Progress In Electromagnetics Research, PIER 102, 267286, 2010

PLANAR TRANSMISSION LINE METHOD FOR CHARACTERIZATION OF PRINTED CIRCUIT BOARD DIELECTRICS
J. Zhang
CISCO Systems, Inc.
CA, USA
M. Y. Koledintseva
Missouri University of Science & Technology
Rolla, MO, USA
G. Antonini
Department of Electrical Engineering
University of LAquila
Poggio di Roio, 67040 AQ, Italy
J. L. Drewniak
Missouri University of Science & Technology
Rolla, MO 65401, USA
A. Orlandi
Department of Electrical Engineering
University of LAquila
Poggio di Roio, 67040 AQ, Italy
K. N. Rozanov
Institute for Theoretical and Applied Electromagnetics
Russian Academy of Sciences
Moscow 125412, Russia
Corresponding author: M. Y. Koledintseva ([email protected]).

268

Zhang et al.

AbstractAn effective approach to characterize frequency-dispersive

sheet materials over a wide RF and microwave frequency range based
on planar transmission line geometries and a genetic algorithm is
proposed. S-parameters of a planar transmission line structure with
a sheet material under test as a substrate of this line are measured
using a vector network analyzer (VNA). The measured S-parameters
are then converted to ABCD matrix parameters. With the assumption
of TEM/quasi-TEM wave propagation on the measured line, as well
as reciprocity and symmetry of the network, the complex propagation
constant can be found, and the corresponding phase constant and
attenuation constant can be retrieved. Attenuation constant includes
both dielectric loss and conductor loss terms. At the same time,
phase term, dielectric loss and conductor loss can be calculated
for a known transmission line geometry using corresponding closedform analytical or empirical formulas. These formulas are used to
construct the objective functions for approximating phase constants,
conductor loss and dielectric loss in an optimization procedure based
on a genetic algorithm (GA). The frequency-dependent dielectric
properties of the substrate material under test are represented as
one or a few terms following the Debye dispersion law.
The
parameters of the Debye dispersion law are extracted using the
GA by minimizing the discrepancies between the measured and the
corresponding approximated loss and phase terms. The extracted data
is verified by substituting these data in full-wave numerical modeling
of structures containing these materials and comparing the simulated
results with experimental.
1. INTRODUCTION
Development of simple and robust methods for wideband extraction
of frequency characteristics of planar sheet materials for various
electromagnetic applications is an important present-day problem. In
particular, characterization of dielectric substrates for printed circuit
boards (PCBs) is vital to achieve the first-pass success in modern highspeed digital system designs. When the on-board data rate is in the
Gbps (gigabits per second) range or higher, traces and discontinuities
including vias, AC coupling pads, and trace bends on a signal path have
to be modeled to catch the channel response accurately [13]. A static
field solver is not sufficient to model these discontinuities and traces,
and full-wave modeling tools have to be used. To build the full-wave
model for a given signal path, the detailed structures are known, but
the well-represented dielectric material properties of the corresponding
substrates are unknown. In general, the dielectric properties (relative

Progress In Electromagnetics Research, PIER 102, 2010

269

permittivity and loss tangent) used in the full-wave model come from a
PCB vendor with only one or two frequency points. However, dielectric
representations with either one or two frequency points for a PCB
substrate are not sufficient for accurate full-wave simulations, since
complex permittivity of a PCB substrate may vary substantially over
the wide frequency range. Besides, dielectric representation with only
one or two points may result in causality issues in full-wave modeling,
which causes the divergence problem in time-domain simulations.
Numerous techniques are known for characterization of dielectric
properties over different frequency bands [413]. Each technique
benefits a different type of materials over a certain frequency range.
The resonance techniques widely used in the past several decades to
characterize dielectric materials are accurate, but are narrowband [4
6]. Reference [7] extends the resonance techniques to a wideband
application by designing a complex structure on a PCB to cover
multi-resonant frequency points. The dielectric properties at the
corresponding frequency points are tuned by matching the numeric
resonant peak to the measurements. The procedure is complicated, and
the numerical tuning is cumbersome. In addition, this approach does
not measure complex permittivity of a material in the frequency range
of interest, since dielectric loss cannot be obtained. As for the coaxial
line techniques, they are good for measuring wideband properties
of materials homogeneously distributed over the cross-section of the
line [8], but they are not suitable for layered materials. Besides, it is
difficult to de-embed port effects in this type of techniques. Though
it is possible to retrieve dielectric constant and loss tangent of layered
materials directly from measurements using an impedance analyzer,
this technique is available only at low frequencies with a relatively
narrow frequency span [11].
A short-pulse propagation time-domain technique is used to obtain
dielectric properties for PCB substrate materials in wide range up
to 30 GHz [12]. However, this procedure is complex, while practical
manufacturing capabilities and an inherent signal-to-noise ratio of
time-domain measurement limit application possibilities as well. A
technique for wideband extraction of one-term Debye or Lorentizian
behavior of permittivity for PCB substrates directly from frequencydomain S-parameter measurement has been proposed in [13]. It is
based on using different planar transmission line structures, and is
applied to extraction of dielectric properties up to 5 GHz. Another
approach to extract dielectric properties [14] is based on measuring
dielectric loss and conductor loss for transmission lines, and it was also
tested up to 5 GHz. For an FR-4 material, the approximation of its
permittivity by single-term Debye frequency dependence at frequencies

