductor in a square
1
Department of Mathematics, Macquarie University, Sydney, NSW, conformal coaxial line are described by the different capaci-
Australia; Corresponding author: [email protected] tance values. Hence, there is a possibility to adjust the charac-
2
Phenomics and Bioinformatics Research Centre, University of teristic impedance of the coaxial line by changing inner
South Australia, SA, Australia
3
Australian Centre for Plant Functional Genomics, University of conductor location or rotation angle. This idea found its practi-
South Australia, SA, Australia cal realization in an invention patented in 1968 [15]. However,
4
Gebze Institute of Technology, Kocaeli, Turkey a weak point of the general theory of the conformal coaxial
transmission lines is an absence of reliable analysis of the
Received 17 January 2014 coaxial lines with adjustable characteristic impedance. The algo-
rithm presented in this article allows to overcome this short-
ABSTRACT: The mathematically rigorous solution of the two- comings, providing a rigorous solution to the two-dimensional
dimensional Laplace equation for the multiple-body systems is applied (2D) problem involving various conductor shapes and relative
for accurate impedance calculations of transmission lines of arbitrary positions.

2066 MICROWAVE AND OPTICAL TECHNOLOGY LETTERS / Vol. 56, No. 9, September 2014 DOI 10.1002/mop
2. SOLUTION BY METHOD OF ANALYTICAL
REGULARIZATION
The method described in this article may be classified as an
analytical-numerical method with generic features of reliability
and stability, in contrast to the purely numerical techniques,
where it is not possible to guarantee convergence or accuracy
when some parameter, such as grid size, is increasingly refined.
The details of this method applied to scattering problems for
canonical structures are presented in [16,17].
Consider N 2 1 arbitrary profiled charged PEC cylinders
embedded into a homogeneous dielectric medium with relative
permittivity Er, where N is a positive integer exceeding 1. To
ensure that the electrostatic potential problem is physically rea-
sonable, we consider an electrically neutral system or a case
Figure 1 View of conformal transmission line with arbitrary contour
when a finite dielectric medium is bounded by an infinitesimally
q(U)
thin grounded cylindrical shell. The TEM mode propagation
problem and potential problem in such system are fully
described by the Dirichlet boundary value problem for Laplace’s
equation
to the integral operator of the initial problem, x is the solution
DU50; vector, and B is a given vector, defined by the boundary condi-
tions. We decompose the operator A in two parts: a singular
with boundary conditions for the potentials given at the surface part A0 and a regular part (containing no singularities) A1, (A 5
boundary Ln(n 5 1, . . ., N) of each of N cylinders: A0 1 A1). In our problem, kernels of integral equations Gss con-
tain singularity of the logarithmic type, so we analytically split
Uj5Vn ; n51; . . . ; N; Green’s function in two parts:

where VN 5 0 for a system with a grounded shield enclosing all 1

G Rsj ðh; sÞ 5 log Rsj ðh; sÞ
other cylinders. To use the regularization procedure, all contours 2p
Ln must be smooth and nonself crossing with a continuous   
1 h2s
parameterization that is twice differentiable at each point of Ln. 5 Lsj ðh; sÞ1log 2sin :
2p 2
When the boundary conditions are fulfilled, using the single-
layer solution representation, the problem reduces to the coupled The function Lsj, s 5 j is a regular function, defined every-
system of integral equations for the unknowns Zj, representing where except at the points h 5 s; the function Lsj, s 6¼ j is
line charge densities: defined everywhere. The exact expression for Lsj, s 5 j at the
points of singularity where h 5 s is obtained analytically:
N ð
X
Gij ðjp2qjÞZj ðpÞdlp 5Vn ; i51; . . . ; N;
j51 Lj Lsj ðh; sÞ5log ls ðhÞ :

where G is Green’s function, Vn is the potential given on the n- The initial operator equation becomes (A0 1 A1)x 5 B.
th boundary. Now, we introduce parameterization of each con- Then, we analytically construct the inverse operator A21 0 of the
tour Ln, singular part and apply it to both sides of the equation to obtain

gn ðhÞ  xn ðhÞ1yn ðhÞ ; ðn51; . . . ; NÞ; h 2 ½2p; p A21 21

0 ðA0 1A1 Þx5A0 B; that is; ^
ðI1HÞx5B;

