close

Вход

Забыли?

вход по аккаунту

?

Feasibility study to increase the sensitivity of a microwave microstripline bandpass filter based biosensor for the detection of bacteria in water

код для вставкиСкачать
UNIVERSITY OF NEVADA, RENO
Feasibility Study to Increase the Sensitivity of a Microwave
Microstripline Bandpass Filter Based Biosensor for the Detection of
Bacteria in Water
A thesis submitted in partial fulfillment of the requirements of the degree of
MASTER OF SCIENCE IN ELECTRICAL ENGINEERING
by
Sasi Kiran Kambavalasa
Dr. Indira Chatterjee / Thesis Advisor
May, 2007
Reproduced with permission of the copyright owner. Further reproduction prohibited without permission.
UMI Number: 1447811
INFORMATION TO USERS
The quality of this reproduction is dependent upon the quality of the copy
submitted. Broken or indistinct print, colored or poor quality illustrations and
photographs, print bleed-through, substandard margins, and improper
alignment can adversely affect reproduction.
In the unlikely event that the author did not send a complete manuscript
and there are missing pages, these will be noted. Also, if unauthorized
copyright material had to be removed, a note will indicate the deletion.
®
UMI
UMI Microform 1447811
Copyright 2007 by ProQuest Information and Learning Company.
All rights reserved. This microform edition is protected against
unauthorized copying under Title 17, United States Code.
ProQuest Information and Learning Company
300 North Zeeb Road
P.O. Box 1346
Ann Arbor, Ml 48106-1346
Reproduced with permission of the copyright owner. Further reproduction prohibited without permission.
THE GRADUATE SCHOOL
University o f Nevada, Reno
S ta te w id e • W orldw ide
We recommend that the thesis
prepared under our supervision by
SASIKIRAN KAMBAVALASA
Entitled
Feasibility Study To Increase The Sensitivity O f A Microwave
Microstripline Bandpass Filter Based Biosensor For The Detection O f Bacteria In Water
be accepted in partial fulfillment of the
requirements for the degree of
MASTER OF SCIENCE
Indira Chatteijee, Ph.D^Advisor
James^Henson, Ph.D., Comhrittee Member
^
'
/I
Gale L Craviso, Ph.D., Graduate School Representative
Marsha H. Read, Ph. D., Associate Dean, Graduate School
May, 2007
Reproduced with permission of the copyright owner. Further reproduction prohibited without permission.
ABSTRACT
This thesis focuses on ways to increase the sensitivity of a microstrip line
bandpass filter based biosensor for the detection of bacteria in water. The biosensor
output is the change in response of the coupled line bandpass filter when bacteria adhere
to the surface of the filter circuit resulting from a change of permittivity of the region
above the filter circuit. This study has shown that the sensitivity of the biosensor can be
increased when the bandwidth of the coupled line bandpass filter is increased. In
addition, the change in response of the biosensor, quantified by a shift in the center
frequency o f the bandpass filter is observed to increase with increasing thickness of a
dielectric overlay that simulates bacteria adhering in multilayers to the biosensor surface.
A further advancement in the design of the biosensor is presented based on a dual band
stepped impedance bandpass filter, wherein the possibility of using this design for the
simultaneous detection of two types of bacteria is presented.
Reproduced with permission of the copyright owner. Further reproduction prohibited without permission.
ACKNOWLEDGEMENTS
I would like to sincerely thank Dr. Indira Chatterjee for helping, supporting, and
guiding me in completing this work. I would always be obliged to her for the way she
helped me during the writing of this work.
I would also like to thank Dr. Gale Craviso of the Department of Pharmacology,
Graduate School representative of my thesis committee, for giving me her valuable time
and for helping me in completing this work.
I would like to thank Dr. James Henson of the Department of Electrical and
Biomedical Engineering, for being on my graduate committee and helping me in
completing this work.
I would like to thank Jihwan Yoon for doing such great work in this area of study
and setting up a foundation for this study.
Reproduced with permission of the copyright owner. Further reproduction prohibited without permission.
iii
TABLE OF CONTENTS
CHAPTER 1
INTRODUCTION
1
CHAPTER 2
COPULED LINE THEORY
4
2.1
INTRODUCTION
4
2.2
MICROSTRIP TRANSMISSION LINE BASICS
4
2.3
COUPLED LINE COUPLERS
6
2.4
COUPLED LINE FILTER
14
2.4.1 Design of Coupled Line Filter
CHAPTER 3
15
MICROSTRIP COUPLED LINE FILTER AS A SENSITIVE
BIOSENSOR
19
3.1
INTRODUCTION
19
3.2
DESIGN OF COUPLED LINE FILTER
19
3.2.1 Design of the Coupled Line Filter Centered at 2 GHz with
10% Bandwidth, using R03003 Substrate (A = 0.1)
21
3.2.2 Design of the Coupled Line Filter Centered at 2 GHz with
15% Bandwidth, using R03003 Substrate (A = 0.15)
25
3.2.3 Design of the Coupled Line Filter Centered at 2 GHz with
20% Bandwidth, using R03003 Substrate (A = 0.2)
27
3.2.4 Design of the Coupled Line Filter Centered at 5 GHz with
10% Bandwidth, using R03003 Substrate (A = 0.1)
29
3.2.5 Design of the Coupled Line Filter Centered at 5 GHz with
15% Bandwidth, using R03003 Substrate (A = 0.15)
30
3.2.6 Design of the Coupled Line Filter Centered at 5 GHz with
20% Bandwidth, using R03003 Substrate (A = 0.2)
3.3
SIMULATION OF THE FILTERS DESIGNED USING R03003
SUBSTRATE WITH DIELECTRIC OVERLAY
3.4
32
34
COUPLED LINE FILTERS USING RT6010 HIGH DIELECTRIC
SUBSTRATE
41
3.4.1 Design of the Coupled Line Filter Centered at 2 GHz with
10% Bandwidth, using RT6010 Substrate (A = 0.1)
Reproduced with permission of the copyright owner. Further reproduction prohibited without permission.
41
iv
3.4.2
Design of the Coupled Line Filter Centered at 2 GHz with
15% Bandwidth, using RT6010 Substrate (A = 0.15)
3.4.3
Design of the Coupled Line Filter Centered at 2 GFIz with
20% Bandwidth, using RT6010 Substrate (A = 0.2)
3.4.4
50
SIMULATION OF THE FILTERS DESIGNED USING RT6010
SUBSTRATE WITH DIELECTRIC OVERLAY
3.6
48
Design of the Coupled Line Filter Centered at 5 GHz with
20% Bandwidth, using RT6010 Substrate (A = 0.2)
3.5
46
Design of the Coupled Line Filter Centered at 5 GHz with
15% Bandwidth, using RT6010 Substrate (A = 0.15)
3.4.6
45
Design of the Coupled Line Filter Centered at 5 GHz with
10% Bandwidth, using RT6010 Substrate (A = 0.1)
3.4.5
43
51
SIMULATION RESULTS FOR INCREASING DIELECTRIC
OVERLAY THICKNESS
58
CHAPTER 4 DISCUSSION
71
4.1
INTRODUCTION
71
4.2
COUPLED LINE BANDPASS FILTER SENSITIVITY WITH
DIELECTRIC OVERLAY
71
4.2.1
Increasing the Fractional Bandwidth of the Coupled Line Bandpass Filter
71
4.2.2
Low Dielectric Constant Substrate vs. High Dielectric Constant Substrate
72
4.2.3
Filters Designed at Higher Center Frequency
73
4.2.4
Increase the Dielectric Overlay Thickness
73
CHAPTER 5 THE COMBLINE DUAL BANDPASS FILTER USING STEPPED
IMPEDANCE RESONATORS AS A BIOSENSOR
75
5.1
INTRODUCTION
75
5.2
COMBLINE FILTER
75
5.2.1
Frequency Response with Dielectric Overlay on the Entire Filter Circuit
78
5.2.2
Frequency Response of the Filter Adding Dielectric Bricks on Various
Parts of the Circuit
Reproduced with permission of the copyright owner. Further reproduction prohibited without permission.
79
V
CHAPTER 6 FUTURE WORK
6.1
Bandpass filters with high bandwidth for biosensor
85
6.2
Choice of center frequency
85
6.3
Choice of dielectric substrate
85
6.4
Dual-band bandpass filters as a biosensor
86
6.5
Immobilization of antibodies
86
Reproduced with permission of the copyright owner. Further reproduction prohibited without permission.
vi
LIST OF FIGURES
2.2.1 Geometry of a microstrip line.
2.2.2
5
Electric (solid lines) and magnetic field (dashed lines) of a
microstrip transmission line.
2.3.1 Coupled transmission line (a), and its equivalent circuit (b).
5
6
2.3.2 Even (a) and odd (b) mode excitations for a coupled line, and the
equivalent capacitance networks.
2.3.3
A single section coupled line coupler, (a) Geometry and ports.
(b) Equivalent schematic circuit.
2.3.4
2.3.5
7
9
Even and odd mode decomposition of the coupled line coupler.
(a) Even mode, (b) Odd mode.
10
Zoe, Zoo vs. coupling coefficient C.
14
2.4.1 Z0J n vs. C for 0 < Z 0J n < 1.
16
2.4.2 A vs. Zoe, Z0o of the four coupled sections.
18
3.2.1 Geometry of the coupled line bandpass filter. P, S, and W are the
length, width, and spacing of the coupled sections respectively.
3.2.2
Geometry of the optimized coupled line bandpass filter designed with
R03003 substrate. Center frequency: 2GHz, bandwidth: 10%.
3.2.3
21
24
S21 response of the coupled line bandpass filter shown in
Figure 3.3.2 using R03003 substrate. Center frequency: 2 GHz,
bandwidth: 10%.
3.2.4
Geometry of the optimized coupled line bandpass filter designed
with R03003 substrate. Center frequency: 2GHz, bandwidth: 15%.
3.2.5
28
S2i response of the designed coupled line bandpass filter using
R03003 substrate. Center frequency: 2GHz, bandwidth: 20%.
3.2.8
26
Geometry of the optimized coupled line bandpass filter designed
with R03003 substrate. Center frequency: 2GHz, bandwidth: 20%.
3.2.7
26
S2i response of the designed coupled line bandpass filter using
R03003 substrate. Center frequency: 2GHz, bandwidth: 15%.
3.2.6
24
Geometry of the optimized coupled line bandpass filter designed
Reproduced with permission of the copyright owner. Further reproduction prohibited without permission.
28
with R03003 substrate. Center frequency: 5GHz, bandwidth: 10%.
3.2.9
S21 response of the designed coupled line bandpass filter using
R03003 substrate. Center frequency: 5GHz, bandwidth: 10%.
3.2.10 Geometry of the optimized coupled line bandpass filter designed with
R03003 substrate. Center frequency: 5GHz, bandwidth: 15%.
3.2.11 S2i response of the designed coupled line bandpass filter using
R03003 substrate. Center frequency: 5GHz, bandwidth: 15%.
3.2.12 Geometry of the optimized coupled line bandpass filter designed with
R03003 substrate. Center frequency: 5GHz, bandwidth: 20%.
3.2.13 S2i response of the designed coupled line bandpass filter using
R03003 substrate. Center frequency: 5GHz, bandwidth: 20%.
3.3.1
S2i response of the coupled line bandpass filter using R03003 substrate.
Center frequency: 2GHz, bandwidth: 10%. Response is shown
with and without the dielectric overlay.
3.3.2
S2i response of the coupled line bandpass filter using R03003 substrate.
Center frequency: 2GHz, bandwidth: 15%. Response is shown
with and without the dielectric overlay.
3.3.3
S2i response of the coupled line bandpass filter using R03003 substrate.
Center frequency: 2GHz, bandwidth: 20%. Response is shown
with and without the dielectric overlay.
3.3.4
Shift in center frequency with increase in bandwidth for a
coupled line filter. R03003 substrate is used. Center frequency: 2 GHz.
3.3.5
S2i response of the coupled line bandpass filter using R03003 substrate.
Center frequency: 5GHz, bandwidth: 10%. Response is shown
with and without the dielectric overlay.
3.3.6
S2i response of the coupled line bandpass filter using R03003 substrate.
Center frequency: 5GHz, bandwidth: 15%. Response is shown
with and without the dielectric overlay.
3.3.7
S2i response of the coupled line bandpass filter using R03003 substrate.
Center frequency: 5GHz, bandwidth: 20%. Response is shown
with and without the dielectric overlay.
with permission of the copyright owner. Further reproduction prohibited without permission.
3.3.8
Shift in center frequency with increase in bandwidth for a
coupled line filter. R03003 substrate is used. Center frequency: 5 GHz.
3.4.1
Geometry of the optimized coupled line bandpass filter designed
with RT6010 substrate. Center frequency: 2GHz, bandwidth: 10%.
3.4.2
S2i response of the designed coupled line bandpass filter using
RT6010 substrate. Center frequency: 2GHz, bandwidth: 10%.
3.4.3
Geometry of the optimized coupled line bandpass filter designed
with RT6010 substrate. Center frequency: 2GHz, bandwidth: 15%.
3.4.4
S2i response of the designed coupled line bandpass filter using
RT6010 substrate. Center frequency: 2GHz, bandwidth: 15%.
3.4.5
Geometry of the optimized coupled line bandpass filter designed
with RT6010 substrate. Center frequency: 2GHz, bandwidth: 20%.
3.4.6
S2i response of the designed coupled line bandpass filter using
RT6010 substrate. Center frequency: 2GHz, bandwidth: 20%.
3.4.7
Geometry of the optimized coupled line bandpass filter designed
with RT6010 substrate. Center frequency: 5GHz, bandwidth: 10%.
3.4.8
S2i response of the designed coupled line bandpass filter using
RT6010 substrate. Center frequency: 5 GHz, bandwidth: 10%.
3.4.9
Geometry of the optimized coupled line bandpass filter designed
with RT6010 substrate. Center frequency: 5GHz, bandwidth: 15%.
3.4.10 S2i response of the designed coupled line bandpass filter using
RT6010 substrate. Center frequency: 5 GHz, bandwidth: 15%.
3.4.11 Geometry of the optimized coupled line bandpass filter designed
with RT6010 substrate. Center frequency: 5GHz, bandwidth: 20%.
3.4.11 S2J response of the designed coupled line bandpass filter using
RT6010 substrate. Center frequency: 5 GHz, bandwidth: 20%.
3.5.1
S2i response of the coupled line bandpass filter using RT6010 substrate.
Center frequency: 2GHz, bandwidth: 10%. Response is shown
with and without the dielectric overlay.
3.5.2
S2i response of the coupled line bandpass filter using RT6010 substrate.