270

Zhang et al.

above 10 GHz may be not accurate. It is known that dielectric

dispersion of polymer materials can be better approximated by ColeCole, Cole-Davidson, or Havriliak-Negami dispersion laws [15, 16]. In
addition, it is known that FR-4 type materials can be approximated by
so-called wideband Debye dependence, proposed in [17]. However,
this dependence contains logarithms of frequency and is not quite
convenient for wideband time-domain numerical modeling, such as
FDTD algorithms. Another known way to fit wideband frequency
characteristics of such materials is to apply multi-term Debye
dependence [1719],
() = +

n
X
i=1

je
i

,
1 + ji 0

(1)

where the static susceptibility i = si is the difference

between the static relative permittivity and the high-frequency relative
permittivity for the ith Debye term, i is the corresponding relaxation
constant, 0 is the free-space permittivity, and e is the effective
conductivity associated with the lowest frequency of interest. This
is convenient for representation in numerical codes using time-domain
representation.
Even if the dispersion law (e.g., the Debye dependence) for
a given dielectric is known, its parameters are typically unknown.
Characterization of dielectric materials then can be formulated as
an experimental determination of the parameters of the dispersion
law without getting the detailed interim information on the values of
material parameters over the frequency range of measurements. This
allows for simplifying the characterization procedure. In addition,
an important requirement for linear passive dielectric materials is
compliance with Kramers-Kroenig causality relations [20].
The present paper is aimed at the development of an effective
and convenient method to extract parameters of one- and multiterm Debye curves from measurements based on transmission line
losses and application of a genetic algorithm (GA). In the past few
years, application of GA for solving various electromagnetic problems
that require optimization or curve-fitting has gained popularity [21
28], including extraction dielectric properties of materials [13, 14, 29
31]. GA is the most reasonable way of curve-fitting when specifically
using rational-fractional functions, such as Debye terms, since it is
easily formulated and programmable, robust, efficiently converging to
a global minimum. It is important that curve-fitting using rational
fractional functions such as the Debye terms provides satisfying
Kramers-Kroenig causality relations.
In this paper, the results of extraction are shown for up to two

Progress In Electromagnetics Research, PIER 102, 2010

271

Debye terms, but an extension for more terms is quite straightforward.

The idea of the approach and the GA application are discussed in
Section 2. Section 3 contains formulations for different transmission
line structures. Three test cases are considered in Section 4. Singleterm Debye parameter extraction is demonstrated in a parallel-plate
structure and a microstrip structure up to 5 GHz. The parameters
of two-term Debye curves are extracted for a stripline structure in
the frequency range up to 20 GHz, where one-term Debye curve is
insufficient to fit an actual behavior of the dielectric substrates. Sparameters of the structures have been also modeled using full-wave
numerical simulations with extracted Debye parameters, and compared
with corresponding measurements. Section 5 contains conclusions.
2. PROPOSED APPROACH AND APPLICATION OF A
GENETIC ALGORITHM
The approach proposed in the paper to determine parameters of the
Debye dispersion law for dielectric substrates includes the following
steps: (a) S-parameter measurements, (b) calculation of phase
constant and loss based on analytical models for the particular
transmission lines, (c) comparison between the measured and modeled
values of and according to some accepted criteria in the
frequency range of interest, and (d) correction of the dispersion law
parameters until these criteria are satisfied. The correction is fulfilled
using a genetic algorithm, which has recently gained popularity for
global optimization [21]. This approach is straightforward to extract
parameters for a single- or a two-term Debye material. It can also
be used for multi-term Debye and more complex dispersion laws [16],
including Lorentzian-type characteristics [19, 32], both for permittivity
and permeability of magnetic and magneto-dielectric materials. In
this method, S-parameters of planar transmission lines (parallel-plate,
stripline, and microstrip) with dispersive dielectric substrates are
measured using a VNA. The measured S-parameters can be converted
into the ABCD matrix parameters, and the complex propagation
constant = + j in a passive reciprocal network can be calculated
as

arccos h A D
=
.
(2)
l
if the network is asymmetrical in the general case [33].
The accuracy of the permittivity extraction strongly depends on
the accuracy of the measured raw S-parameters, length of the line l,
and correct separation of dielectric loss d from conductor loss c ,
since total loss is = c + d . If a zero Through-Reflect-Line (TRL)

272

Zhang et al.