where H5A21 ^ 21 21
and some new notations 0 A1 and B5A0 B;, and I5A0 A0 is an identity
operator. The inverse operator for this problem is constructed
with the use of the exponential Fourier transform. A0 is chosen
Zj 5lj ðhÞZj gj ðhÞ ;
such that H is a compact operator on a suitable solution space,
0 0 1=2 and B^ is a bounded vector, that belongs to the same class as the
lj ðhÞ5f½xj ðhÞ2 1½yj ðhÞ2 g solution vector, therefore, the matrix equation is a well-
conditioned second kind Fredholm equation. Such equations can
where xn(h), yn(h) are coordinates of some point q at the contour be reliably solved by truncation methods, with guaranteed con-
Ln, and ln(h)) is an arc length at the point q. We thus obtain the vergence as truncation number Ntr ! 1.
system of N integral equations: Following the main steps, described above, we obtain the
final system of linear algebraic equations:
N ðp
X
G Rsj ðh; sÞ Zj ðsÞds5Vs ðhÞ; s51; . . . ; Nð1Þ
j51 2p nsn ð12dn0 Þ1rn rm lsjn;2m njm 5rn vsn ;
n50; 61; 62 . . . ; s51; 2; . . . ; N:
where Rsj denotes the distance between some point h on the s-th
boundary and a point s on the j-th boundary. Here, the unknown Fourier coefficients of the charge density
In operator formulation, each equation may be represented as function nsn are rescaled to ensure the compactness of
follows: Ax 5 B, where A is the matrix operator corresponding operator H:

DOI 10.1002/mop MICROWAVE AND OPTICAL TECHNOLOGY LETTERS / Vol. 56, No. 9, September 2014 2067
 Figure 4 Condition number graph for the square transmission line

The above regularized solution is only valid for the contours

with the smooth boundaries, so we replace the shapes with sharp
edges with contours generated by the superellipse equation:
Figure 2 Conformal coaxial transmission lines: (a) circular (n 5 n1 5  1 !21
n2 n3
n2 5 n3 5 2, m 5 4); (b) intermediate shape (n 5 4, m 5 4); (c) “near 1 m 1 m n1

square” (n 5 40, m 5 4); and (d) rotated inner conductor n~ð/Þ5 cos / 1 sin /
a 4 b 4

The results of truncation error calculation (see [17]) are

nsn shown in Figure 3 for the solution to the Laplace equation for
nsn 5 ; a structure with a square shield with side a equal to 1 and a
rn
square inner conductor with a side b equal to 0.2, embedded in
where rn 5jnj1=2 , when n 6¼ 0 and r0 51. More detailed descrip- the center of the shield (see Fig. 2). Figure 4 shows the condi-
tion of the algorithm can be found in [18]. tion number behavior in the same case. It should be noted that
the results are very accurate and stable: the condition number
3. NUMERICAL RESULTS has reached a limiting value even for small values of the trunca-
tion number. The solutions corresponding to the different struc-
The obtained solution is applied for the analysis of conformal
ture parameters (side length and rotation angle) have the similar
coaxial lines consisting of a central conductor and an outer con-
convergence rates.
ductor chosen so that the ratio between the distances from the
In this article, we briefly describe the applications of the
center to the outer and inner conductor along any radial line is a
developed algorithm to the problems involving noncanonical
constant k > 1 (Fig. 1).
shapes and a rotationally displaced inner conductor.
The numerical code was validated by comparing the obtained
First, for the illustrative purposes, we calculate the character-
results with known analytical and numerical solution for the
istic impedance of the conformal transmission line of the equi-
coaxial transmission lines [5] and square transmission lines with
lateral triangle cross-section, shown in Figure 5.
the square inner conductors [6,10]. Now let us illustrate the
The parameters which generate our approximation (with
above assertions about accuracy and convergence.
rounded corners) to the equilateral triangle are following: a 5
b, m 5 3, n1 5 26, n2 5 n3 5 48. The aspect ratio q 5 a/b
characterizes the degree of the filling the interior of the external
triangle by the inner conductor; from here on we refer to this
parameter as a filling coefficient. The dependence of the

Figure 5 Geometry of conformal transmission line of the equilateral

Figure 3 Truncation error graph for the square transmission line triangle cross-section: (a) conformal TL; (b) rotated inner conductor