Center frequency: 2GHz, bandwidth: 15%. Response is shown
with and without the dielectric overlay.
Reproduced with permission of the copyright owner. Further reproduction prohibited without permission.
IX
3.5.3
S21 response of the coupled line bandpass filter using RT6010 substrate.
Center frequency: 2GHz, bandwidth: 20%. Response is shown
with and without the dielectric overlay.
3.5.4
Shift in center frequency with increase in bandwidth for a
coupled line filter. RT6010 substrate is used. Center frequency: 2 GHz.
3.5.5
53
54
S21 response of the coupled line bandpass filter using RT6010 substrate.
Center frequency: 5GHz, bandwidth: 10%. Response is shown
with and without the dielectric overlay.
3.5.6
55
S2i response of the coupled line bandpass filter using RT6010 substrate.
Center frequency: 5GHz, bandwidth: 15%. Response is shown
with and without the dielectric overlay.
3.5.7
55
S2i response of the coupled line bandpass filter using RT6010 substrate.
Center frequency: 5GHz, bandwidth: 20%. Response is shown
with and without the dielectric overlay.
3.5.8
Shift in center frequency with increase in bandwidth for a
coupled line filter. RT6010 substrate is used. Center frequency: 5 GHz.
3.6.1
56
57
S21 response of the filter with two overlay thicknesses at 2 GHz
center frequency designed with R03003 substrate for A = 0.1.
The response with and without overlay is shown for comparison.
3.6.2
58
S2i response of the filter with two overlay thicknesses at 2 GHz
center frequency designed with R03003 substrate for A = 0.15.
The response with and without overlay is shown for comparison.
3.6.3
59
S21 response of the filter with two overlay thicknesses at 2 GHz
center frequency designed with R03003 substrate for A = 0.2.
The response with and without overlay is shown for comparison.
3.6.4
59
S21 response of the filter with two overlay thicknesses at 5 GHz
center frequency designed with R03003 substrate for A = 0.1.
The response with and without overlay is shown for comparison.
3.6.5
60
S2i response of the filter with two overlay thicknesses at 5 GHz
center frequency designed with R03003 substrate for A = 0.15.
The response with and without overlay is shown for comparison.
Reproduced with permission of the copyright owner. Further reproduction prohibited without permission.
60
3.6.6
S21 response of the filter with two overlay thicknesses at 5 GHz
center frequency designed with R03003 substrate for A = 0.2.
The response with and without overlay is shown for comparison.
3.6.7
S2i response of the filter with two overlay thicknesses at 2 GHz
center frequency designed with RT6010 substrate for A = 0.1.
The response with and without overlay is shown for comparison.
3.6.8
S2i response of the filter with two overlay thicknesses at 2 GHz
center frequency designed with RT6010 substrate for A = 0.15.
The response with and without overlay is shown for comparison.
3.6.9
S21 response of the filter with two overlay thicknesses at 2 GHz
center frequency designed with RT6010 substrate for A = 0.2.
The response with and without overlay is shown for comparison.
3.6.10 S2i response of the filter with two overlay thicknesses at 5 GHz
center frequency designed with RT6010 substrate for A = 0.1.
The response with and without overlay is shown for comparison.
3.6.11 S2i response of the filter with two overlay thicknesses at 5 GHz
center frequency designed with RT6010 substrate for A = 0.15.
The response with and without overlay is shown for comparison.
3.6.12 S2i response of the filter with two overlay thicknesses at 5 GHz
center frequency designed with RT6010 substrate for A = 0.2.
The response with and without overlay is shown for comparison.
3.6.13 Frequency shifts at the 3dB point for the coupled line filter centered
at 2GHz, using R03003 substrate.
3.6.14 Frequency shifts at the 1OdB point for the coupled line filter centered
at 2GHz, using R03003 substrate.
3.6.15 Frequency shifts at the 3dB point for the coupled line filter centered
at 5GHz, using R03003 substrate.
3.6.16 Frequency shifts at the 1OdB point for the coupled line filter centered
at 5GHz, using R03003 substrate.
3.6.17 Frequency shifts at the 3dB point for the coupled line filter centered
at 2GHz, using RT6010 substrate.
3.6.18 Frequency shifts at the 1OdB point for the coupled line filter centered
Reproduced with permission of the copyright owner. Further reproduction prohibited without permission.
xi
at 2GHz, using RT6010 substrate.
69
3.6.19 Frequency shifts at the 3dB point for the coupled line filter centered
at 5GHz, using RT6010 substrate.
70
3.6.20 Frequency shifts at the lOdB point for the coupled line filter centered
at 5GHz, using RT6010 substrate.
70
4.2.1
Schematic for the regulation of biofilm formation
74
5.2.1
Stepped impedance resonator.
76
5.2.2
Geometry of the combline dual-band bandpass filter generated in Sonnet.
77
5.2.3
S2i response of the combline dual-band bandpass filter.
78
5.2.4
S2i response of the combline filter with and without dielectric overlay
on top of the entire filter circuit. Two overlay thicknesses are used.
5.2.5
Geomehy of the combline bandpass filter with a dielectric brick on the
first resonator.
5.2.6
81
S21 responses of the combline filter with 0.2mil thick dielectric brick
on each of the resonators as shown in Figures 5.2.5 and 5.2.6.
5.2.9
80
S2] responses of the combline filter with O.lmil thick dielectric brick
on each of the resonators as shown in Figures 5.2.5 and 5.2.6.
5.2.8
80
Geomehy of the combline bandpass filter with a dielectric brick on the
second resonator.
5.2.7
79
81
S2j responses of the combline filter with O.lmil thick dielectric brick
on each of the resonators as shown in Figures 5.2.5 and 5.2.6.
82
5.2.10 Geometry of the combline bandpass filter with a dielectric brick on
both the resonators.
5.2.11
82
S2i response of the combline filter with dielectric overlay bricks on
both the resonators.
83
5.2.12 S2i response of the combline filter with different overlay thicknesses
applied on the left resonator.
Reproduced with permission of the copyright owner. Further reproduction prohibited without permission.
84
xii
LIST OF TABLES
2.1a
Zoe, Z0ovalues of the coupled lines for A = 0.1
17
2.1b
Z^, Zoo values of the coupled lines for A = 0.15
17
2.1c
Zoe, Zoo values of the coupled lines for A = 0.2
17
3.1
Filter parameters for the third order Chebeshev 0.5 dB equal ripple
coupled line filter. Zoe and Z^ are the even and odd mode characteristic
impedances for 10% bandwidth.
3.2
23
Optimized dimensions of the four coupled sections of the filter designed
with R03003 substrate. Center frequency: 2GHz, bandwidth: 10%.
23
3.3
Zoe and Zo0values for 15% bandwidth.
25
3.4
Optimized dimensions of the four coupled sections of the filter designed
with R03003 substrate. Center frequency: 2GHz, bandwidth: 15%.
25
3.5
Z0e and Z0ovalues for 20% bandwidth.
27
3.6
Optimized dimensions of the four coupled sections of the filter designed
with R03003 substrate. Center frequency: 2GHz, bandwidth: 20%.
3.7
Optimized dimensions of the four coupled sections of the filter designed
with R03003 substrate. Center frequency: 5GHz, bandwidth: 10%.
3.8
31
Optimized dimensions of the four coupled sections of the filter designed
with R03003 substrate. Center frequency: 5GHz, bandwidth: 20%.
3.10
29
Optimized dimensions of the four coupled sections of the filter designed
with R03003 substrate. Center frequency: 5GHz, bandwidth: 15%.
3.9
27
32
Frequencies corresponding to 3dB and lOdB points on the low frequency
side of the S2i response, and frequency shifts with and without overlay.
Substrate used is R03003 and filter is centered at 2GHz.
3.11
36
Frequencies corresponding to 3dB and lOdB points on the low frequency
side of the S2i response, and frequency shifts with and without overlay.
Substrate used is R03003 and filter is centered at 5GHz.
3.12
39
Optimized dimensions of the four coupled sections of the filter designed
with RT6010 substrate. Center frequency: 2GHz, bandwidth: 10%.
Reproduced with permission of the copyright owner. Further reproduction prohibited without permission.
42
3.13
Optimized dimensions of the four coupled sections of the filter designed
with RT6010 substrate. Center frequency: 2GHz, bandwidth: 15%.
3.14
Optimized dimensions of the four coupled sections of the filter designed
with RT6010 substrate. Center frequency: 2GHz, bandwidth: 20%.
3.15
Optimized dimensions of the four coupled sections of the filter designed
with RT6010 substrate. Center frequency: 5GHz, bandwidth: 10%.
3.16
Optimized dimensions of the four coupled sections of the filter designed
with RT6010 substrate. Center frequency: 5GHz, bandwidth: 15%.
3.17
Optimized dimensions of the four coupled sections of the filter designed
with RT6010 substrate. Center frequency: 5GHz, bandwidth: 20%.
3.18
Frequencies corresponding to 3dB and lOdB points on the low
frequency side of the S2i response, and frequency shifts with and
without overlay. Substrate used is RT6010 and filter is centered at 2GHz.
3.19
Frequencies corresponding to 3dB and lOdB points on the low
frequency side of the S2i response, and frequency shifts with and
without overlay. Substrate used is RT6010 and filter is centered at 5GHz.
3.20
Frequency shift at the 3dB and lOdB points for R03003 substrate at
2GHz and 5GHz for A = 0.1, 0.15, and 0.2, with 0.1 and 0.2mil
overlay thicknesses.
3.21
Frequency shift at the 3dB and lOdB points for RT6010 substrate at
2GHz and 5GHz for A = 0.1, 0.15, and 0.2, with 0.1 and 0.2mil
overlay thicknesses.
5.1
Dimensions of the stepped impedance resonators.
Reproduced with permission of the copyright owner. Further reproduction prohibited without permission.
CHAPTER 1
INTRODUCTION
A biosensor is defined as a compact analytical device which incorporates a
biological or a biologically derived sensing element, either integrated or intimately
associated with a physiochemical transducer [1]. The common function of a biosensor is
to produce either a digital or analog signal which is proportional to the amount of
detected analyte. A biosensor typically consists of a bioreceptor and a transducer. The
bioreceptor is a biomolecule that senses the target analyte which is to be detected and the
transducer converts this sensing event into a recordable signal [2]. By choosing an
appropriate bioreceptor and transducer, a sensitive biosensor could be developed. The
form o f the output data from the biosensor may be digital or analog depending on the
method of measurement and type of biosensor used for measurement. Most modern
biosensors can be controlled by computers which acquire, save and process the data
obtained from the biosensor. Once the target analyte has been determined then a suitable
bioreceptor can be chosen. The main bioreceptors used to date in biosensors are enzymes,
antibodies and DNA [2],
A key part o f the biosensor is the transducer and it is critical in determining the
sensitivity of the biosensor. Various types of physical transducers like electrodes, optical
fiber probes, gas probes, and piezoelectric probes have been used to date for the purpose
of sensing [2-7].
Biosensors are generally used in the healthcare industry, food processing, bio­
processing, agriculture, military and homeland security applications and environmental
monitoring [2]. For instance, in the context of healthcare, the glucose sensor has been
Reproduced with permission of the copyright owner. Further reproduction prohibited without permission.
2
used with a great deal of success, becoming one of the most powerful commercial
biosensors designed for monitoring blood sugar and glucose levels in diabetics [3].
Biosensors are also very useful in environmental analysis due to increasing pollution by
toxic biological contaminants, chemical vapors and biological warfare. Such sensors are
being utilized for determining the quality of water used for human consumption and
water that is discharged into the environment [2]. These sensors must have high
sensitivity, high reliability, be user friendly, portable and should also be capable of online
monitoring. Commercial biosensor technology has taken off over the past few years, and
the number of applications has been rapidly increasing [7].
Previously, a biosensor based on common microwave bandpass microstrip filters
had been designed, fabricated and tested for environmental monitoring, mainly, for the
detection of bacteria in water [8]. The sensor was based on the change in dielectric
permittivity on the surface of the sensor due to the presence of bacteria. The above work
describes the required characteristics of biosensors like sensitivity, calibration, linearity,
detection limit, hysteresis, specificity, selectivity, repeatability, reproducibility, and
response time. Various types of microstrip bandpass filters like the coupled line, hairpin
type, and interdigital bandpass filter were designed and numerically simulated using the
RF/microwave design software package Sonnet EM Suite v. 8.52 to investigate the
feasibility of using them as a biosensor for detecting bacteria in water. The designed
filters responded to the presence of bacteria due to the change in dielectric permittivity of
the region above the circuits via a shift in the response. Various factors that might affect
the behavior of the biosensor, like absorption o f water by the dielectric substrate material,
toxic effects o f the circuits on the growth of bacteria, unreliable numerical simulations,
Reproduced with permission of the copyright owner. Further reproduction prohibited without permission.
effects of autoclaving on the circuit, and residual culture medium on the circuit were
investigated. An interdigital filter was fabricated and tested successfully for the detection
of Pseudomonas Aeruginosa bacteria.
This thesis uses the method of detection described in [8] and primarily focuses on
improving the sensitivity of the biosensor. A bandpass coupled line filter was designed
using well known techniques in microwave engineering [9]. Numerical simulations using
the RF/microwave design software package Sonnet EM Suite v. 9.52 were used to
investigate the effect o f varying (1) substrate dielectric constant, (2) bandwidth, and (3)
bacterial overlay thickness on the sensitivity of the biosensor.
In this thesis, chapter 2 presents microstrip transmission line basics, the basic
theory o f coupled line couplers, and a method to increase the sensitivity of a coupled line
bandpass filter in the presence o f a dielectric overlay that simulates a layer of bacteria
adhering to the circuit. Chapter 3 presents the simulation results for the coupled line band
pass filters with three different values of fractional bandwidth (0.1, 0.2, and 0.3) at center
frequencies o f 2 GHz and 5GHz, and using two different dielectric substrates, R03003
(er = 3.0) and RT6010 (er = 10.2). These results are discussed in chapter 4. Chapter 5
presents the possibility o f using a dual-band bandpass combline stepped impedance filter
as a potential candidate for a biosensor. The simulation results for this filter are presented
with dielectric overlay on various parts of the filter circuit and the results are discussed in
this chapter. In chapter 6, suggestions are made for future work.
Reproduced with permission of the copyright owner. Further reproduction prohibited without permission.
4
CHAPTER 2
COUPLED LINE THEORY
2.1 INTRODUCTION
As introduced in the previous chapter, the aim of this work is to explore ways to
increase the sensitivity o f a previously designed microstrip coupled line bandpass filter
based biosensor. The design of microstrip bandpass filters and their responses with
dielectric overlay have been widely explored in [8].