calibration is used in the measurement to remove the port effects,

the length l is obtained by subtracting the through length from the
actual length of the test line.
The proposed extraction technique uses a GA optimization. In
this technique, the optimization goal is the restoration of the total
attenuation and the propagation constant , obtained from the
measured S-parameters. The evaluated at each frequency point
attenuation and propagation constants are related with dielectric
properties of substrate through analytical and/or semi-empirical
formulas for all the planar transmission lines (parallel-plate, microstrip,
and stripline) with a single TEM, or quasi-TEM mode. Interim Debye
dielectric parameters are used in each iteration cycle of the GA search.
The objective function for optimization is calculated through the root
mean square value with respect to all N frequency points
v
uN n
o
X
1u
= t
[c ]2 + [d ]2 + [ ]2 ,
(3)
N
i=1

where c , d , and are the normalized deviations between the

measured (with superscript m) and evaluated values (with superscript
e)
c =

|cm ce |
,
max |cm |

d =

|dm de |
,
max dm

and =

| m e |
.
max | m |

(4)

A fitness index p is assigned to each set of parameters under GA

evaluation at that iteration [21]. This index distinguishes how well
each individual taken from a solution pool competes with its peers.
An individual with a higher p value is much closer to the real solution,
and has a higher chance of remaining in the search pool. Based on
the fitness index, only good individuals are allowed to generate new
offspring with higher fitness indices. To maintain diversities in the
GA search pool, a small perturbation, or mutation parameter (0.7%)
is applied to the new offspring to avoid missing the possible good
genes. As soon as all the chosen criteria are satisfied, the global
optimal solutions are reached. Herein, the fitness index is chosen as
1/3
1
p=
.
(5)

The power (1/3) in (5) is used to shrink the dynamic range of p,

and has been found by extensive numerical experimenting to be a
reasonable one for all three geometries stripline, microstrip, and
parallel-plate. We have found that reasonable population size is in the

Progress In Electromagnetics Research, PIER 102, 2010

273

range of 240400, and the median value of 320 has been chosen as the
optimal population size. The cross-over parameter is chozen as 75%.
Another way to simulate dielectric loss and conductor loss is to
assume that the dielectric part is proportional
to frequency (d ),

while skin-effect part behaves as c . This is valid only for

perfectly smooth surfaces, while taking into account rough surfaces
requires some frequency correction [33, 34]. However, assuming that
metal surfaces are basically smooth, the total loss can be approximated
as

(6)
= a + b .
The coefficients a and b can be retrieved using another genetic
algorithm to approximate the dependence () retrieved from Sparameter measurements. If surface roughness is included in conductor
loss, the frequency dependence of total loss is more complex than (6),
and roughness
may contribute to dielectric -term as well as smooth
conductor -term, and higher powers of frequency [35]. How to
correctly split conductor loss contributions from dielectric loss in a
rough conductor is a serious separate problem, and it is beyond the
scope of the present paper.
3. FORMULATION FOR PLANAR TRANSMISSION
LINES
Analytical or semi-empirical formulas known from literature are used
for conversion of complex propagation constant to dielectric parameters
of parallel-plate, stripline, and microstrip structures. Though these
models are generally approximate, they are accurate enough for
dielectric parameter extraction in the frequency range of interest,
where TEM (or quasi-TEM) propagation takes place. Limitations
of parameter extraction for the transmission line structures under
consideration are discussed.
It should be mentioned that though the types of lines, other than
those with TEM (quasi-TEM) modes, have not been considered in this
particular paper, the presented methodology can be extended to the
other regular waveguide structures. It is important that and are
extracted through measurements, and an adequate model correlating
these propagation parameters with dielectric properties of the media
under study should be available [36, 37].
3.1. Parallel-plate Structure
A parallel-plate structure shown in Fig. 1 is the simplest transmission
line. For the TM0 mode in the parallel-plate waveguide, Ez = 0,

274

Zhang et al.

Figure 1. Cross-section of the parallel-plate structure.

and both electric and magnetic fields are transverse to the guidance
direction. Therefore, TM0 mode is also the TEM mode. Since the
cut-off frequency of TEM mode is zero, it is often referred to as the
lowest (dominant, or fundamental) mode [38].
Formulas for and are based on the assumption that the higher
order modes and fringing fields are ignored. This is true for a parallelplate structure only over a limited frequency range, depending on its
dimensions and the substrate dielectric. Hence, a set of parallel-plate
structures for studying dielectrics in each specific frequency range may
be needed. The assumptions given herein imply two rules: (1) the
ratio w/d must be large enough, so that the perfect magnetic boundary
condition is applicable for neglecting the fringing fields; (2) the cutoff frequency of the first higher-order mode associated with perfect
electrical conductor boundary condition limits the thickness d of the
dielectric medium between two plates. Thus, the first higher-order
modes TE1 and TM1 have the cut-off frequency fcutof f = c /(2d),
where c is the wave velocity in the dielectric.
The phase constant for the TEM wave in a parallel-plate
transmission line is
p