2068 MICROWAVE AND OPTICAL TECHNOLOGY LETTERS / Vol. 56, No. 9, September 2014 DOI 10.1002/mop
 Figure 8 Characteristic impedance Zc versus filling coefficient q 5 a/
Figure 6 Characteristic impedance Zc versus the filling coefficient q b: circular coaxial line (black); conformal coaxial line with an intermedi-
ate cross-section (blue); near square conformal coaxial line (red). [Color
figure can be viewed in the online issue, which is available at wileyonli-
characteristic impedance Zc on the filling coefficient q in the nelibrary.com]
range 0.05  q  0.99 is shown in Figure 6.
The graph in Figure 6 reveals the qualitatively expected
behavior, usually explained by capacitance value increasing as previous cases, we use the superellipse Eq. (6) to approximate
filling coefficient grows. the square shape. The parameters we set are a 5 b, n 5 n1 5
The dependence of characteristic impedance on the rotation n2 5 n3 >> 1, m 5 4. The approximating accuracy entirely
angle a is plotted in Figure 7 for the three values of the filling depends on the parameter n: the bigger n is, the better a para-
coefficient: q 5 0.015, 0.3, 0.49. The triangle in triangle geome- metric curve matches with the square contour. It should be
try possesses a threefold symmetry, hence, it is enough to set noted that the parametric curve stays smooth for any n. For cal-
the range of rotation angle change as follows: 0  a  120 . culation of the capacitance, we used the parameter n 5 40 [Fig.
One can observe from the graph in Figure 7 that rotation of a 2(c)], that with Ntr 5 512 guaranteed the correct 4–5 decimal
small inner conductor has almost no influence on the character- places compared to the exact value for an ideal (with sharp cor-
istic impedance value (Zc is almost a straight line for q 5 0.15). ners) square conformal coaxial linencite {Bowman}. The low
For the larger q values, the impact of the relative positions of values of n produce the cylindrical coaxial line [Fig. 2(a), n 5
the two triangles on the characteristic impedance Zc becomes 2] and conformal coaxial line with an intermediate shape
more pronounced, reaching the maximum magnitude at q 5 between a circle and a square cross-section [Fig. 2(b), n 5 4].
0.49. The characteristic impedance for all these three types of
Now, we examine the conformal rectangular coaxial line
with the rotating inner conductor, shown in Fig. 2(d). As is in

Figure 7 Characteristic impedance Zc versus rotation angle a (triangle Figure 9 Square coaxial TL: characteristic impedance Zc versus rota-
in triangle TL). [Color figure can be viewed in the online issue, which is tion angle a. [Color figure can be viewed in the online issue, which is
available at wileyonlinelibrary.com] available at wileyonlinelibrary.com]

DOI 10.1002/mop MICROWAVE AND OPTICAL TECHNOLOGY LETTERS / Vol. 56, No. 9, September 2014 2069
 The solution to the final matrix system with a correct choice
of the truncation number can be obtained with any prespecified
accuracy. The composition of the solution guarantees the accu-
rate and stable calculation of the electrostatic characteristics of
the systems. The computational code (constructed in MATLAB)
is applicable for studies on a range of problems involving vari-
ous numbers of cylinders of different cross-sections and relative
positioning.
In this article, the developed method is applied to the
analysis of infinitely long conformal transmission lines of
various cross-sections. The impedance dependences on the
inner conductor relative size and rotation angle are
considered.