This chapter presents some
microstrip transmission line basics (section 2.2), the basic theory of coupled line couplers
and ways to increase the coupling between the coupled lines (section 2.3), as well as the
implementation o f the method described in section 2.3 to a coupled line bandpass filter
(section 2.4).
2.2 MICROSTRIP TRANSMISSION LINE BASICS
The microstrip line is one of the most popular types of planar transmission lines
which is widely used in RF/microwave engineering since it can be easily fabricated by
the photolithographic process and can be easily integrated with other active and passive
microwave devices as well as discrete components. The basic geometry of a microstrip
line is shown in Figure 2.2.1 and consists of a strip conductor of width “w” and thickness
“t” on a thin dielectric substrate of thickness “h” and a relative dielectric constant o f ‘V ’
backed by a ground plane.
Reproduced with permission of the copyright owner. Further reproduction prohibited without permission.
5
Conductor
w
Dielectric Substrate
Ground Plane
Figure 2.2.1: Geometry of a microstrip line.
The presence o f conducting strip over the dielectric and the fact that the dielectric
does not fill the air region above the strip makes the analysis of a microstrip line
complicated. A microstrip line has most of the electric and magnetic field lines
concentrated in the dielectric region between the ground plane and the conducting strip,
with some fraction o f the field lines concentrated in the air above the conducting strip as
shown in Figure 2.2.2. This open structure of the microstrip line makes shielding of the
circuit difficult, and also results in more complicated modes o f propagation.
Dielectric Substrate
Microstrip Conductor
Ground plane
Figure 2.2.2: Electric (solid lines) and magnetic field (dashed lines) of a microstrip
transmission line.
Reproduced with permission of the copyright owner. Further reproduction prohibited without permission.
Since some o f the fields are in air and some of them are confined to within the
dielectric substrate the microstrip line is assigned an effective dielectric constant eeg
which takes into account both the air and substrate dielectric constants [9]. As expected,
it was shown in [8] that, if the microstrip line is covered with a dielectric overlay
representing a layer of bacteria, the effective permittivity of the microstrip line changes,
and this change results in a change in response of a microstrip line based microwave filter
circuit.
2.3 COUPLED LINE COUPLERS [9, Chapter 7, Section 7.6]
For a pair of transmission lines close to each other, coupling of fields takes place
between the two lines. Such lines are called coupled transmission lines. (Figure 2.3.1).
Coupled transmission lines are generally assumed to operate in the Transverse Electro
Magnetic (TEM) mode. These coupled lines can be used to implement devices like filters
and directional couplers.
wwwwwwwwwmwwmww
fat
(b)
Figure 2.3.1: Coupled transmission line (a), and its equivalent circuit (b).
Reproduced with permission of the copyright owner. Further reproduction prohibited without permission.
7
The coupled line structure shown in Figure 2.3.1 (a) can be represented by the
equivalent circuit shown Figure 2.3.1 (b). If TEM type propagation is assumed, the
electrical characteristics of the coupled lines can be completely determined from the
effective capacitances C n, C2 2 , C 12 and the velocity of propagation on the line. Here, Cn
and C2 2 represent the capacitance between one strip conductor and ground in the absence
of the other strip conductor, while C 12 represents the capacitance between the two strip
conductors in the absence o f the ground plane. If the strip conductors are identical in size
and location with respect to the ground plane, then Cn = C2 2 .
The coupled lines can be excited in two ways, even mode excitation and odd
mode excitation. When currents in the strip conductors are equal in amplitude and in the
same direction then it is considered as even mode excitation, while if the currents are
equal in amplitude and opposite in direction then it is considered as odd mode excitation.
Figure 2.3.2 shows the even and odd mode excitations and the corresponding equivalent
circuits in both cases.
o „
I
T
7/ / / / /
-V
2C,2
H I _i
—
20,
I h-
1 ,
T
X
IT
Figure 2.3.2: Even (a) and odd (b) mode excitations for a coupled line, and the
equivalent capacitance networks.
Reproduced with permission of the copyright owner. Further reproduction prohibited without permission.
8
For even mode excitation the electric field has an even symmetry about the center
line, and there is no current flow between the strip conductors. In the equivalent
capacitance circuit for this type of excitation (Figure 2.3.2 (a)), capacitance C 12 is
effectively open circuited. Assuming that the two strip conductors are identical in size
and location with respect to the ground plane, the effective capacitance of either line to
ground is
(2.1)
Then the characteristic impedance for the even mode is given by
(2.2)
where vp is the velocity o f propagation on the line and L is the line inductance .
For odd mode excitation, there is an odd symmetry o f the electrical field about the
center line, and a voltage null exists between the two strip conductors, or in other words,
there is a ground plane between the two strip conductors. The resultant equivalent
capacitive circuit is shown in Figure 2.3.2 (b). The effective capacitance between either
strip conductor and the ground plane is given by
C + Z^12
2C
^C o = ^C u + 2C 12 = '-'22
(2.3)
The characteristic impedance of the odd mode is given by
(2.4)
An arbitrary excitation of the coupled line can be considered as a superposition of even
and odd modes.
Reproduced with permission of the copyright owner. Further reproduction prohibited without permission.
9
Using the definitions for even and odd mode characteristic impedances, even and
odd mode analysis can be applied to analyze a single section coupled line coupler shown
in Figure 2.3.3.
Coupled
Isolated
V ®
©
Through
X -©
(»>
0
•
Z0
h--------------------------------1
h
^
I4
—V y V W - j /
y -^ V v V W
1
— 1
+^4
Z
2
^•0e«**0o
—
" z,
T ^ W W V —I
+v2
(b)
Figure 2.3.3: A single section coupled line coupler, (a) Geometry and ports, (b)
Equivalent schematic circuit.
Even and odd mode analysis technique is applied in conjunction with the input
impedances of the line. Figure 2.3.4 gives the decomposition of the coupler into even and
odd mode excitations.
Reproduced with permission of the copyright owner. Further reproduction prohibited without permission.
Figure 2.3.4: Even and odd mode decomposition of the coupled line coupler, (a) Even
mode (b) Odd mode
From Figure 2.3.4, it can be said that, 7* = 7* ,7* = I e2 , V, = V3 , and V e4 = V2 for
even mode excitation. Similarly for odd mode excitation it is, 1° = -7", 1°2 = -1°,
v;=-v; ,mdv; =-v;.
The input impedance of the coupler at port 1 of the coupler of Figure 2.3.3 can be
written as
^
Vj v;+v;
Reproduced with permission of the copyright owner. Further reproduction prohibited without permission.
11
If the input impedance o f the even mode at port 1 is Zine, and Zin° is the input impedance
of the odd mode, then
Z„ + j Z n. tan 8
..........,
Z e = Z„ 0
m
Z 0e + j Z 0 tand
(2.6a)
Z n + jZ 0ntand
(2<Sb)
since for each mode, the line looks like a transmission line of characteristic impedance
Zoe or Zoo, terminated with a load impedance of Zo. Here 8 = electrical length of the
coupled section. As a result, by voltage division,
F/ =F—
,
(2.7a)
in
v
;
=v z ^ T z '’
A in
( 2
’7 b )
,
(2.7c)
.
(2.7d)
-^ 0
Z ine + Z
Using the results in 2.7 in 2.5 gives,
2 ( Z ° Z ei - Z 2)
z = z + -2—2L
+ - 2—
+ 27
^ in
7 « x
01
7 0x 0 - 7
in T Z Z , 0
ro r'i
VZ -6 1
if the characteristic impedance is
z c = jz iz ° .
Using (2.9) in (2.6a, b)
Reproduced with permission of the copyright owner. Further reproduction prohibited without permission.
(2.9)
12
Ze =Z
J^ z 0o + -jJ> ZV 0e_------0 tan 9
(21Q a)
^ e +j 4 ^ tan9
Z o =
z
JS 0ze7_ + jJ
Z 0o tan 9
J S O o _
<JzZ +
^
( 2
1
0
b
)
tanO
From (2.10a, b)
Z?„Z°n=Z0eZoo=Z02,
(2.11)
Z (, = Z 0
(2.12)
and using 2.9 in yields,
So, as long as (2.9) is satisfied and is true, all the ports of the coupler will be
matched. Also in this case(2.12) will be true, and hence from voltage division we can say
that V[=V. Voltage at port 3 is the quantity of interest, because port 3 is the coupling port
of the coupler. Using (2.7) the voltage at port 3 is given by
v3=v;+v;=v;-v;=v
<2 . 13)
in
0
Z j,,
+
Z
q
Using (2.9) and (2.6), equation (2.13) can be reduced to
j ( Z (. - Z nn)ta
00/-n d
y =y
tl—2£-------5
2Z0 + j( Z 0e + Z Oo)ta n 0
f? 141
>
Define C as
C = 7°e ~Z °°
(2.15)
^ 0 e + ^O o
Then from (2.15),
■Jl^CT = 7 2Z°
^O e
(2.16)
Oo
Reproduced with permission of the copyright owner. Further reproduction prohibited without permission.
Using (2.15) and (2.16), equation (2.14) can be written as
jCtanO
V s= V r~ ~~7 ~ r
4 l - C 2 + jtanO
(2.17)
For# = icj2 , the coupler is 7J4 long and (2.17) can be reduced to
(2.18)
which shows that C < 1. C is called the voltage coupling factor at the design frequency,
and at 6 = n f2 .
So, if the characteristic impedance Z0, and the voltage coupling coefficient C, are
specified, the even and odd mode characteristic impedances could be calculated using the
equations below.
(2.19)
(2.20)
Equations (2.19) and (2.20) will be true, when even and odd modes of the coupled line
structure have the same velocity of propagation, so that the line has the same electrical
length for both modes. For coupled microstrip, or other than non-TEM line this condition
will generally not be satisfied, and the coupler will have poor directivity. These types of
couplers are best used for weak coupling, since tight coupling requires the spacing
between the lines to be very small. This is difficult to achieve practically as well as to
obtain the combination of even and odd mode impedances required for higher coupling.
Using the equations (2.19) and (2.20), and for Zo = 50Q, a graph is plotted of C
vs. Z0e,Z0o. (Figure 2.3.5).
Reproduced with permission of the copyright owner. Further reproduction prohibited without permission.
14
250
Even mode impedance (ZOe)
Odd mode impedance (ZOo)
200
O 150
9 100
0.2
0.3
0.4
0.5
0.6
0.7
0.8
0.9
COUPLING COEFFICIENT (C)
Figure 2.3.5: Zoe, Zo0 vs. coupling coefficient C.
From equations (2.19), (2.20) and from Figure 2.3.5, it can be said that, as C tends
to 1, Zoe tends to infinity and Zoo tends to zero. So, the C of a coupler is directly
proportional to Zoe and inversely proportional to Z q0.
2.4 COUPLED LINE FILTER
Using the parallel coupled line sections discussed in the previous section, the
coupled line filter is designed. This thesis uses the coupled line bandpass filter as the
sensing element. This section focuses on how the sensitivity of the coupled line filter can
be increased in the presence of a dielectric overlay. As the coupling between the strip
Reproduced with permission of the copyright owner. Further reproduction prohibited without permission.
15
conductors increases, the sensitivity of the circuit to an overlay increases [8, chapter 3,
section 3.4.4]. The previous section explains the relationship between the coupling and
the even-odd mode impedances of the coupler. Since the coupled line bandpass filter is a
cascade of several coupled line sections, the above theory can be applied to the coupled
line filter.
2.4.1 Design of coupled line filter [9, Chapter 8, Section 8.7]
The design of the coupled microstrip lines used in this filter depends on the evenodd mode impedances o f the coupled lines of the filter, and these impedances in turn
depend on the fractional bandwidth A. With larger A, Zoe increases and Zo0 decreases,
making the coupling coefficient C larger and hence resulting in more coupling between
the lines.
The Zoe and Zo0 of the coupled sections can be determined using the equations
where
Z 0e ~ Z0[ l + J„Z0 + ( J nZ0) 2] ,
(2.21)
Z0o = Z0[ l - J „ z 0 + ( J nZ0) 2] ,
(2.22)
iz $
<2-23)
7
i= t^ := c
which is the coupling coefficient C, from (2.15).
The Jn s are the admittance inverters. For a coupled line filter with N+l coupled sections,
the design equations for ZoJ„'s are
I kA
=
ite r
’
Reproduced with permission of the copyright owner. Further reproduction prohibited without permission.
< 2 -2 4 )
16
nA
Z qJ n
(2.25)
2Jgn-lS» ’
_ I
kA
(2.26)
where the g„’s are the lowpass filter prototype circuit element values. Using equation
2.23 a graph is plotted of C as a function of ZoJ„. (Figure 2.4.1).
0.5
0.4
o
I 0.3
ao>
O
c
CL
zs
a
O
0.1
0.2
0.6
0.4
0.8
ZoJn
Figure 2.4.1: Z 0J n vs. C for 0 < Z 0J n < 1.
It can be concluded from equations (2.21-2.26) and Figure 2.4.1 that, an increase in A
increases ZoJ„'s, and this results in an increase in the C of the coupled lines.
Hence, since more coupling is achieved by this technique (increase in A), the
response o f the coupled line filter would be more sensitive to the presence of a dielectric
overlay on the filter circuit. This basic principle was used in increasing the sensitivity of
the coupled line filters with a dielectric overlay on the filter circuit.
Reproduced with permission of the copyright owner. Further reproduction prohibited without permission.
The Zoe and Zo0 for a coupled line bandpass filter were calculated using the above
theory for three different A values and they are given in Tables (2.1 a-c)
A = 0.1
n
1
2
3
4
gn
1.5963
1.0967
1.5963
1.0000
ZflJn
0.3137
0.1187
0.1187
0.3137
Zoe (ft)
70.6047
56.6406
56.6406
70.6047
Zoo (ft)
39.2355
44.7688
44.7688
39.2355
Table 2.1a: Zoe, Zo0 values o f the coupled lines for A = 0.1.
A = 0.15
n
1
2
3
4
gn
1.5963
1.0967
1.5963
1.0000
ZoJB
0.3842
0.1781
0.1781
0.3842
Zoe (ft)
76.5898
60.4895
60.4895
76.5898
Zoo (ft)
38.1706
42.6817
42.6817
38.1706
Table 2.1b: Zoe, Zo0 values of the coupled lines for A = 0.15.
A = 0.2
n
1
2
3
4
gn
1.5963
1.0967
1.5963
1.0000
ZoJB
0.4436
0.2374
0.2374
0.4436
Zoe (ft)
82.0216
64.6907
64.6907
82.0216
Zoo (ft)
37.6589
40.9470
40.9470
37.6589
Table 2.1c: Zoe, Zo0 values of the coupled lines for A = 0.2.