= 0 0 r 0r
(7)
where r = 1 is the relative permeability of the non-magnetic substrate
material, and 0r is the real part of r in (1), which is an interim value
during the GA extraction. If conductors of the parallel-plate line are
smooth, and if there is the only TEM mode propagating, then the
conductor loss is [39]
Rs
c =
,
(8)
d
q
p
where = 120 0r is the TEM wave impedance, Rs = 0 /(2c ) is
r

the surface resistance of conductors, and d is the thickness of dielectric

Progress In Electromagnetics Research, PIER 102, 2010

275

substrate. Assuming that the substrate dielectric is low-dispersive and

low-loss, the dielectric loss is [39]
d =

tan
00
, where tan = r0 .
2
r

(9)

If loss and dispersion in a dielectric substrate is considerable loss, then

the attenuation constant can be calculated as
p
p

d = 0 0 0r 4 1 + (tan )2 sin(/2).
(10)
At the same time, the propagation constant will be calculated as
p
p

= 0 0 0r 4 1 + (tan )2 cos(/2).
(11)
These formulas are derived from the rigorous expressions for complex
propagation constant for TEM wave propagating in a dielectric
medium.
3.2. Microstrip Transmission Line
The calculation of and for a microstrip line (Fig. 2), is analogous
to that for the parallel-plate geometry.
Strictly speaking, the
electromagnetic field in a microstrip is a hybrid TE-TM mode, and
wave propagation is not completely contained within a substrate.
However, it can be considered as a quasi-TEM mode for the structures
with electrically thin dielectric substrates (h/diel 1). The phase
term for the microstrip line filled with a comparatively low-loss
dielectric is
p
= 0 0 0e ,
(12)
where the effective permittivity 0e is used instead of real part of
permittivity 0r for the substrate dielectric. The expression for effective

Figure 2. Cross-section of the microstrip structure.

276

Zhang et al.

permittivity in a microstrip line can be found in textbooks, for

example, [39, P. 162], or [40].
0 1
0 + 1
0e = r
+ qr
.
(13)
2
2 1 + 12h
w

The attenuation of the microstrip line due to the finite

conductivity in the smooth conductor is
Rs
c =
,
(14)
wZw
p
where Rs = 0 /(2c ) is the surface resistance of the conductor, and
Zw is the wave impedance of the line [39]. The dielectric attenuation
is calculated for the structure as [39]

0 0 0e 1
0
d =
0 r tan ,
p
(15)
2
0e r 1
00

where tan = r0 is the loss tangent of the dielectric material. If the

r
dimensions of the structure are known, the total loss = c + d
and can be calculated using the above formulas, or through (6). If
the material under study is substantially lossy and dispersive, then
formulas (10) and (11) should be applied for calculating and , but
substituting 0r with 0e in tan . Higher-order modes, surface waves
in the metal-dielectric-air structure, and radiation effects in the open
dielectric structure are not taken into account, and these factors limit
frequency range for permittivity extraction [41].
3.3. Stripline Structure
Equations (7)(11) can also be used to calculate and for a stripline
(Fig. 3). If dimensions of a stripline are given, the conductor loss in a

Figure 3. Cross-section of the stripline structure.

Progress In Electromagnetics Research, PIER 102, 2010

277

smooth conductor can be estimated, for example, using the incremental

inductance rule proposed by Wheeler [42, 43]. Wheelers formulas are
valid only for a single TEM mode in a stripline with the assumption of
fringing fields and edge coupling negligible. In reality, the higher-order
modes can be suppressed by limiting the spacing between the reference
plates of the strip to the quarter wavelength (/4). These assumptions
are often true in multilayer PCBs, where t b and b /4 (see
Fig. 3).
4. MEASUREMENTS AND CASE STUDIES
Three structures have been built and tested. Two structures, parallelplate and microstrip, made of the same double-sided copper-clad FR-4
sheet have been tested in the frequency range of 100 MHz5 GHz. This
is needed for verifying the consistency while the proposed approach
is under validation. Another study is a stripline embedded in an 8layer PCB. It is shown that an appropriate TRL calibration allows for
accurate extracting of two-term Debye curves up to 20 GHz.
4.1. Parallel-plate and Microstrip Transmission Lines
The microstrip and the parallel-plate were made of the copper-clad FR4 and cut from the same sheet sample. The parallel-plate structure had
the dimensions of 71.36 mm (length) 19.80 mm (width) 1.25 mm
(height). The dielectric spacing (FR-4) was 1.05 mm, and the thickness
of the copper was 0.1 mm. Two SMA connectors were symmetrically
mounted at the both ends of the structure in the long direction, and
the distance between the centre conductors of the SMA was 63.4 mm.
The dimensions of the microstrip line with the identical FR-4 material
were 69.00 mm 19.80 mm 1.25 mm, and the distance between the
centre conductors of the two SMA connectors was 61 mm.
S-parameters were measured using an HP 8753D VNA in the
frequency range of 100 MHz5 GHz with 1601 sampling frequency
points.
Prior to the measurements, SOLT (Short-Open-LoadThrough) calibration was implemented. The impact of the electrical
length of the SMA connectors upon the measured S-parameters
was removed by the port extension after the SOLT calibration.
However, the discontinuities due to the SMA transitions still affected
measurements. The measured S-parameters were converted into the
ABCD parameters, and and were calculated as real and imaginary
parts of the (2). The Debye parameters (Table 1) were extracted
for both lines using the GA procedures. The real and the imaginary
parts of the corresponding relative permittivity, including effective
conductivity of the dielectrics, are plotted in Fig. 4.