REFERENCES
1. R. Schinzinger and P.A.A. Laura, Conformal mapping: Methods and
Figure 10 Rectangular coaxial TL: characteristic impedance versus applications, Courier Dover Publications, New York, NY, 2003.
rotation angle. [Color figure can be viewed in the online issue, which is 2. M. Lukic, S. Rondineau, Z. Popovic, and D.S. Filipovic, Modelling
available at wileyonlinelibrary.com] of realistic rectangular coaxial line, IEEE Trans Microwave Theory
Tech 4/5 (2006), 2068–2076.
3. W. Lin and W. Changhua, Solutions of some transmission lines,
conformal coaxial lines depending on the filling coefficient is Microwave Opt Technol Lett 1 (1988), 163–166.
shown in Figure 8. 4. W.P. Calixto, B. Alvarenga, J.C. da Mota, et al., Electromagnetic
problems solving by conformal mapping: A mathematical operator
We conclude that when the inner conductor is small com-
for optimization, Math Probl Eng 2010.
pared to the outer conductor (q << 1), all three values Zcare 5. J.D. Jackson, Classical electrodynamics, Wiley, New York, NY,
very close to each other. The same behavior occurs at the other 1999, 808 p.
extreme, when 1 2 q << 1. The maximum deviation between 6. Y.M. Yakover, General analysis of shielded transmission lines with
all three graphs does not exceed 8% on the entire range (0 < q cylindrical geometry conductors, In: 27th European Microwave Con-
< 1). ference, Jerusalem, Israel, Vol. 2, 1997, pp. 955–958.5
Now, we examine the square in a square conformal coaxial 7. F. Bowman, Introduction to elliptic functions, Dover, New York,
line with rotating inner conductor [Fig. 2(d)]. 1961.
Similar to the case of the triangle in a triangle TL, rotation 8. H. Guckel, Characteristic impedances of generalized rectangular
of a relatively small (q 5 0.2) inner conductor in a square TL transmission lines, IEEE Trans Microwave Theory Tech 13/3 (1965),
270–274.
does not influence the characteristic impedance Zc across the
9. H.J. Riblet, Expansions for the capacitance of the Bowman squares,
whole range 0  a  180 (Fig. 9). Rotation of the relatively IEEE Trans Microwave Theory Tech 36/7 (1988), 1216–1219.
large inner conductor of the critical sizep(qffiffiffi 5 0.7), that passes 10. H.J. Riblet, Expansions for the capacitance of a square in a square
the wall at the minimum distance D512 2q  0:01 at a 5 45 with a comparison, IEEE Trans Microwave Theory Tech 44/2
(and 135 ), significantly impacts the characteristic impedance (1996), 338–340.
Zc. The minimum values of Zc occur exactly at the minimum 11. Q. Zheng W. Lin, F. Xie, and M. Li, Multipole theory analysis of a
distance D between the corner of the inner conductor and the rectangular transmission line family, Microwave Opt Technol Lett
wall of the external square (“diamond” positioning of the inner 18/6 (1998), 382–384.
conductor). 12. S. Seghier, N. Benabdallah, N. Benahmed, N. Benmostefa, and R.
In conclusion, we calculate the characteristic impedance Bouhmidi, Accurate closed-form formulas for the electromagnetic
parameters of squared coaxial lines, Int J Electron Commun 62/5
Zcfor the rectangular conformal coaxial line with aspect ratio a/
(2008), 395–400.
b 5 0.5 for both the external conductor and the rotating inner 13. M. Xikui, Analysis of the capacitance of rectangular shielded line
conductor (Fig. 10). In general, the dependence Zc(a) for rectan- with offset inner conductor, J Electron (China) 6/2 (1989), 140–147.
gular conformal coaxial line is similar to that of a square con- 14. M. Xikui, Characteristic impedance of specific transmission lines with
formal line, although one can notice that at the small values of offset inner conductors, J Electron (China) 7/3 (1990), 223–231.
the filling coefficient q 5 w/b << 1 Zc is more sensitive to 15. J.E. Cruz, Variable impedance coaxial device with relative relation
rotation of the inner conductor than in the case of the square between conductors, U S Patent 3,417,350, December 17, 1968.
conformal line. 16. S.S. Vinogradov, P.D. Smith, and E.D. Vinogradova, Canonical
problems in scattering and potential theory, part 1: Canonical struc-
4. CONCLUSION tures in potential theory, Chapman & Hall/CRC, London, UK, 2001.
17. S.S. Vinogradov, P.D. Smith, and E.D. Vinogradova, Canonical
The rigorous method for solution of two-dimensional potential problems in scattering and potential, theory part II: Acoustic and
problems for arbitrarily profiled cylinders is developed. In the electromagnetic diffraction by canonical structures, Chapman &
core of our approach, lies the idea of analytical regularization of Hall/CRC, London, UK, 2001.
integral equations containing singularities. The proposed regula- 18. G. Safonova, E. Vinogradova, Rigorous approach to analysis of 2-D
rization technique reduces them to the well-conditioned infinite electrostatic field problems for multi-conductor systems, In: Proceed-
system of algebraic equations of the second kind. The computa- ings of the XXXth URSI General Assembly and Scientific Sympo-
tion of the Fourier coefficients defining the matrix elements was sium (URSI GASS), 2012.
implemented using single and double fast Fourier transforms,
resulting in fast computational time. C 2014 Wiley Periodicals, Inc.
V

2070 MICROWAVE AND OPTICAL TECHNOLOGY LETTERS / Vol. 56, No. 9, September 2014 DOI 10.1002/mop

View publication stats