From these tables it is concluded that, as A increases, Zoe increases and Zoo
decreases. Thus C increases (Equation (2.23)). A graph is plotted for A vs. Z0e, Z0o
(Figure 2.4.2.).
Reproduced with permission of the copyright owner. Further reproduction prohibited without permission.
18
100
ZOe first and third coupled sections
20a first and third coupled sections
ZOe second and Fourth coupled sections
ZOo second and fourth coupled sections
co
5
X
o
LU
o
0.12
0.14
0.16
0.18
0.2
FRACTIONAL BANDWIDTH
Figure 2.4.2: Zoe, Zq0 v s . A of the four coupled sections.
Applying the theory presented in this chapter, coupled line bandpass filters are
designed and simulated for three different A values (0.1, 0.2, and 0.3) at 2 GHz and 5
GHz center frequencies, using two different dielectric substrates, R03003 (er - 3.0) and
RT6010 (er = 10.2), and the results are presented in Chapter 3.
Reproduced with permission of the copyright owner. Further reproduction prohibited without permission.
19
CHAPTER 3
MICROSTRIP COUPLED LINE FILTER AS A SENSITIVE BIOSENSOR
3.1 INTRODUCTION
It was shown in [8] that adding a dielectric overlay on the filter circuit changes
the effective permittivity o f the medium above the circuit, thus resulting in a shift in
frequency response towards lower frequencies. The sensitivity of the coupled line
bandpass filter biosensor could be increased with increase in bandwidth of the pass-band
o f the filter in the presence of a dielectric overlay. As discussed in Chapter 2 the coupling
between the coupled lines increases with increase in bandwidth of the coupled line filter,
and hence the filter circuit will become more sensitive to the dielectric overlay. In this
chapter coupled line bandpass filters with different bandwidths are designed and
numerically simulated using the RF/microwave simulation software package Sonnet 9.52
with and without dielectric overlay on the surface of the filter. These filters are designed
at two separate center frequencies, 2 GHz and 5 GHz, and 10%, 15%, and 20%
bandwidths. Two different dielectric substrates, R03003 (er = 3.0) and RT6010 (er 10.2) {Rogers Corporation, CT, USA), are used for these simulations. The frequency
shifts in the S21 responses at the 3 dB and 10 dB points are computed in all cases, and the
amount of shifts are compared in presence of dielectric overlay.
3.2 DESIGN OF THE COUPLED LINE FILTER [2]
The design equations for a coupled line band pass filter with N+l coupled
sections are given below.
Reproduced with permission of the copyright owner. Further reproduction prohibited without permission.
20
icA
z0Jn= J '—
2gI
(3-D
Z 0J„ =
(3.2)
3!lL = ,f o r n = 1 ,2 ,3 ....,N
yg n -iS n
(3 3 )
where Zo is the characteristic impedance, A is the fractional bandwidth, J„’s are the
admittance inverter constants, and gn’s are the lowpass filter prototype circuit element
values. The fractional bandwidth A is given by
A =^ L ,
O)0
(3.4)
where col and co2 are the lower and upper angular frequencies of the passband and o)0 is
the center frequency of the filter.
The even and odd mode characteristic impedances for the coupled lines are found using
the equations given below.
z 0e = Z0[ l + J nZ0 +( J „Z0) 2J
(3.5)
Z0o = Z 0[ 1 - J „ Z 0 + ( J nZ0) 2]
(3.6)
The calculated even and odd mode impedances are used to design the widths and
spacings of the coupled lines. The geometry of the coupled line bandpass filter with four
coupled sections is shown in Figure 3.2.1.
Reproduced with permission of the copyright owner. Further reproduction prohibited without permission.
21
Conductor
Ground Plane
Substrate
Figure 3.2.1: Geometry of the coupled line bandpass filter. P, S, and W are the length,
width, and spacing of the coupled sections respectively.
The calculated even and odd mode impedances are used to design the widths (W)
and spacings (S) o f the coupled lines. Once the dimensions of the filter circuit are
obtained the filter geometry is input into Sonnet and simulated to obtain the frequency
response.
3.2.1 Design of the coupled line filter centered at 2 GHz with 10% bandwidth, using
R03003 substrate (A = 0.1)
A 0.5 dB equal ripple Chebyshev coupled line filter was designed to meet the
following specifications:
Center Frequency: 2 GHz.
Bandwidth: 10%
Number o f Coupled Line sections: 4 (N=3)
Reproduced with permission of the copyright owner. Further reproduction prohibited without permission.
22
Substrate used: R03003
Relative dielectric constant: 3.0
Dielectric Loss tangent: 0.0013
Thickness o f the substrate: I5m.il
The filter was designed and optimized using the RF/microwave design software Ansoft
Serenade 8.5. It was modeled and simulated using Sonnet. The lowpass filter prototype
circuit element values for this filter are given below [9]:
g} = 1.5963
g2 = 1.0967
g3 = 1.5963
g4 = 1.0000
Equations (3.1), (3.2), (3.3), and a characteristic impedance Zo of 50 Q were used
to calculate the even and odd mode characteristic impedances. Equations (3.7) - (3.10)
explain the design procedure.
I nA
,
^
-
f
e
I n x 0.1
-
f
e
j s
r
* 3137
(3 J)
^ T
nA
it* 0.1
Z 0J 2 = — T= = ----- , =0.1187
2 jg ^
2* V1.5963*1.0967
(3.8)
V ’
_ ,
nA
it* 0.1
Z 0J 2 = — = = = ----- = = = = 0 . 1 1 8 7
2 ^[g &
2 1.5963 *1.0967
(3.9)
J
I nA
z »J ‘ = f e
( 3 -, 0 )
I
= f
it *0.1
e
^
r
0'3137
By substituting the ZoJn values obtained from (3.7) - (3.10) in equations (3.5) and
(3.6), the Z0e and Z0o are calculated. The results are summarized in Table 3.1.
Reproduced with permission of the copyright owner. Further reproduction prohibited without permission.
23
n
1
2
3
4
gn
1.5963
1.0967
1.5963
1.0000
ZoJn
0.3137
0.1187
0.1187
0.3137
Zoe ( ^ )
70.6047
56.6406
56.6406
70.6047
Zoo (Hi)
39.2355
44.7688
44.7688
39.2355
Table 3.1: Filter parameters for the third order Chebyshev 0.5 dB equal ripple coupled
line filter. Zoe and Zo0 are the even and odd mode characteristic impedances for 10%
bandwidth.
Using the Z0e and Z0o values and the design specifications, the widths (W),
spacings (S), and lengths (P) of the coupled lines are calculated using Ansoft Serenade.
To calculate the W, S, and P of the coupled lines, (1) dielectric substrate material
properties, (2) thickness o f the dielectric substrate (3) center frequency of the desired
filter and (4) Zoe, and Zo0 values, are input into the software package. After calculating
the dimensions of the coupled lines, the filter geometry is created and optimized using the
same software. The final widths, spacings and lengths o f the coupled lines after
optimization are given in Table 3.2.
n
1
2
3
4
Width W (mils)
19
29.5
29.5
19
Spacing S (mils)
8.5
16
16
8.5
Length P (mils)
967
951
951
967
Table 3.2: Optimized dimensions of the four coupled sections of the filter designed with
R03003 substrate. Center frequency: 2 GHz, bandwidth: 10%.
The dimensions from Table 3.2 are used to create the coupled line filter geometry
with four sections in Sonnet (Figure 3.2.2).
Reproduced with permission of the copyright owner. Further reproduction prohibited without permission.
24
A
30T/I
*— Air...... 3hC
kvwwwWVwVVVwWV
C1I
?*V —"SA
O
V
. .*
3^
OF?
*W "301
?s
an
311
......... I
\vT
A
IQw'vXWwV’AVV
-JU
1—
*
18
T
8.5
**f
Iffi
S
|
28.5 T ~
37.99989
Figure 3.2.2: Geometry of the optimized coupled line bandpass filter designed with
R03003 substrate. Center frequency: 2 GHz, bandwidth: 10%. All dimensions are in
mils.
The conductors for the coupled lines used in the geometry shown in Figure 3.2.2
were assumed to be perfectly conducting. The S21 response of the filter is plotted in
Figure 3.2.3.
S21
-
10-
-20
-
-30 ■
-40 -50 -60 (dB) -70 -80 1
1.2
1.4
1.6
1.8
2
2.2
2.4
2.6
2.8
3
Frequency (GHz)
Figure 3.2.3: S21 response of the coupled line bandpass filter shown in Figure 3.3.2 using
R03003 substrate. Center frequency: 2 GHz, bandwidth: 10%.
The above design procedure was also used to design coupled line bandpass filters
with 15% and 20% bandwidths.
Reproduced with permission of the copyright owner. Further reproduction prohibited without permission.
25
3.2.2 Design of the coupled line filter centered at 2GHz with 15% bandwidth, using
R03003 substrate (A = 0.15)
The calculated values of Zoe and Z q0 are given in Table 3.3.
n
1
2
3
4
gn
1.5963
1.0967
1.5963
1.0000
ZoJn
0.3842
0.1781
0.1781
0.3842
Zoe (£2)
76.5898
60.4895
60.4895
76.5898
Zoo (£2)
38.1706
42.6817
42.6817
38.1706
Table 3.3: Z0e and Z0o values for 15% bandwidth.
The dimensions of the coupled line filter were calculated and the filter was optimized in
Ansoft Serenade to meet the specifications. The dimensions of the designed coupled line
filter are summarized in Table 3.4.
n
1
2
3
4
Width W (mils)
17.5
26.5
26.5
17.5
Spacing S (mils)
6
10
10
6
Length P (mils)
972
956
956
972
Table 3.4: Optimized dimensions of the four coupled sections of the filter designed with
R03003 substrate. Center frequency: 2 GHz, bandwidth: 15%.
Using the dimensions shown in Table 3.4, the coupled line filter geometry was created in
Sonnet (Figure 3.2.4).
Reproduced with permission of the copyright owner. Further reproduction prohibited without permission.
t
m
'i ('
j
■
1 7J ?i t
'ild
'Tj
T
O
JJw
4d
nr
T
O
J jv
1r
\
^
47?
3f t
*
Air
u
* *------- Mi
tii.u'i vu1 \ \ mm.uuAi vVU\ u UvV\,\„ \ va \ U
mTTxYYVVm'UfcTYV
\ WTTOV\TOTT 1
«•>
, ., ,, ■, , .'tj
8924
lf
v
?
r
26
it
Figure 3.2.4: Geometry of the optimized coupled line bandpass filter designed with
R03003 substrate. Center frequency: 2 GHz, bandwidth: 15%. All dimensions are in
mils.
The filter was simulated and the S21 response was obtained (Figure 3.2.5).
80 H-------1-------1------- 1-------1-------1------ 1------ 1-------1-------1-----1
1.2 1.4 1.6 1.8
2
2.2 2.4 2.6 2.8
3
Frequency (GHz)
Figure 3.2.5: S21 response of the designed coupled line bandpass filter using R03003
substrate. Center frequency: 2 GHz, bandwidth: 15%.
Reproduced with permission of the copyright owner. Further reproduction prohibited without permission.
27
3.2.3 Design of the coupled line filter centered at 2 GHz with 20% bandwidth, using
R03003 substrate (A = 0.2)
Zoe and Z q0 for the bandpass filter with 20% bandwidth shown Table 3.5.
n
1
2
3
4
gn
1.5963
1.0967
1.5963
1.0000
ZoJn
0.4436
0.2374
0.2374
0.4436
Zoe (« )
82.0216
64.6907
64.6907
82.0216
Zo„(ii)
37.6589
40.9470
40.9470
37.6589
Table 3.5: Zoe and Zq0 values for 20% bandwidth.
The corresponding optimized filter dimensions are given in Table 3.6
n
1
2
3
4
Width W (mils)
18.5
16
16
18.5
Spacing S (mils)
1.5
4.5
4.5
1.5
Length P (mils)
976
961
961
976
Table 3.6: Optimized dimensions of the four coupled sections of the filter designed with
R03003 substrate. Center frequency: 2 GHz, bandwidth: 20%.
Figures 3.2.6 and 3.2.7 show the geometry of the filter and its S21 response
respectively.
Reproduced with permission of the copyright owner. Further reproduction prohibited without permission.
28
t
‘lid
nIt
1??
J it
}
t y............... J M " " .......... ?
JdU
nw
v------------j | £
v ■<
? \
Off
Ml
X
f
TTTTTrnrnTTTTTOTTI
:m T m x T m rm m T
mUUlUUHVUWU
l u \ U W \ U n n \ « n l \ HUIIIUUIUHHIUI
m m
6!i
l
j
»
a
Figure 3.2.6: Geometry of the optimized coupled line bandpass filter designed with
R03003 substrate. Center frequency: 2 GHz, bandwidth: 20%. All dimensions are in
mils.
S21 -10
-
-20
-
-30 -40 (dB) -50 -60
1
1.2
1.4
1.6
1.8
2
2.2
2.4
2.6
2.8
3
Frequency (GHz)
Figure 3.2.7: S21 response of the designed coupled line bandpass filter using R03003
substrate. Center frequency: 2 GHz, bandwidth: 20%.
Reproduced with permission of the copyright owner. Further reproduction prohibited without permission.
29
3.2.4 Design of the coupled line filter centered at 5 GHz with 10% bandwidth, using
R03003 substrate (A = 0.1)
Using the same dielectric substrate (R03003) third order Chebyshev 0.5 dB equal
ripple coupled line bandpass filters were designed at the higher center frequency of 5
GHz, having 10, 15 and 20% bandwidths. The higher frequency bandpass filter was
chosen to compare its sensitivity to that of a bandpass filter centered at a lower frequency
(2 GHz). The design specifications of the filter are given below.
Center Frequency: 5 GHz.
Number o f Coupled Line Sections: 4 (N=3)
Substrate used: R03003
Relative dielectric constant: 3.0
Dielectric Loss tangent: 0.0013
Thickness o f the substrate: 15mil
The optimized filter dimensions are given in Table 3.7.
n
1
2
3
4
W idth W (mils)
15
23
23
15
Spacing S (mils)
9
18
18
9
Length P (mils)
386
380
380
386
Table 3.7: Optimized dimensions of the four coupled sections of the filter designed with
R03003 substrate. Center frequency: 5 GHz, bandwidth: 10%.
Figures 3.2.8 and 3.2.9 show the geometry and the S21 response of the filter respectively.
Reproduced with permission of the copyright owner. Further reproduction prohibited without permission.
30
386
380
386
380
378
Figure 3.2.8: Geometry of the optimized coupled line bandpass filter designed with
R03003 substrate. Center frequency: 5 GHz, bandwidth: 10%. All dimensions are in
mils.