278

Zhang et al.
4.56
Parallel-plate
Microstrip

4.54

4.5

'

4.52

4.48
4.46

4.42
0.1

0.5
Parallel-plate
Microstrip
0.4

Frequency (GHz)

0.3
0.2

'' including
r
e

4.44

0.1

0.1

0
5

Frequency (GHz)

Figure 4. Real and imaginary parts of the extracted permittivity

from the parallel-plate and the microstrip structures.
Table 1. Extracted Debye parameters for parallel-plate and microstrip
transmission lines.
Structure
Parallel-plate
Microstrip

s
4.504
4.530

4.420
4.398

(ps)
46.37
57.22

e (mS/m)
2.531
2.351

The difference between the extracted permittivity (both real and

imaginary parts) for two structures is less than 0.025 in the frequency
range of 100 MHz5 GHz. The differences in the extracted parameters
can be explained by some tolerances on geometrical parameters, while
building test structures. Besides, though FR-4 samples are very close
to each other in their dielectric parameters, they might be not identical
because of the inhomogeneity of FR-4. Another source of discrepancy
for the extracted dielectric parameters may be associated with the fact
that in the extraction procedure the conductor surface roughness has
been neglected, and the contribution of conductor roughness depends
on the geometry of the line.
This comparison verifies the consistency of the proposed method
since both the test structures are cut from the same FR-4 sample sheet,
and validates the Debye parameter extraction. The extracted Debye
parameters (Table 1) are then used in the FDTD (finite-difference
time-domain) numerical model for the corresponding parallel-plate and

Progress In Electromagnetics Research, PIER 102, 2010

279

0
-2

-1
-2

-6

|S21| (dB)

|S 21 | (dB)

-4

-8

-3

-10

-4

-12

-5

-14
-16

FDTD modeled
Measured
0.5

1.5

2
2.5
3
3.5
Frequency (GHz)

FDTD modeled
Measured
4

4.5

Figure 5.
|S21 | comparison
between the measurement and
the full-wave modeling for the
parallel-plate transmission line.

-6

0.5

1.5

2
2.5
3
3.5
Frequency (GHz)

4.5

Figure 6.
|S21 | comparison
between the measurement and
the full-wave modeling for the
microstrip transmission line.

microstrip lines. The SMAs are modeled as thin-wires [44], and the
surface impedance boundary condition algorithm is used to model the
conductor loss [45]. Figs. 5 and 6 show that the maximum difference
between the FDTD simulated and measured |S21 | for both the parallelplate structure and the microstrip structure in the frequency range
from 100 MHz to 5 GHz is less than 1 dB. The SMA port effects
are partially included in the extracted Debye parameters, and this
may lead to discrepancy between the full-wave modeling and the
measurements.
4.2. Stripline
A TRL calibration pattern and a test line for the study of stripline
structure are designed in an 8-layer PCB on layer 5 with solid reference
plane on layers 4 and 6. Three line standards are built to support three
different frequency bands of 200 MHz930 MHz, 930 MHz4.3 GHz, and
4.3 GHz20 GHz. The PCB board dimensions are 264 mm (length)
248 mm (width) 2.69 mm (thickness). The total length of the
stripline after moving the TRL calibration reference plane back into
the test line is 202.6 mm. The cross-sectional dimensions of the test
line, referring to Fig. 3, are t = 0.03 mm, b = 0.75 mm, w = 0.32 mm,
d = 7.3 mm and s = 0.007 mm. The frequency range of interest is
from 200 MHz to 20 GHz. According to [46, 47], the calculated cut-off
frequency of the first higher-order mode of the stripline is 82 GHz, and
the stripline supports TEM wave propagation over the entire frequency
range of interest.

280

Zhang et al.

Table 2. Extracted two-term Debye parameters for the stripline.

s1
4.081

s2
4.068

1 (ps)
82.12

2 (ps)
5.712

3.95

e (mS/m)
1.136

4.25
Stripline
4.2

r'

4.15
4.1
4.05
4
Stripline 0.2

r''

0.15

0.1

0.1

1
Frequency (GHz)

10

0.05
20

Figure 7. Real and imaginary parts of the extracted permittivity

from the stripline structure.
The measurement was performed on an HP 8720ES VNA
with ATN-4112 S-parameter test set. The TRL calibration was
implemented before the measurement. The number of sampling
points was 201, 801, and 1601, for the frequency spans of 200 MHz
930 MHz, 930 MHz4.3 GHz, and 4.3 GHz20 GHz, respectively. The
total number of sampling points was 2601 over the entire frequency
range of interest, which was sufficient for GA extraction. Since a TRL
calibration is based on the standards of Through, Reflect, and
Line to characterize the error model including both VNA and the
test structure, errors due to the imperfections of Short, Open, and
Load used in the SOLT calibration are excluded from measurements.
Moreover, the TRL calibration moves the measurement reference plane
inside the structure under test. The higher-order modes and port
parasitics are eliminated from the measurements. For the stripline
case, the measurement reference plane is moved 0.5 inch inside of the
test line at each end from the SMA centre conductor. The extracted
two-term Debye parameters and the effective conducting e for the

