Beruflich Dokumente
Kultur Dokumente
A. Rashidian et al.: Dielectric Characterization of Materials using a Modified Microstrip Ring Resonator Technique
ABSTRACT
The goal of this study is to present a simple model based on the ring-resonator
technique to measure nondestructively the permittivity and loss tangent of dielectric
materials. The proposed measurement model utilizes a modified ring-resonator
technique in one-layer and two-layer microstrip configurations. This method eliminates
the requirement to metalize the samples and enables characterization of permittivity
and dielectric loss from 2 to 40 GHz. The effects of conductor and radiation losses that
may introduce significant errors in the calculation of the loss tangent, especially at very
high frequencies, are minimized. The measurement precision is evaluated by
comparing the results with those obtained by using two well-known standard
techniques. Uncertainties associated with the proposed model are addressed.
Index Terms Dielectric materials, dielectric measurements, ring resonators,
microstrip, permittivity measurement, dielectric loss tangent measurement.
1 INTRODUCTION
EXTENSIVE research was performed to measure dielectric
parameters and consequently introduce novel functional
materials in the past few years [1- 3]. Dielectric characterization
is a critical part of the qualification of materials for a wide range
of applications. The knowledge of the dielectric constant and loss
tangent is necessary for the design, analysis, and fabrication of
antennas and microwave circuits [4, 5]. By measuring dielectric
properties of agricultural and food products one can make very
useful interpretations regarding the quality of these products [6,
7]. Accurate determination of soil moisture distribution, which is
realized by measuring dielectric constant of soil, plays an
important role in the hydrologic, ecologic, and climatic models
[8]. Because of the variation of tissue permittivity, especially
between malignant and normal tissues, dielectric characterization
is also used in medical applications [9, 10].
Proposed by Troughton to measure dispersion of microstrip
transmission lines [11], microstrip ring resonators have attracted
considerable interest for electrical properties measurements of
dielectrics even at very high frequencies [12, 13]. The advantage
of this technique, which uses the resonance frequency and the Q
factor of the resonator to calculate the permittivity and the loss
tangent of the substrate, is its simplicity and accuracy due to the
absence of end-effects and high Q factor. Numerous analytical
and numerical models are presented in the literature to increase
Manuscript received on 14 October 2011, in final form 15 February 2012.
1393
2 THEORY OF MEASUREMENT
TECHNIQUE
Figure 1 illustrates the two-port microstrip ring resonator
configuration. The ring resonator with mean radius of r and
width of w is printed on a microwave substrate with thickness of
hs. The slab material with thickness of hd is superimposed on the
ring resonator and the differences in Q factor and resonant
frequency before and after slab insertion are used to determine
the dielectric constant and loss tangent of the material.
eff , 2
f r ,1
f r ,2
eff ,1
(1)
here eff,1 and eff,2 are the effective permittivity of the one- and
two-layer microstrip ring resonator, and fr,1 and fr,2 can be
extracted from the measured insertion loss of the ring
resonator with and without the slab, as shown in Figure 2.
Based on the analytical formulations of the multilayer
microstrip circuits [20], the relative permittivity of the slab,
r,2, can be found by using the effective permittivity of the
two-layer microstrip configuration and the relative
permittivity of the substrate, r,1,
r ,2
q2
q1 q 2 1
1 q1 2
(2)
eff , 2 q1 r ,1
QD, 2 QD,1 QU , 2 QU ,1
Figure 2. Insertion loss measurement for the 6 GHz ring resonator (r = 5.84
mm, hs = 0.254 mm, w = 0.71 mm) on TLY5 substrate (r = 2.2, tan =
0.0009) before and after insertion of SU-8 slab (hd = 1.73 mm).
where QU,1 and QU,2 are the unloaded Q factors of one- and twolayer configurations, computed using the measured BW [21]
QU
fr
BW
1
1 10
L
20
(4)
QD,1 and QD,2 are the dielectric Q factors of one- and two-layer
configurations, and L is the measured insertion loss in dB of
the ring at resonance. The dielectric Q factor of an N-layer
microstrip configuration, QD,N, is related to the loss tangents
of different layers [22]
N
i, N
tan i
i 1
1
Q D, N
(5)
1
p 2, 2
1
1 p1,1 p1, 2
tan 1
p 2, 2
D , 2 Q D ,1
(6)
1394
A. Rashidian et al.: Dielectric Characterization of Materials using a Modified Microstrip Ring Resonator Technique
Table 1. Physical parameters of ring resonators. The relative permittivity and loss tangent of the substrates are r = 6.15, tan = 0.0027 for RT/duroid
6006 and r = 2.2, tan = 0.0009 for TLY-5 with a few percents variation from 2 to 40 GHZ.