S21 . 1 0
M
a
g -20
n
i
-30
t
u
d
-40
e
(dB) '50
-60
3.5
4
4.5
5
5.5
6
6.5
Frequency (GHz)
Figure 3.2.9: S2i response o f the designed coupled line bandpass filter using R03003
substrate. Center frequency: 5 GHz, bandwidth: 10%.
3.2.5 Design of the coupled line filter centered at 5 GHz with 15% bandwidth, using
R03003 substrate (A = 0.15)
The optimized dimensions of the coupled line filter are given in table 3.8.
Reproduced with permission of the copyright owner. Further reproduction prohibited without permission.
31
n
1
2
3
4
W idth W (mils)
15
21
21
15
Spacing S (mils)
9
18
18
9
Length P (mils)
388
382
382
388
Table 3.8: Optimized dimensions of the four coupled sections of the filter designed with
R03003 substrate. Center frequency: 5 GHz, bandwidth: 15%.
Figures 3.2.10 and 3.2.11 show the geometry and the S21 response of the filter
respectively.
-378-
9.888833
<s- -382
-388-
18
- > <r~
IT
21
-378-
-382-
*
15
Figure 3.2.10: Geometry of the optimized coupled line bandpass filter designed with
R03003 substrate. Center frequency: 5 GHz, bandwidth: 15%. All dimensions are in
mils.
Reproduced with permission of the copyright owner. Further reproduction prohibited without permission.
32
S21
M
a
-10
g -20
n
i
t
-30
U
d
-40
(dB) ' 50
-60
3.5
4
4.5
5
5.5
6
6.5
Frequency (GHz)
Figure 3.2.11: S21 response of the designed coupled line bandpass filter using R03003
substrate. Center frequency: 5 GHz, bandwidth: 15%.
3.2.6 Design of the coupled line filter centered at 5 GHz with 20% bandwidth, using
R03003 substrate (A = 0.2)
The optimized dimensions of the designed filter are given in Table 3.9.
n
1
2
3
4
Width W (mils)
12
19
19
12
Spacing S (mils)
9
11
11
9
Length P (mils)
390
384
384
390
Table 3.9: Optimized dimensions of the four coupled sections of the filter designed with
R03003 substrate. Center frequency: 5 GHz, bandwidth: 20%.
Reproduced with permission of the copyright owner. Further reproduction prohibited without permission.
The geometry of the created coupled line filter and its S21 response are given in Figures
3.2.12 and 3.2.13 respectively.
«----- 378------
I
X s
9.00002
Figure 3.2.12: Geometry of the optimized coupled line bandpass filter designed with
R03003 substrate. Center frequency: 5 GHz, bandwidth: 20%. All dimensions are in
mils.
S21
M
-1 0
a
-15
I i
-25
*
-30
U
d
-35
e
.4 0
(dB )
'
'
-45
-50
3.5
4
4.5
5
55
6
6.5
F r e q u e n c y (G H z)
Figure 3.2.13: S21 response of the designed coupled line bandpass filter using R03003
substrate. Center frequency: 5 GHz, bandwidth: 20%.
Reproduced with permission of the copyright owner. Further reproduction prohibited without permission.
34
3.3 SIMULATION OF THE FILTERS DESIGNED USING R03003 SUBSTRATE
WITH DIELECTRIC OVERLAY
This section presents the frequency responses (Figures 3.3.1, 3.3.2, 3.3.3, 3.3.5,
3.3.6, 3.3.7) of all the coupled line bandpass filters centered at 2 GHz and 5 GHz with 3
different bandwidths (10%, 15%, and 20%) designed in section 3.2 with a dielectric
overlay of 0.1 mil thickness on the entire filter circuit. The dielectric overlay is assumed
lossless and has a relative dielectric constant of 80.
021
kJ
M
-10
a
-20
n
-30
i
-40
t
u
-50
d
-60
—0 —No overlay
—Q —0.imil overlay
e
(dB) -70
-80
1
1.2
1.4
1.6
1.8
2
2.2
2.4
2.6
2.8
3
Frequency (GHz)
Figure 3.3.1: S21 response of the coupled line bandpass filter using R03003 substrate.
Center frequency: 2 GHz, bandwidth: 10%. Response is shown with and without the
dielectric overlay.
Reproduced with permission of the copyright owner. Further reproduction prohibited without permission.
35
0
S21
M
-10
a
-20
g
-30
n
i
t
u
d
—O —No overlay
—Q —0.1mil overlay
-40
-50
-60
e
-70
(dB)
-80
1.2
1
1.4
1.6
1 .8
2
2.2
2.4
2.8
2.6
3
Frequency (GHz)
Figure 3.3.2: S21 response o f the coupled line bandpass filter using R03003 substrate.
Center frequency: 2 GHz, bandwidth: 15%. Response is shown with and without the
dielectric overlay.
—O —No overlay
- O aim H overlay
S21
-1 0
M
Q
-20
n
-30
g
i
t
-40
u
d
-50
(dB) -60
-70
1
1.2
1.4
1.6
1.8
2
2.2
2.4
2.6
2.8
3
Frequency (GHz)
Figure 3.3.3: S21 response o f the coupled line bandpass filter using R03003 substrate.
Center frequency: 2 GHz, bandwidth: 20%. Response is shown with and without the
dielectric overlay.
Reproduced with permission of the copyright owner. Further reproduction prohibited without permission.
36
Figures 3.3.1,3.3.2 and 3.3.3 indicate that for all values of bandwidth, a dielectric
overlay on top o f the entire filter circuit shifts the S21 response towards lower
frequencies. The frequencies corresponding to the 3dB and lOdB points on the low
frequency side of the graphs, with and without the dielectric overlay are summarized in
Table 3.10 together with the amount of center frequency shift when the overlay is added.
Bandwidth
(%)
3dB
frequency
(GHz) [No
Overlay]
10
15
20
1.8916
1.8646
1.7824
3dB
frequency
(GHz)
[0.1 mil
thick
overlayl
1.7619
1.7277
1.6118
Shift in
frequency
at3dB
(MHz)
lOdB
frequency
(GHz)
[No
Overlay]
129.7
136.9
170.6
1.8649
1.8357
1.7319
lOdB
frequency
(GHz)
[O.lmil
thick
overlay]
1.7252
1.6902
1.5386
Shift in
frequency
atlOdB
(MHz)
139.7
145.5
193.3
Table 3.10: Frequencies corresponding to 3dB and lOdB points on the low frequency
side of the S21 response, and frequency shifts with and without overlay. Substrate used is
R03003 and filter is centered at 2 GHz.
The center frequency shift as a function of bandwidth is plotted in Figure 3.3.4,
which indicates that the frequency shift increases with increase in bandwidth, in the
presence of a dielectric overlay on the entire filter circuit.
Reproduced with permission of the copyright owner. Further reproduction prohibited without permission.
37
Figure 3.3.4: Shift in center frequency with increase in bandwidth for a coupled line
filter. R03003 substrate is used. Center frequency: 2 GHz.
Figure 3.3.4 also shows that the center frequency shift is higher for the 10 dB point.
The filters that were designed in section 3.2 using R03003 substrate at 5 GHz
center frequency were simulated adding a dielectric overlay of
0 .1
mil thick over the
entire circuit. Figures 3.3.5, 3.3.6, and 3.3.7 show the shift in S21 response towards lower
frequency.
Reproduced with permission of the copyright owner. Further reproduction prohibited without permission.
38
—O —No overlay
- O —O.Imil overlay
M -10
a
g -20
n
i -30
t
u
-40
d
0 -50
(dB )
-60
3.5
4
5
5.5
F re q u en cy (GHz)
4.5
6
6.5
Figure 3.3.5: S21 response of the coupled line bandpass filter using R03003 substrate.
Center frequency: 5 GHz, bandwidth: 10%. Response is shown with and without the
dielectric overlay.
—O — No overlay
HD—O.Imil overlay
M -10 a
g
n
-20 -
i
t
-30 -
U
.40 -
d
e -50 (dB )
-60
3.5
4
4.5
5
5.5
6
6.5
F re q u en cy (GHz)
Figure 3.3.6: S21 response of the coupled line bandpass filter using R03003 substrate.
Center frequency: 5 GHz, bandwidth: 15%. Response is shown with and without the
dielectric overlay.
Reproduced with permission of the copyright owner. Further reproduction prohibited without permission.
39
—O —No overlay
—
O.Imil overlay
S21
M -10
a
g -20
n
i
-30
t
U
-40
d
e
(dB)
-50
-60
3.5
4
4.5
5
5.5
6
6.5
Frequency (GHz)
Figure 3.3.7: S21 response of the coupled line bandpass filter using R03003 substrate.
Center frequency: 5 GHz, bandwidth: 20%. Response is shown with and without the
dielectric overlay.
The frequencies corresponding to the 3 dB and 10 dB points on the low frequency side of
the graphs, with and without the dielectric overlay are summarized in Table 3.11 along
with the amount of center frequency shift when the overlay is present.
Bandwidth
(%)
3dB
frequency
(GHz)
[No
Overlay]
10
15
20
4.7236
4.6222
4.5445
3dB
frequency
(GHz)
[O.Imil
thick
overlay]
4.3858
4.2691
4.1238
Shift in
frequency
at3dB
(MHz)
lOdB
frequency
(GHz) [No
Overlay]
337.8
353.1
420.7
4.6543
4.6653
4.4961
lOdB
frequency
(GHz)
[0.1mil
thick
overlay]
4.2944
4.2821
4.0426
Shift in
frequency
at lOdB
(MHz)
359.9
383.2
453.5
Table 3.11: Frequencies corresponding to 3dB and lOdB points on the low frequency
side of the S21 response, and frequency shifts with and without overlay. Substrate used is
R03003 and filter is centered at 5 GHz.
Reproduced with permission of the copyright owner. Further reproduction prohibited without permission.
40
The center frequency shift as a function of bandwidth when a dielectric overlay is present
on the entire filter circuit is plotted in Figure 3.3.8.
500
453.5
450
^ 400
N
S 350
£ 300
13dB Shift
* 250
H O d B S h ift
ca> 200
” 150
£
100
50
0
10
15
20
Bandwidth (°Z)
Figure 3.3.8: Shift in center frequency with increase in bandwidth for a coupled line
filter. R03003 substrate is used. Center frequency: 5 GHz.
The results from this section show that increase in bandwidth increases the center
frequency shift in the S21 response when a dielectric overlay is present on the entire filter
circuit. However, fabricating filters having bandwidths more than 20% is practically
difficult, because they require tight coupling [2 ].
Reproduced with permission of the copyright owner. Further reproduction prohibited without permission.
41
3.4
COUPLED
LINE
FILTERS
USING
RT6010
HIGH
DIELECTRIC
CONSTANT SUBSTRATE
This section discusses the design of coupled line bandpass filters using a high
dielectric constant substrate, RT6010 (er = 10.2). Although migrating to high dielectric
substrates would decrease the sensitivity of the filter, the physical size of the filter would
be reduced. When the physical size of the biosensor is a constraint, higher dielectric
substrates can be used at the expense of sensitivity. However, as the dielectric constant of
the substrate increases, filters can be designed for much higher bandwidths to maintain
sensitivity to a dielectric overlay.
3.4.1 Design of the coupled line filter centered at 2 GHz with 10% bandwidth, using
RT6010 substrate (A = 0.1)
The design specifications o f the filter are given below.
Center Frequency: 2 GHz.
Bandwidth: 10%
Number o f Coupled Lines: 4 (N=3)
Substrate used: RT6010
Dielectric constant: 10.2
Dielectric Loss tangent: 0.0023
Thickness o f the substrate: 15mil
The optimized dimensions of the coupled line bandpass filter for 10% bandwidth are
given in Table 3.12.
Reproduced with permission of the copyright owner. Further reproduction prohibited without permission.
42
n
Width W (mils)
11
6
13.5
13.5
17.5
17.5
11
6
1
2
3
4
Spacing S (mils)
Length P (mils)
577
568
568
577
Table 3.12: Optimized dimensions of the four coupled sections o f the filter designed with
RT6010 substrate. Center frequency: 2 GHz, bandwidth: 10%.
The geometry created in Sonnet and the S21 response of the filter are shown in Figure
3.4.1, and Figure 3.4.2, respectively.
565
577
568
568
17.5
577
565
13.5
Figure 3.4.1: Geometry of the optimized coupled line bandpass filter designed with
RT6010 substrate. Center frequency: 2 GHz, bandwidth: 10%. All dimensions are in
mils.
Reproduced with permission of the copyright owner. Further reproduction prohibited without permission.
43
S21
-10
M
a
-20
9
n
i
t
U
-30
-40
-50
d -60
e
(dB) -70
-80
1
1.2
1.4
1.6
1.8
2
2.2
2.4
2.6
2.8
3
Frequency (GHz)
Figure 3.4.2: S21 response of the designed coupled line bandpass filter using RT6010
substrate. Center frequency: 2 GHz, bandwidth: 10%.
3.4.2 Design of the coupled line filter centered at 2 GHz with 15% bandwidth, using
RT6010 substrate (A = 0.15)
The optimized dimensions of the coupled line bandpass filter for 15% bandwidth
are given in Table 3.13.
n
1
2
3
4
Width W (mils)
10
13
13
10
Spacing S (mils)
4
12
12
4
Length P (mils)
581
570
570
581
Table 3.13: Optimized dimensions of the four coupled sections of the filter designed with
RT6010 substrate. Center frequency: 2 GHz, bandwidth: 15%.
Reproduced with permission of the copyright owner. Further reproduction prohibited without permission.
44
The geometry and the S21 response of the filter are shown in Figures 3.4.3 and 3.4.4,
respectively.
565
581
578
578
565
581
4.888089
Figure 3.4.3: Geometry o f the optimized coupled line bandpass filter designed with
RT6010 substrate. Center frequency: 2 GHz, bandwidth: 15%. All dimensions are in
mils.
10 -20
-
-30 -40 -50 -
-70
1
1.2
1.4
1.6
1.8
2
2.2
2.4
2.6
2.8
3
Frequency (GHz)
Figure 3.4.4: S21 response of the designed coupled line bandpass filter using RT6010
substrate. Center frequency: 2 GHz, bandwidth: 15%.
Reproduced with permission of the copyright owner. Further reproduction prohibited without permission.
45
3.4.3 Design of the coupled line filter centered at 2 GHz with 20% bandwidth, using
RT6010 substrate (A = 0.2)
The optimized dimensions of the filter designed for 20% bandwidth centered at 2GHz are
given in Table 3.14.
n
1
Width W (mils)
9
2
12
3
4
12
9
Length P (mils)
583
573
573
583
Spacing S (mils)
3.5
8.5
8.5
3.5
Table 3.14: Optimized dimensions of the four coupled sections of the filter designed with
RT6010 substrate. Center frequency: 2 GHz, bandwidth: 20%.