Progress In Electromagnetics Research, PIER 102, 2010

281

substrate material are given in Table 2, and the real and the imaginary
parts of the extracted permittivity are plotted in Fig. 7.
These Debye parameters are used in a full-wave simulation tool,
which is the CST Microwave Studio realized on the finite integration
technique (FIT) [48]. The magnitude and phase of S21 obtained
by the numerical simulation and measurements are shown in Figs. 8
and 9, respectively. The maximum difference of the |S21 | is less than
0.7 dB over the frequency range up to 20 GHz, and the phases almost
coincide. This comparison validates the correctness of the extracted
Debye parameters and confirms that the proposed method works well.
In this extraction, the port effects are removed.
0

200

-2

150
100
S21 phase ( o )

|S21| (dB)

-4
-6
-8

-10
-12
-14

FIT Modeled
Measured

50
0
-50

-100
FIT Modeled
Measured
2
4
6

-150
8
10
12
14
Frequency (GHz)

16

18

20

Figure 8.
|S21 | comparison
between the measurement and
the full-wave modeling for the
stripline structure.

-200

8
10
12
14
Frequency (GHz)

16

18

20

Figure 9. S21 phase comparison

between the measurement and
the full-wave modeling for the
stripline structure.

5. CONCLUSIONS
The presented approach to extract Debye parameters for dispersive
dielectric substrates in planar transmission line structures is based
on approximating complex propagation constant by tuning the
Debye parameters in the analytical/empirical models for a dielectric.
Dielectric and conductor loss, obtained from measured S-parameters,
serve as target data to be approximated in a genetic algorithm.
Parameter extraction for both one- and two-term Debye dependencies
has been tested in the study.
Full-wave FDTD/FIT modeling
that used the extracted Debye terms and the measurements were
compared, and good agreement was achieved. The proposed approach
is straightforward and convenient to use. However, the accuracy of the
extracted Debye parameters is directly related to the accuracy of the

282

Zhang et al.

S-parameters measurement, which can be seen from the 5-GHz and the
20-GHz test cases. In the 5-GHz case (parallel-plate and microstrip),
port effects are partially embedded in the extracted Debye parameters,
and the maximum difference between the measured and the full-wave
modeled |S21 | is of 1 dB. For the 20-GHz case (stripline), port effects are
de-embedded from the Debye parameters, and the maximum difference
seen is 0.7 dB up to 20 GHz for |S21 |.
REFERENCES
1. Nikellis, K., N. K. Uzunoglu, Y. Koutsoyannopoulos, and
S. Bantas, Full-wave modeling of stripline structures in multilayer
dielectrics, Progress In Electromagnetics Research, PIER 57,
253264, 2006.
2. Wu, B. and L. Tsang, Full-wave modeling of multiple vias
using differential signaling and shared antipad in multilayered
high speed vertical interconnects, Progress In Electromagnetics
Research, PIER 97, 129139, 2009.
3. Bernardi, P., R. Cicchetti, G. Pelosi, A. Reatti, S. Selleri,
and M. Tatini, An equivalent circuit for EMI prediction in
printed circuit boards featuring a straight-to-bent microstrip line
coupling, Progress In Electromagnetics Research B, Vol. 5, 107
118, 2008.
4. Saito, S. and K. Kurokawa, A precision resonance method for
measuring dielectric properties of low-loss solid materials in the
microwave region, Proceedings of the IRE, Vol. 44, No. 1, 3542,
1956.
5. Du, S., A new method for measuring dielectric constant using
the resonant frequency of a patch antenna, IEEE Trans. Microw.
Theory Tech., Vol. 34, No. 9, 923931, Sep. 1986.
6. Abdulnour, J., C. Akyel, and K. Wu, A generic approach
for permittivity measurement of dielectric materials using a
discontinuity in a rectangular waveguide or a microstrip line,
IEEE Trans. Microw. Theory Tech., Vol. 43, No. 5, 10601066,
1995.
7. Holzman, E. L., Wideband measurement of the dielectric
constant of an FR4 substrate using a parallel-coupled microstrip
resonator, IEEE Trans. Microw. Theory Tech., Vol. 54, No. 7,
31273130, 2006.
8. Huang, J., K. Wu, and C. Akyel, Characterization of highly
dispersive materials using composite coaxial cells, electromagnetic

Progress In Electromagnetics Research, PIER 102, 2010

9.

10.
11.

12.

13.

14.

15.
16.
17.

18.

283

analysis and wideband measurement, IEEE Trans. Microw.