Substrate
Ring
Radius (r)
Gap (g)
Microstrip
Width (w)
Fundamental Resonant
Frequency
Material
Thickness (hs)
Copper Cladding
RT/duroid 6006
0.625 mm
17.8 m
11.2 mm
120 m
0.91 mm
2 GHz
TLY-5
0.381 mm
17.5 m
8.66 mm
100 m
1.16 mm
4 GHz
TLY-5
0.254 mm
17.5 m
5.84 mm
70 m
0.71 mm
6 GHz
where p1,1 can be found directly [22] and p1,2 and p2,2 are
derived by manipulating formulations reported elsewhere [20,
22]
p1, 2
r ,1
q1
eff , 2
p 2, 2
r ,2
q 2 1 q1 2
eff ,2 r , 2 1 q1 q 2 q 2 2
(7)
(8)
(10)
p1,1 f
2 P f 2 eff ,1 f
hS
r ,1q1 r , 2
r , 2 1 q1 q 2 q 2
1 G
2
Substituting equation (11) and the quasi TEM effective
permittivity of the two-layer structure into equation (9), the
1 q1 2
(12)
eff , 2 f A r ,1q1 r , 2
B
r , 2 1 q1 q 2 q 2
1 G P f . This
2 P f 1 G
where A
and B
1 P f 1 G
1 P f 1 G
equation is incorporated to derive frequency dependent filling
factors
p1,2 f Aq1
r ,1
(13)
eff ,2 f
q 2 1 q1 2
p 2, 2 f A
1 q q q 2
1
2
2
r ,2
r ,2
eff , 2 f
(14)
3 EXPERIMENTAL RESULTS
The two-layer microstrip ring resonator technique was used
to measure the dielectric properties of materials in this section.
The experimental results are compared with other
measurement methods in some cases to evaluate the accuracy
of the proposed method. Since the ring resonator parameters,
such as substrate thickness and permittivity, line width, ring
diameter, and coupling gaps may affect the accuracy of the
results [14], microstrip ring resonators of varying diameters
were designed and fabricated on different substrates. Table 1
shows the physical parameters of different rings. In all designs
the line width to ring radius ratio (w/r) of the ring resonator
structure is kept smaller than 0.2 to avoid excitation of
possible higher-order modes [21]. The first ten resonances of
the large rings (2 GHz and 4 GHz designs) and the first six
resonances of the small ring (6 GHz design) are within the
desired frequency range of measurement (up to 40 GHz), have
clear and sharp peaks (Q>100), and are suitable for
calculations to determine the electrical properties of the slab
materials. The slabs are superimposed on the rings and the
whole circuit is mechanically stabilized with small plastic
clips installed electrically far from the ring resonator. Figure 3
shows the experimental setup and the ring resonator circuit.
An Agilent 8722ES Vector Network Analyzer was used to
measure the insertion loss of the two-port ring resonators from
1 to 40 GHz. The universal substrate test fixture WK-3001-G
from Inter-continental Microwave was directly used to
1395
Relative Permittivity
3.6
3.4
3.2
(a)
3
10
15
20
25
Frequency (GHz)
30
35
40
30
35
40
(a)
0.05
(b)
Figure 3. (a) Microstrip ring resonators with different designs and fabricated
on different substrates; (b) experimental setup when transparent PMMA is
superimposed on the 2 GHz microstrip ring resonator.
0.04
0.03
0.02
0.01
10
15
20
25
Frequency (GHz)
(b)
Figure 4. (a) Dielectric constant and (b) loss tangent vs. frequency of SU-8.
The error bars are calculated based on uncertainties in resonant frequency,
Q-factor, microstrip width, substrate thickness, and substrate permittivity.
1396
A. Rashidian et al.: Dielectric Characterization of Materials using a Modified Microstrip Ring Resonator Technique
2.6
Relative Permittivity
2.8
Relative Permittivity
2.7
2.6
2.4
2.2
2.5
2.4
10
15
20
25
Frequency (GHz)
30
35
10
20
Frequency (GHz)
30
40
(a)
40
(a)
0.020
0.010
0.015
0.010
0.008
0.006
0.004
0.002
0.005
0
0
10
20
Frequency (GHz)
30
40
(b)
5
10
15
20
25
Frequency (GHz)
30
35
40
(b)
Figure 5. (a) Dielectric constant and (b) loss tangent vs. frequency of
PMMA. The error bars are calculated based on uncertainties in resonant
frequency, Q-factor, microstrip width, substrate thickness, and substrate
permittivity.
1397
Relative Permittivity
5
10
20
30
Frequency (GHz)
40
50
(a)
0.008
0.006
4
0.004
0.002
0
0
10
20
30
Frequency (GHz)
40
50
(b)
QC
for a different ring. The error bars are shown for all
measured points in Figures 6 and 7.
The errors for loss tangents are calculated based
on the uncertainties in the above mentioned parameters,
uncertainties in Q-factor measurements (5%), and also the
calculated uncertainty in slab permittivity (2-6%). The
errors are calculated for all measured points, and the
overall cumulative errors were on the order of 10% for all
samples. For instance, the minimum and maximum loss
tangent at 36 GHz are 0.034 and 0.042 for SU-8
(equivalent to 19% variation), 0.013 and 0.018 for PMMA
(equivalent to 28% variation), 0.0018 and 0.0033 for
RT/duroid 5880 (equivalent to 45% variation), and 0.0007
and 0.0013 for RO3006 (equivalent to 46% variation).
The errors are less for SU-8 than for PMMA and the other
two low loss materials, which can be attributed to the
1
QR
1
QU
1
QD
(15)
1398
A. Rashidian et al.: Dielectric Characterization of Materials using a Modified Microstrip Ring Resonator Technique
REFERENCES
[1]
[2]
[3]
[4]
[5]
[6]
[7]
[8]
[9]
[10]
[11]
[12]
[13]
[14]
[15]
[16]
[17]
[18]
[19]
[20]
ACKNOWLEDGMENT
[21]
[22]
1399