The geometry and the S21 response of the filter are shown in Figures 3.4.5 and 3.4.6
respectively.
<______ 515______ J U___ ..$Q3................J ft
U—J
3.499937
8*5
573
j,
<
583
9 «--------- 565---------->
T " ......‘.................... f
12
9
Figure 3.4.5: Geometry o f the optimized coupled line bandpass filter designed with
RT6010 substrate. Center frequency: 2 GHz, bandwidth: 20%. All dimensions are in
mils.
Reproduced with permission of the copyright owner. Further reproduction prohibited without permission.
46
o
a
n
i
-30 -
t
U
-40 -
d
-60
1
1.2
1.4
1.6
1.8
2
2.2
2.4
2.6
2.8
3
Frequency (GHz)
Figure 3.4.6: S21 response of the designed coupled line bandpass filter using RT6010
substrate. Center frequency: 2 GHz, bandwidth: 20%.
3.4.4 Design nf the coupled line filter centered at 5 GHz with 10% bandwidth, using
RT6010 substrate (A = 0.1)
The design specifications of the filter are given below.
Center Frequency: 5 GHz.
Number o f Coupled Lines: 4 (N=3)
Substrate used: RT6010
Dielectric constant: 10.2
Dielectric Loss tangent: 0.0023
Thickness o f the substrate: 1 mil
The optimized dimensions of the coupled line filter are given in Table 3.15.
Reproduced with permission of the copyright owner. Further reproduction prohibited without permission.
47
n
W idth W (mils)
Spacing S (mils)
1
11
6
13.5
13.5
17.5
17.5
11
6
2
3
4
Length P (mils)
230
226
226
230
Table 3.15: Optimized dimensions of the four coupled sections of the filter designed with
RT6010 substrate. Center frequency: 5 GHz, bandwidth: 10%.
The geometry and response of the filter are shown in Figures 3.4.7 and 3.4.8 respectively.
*
6.00001'1
1
*
17.5
+
13.5
_____
wWWWXwvxWvxXvVxx
___ ______ N \\\\N \\\\\\N V .\N N \\N N L w w v w w w x w w w w i l
*
11
Figure 3.4.7: Geometry of the optimized coupled line bandpass filter designed with
RT6010 substrate. Center frequency: 5 GHz, bandwidth: 10%. All dimensions are in
mils.
Reproduced with permission of the copyright owner. Further reproduction prohibited without permission.
48
0
S21
-10
-20
-30
-40
(dB)
' 50
-60
4
3.5
4.5
5
5.5
6
6.5
Frequency (GHz)
Figure 3.4.8: S21 response of the designed coupled line bandpass filter using RT6010
substrate. Center frequency: 5 GHz, bandwidth: 10%.
3.4.5 Design of the coupled line filter centered at 5 GHz with 15% bandwidth, using
RT6010 substrate (A = 0.15)
The optimized dimensions o f the filter are given in Table 3.16.
n
Width W (mils)
1
10
Spacing S (mils)
4
13
13
12
10
4
2
3
4
12
Length P (mils)
231.5
227
227
231.5
Table 3.16: Optimized dimensions of the four coupled sections of the filter designed with
RT6010 substrate. Center frequency: 5 GHz, bandwidth: 15%.
Reproduced with permission of the copyright owner. Further reproduction prohibited without permission.
49
The geometry and the S21 response of the filter are shown in Figures 3.4.9 and 3.4.10,
respectively.
■225
■231.5
■227
12
'
231.5
225
kwwvwww'A.wwww'
Figure 3.4.9: Geometry of the optimized coupled line bandpass filter designed with
RT6010 substrate. Center frequency: 5 GHz, bandwidth: 15%. All dimensions are in
mils.
-20
-
-25 u -30 d
e _35'
(dB) _4 0 .
-45 3.5
4
4.5
5
5.5
6
6.5
Frequency (GHz)
Figure 3.4.10: S21 response of the designed coupled line bandpass filter using RT6010
substrate. Center frequency: 5 GHz, bandwidth: 15%.
Reproduced with permission of the copyright owner. Further reproduction prohibited without permission.
50
3.4.6 Design of the coupled line filter centered at 5 GHz with 20% bandwidth, using
RT6010 substrate (A = 0.2)
The optimized dimensions of the filter are given in Table 3.17.
n
1
Width W (mils)
9
2
12
3
4
12
9
Spacing S (mils)
3.5
8.5
8.5
3.5
Length P (mils)
232.5
228.5
228.5
232.5
Table 3.17: Optimized dimensions of the four coupled sections of the filter designed with
RT6010 substrate. Center frequency: 5 GHz, bandwidth: 20%.
The geometry and the S21 response of the filter are shown in Figures 3.4.11 and 3.4.12,
respectively.
-228.5-
1.5
232.5
225
T l V \ V v s X S \ \ V s \\ \\ \ \V ^
Figure 3.4.11: Geometry of the optimized coupled line bandpass filter designed with
RT6010 substrate. Center frequency: 5 GHz, bandwidth: 20%. All dimensions are in
mils.
Reproduced with permission of the copyright owner. Further reproduction prohibited without permission.
51
S21
-15
-20
-25
u -30
e
(dB) -35
-40
3.5
4
4.5
5
5.5
6
6.5
Frequency (GHz)
Figure 3.4.12: S21 response of the designed coupled line bandpass filter using RT6010
substrate. Center frequency: 5 GHz, bandwidth: 20%.
3.5 SIMULATION OF THE FILTERS DESIGNED USING RT6010 SUBSTRATE
WITH DIELECTRIC OVERLAY
This section presents the frequency responses of all the coupled line filters designed in
section 3.4 with a dielectric overlay on the entire filter circuit. The S21 responses of the
filters centered at 2 GHz frequency with fractional bandwidths of 0.1, 0.15, and 0.2, with
and without a dielectric overlay are shown in Figures 3.5.1, 3.5.2 and 3.5.3, respectively.
Reproduced with permission of the copyright owner. Further reproduction prohibited without permission.
52
—0 —No overlay
—□ —O.Imil overlay
-10
-20
-30
-40
-50
-60
(dB) -70
-80
1
1.2
1.4
1.6
1.8
2
2.2
2.4
2.6
2.8
3
Frequency (GHz)
Figure 3.5.1: S21 response of the coupled line bandpass filter using RT6010 substrate.
Center frequency: 2 GHz, bandwidth: 10%. Response is shown with and without the
dielectric overlay.
- 0 —No overlay
—D —O.Imil overlay
-20
-30
-40
-50
(dB) ' 60
-70
1
1.2
1.4
1.6
1.8
2
2.2
2.4
2.6
2.8
3
Frequency (GHz)
Figure 3.5.2: S21 response of the coupled line bandpass filter using RT6010 substrate.
Center frequency: 2 GHz, bandwidth: 15%. Response is shown with and without the
dielectric overlay.
Reproduced with permission of the copyright owner. Further reproduction prohibited without permission.
53
—O —No overlay
—D ~ 0.1mil overlay
M
M -10
a
g
-20
n
i
-30
t
u
-40
d
(dB) ■5°
-60
1
1.2
1.4
1.6
1.8
2
2.2
2.4
2.6
2.8
3
Frequency (GHz)
Figure 3.5.3: S21 response of the coupled line bandpass filter using RT6010 substrate.
Center frequency: 2 GHz, bandwidth: 20%. Response is shown with and without the
dielectric overlay.
The frequencies corresponding to the 3 dB and 10 dB points on the low frequency
side of the graphs, for the bandpass filters centered at 2 GHz, with and without the
dielectric overlay, are summarized in Table 3.18 along with the center frequency shift
when the overlay is present.
Reproduced with permission of the copyright owner. Further reproduction prohibited without permission.
54
Bandwidth
(%)
3dB
frequency
(GHz)
[No
Overlay]
10
15
20
1.8808
1.8250
1.7711
3dB
frequency
(GHz)
[O.Imil
thick
overlay]
1.7979
1.7366
1.6777
Shift in
frequency
at3dB
(MHz)
lOdB
frequency
(GHz)
[No
Overlay]
82.9
88.4
93.4
1.8525
1.7818
1.7225
lOdB
frequency
(GHz)
[O.Imil
thick
overlay!
1.7654
1.6861
1.621
Shift in
frequency
at lOdB
(MHz)
87.1
95.7
101.5
Table 3.18: Frequencies corresponding to 3 dB and 10 dB points on the low frequency
side of the S21 response, and frequency shifts with and without overlay. Substrate used is
RT6010 and filter is centered at 2 GHz.
The center frequency shift as a function of bandwidth is plotted in Figure 3.5.4.
120
100
U
X
*
80
m
60
o>.
c
V
3
40
or
£
U20
0Bandwidth WA
Figure 3.5.4: Shift in center frequency with increase in bandwidth for a coupled line
filter. RT6010 substrate is used. Center frequency: 2 GHz.
The filters that were designed in section 3.4 using RT6010 substrate at 5 GHz
center frequency were simulated adding a dielectric overlay of 0.1 mil thickness on the
Reproduced with permission of the copyright owner. Further reproduction prohibited without permission.
55
entire filter circuit. Figures 3.5.5, 3.5.6, and 3.5.7 show the S21 responses with and
without dielectric overlay.
—O —No overlay
—Q —O.Imil overlay
-60
1------------- 1
4
4.5
3.5
1
5
1------------- 1------------5.5
6
6.5
Frequency (GHz)
Figure 3.5.5: S21 response of the coupled line bandpass filter using RT6010 substrate.
Center frequency: 5 GHz, bandwidth: 10%. Response is shown with and without the
dielectric overlay.
—O —No overlay
—□ —O.Imil overlay
M _10 .
a
g -15n
-20 -
t
-25 -
u
-30 -
d
e -35 (dB) _40 .
-45
3.5
4
4.5
5
5.5
6
6.5
Frequency (GHz)
Figure 3.5.6: S21 response of the coupled line bandpass filter using RT6010 substrate.
Center frequency: 5 GHz, bandwidth: 15%. Response is shown with and without the
dielectric overlay.
Reproduced with permission of the copyright owner. Further reproduction prohibited without permission.
56
—O —No overlay
—
O.lmii overlay
S21
M '5 '
a -10 -
?
-
^
-30 -
i -20 t
U -25-
e
(dB) -35 ’
-40
3.5
4
4.5
5
5.5
6
6.5
Frequency (GHz)
Figure 3.5.7: S21 response of the coupled line bandpass filter using RT6010 substrate.
Center frequency: 5 GHz, bandwidth: 20%. Response is shown with and without the
dielectric overlay.
The frequencies corresponding to the 3 dB and 10 dB points on the low frequency
side of the graphs, for the bandpass filters centered at 5GHz, with and without the
dielectric overlay are summarized in Table 3.19 along with the amount of center
frequency shift when the overlay is added.
Reproduced with permission of the copyright owner. Further reproduction prohibited without permission.
57
Bandwidth
(%)
3dB
frequency
(GHz)
[No
Overlay]
10
15
20
4.6593
4.5184
4.3788
3dB
frequency
(GHz)
[O.Imil
thick
overlay]
4.4467
4.2839
4.1211
Shift in
frequency
at 3dB
(MHz)
lOdB
frequency
(GHz)
[No
Overlay]
212.6
234.5
257.7
4.5904
4.4116
4.2483
lOdB
frequency
(GHz)
[O.Imil
thick
overlay]
4.3682
4.1581
3.9659
Shift in
frequency
at lOdB
(MHz)
222.2
253.5
282.4
Table 3.19: Frequencies corresponding to 3 dB and 10 dB points on the low frequency
side of the S21 response, and frequency shifts with and without overlay. Substrate used is
RT6010 and filter is centered at 5 GHz.
The center frequency shift as a function of bandwidth is plotted in Figure 3.5.8.
Bandwidth (%)
Figure 3.5.8: Shift in center frequency with increase in bandwidth for a coupled line
filter. RT6010 substrate is used. Center frequency: 5 GHz.
Reproduced with permission of the copyright owner. Further reproduction prohibited without permission.
58
3.6 SIMULATION RESULTS FO R INCREASING DIELECTRIC OVERLAY
THICKNESSES
By increasing the dielectric overlay thickness a larger shift in center frequency
response can be obtained. This section presents the simulation results, with two different
overlay thicknesses on the entire filter circuit.
Figures (3.6.1) - (3.6.3) show the shifts in S21 response with 0.1 mil and 0.2 mil
dielectric overlay thicknesses for filters designed using R03003 substrate, centered at 2
GHz, with A = 0.1, 0.15, and 0.2 respectively. Also shown are the responses without an
overlay.
M
a
- O — No overlay
—[3— O.lmM
—O— 0.2mil
.10
-20
9
n
i
-30
-40
t
u
-50
d
-60
e
-70
(dB ) -so
-90
1
1.2
1.4
1.6
1.8
2
2.2
2.4
2.6
2.8
3
F re q u en cy (GHz)
Figure 3.6.1: S21 response of the filter with two overlay thicknesses at 2 GHz center
frequency designed with R03003 substrate for A = 0.1. The response with and without
overlay is shown for comparison.
Reproduced with permission of the copyright owner. Further reproduction prohibited without permission.
59
S 21
o
M
-io
—0 ~ No overlay
- Q - O.Imil
—0 — O.Zmil
a
g
-20
n
-30
’
-40
t
u
-50
d
-60
e
-70
(dB ) -8D
-90
1
1.2
1.4
1.6
1.8
2
2.2
2.4
2.6
2.8
3
F re q u en cy (GHz)
Figure 3.6.2: S21 response of the filter with two overlay thicknesses at 2 GHz center
frequency designed with R03003 substrate for A = 0.15. The response with and without
overlay is shown for comparison.
—G — No overlay
- Q — 0.1mil
—0 — 0.2mil
M
a
g
n
i
-20
t
-40
u
d
e
-so
(dB)
-80
1
1.2
1.4
1.6
1 .8
2
2.2
2.4
2.6
2.8
3
F re q u en cy (GHz)
Figure 3.6.3: S21 response of the filter with two overlay thicknesses at 2 GHz center
frequency designed with R03003 substrate for A = 0.2. The response with and without
overlay is shown for comparison.
Reproduced with permission of the copyright owner. Further reproduction prohibited without permission.
60
Figures (3.6.4) - (3.6.6) show the shifts in S21 response centered at 5 GHz, with A = 0.1,
0.15, and 0.2 respectively.]