Theory Tech., Vol. 44, No. 5, 770777, 1996.
Baker-Jarvis, J., M. D. Janezic, B. F. Riddle, R. T. Johnk,
P. Kabos, C. L. Holloway, R. G. Geyer, and C. A. Grosvenor,
Measuring the permittivity and permeability of lossy materials:
Solid, metals, building materials, and negative-index materials,
NIST Technical Note 1536, Boulder, CO, USA, Dec. 2004.
Afsar, M. N., J. B. Birch, and R. N. Clarke, The measurement of
the properties of materials, Proc. IEEE, Vol. 74, No. 1, 183199,
1986.
Klein, K. and J. C. Santamarina, Methods for broad-band
dielectric permittivity measurements (Soil-Water Mixtures, 5 Hz
to 1.3 GHz), Geotechnical Testing Journal, Vol. 20, No. 2,
Jun. 1997.
Deutsch, A., T. Winkel, G. V. Kopcsay, C. W. Surovic, B. J. Rubin, G. A. Katopis, B. J. Chamberlin, and R. S. Krabbenhoft,
Extraction of r (f ) and tan (f ) for printed circuit board insulatiors up to 30 GHz using the short-pulse propagation technique,
IEEE Trans. Adv. Packag., Vol. 28, No. 1, 412, Feb. 2005.
Zhang, J., M. Y. Koledintseva, J. L. Drewniak, D. J. Pommerenke,
R. E. DuBroff, Z. Yang, W. Chen, K. N. Rozanov, G. Antonini,
and A. Orlandi, Reconstruction of dispersive dielectric properties
for PCB substrates using a genetic algorithm, IEEE Trans.
Electromagn. Compat., Vol. 50, No. 3, 704714, Aug. 2008.
Zhang, J., M. Y. Koledintseva, D. P. Pommerenke, J. L. Drewniak,
K. N. Rozanov, G. Antonini, and A. Orlandi, Extraction of
dispersive material parameters using vector network analysers
and genetic algorithms, Proc. IEEE Instrumentation and
Measurement Technology Conference, 462467, Sorrento, Italy,
Apr. 2006.
Hilfer, R., H-function representation for stretched exponential
relaxation and non-Debye susceptibilities in glassy systems, Phys.
Rev. E, Vol. 65, 061510, 2002.
Jonscher, A. K., Dielectric Relaxation in Solids, Chelsea Dielectric
Press, 1983.
Djordjevic, A. R., R. M. Biljic, V. D. Likar-Smiljanic, and
T. K. Sarkar, Wideband frequency-domain characterization of
FR-4 and time-domain causality, IEEE Trans. on Electromag.
Compat., Vol. 43, No. 4, 662667, Nov. 2001,
Koledintseva, M. Y., J. Wu, J. Zhang, J. L. Drewniak, and
K. N. Rozanov, Representation of permittivity for multi-phase
dielectric mixtures in FDTD modeling, Proc. IEEE Symp.

284

19.

20.
21.
22.
23.

24.

25.

26.

27.
28.
29.

Zhang et al.

Electromag. Compat., Vol. 1, 309314, Santa Clara, CA, USA,

Aug. 2004.
Koledintseva, M. Y., K. N. Rozanov, A. Orlandi, and
J. L. Drewniak, Extraction of the Lorentzian and Debye
parameters of dielectric and magnetic dispersive materials for
FDTD modeling, J. Electr. Eng., IEE Slovak, Vol. 53, No. 9,
97100, 2002.
Landau, L. D. and E. M. Lifshitz, Electrodynamics of Continuous
Media, 256265, Pergamon Press, 1960.
Rahmat-Samii, Y. and E. Michielssen, Electromagnetic Optimization by Genetic Algorithms, 193, Wiley, 1999.
Haupt, R. L. and D. H. Werner, Genetic Algorithms in
Electromagnetics, IEEE Press, Wiley, 2007.
Mittal, G. and D. Singh, Critical analysis of microwave specular
scattering response on roughness parameter and moisture content
for bare periodic rough surfaces and its retrieval, Progress In
Electromagnetics Research, PIER 100, 129152, 2010.
Liu, B., L. Beghou, L. Pichon, and F. Costa, Adaptive genetic
algorithm based source identification with near-field scanning
method, Progress In Electromagnetics Research B, Vol. 9, 215
230, 2008.
Chen, H. T., G.-Q. Zhu, and S.-Y. He, Using genetic algorithm
to reduce the radar cross section of three-dimensional anisotropic
impedance object, Progress In Electromagnetics Research B,
Vol. 9, 231248, 2008.
Agastra, E., G. Bellaveglia, L. Lucci, R. Nesti, G. Pelosi,
G. Ruggerini, and S. Selleri, Genetic algorithm optimization of
high-efficiency wide-band multimodal square horns for discrete
lenses, Progress In Electromagnetics Research, PIER 83, 335
352, 2008.
Xu, Z., H. Li, Q.-Z. Liu, and J.-Y. Li, Pattern synthesis
of conformal antenna array by the hybrid genetic algorithm,
Progress In Electromagnetics Research, PIER 79, 7590, 2008.
Meng, Z., Autonomous genetic algorithm for functional
optimization, Progress In Electromagnetics Research, PIER 72,
253268, 2007.
Zhang, Z., T. Hagfors, E. Nielsen, G. Picardi, A. Mesdea, and
J. J. Plaut, Dielectric properties of the Martian south polar
layered deposits: MARSIS data inversion using Bayesian inference
and genetic algorithm, J. Geophys. Res., Vol. 113, E05004, May
2008.

Progress In Electromagnetics Research, PIER 102, 2010

285

30. Diaz-Morcillo, A., J. Monzo-Cabrera, M. E. Requena-Perez,

and A. Lozano-Guerrero, Application of genetic algorithms
in the determination of dielectric properties of materials at
microwave frequencies, Lecture Notes in Computer Science
(LNCS), Vol. 4528, 608617, Nature Inspired Problem-Solving
Methods in Knowledge Engineering, Springer, Berlin, 2007.
31. Oswald, B., D. Erni, H. R. Benedickter, W. Bachtold, and
H. Fluhler, Dielectric properties of natural materials, IEEE Int.
Symp. Antennas and Propagation Society (APS), Vol. 4, 2002
2005, Jun. 2126, 1998.
32. Koledintseva, M. Y., J. L. Drewniak, D. J. Pommerenke,
K. N. Rozanov, G. Antonini, and A. Orlandi, Wide-band
Lorentzian media in the FDTD algorithm, IEEE Trans. on
Electromag. Compat., Vol. 47, No. 2, 392398, May 2005.
33. Koul, A., P. K. R. Anmula, M. Y. Koledintseva, J. L. Drewniak,
and S. Hinaga, Improved technique for extracting parameters of
low-loss dielectrics on printed circuit boards, Proc. IEEE Symp.
Electromag. Compat., 191196, Austin, TX, Aug. 1721, 2009.
34. Braunisch, H., X. Gu, A. Camacho-Bragado, and L. Tsang,
Off-chip rough-metal-surface propagation loss modeling and
correlation with measurements, IEEE Electronic Components
and Technology Conference, 785791, 2007.
35. Koledintseva, M. Y., A. Koul, P. K. R. Anmula, J. L. Drewniak,
S. Hinaga, E. Montgomery, and K. N. Rozanov, Separating
dielectric and conductor loss for rough striplines in printed circuit
boards, Progress In Electromagnetics Research Symposium
Abstracts, 213, Moscow, Russia, Aug. 1821, 2009.
36. Janezic, M. D. and J. A. Jargon, Complex permittivity
determination from propagation constant measurements, IEEE
Microwave and Guided Wave Letters, Vol. 9, No. 2, 7678,
Feb. 1999.
37. Baker-Jarvis, J., E. Vanzura, and W. Kissick, Improved technique for determining complex permittivity with the transmission/reflection method, IEEE Trans. Microw. Theory Techn.,
Vol. 38, 10961103, Aug. 1990.
38. Hoffman, R. K., Handbook of Microwave Integrated Circuits,
Artech House, 1987.
39. Pozar, M., Microwave Engineering, 2nd edition, Wiley, 1998.
40. Schneider, M. V., Microstrip lines for microwave integrated
circuits, The Bell System Technical Journal, Vol. 48, No. 56,
14211444, 1969.

286

Zhang et al.

41. Gardiol, F. and K. Chang, Microstrip Circuits, 3358, Wiley,

1994.
42. Wheeler, H. A., Transmission line properties of a strip on a
dielectric sheet on a plane, IEEE Trans. Microw. Theory Tech.,
Vol. 25, No. 8, 631647, 1977.
43. Wadell, B. C., Transmission Line Design Handbook, 129131,
Artech House, 1991.
44. Hockanson, D. M., J. L. Drewniak, T. H. Hubing, and
T. P. van Doren, FDTD modeling of thin wires simulating
common-mode radiation from structures with attached cables,
Proc. IEEE Symp. Electromag. Compat., 168173, Atlanta, GA,
USA, Aug. 1995.
45. Wang, C., J. L. Drewniak, and M. Li, FDTD modeling of skin
effect, Proc. IEEE Symp. Electromag. Compat., 246249, Beijing,
China, May 2002.
46. Mongia, R., I. Bahl, and P. Bhartia, RF and Microwave CoupledLine Circuits, 8390, Artech House, 1999.
47. Zhang, J., J. L. Drewniak, D. P. Pommerenke, R. E. DuBroff,
Z. Yang, W. Cheng, J. Fisher, and S. Camerlo, Signal
link-path characterization up to 20 GHz based on a stripline
structure, Proc. of the 2006 IEEE International Symposium on
Electromagnetic Compatibility, Vol. 2, 356361, Portland, OR,
Aug. 2006.
48. Clemens, M., S. Drobny, and T. Weiland, Time integration
of slowly-varying electromagnetic field problems using the
finite integration technique, Scientific Computing in Electrical
Engineering, U. van Reinen, M. Guenther, and Hecht (eds.), 63
70, Springer Verlag, 2001.