S21
M
—O — No overlay
—Q — 0.1 mil
- O — 0.2mil
-10
a
9
-20
n
i
t
y
d
-30
-50
e
-60
(d B )
-70
3.5
4
4.5
5
5.5
6
6.5
Frequency (GHz)
Figure 3.6.4: S21 response of the filter with two overlay thicknesses at 5 GHz center
frequency designed with R03003 substrate for A = 0.1. The response with and without
overlay is shown for comparison.
S21
M
a
9
n
i
t
u
d
e
(dB)
0
—O — No overlay
—Q — 0.1 mil
—O— 0.2mil
-10
-20
-30
-50
-60
-70
3.5
4
4.5
5
5.5
6
6.5
Frequency (GHz)
Figure 3.6.5: S21 response o f the filter with two overlay thicknesses at 5 GHz center
frequency designed with R03003 substrate for A = 0.15. The response with and without
overlay is shown for comparison.
Reproduced with permission of the copyright owner. Further reproduction prohibited without permission.
61
M
a
g
—O — No overlay
—Q —0.1mil
—O — 0.2mil
-10
-20
n
-30
U
d
e
-50
-60
(dB)
-70
3.5
4
4.5
5
5.5
6
6.5
Frequency (GHz)
Figure 3.6.6: S2 1 response of the filter with two overlay thicknesses at 5 GHz center
frequency designed with R03003 substrate for A = 0.2. The response with and without
overlay is shown for comparison.
Figures (3.6.7) - (3.6.9) show the shifts in S21 response with 0.1 mil and 0.2 mil
overlay thicknesses for filters designed using RT6010 substrate, centered at 2 GHz, with
A = 0.1,0.15, and 0.2, respectively.
Reproduced with permission of the copyright owner. Further reproduction prohibited without permission.
62
S 21
M
a
-10
g
-20
n
i
-30
—O — No overlay
—Q — 0.1mil
—0 — 0.2mil
t
u
d
e
(dB)
-50
-60
-70
-80
1
1.2
1.4
1.6
1.8
2
2.2
2.4
2.6
2.8
3
Frequency (GHz)
Figure 3.6.7: S21 response of the filter with two overlay thicknesses at 2 GHz center
frequency designed with RT6010 substrate for A = 0.1. The response with and without
overlay is shown for comparison.
S21
M
a
g
—Q — No overlay
—Q — 0.1mil
—O— 0.2mil
-10
-20
n
-30
t
u
d
e
-40
-50
-60
(dB)
-70
1
1.2
1.4
1.6
1.8
2
2.2
2.4
2.6
2.8
3
Frequency (GHz)
Figure 3.6.8: S21 response of the filter with two overlay thicknesses at 2 GHz center
frequency designed with RT6010 substrate for A = 0.15. The response with and without
overlay is shown for comparison.
Reproduced with permission of the copyright owner. Further reproduction prohibited without permission.
63
S 21
o
—O — No overlay
—Q —O.lmil
- O — 0.2mll
M
a
-10
g
n
i
-20
t
-30
u
d
-40
e
-50
(dB )
-60
1
1.2
1.4
1.6
1.8
2.6
2
2.8
3
F re q u en cy (GHz)
Figure 3.6.9: S21 response of the filter with two overlay thicknesses at 2 GHz center
frequency designed with RT6010 substrate for A = 0.2. The response with and without
overlay is shown for comparison.
Figures (3.6.10) - (3.6.12) show the shifts in S21 response centered at 5 GHz, with A =
0.1,0.15, and 0.2, respectively.
S21
M
a
—O — No overlay
- Q —O.lmil
-O — 0.2mil
-ID
9
n
i
t
-30
u
d
-40
e
-50
(dB)
-60
3.5
4
4.5
5
5.5
6
6.5
F req u en cy (GHz)
Figure 3.6.10: S21 response of the filter with two overlay thicknesses at 5 GHz center
frequency designed with RT6010 substrate for A = 0.1. The response with and without
overlay is shown for comparison.
Reproduced with permission of the copyright owner. Further reproduction prohibited without permission.
64
S 21
D
M
—O — No overlay
—Q —O.lmil
—0 — OJhiil
-to
a
g
-20
n
i
t
u
-30
d
-50
e
-SO
(dB)
-70
3.5
4
4.5
5
6
5.5
6.5
Frequency (GHz)
Figure 3.6.11: S21 response of the filter with two overlay thicknesses at 5 GHz center
frequency designed with RT6010 substrate for A - 0.15. The response with and without
overlay is shown for comparison.
0
No overlay
0.1mtt
O.ZmH
S21
M
a
9
-15
n
i -20
t
u
d -30
e
(dB)
-40
3.5
4
4.5
5
5.5
6
6.5
Frequency (GHz)
Figure 3.6.12: S21 response of the filter with two overlay thicknesses at 5 GHz center
frequency designed with RT6010 substrate for A = 0.2. The response with and without
overlay is shown for comparison.
Reproduced with permission of the copyright owner. Further reproduction prohibited without permission.
65
Tables 3.20 and 3.21 present the frequency shift at the 3 dB and 10 dB points for
A values o f 0.1, 0.15, and 0.20, at 2 GHz and 5 GHz with two different overlay
thicknesses, for R03003 and RT6010 substrates, respectively.
Center
Frequency
(GHz)
Fractional
Bandwidth
(A)
Overlay
thickness
(mil)
2
2
2
2
2
2
5
5
5
5
5
5
0.1
0.1
0.15
0.15
0.2
0.2
0.1
0.1
0.15
0.15
0.2
0.2
0.1
0.2
0.1
0.2
0.1
0.2
0.1
0.2
0.1
0.2
0.1
0.2
3dB
frequency
Shift
(MHz)
129.7
207.0
136.9
233.5
170.6
287.6
337.8
571.1
353.1
590.7
420.7
696.9
10dB
frequency
Shift
(MHz)
139.7
223.1
145.5
250.1
193.3
328.7
359.9
613.3
383.2
639.4
453.5
757.2
Table 3.20: Frequency shift at the 3 dB and 10 dB points for R03003 substrate at 2 GHz
and 5 GHz for A = 0.1, 0.15, and 0.2, with 0.1 and 0.2 mil overlay thicknesses.
Reproduced with permission of the copyright owner. Further reproduction prohibited without permission.
66
Center
Frequency
(GHz)
Fractional
Bandwidth
(A)
Overlay
thickness
(mil)
2
2
2
2
2
2
5
5
5
5
5
5
0.1
0.1
0.15
0.15
0.2
0.2
0.1
0.1
0.15
0.15
0.2
0.2
0.1
0.2
0.1
0.2
0.1
0.2
0.1
0.2
0.1
0.2
0.1
0.2
3dB
frequency
Shift
(MHz)
82.9
142.3
88.4
152.0
93.4
164.8
212.6
364.2
234.5
395.9
257.7
436.3
lOdB
frequency
Shift
(MHz)
87.1
149.9
95.7
165.6
101.5
179.6
222.2
382.5
253.5
431.1
282.4
481.1
Table 3.21: Frequency shift at the 3 dB and 10 dB points for RT6010 substrate at 2 GHz
and 5 GHz for A = 0.1,0.15, and 0.2, with 0.1 and 0.2 mil overlay thicknesses.
Using the results from Table 3.20 and Table 3.21, Figures 3.6.13 - 3.6.20 were
plotted to show the shifts in frequency response, for filters using R03003 and RT6010
substrates, at the 3dB and lOdB points, respectively.
Reproduced with permission of the copyright owner. Further reproduction prohibited without permission.
67
■ 0. lmil Overlay
■ 0.2mil Overlay
Fractional Bandwidth
Figure 3.6.13: Frequency shifts at the 3 dB point for the coupled line filter centered at 2
GHz, using R03003 substrate and two dielectric overlay thicknesses.
■ O.lmil Overlay
■ 0.2mil Overlay
Fractional Bandwidth
Figure 3.6.14: Frequency shifts at the 10 dB point for the coupled line filter centered at
2 GHz, using R03003 substrate two dielectric overlay thicknesses.
Reproduced with permission of the copyright owner. Further reproduction prohibited without permission.
68
■ O.lmil Overlay
■ 0.2mil Overlay
Fractional Bandwidth
Figure 3.6.15: Frequency shifts at the 3 dB point for the coupled line filter centered at 5
GHz, using R03003 substrate two dielectric overlay thicknesses.
■ O.lmil Overlay
■ 0.2mil Overlay
Fractional Bandwidth
Figure 3.6.16: Frequency shifts at the 10 dB point for the coupled line filter centered at
5 GHz, using R03003 substrate two dielectric overlay thicknesses.
Reproduced with permission of the copyright owner. Further reproduction prohibited without permission.
69
■ O.lmil Overlay
■ 0.2mil Overlay
Fractional Bandwidth
Figure 3.6.17: Frequency shifts at the 3 dB point for the coupled line filter centered at 2
GHz, using RT6010 substrate two dielectric overlay thicknesses.
■ 0. lmil Overlay
■ 0.2mil Overlay
Fractional Bandwidth
Figure 3.6.18: Frequency shifts at the 10 dB point for the coupled line filter centered at
2 GHz, using RT6010 substrate two dielectric overlay thicknesses.
Reproduced with permission of the copyright owner. Further reproduction prohibited without permission.
70
■ O.lmil Overlay
■ 0.2mil Overlay
Fractional Bandwidth
Figure 3.6.19: Frequency shifts at the 3 dB point for the coupled line filter centered at 5
GHz, using RT6010 substrate two dielectric overlay thicknesses.
■ 0.1 mil Overlay
■ 0.2mil Overlay
Fractional Bandwidth
Figure 3.6.20: Frequency shifts at the lOdB point for the coupled line filter centered at 5
GHz, using RT6010 substrate two dielectric overlay thicknesses.
The results presented in this chapter are discussed in the next chapter.
Reproduced with permission of the copyright owner. Further reproduction prohibited without permission.
CHAPTER 4
DISCUSSION
4.1 INTRODUCTION
A biosensor based on microstrip line coupled line bandpass filters with 10%,
15%, and 20% bandwidths was simulated at center frequencies of 2 GHz and 5 GHz,
using R03003 (er = 3.0) and RT6010 (sr = 10.2) substrates, with different dielectric
overlay thicknesses (Chapter 3). In this chapter these results are discussed. In addition, a
different method is explored to improve the sensitivity of the biosensor.
4.2 COUPLED LINE BANDPASS FILTER SENSITIVITY WITH DIELECTRIC
OVERLAY
4.2.1 Increasing the fractional bandwidth of the coupled line bandpass filter
The coupled line bandpass filters with three different bandwidths were simulated
in the previous chapter. In chapter 2 it is shown that the coupling between the coupled
lines can be increased by increasing A. With more coupling, more electric fields are
concentrated on top o f the filter circuit. Hence, as the bandwidth (A) of the filter
increased, the sensitivity o f the response increased, when a dielectric overlay is present
on the filter circuit. However, the widths of the coupled lines and spacings between them
became smaller with increase in A. Fabricating filters having very small dimensions is
practically difficult using conventional printed circuit board cutting devices. However,
innovative devices like the LKPF ProtoLaser 20 (LPKF Laser & Electronics North
America, Wilsonville, OR) are capable of fabricating microstrip filter circuits with widths
as small as 2 mils and spacings as small as 1 mil [10].
Reproduced with permission of the copyright owner. Further reproduction prohibited without permission.
72
4.2.2 Low dielectric constant substrate vs. high dielectric constant substrate
In chapter 3, coupled line bandpass filters with two different dielectric substrates
(R03003 and RT6010) were designed and simulated. These filters were centered at two
different frequencies, 2 GHz and 5 GHz. From the results, the amount of frequency shift
in the S21 response is higher in the case of the low dielectric constant substrate (er = 3 .0
for R03003), as compared to a high dielectric constant substrate (er = 10.2 for RT6010).
The amount of frequency shift depends on the change in effective permittivity (seff) of the
circuit. In the absence of an overlay,
(4-1)
1 < £ eff< £ r
where er is the relative dielectric constant of the substrate.
However, when an overlay of dielectric constant sov is present, the new effective
permittivity o f the microstrip line with overlay ( s'eff ) lies between eeff and eov.
S eff
< £ eff
<
£ov
(4-2)
Moreover, in the case of a coupled line bandpass filter, the higher the difference between
er and eov, the higher the amount of shift in center frequency in the presence of a dielectric
overlay. Hence, using a low dielectric constant substrate (er = 3.0) produces a higher shift
in center frequency than a high dielectric constant substrate (sr = 10.2). However, since
the length of the coupled lines is 1 /4 , where Xg = guide wavelength in the microstrip line
and Xg = X0j
, where lo is the free space wavelength, the length of the coupled lines
depends on the effective permittivity of the circuit. Coupled line sections for a filter
designed using a high dielectric constant substrate will have lengths less than those for
designed using a low dielectric constant substrate , which was observed in the previous
Reproduced with permission of the copyright owner. Further reproduction prohibited without permission.
73
chapter. Hence, using a substrate material with a higher value of sr in the filter design will
result in a physically smaller biosensor, in cases where the size of the biosensor is a
constraint, at the expense of a reduction in sensitivity of the biosensor.
4.2.3 Filters designed at higher center frequency
As discussed in section 4.2.1, the sensitivity of the filter increases with increase in
A of the filter. At higher frequencies, for the same value of A, the amount of bandwidth in
GHz is higher compared to the lower frequency design. So the filter responses appear
wider than those designed at lower frequencies. Hence, using the coupled line filter
designed at a higher frequency would give more shifts compared to one designed at a
lower frequency for the same dielectric overlay thickness.
4.2.4 Increase in dielectric overlay thickness
The effective permittivity of the microstrip circuit depends on both the thickness
and permittivity of the dielectric overlay. A higher center frequency shift and hence
higher sensitivity is obtained by increasing the thickness of the overlay, since the
effective permittivity increases with increase in dielectric overlay thickness. Although the
typical dimension of bacteria is of the order of 0.1 mil [11], bacteria are known to form
multilayers, when they adhere to surfaces, i.e., they forms biofilms.
A slimy glue like substance that can be formed on any type o f material such as
metal, plastic, and soil particles due to the adherence of bacteria on the surface of the
material is called a biofilm [12]. Extensive research is going on at Center for Biofilm
Engineering, located at Montana State University, Bozeman, MT, to understand and
Reproduced with permission of the copyright owner. Further reproduction prohibited without permission.
74
control biofilm processes. Biofilms are found on surfaces of drinking water distribution
systems. Pathogenic bacteria in the water may be trapped by these biofilms making the
water harmful for human usage. So, detection of pathogenic bacteria at some selected
points along the water distribution network is very critical [13]. The bacteria start
growing into microcolonies and eventually into a biofilm as shown in Figure 4.2.1 [14].
Mature biofilm
Bacteria
Microcolony
Attached
monolayer
Figure 4.2.1: Schematic for the regulation of biofilm formation [14].
Biofilm properties and thickness were estimated using an acoustic microscope,
and the thickest portions of the biofilm were found to be about 145 pm [15].
As the overlay on the filter circuit becomes thicker, the filter circuit will become
more sensitive producing more shift in the center frequency of the filter. The concept of
biofilms explains that there is a tendency for bacteria to stick to each other forming into
multiple layers. Moreover, there is also a probability for the bacteria layers to form in non
uniform thicknesses.
Reproduced with permission of the copyright owner. Further reproduction prohibited without permission.
75
CHAPTER 5
THE COMBLINE DUAL BANDPASS FILTER USING STEPPED IMPEDANCE
RESONATORS AS A BIOSENSOR
5.1 INTRODUCTION
In this chapter a dual-band bandpass combline stepped impedance filter is
presented as a potential candidate for a biosensor. A combline dual bandpass filter using
stepped impedance resonators was designed, fabricated and tested in [16]. This filter was
designed for multimode wireless LANs to operate over two frequency bands. The dual
passband effect was obtained using two stepped impedance resonators inter-coupled in a
single topology as a combline filter. The feature of dual passband is obtained by multiple
resonances of the stepped impedance resonators. This filter is tested for feasibility of use
as a biosensor for detection o f bacteria in water. The filter circuit is created in Sonnet and
a dielectric overlay is applied on top of various parts of the circuit. The response of the
filter due to the dielectric overlay is presented in the following section.
5.2 COMBLINE FILTER
Generally stepped impedance resonators can be used to implement lowpass filters
in microstrip or stripline circuits. These resonators use altering sections of very high and
very low characteristic impedance lines. Filters designed using these resonators are
generally referred to as stepped impedance filters or hi-Z, low-Z filters. However,
because o f the approximations involved in their implementation, their electrical
performance is generally not as good as filters designed using the insertion loss method,
Reproduced with permission of the copyright owner. Further reproduction prohibited without permission.
76
so they are generally used in applications where a sharp cutoff is not required. Figure
5.2.1 shows the geometry o f a stepped impedance resonator [9].
h’ h
Figure 5.2.1: Stepped impedance resonator.
Here, Z/ and Zh are the characteristic impedances of the low and high impedance sections
and, Oi and Oh are their electrical lengths respectively. The combline filter is designed
using the optimized design specifications from [16]. The optimized design specifications
o f the combline topology and the dielectric substrate details are given below.
Z h = 80.6Q ,6h = 72.6°.
Z, = 55.8Q ,6l =36.9°.
Relative dielectric constant: 3.25
Thickness o f the substrate: 29.5m.il
The widths and lengths of the resonators are obtained by inputting the above values into
Ansoft Serenade. The optimized dimensions of the resonators are given in Table 5.1.
Reproduced with permission of the copyright owner. Further reproduction prohibited without permission.
77
Length (mils)
308
621
Resonator
Low Impedance
High Impedance
W idth (mils)
59
30
Table 5.1: Dimensions of the stepped impedance resonators.
Using the above dimensions the stepped impedance combline filter is created in Sonnet
and the geometry o f the filter is shown in Figure 5.2.2.
<24
-6 2 1 -
&
an
a
1
I
51
*1184
Figure 5.2.2: Geometry o f the combline dual-band bandpass filter generated in Sonnet.
All dimensions are mils.
The spacings (3 mm =118 mils, 0.6 mm = 24 mils) between the resonators were obtained
from [16]. The filter was simulated in Sonnet and the S21 response is shown (Figure
5.2.3).
Reproduced with permission of the copyright owner. Further reproduction prohibited without permission.
1
2
3
4
5
6
7
Frequency (GHz)
Figure 5.2.3: S2 1 response of the combline dual-band bandpass filter.
The filter response has two bands centered at 2.7 GHz and 6.4 GHz and a transmission
zero exists at 3.05 GHz. This is slightly different as compared to the response given in
[16], where the two bands were centered at 2.45 GHz and 5.75 GHz and transmission
zero exists at 3.4 GHz. This difference may be due to a different version of the software
used for the simulation or due to a different cell size used in the simulation software.
5.2.1 Frequency response with dielectric overlay on the entire surface of the filter
The combline filter is simulated in Sonnet with a dielectric overlay (sr = 80) on
top of the entire circuit and the S21 response of the filter is shown in Figure 5.2.4.
Reproduced with permission of the copyright owner. Further reproduction prohibited without permission.
79
o
_ Q _ No overlay
- Q - O.lmil thick overlay
—O— 0.2mH thick overlay
- 1 2 0
*1----------------------- ' i '
1
2
1-------------------------- 1
3
4
t
5
" i " ...............................
6
7
F r e q u e n c y (G H z)
Figure 5.2.4: S21 response of the combline filter with and without dielectric overlay on
the entire filter circuit. Two overlay thicknesses are used.
The results show that the S21 responses with dielectric overlay shift towards lower
frequencies as expected from results shown in chapter 3.
5.2.2 Frequency response of the filter with dielectric bricks on various parts of the
circuit
The combline filter is simulated in Sonnet by adding a dielectric overlay of
varying thicknesses on various parts of the circuit and the results are plotted using
Sonnet. The overlay in this case was added using the dielectric brick feature of Sonnet.
Figures 5.2.5 and 5.2.6 show the filter circuit with a dielectric brick on each of the two
resonators.
Reproduced with permission of the copyright owner. Further reproduction prohibited without permission.
80
Figure 5.2.5: Geometry of the combline bandpass filter with a dielectric brick on the first
resonator.
Figure 5.2.6: Geometry o f the combline bandpass filter with a dielectric brick on the
second resonator.
The S21 responses of the above two circuits are shown in Figures 5.2.7, 5.2.8 and 5.2.9
with various overlay thicknesses simulating layers of bacteria.
Reproduced with permission of the copyright owner. Further reproduction prohibited without permission.
81
-O - No overlay
-a —O.lmil 1st resonator
- O - O.lmil 2nd resonator
-120
1
2
3
4
5
6
7
Frequency (GHz)
Figure 5.2.7: S21 responses of the combline filter with 0.1 mil thick dielectric brick on
each of the resonators as shown in Figures 5.2.5 and 5.2.6.
-C—No overlay
0.2mil 1st resonator
< > 0.2mil 2nd resonator
1................1------------1------------ 1------------ 1-----------1
2
3
4
5
6
7
Frequency (GHz)
Figure 5.2.8: S21 responses of the combline filter with 0.2 mil thick dielectric brick on
each o f the resonators as shown in Figures 5.2.5 and 5.2.6.
Reproduced with permission of the copyright owner. Further reproduction prohibited without permission.
82
-20
-40
-60
-80
-C—No overlay
-a - 4mH 1st resonator
<>- 4mil 2nd resonator
-100
(dB)
-120
1
2
3
4
5
6
7
Frequency (GHz)
Figure 5.2.9: S21 responses o f the combline filter with 4 mil thick dielectric brick on each
of the resonators as shown in Figures 5.2.5 and 5.2.6.
Figures 5.2.7, 5.2.8, and 5.2.9 show that the circuit S21 response exhibits higher
attenuation as the dielectric brick overlay thickness increases on the individual
resonators. Identical S21 responses are obtained when either of the resonators is covered
with the overlay. Figure 5.2.10 shows the geometry of the combline filter with dielectric
bricks on both the resonators.
Figure 5.2.10: Geometry o f the combline bandpass filter with a dielectric brick on both
the resonators.
The S21 response o f the above circuit is shown in Figure 5.2.11.
Reproduced with permission of the copyright owner. Further reproduction prohibited without permission.
83
o
—O —No overlay
—Q —0.1 mil thick overlay
—O—0.2mil thick overlay
- A —4mil thick overlay
- 1 2 0
-------------------------------.------------------------------ 1------------------------------r .......
1
2
3
■'
4
-
- ■ ■ ■ ■ ■ ■ i - F . '—
5
......=
6
H
7
Frequency (GHz)
Figure 5.2.11: S21 response of the combline filter with dielectric overlay bricks on both
the resonators.
Figure 5.2.12 summarizes the combline filter when one of the resonator (left) is
covered with dielectric overlay of varying thicknesses.
Reproduced with permission of the copyright owner. Further reproduction prohibited without permission.
84
o
—O —No overlay
- Q —0.1m l (hide overlay
-<>—0.2m l thick overlay
- A —4m l thick overlay
-1 2 0
1-------------------------- 1
1
2
3
................. -
- i--------------------------- 1--------------------------- 1--------------------------
4
5
6
7
Frequency (GHz)
Figure 5.2.12: S21 response o f the combline filter with different overlay thicknesses
applied on the left resonator.
When the dielectric overlay is applied on either one of the resonators of the filter,
the attenuation in S21 response increased as the overlay thickness is increased on the
resonator (Figure 5.2.12). However, when overlay is applied on both resonators this
effect was not present, though there was a shift in the frequency response.
The above results point towards the feasibility of using the increase in attenuation
exhibited by the combline dual-band bandpass filter in addition to the center frequency
shift as the measured output signal in presence of bacteria adhering to one resonator only.
In addition there is a potential of using this type of filter to detect two different types of
bacteria by making each resonator specific to a single type of bacteria via the use of
antibodies (see chapter 6).
Reproduced with permission of the copyright owner. Further reproduction prohibited without permission.
85
CHAPTER 6
FUTURE WORK AND CONCLUSIONS
Based on the results presented in this thesis the following directions are suggested
for future work towards making the microstrip line bandpass filter based biosensor for
detection of bacteria in water a realizable device.
6.1 Bandpass filters with high bandwidth for biosensor
The results from chapter 3 show that increase in bandwidth of the filter increases
the sensitivity of the biosensor to the presence of a dielectric overlay. Hence, while
designing a microstrip bandpass filter based biosensor, the bandwidth is an important
specification. It is suggested that various types of microstrip bandpass filters should
be explored to achieve the highest possible bandwidth and hence the effect on
sensitivity of the biosensor.
6.2 Choice of center frequency
Results presented in chapter 3 indicated that a filter designed at a higher center
frequency exhibited a higher sensitivity to the presence of a dielectric Overlay.
However, since the dielectric constant of bacteria may decrease at higher frequencies,
this aspect needs to be further exposed by managing dielectric constant measurements
of bacterial suspensions at these higher frequencies. The author of this thesis could
not find any information on dielectric constants of bacterial suspensions in the
literature.
6.3 Choice of dielectric substrate
Results from chapter 3 show that a low dielectric constant substrate produces
more center frequency shift than a high ^ electric constant substrate for the same
Reproduced with permission of the copyright owner. Further reproduction prohibited without permission.
86
circuit design and dielectric overlay thickness. While designing the microstrip line
based biosensor, choosing a dielectric substrate with low dielectric constant is
recommended. However, if the size of the biosensor is a constraint for a particular
application, higher dielectric constant substrates may be used at the expense of
sensitivity.
6.4 Dual-band bandpass filters as a biosensor
In this thesis a stepped impedance based dual-band combline bandpass filter was
introduced as a potential choice for the biosensor. This dual-band filter showed
attenuation in the S21 response in addition to the center frequency shift when
dielectric bricks are added on either of its resonators. It is suggested that future work
involving other similar types of filters should be conducted. More work should be
done in designing the dual-band bandpass filters using resonators which resonate at
different frequencies.
6.5 Immobilization of antibodies
Future work should involve the immobilization of antibodies that
recognize specific bacteria. This would add selectivity to the biosensor design [17].
Reproduced with permission of the copyright owner. Further reproduction prohibited without permission.
87
REFERENCES
1. E. Kress-Rogers, ed., Handbook of biosensors and electronic noses, CRC Press,
Boca Raton, FL, 1997.
2. J.S. Wilson, Sensor technology handbook, Elsevier, Boston, MA, 2005.
3. H. Xue, D. Shan, M. Zhu, S. Cosnier, “Development of amperometric biosensor
for glucose based on a novel attractive enzyme immobilization matrix: Calcium
carbonate nanoparticles”, Biosensors & Bioelectronics, vol.22, no.8, pp. 1612-17 :
Elsevier, Journal Paper, March 2007.
4. X. Yang, A. Li; F. Yang, Y. Ma, “Electrochemical impedance detection of DNA
hybridization
based
on
dendrimer
modified
electrode”,
Biosensors
&
Bioelectronics, vol.22, no.8, pp. 1716-22 : Elsevier, Journal Paper, March 2007.
5. Y.K. Hong, K.S. Moon, M. Levy, R.R. Vanga, “Single-crystal film piezoelectric
sensors,
actuators
and
energy harvesters
with
interdigitated
electrodes,
Ferroelectrics”, vol.342, pp. 11-13 : Gordon & Breach, Journal Paper, 2006.
6. R.O. Claus, F.J. Arregui, I.R. Matias, R. Hang, K. Yuhong, T. Distler, J. Mecham,
B.Davis, “Improvements in self-assembled optical fiber-based biosensors”,
Proceedings of the SPIE - The International Society for Optical Engineering,
vol.5691, no.l, pp. 172-7 : SPIE-Int. Soc. Opt. Eng, Conference Paper; Journal
Paper, 2005.
7. U. Bilitewski, A.P.F. Turner, eds., Biosensors for environmental monitoring
Harwood Academic Publishers, Netherlands, 2000.
8. J. Yoon, A microstrip-based radio frequency biosensor for the detection of
bacteria in water, M.S. Thesis, University o f Nevada, Reno, 2003.
Reproduced with permission of the copyright owner. Further reproduction prohibited without permission.
88
9. D.M. Pozar, Microwave engineering, Addison-Wesley Publishing Company
1990.
10. “LPKF ProtoLaser 200 data sheet”, LPKF Laser & Electronics AG, Wilsonville,
OR, 2007.
Link: http://www.lpkfusa.com/datasheets/prototyping/protolaser200.pdf.
11. A. Katz, A. Alimova, X. Min, E. Rudolph, M.K. Shah, H.E. Savage, R.B. Rosen,
S.A. McCormick, R.R. Alfano, “Bacteria size determination by elastic light
scattering”, IEEE Journal of Selected Topics in Quantum Electronics, vol.9, no.2,
pp. 277-87, April 2003.
12. “A Friendly Guide to Biofilm Basics & the CBE”, Center for Biofilm
Engineering, Montana State University, Bozeman, MT, 2007.
Link: http://www.erc.montana.edu/CBEssentials-SW/bf-basics-99/default.htm
13. M. Burr, CBE Technical Advisory Conference .2004, Center for Biofilm
Engineering, Montana State University, Bozeman, May 2004.
14. W.E. Krumbein, D.M. Paterson, G.A. Zavarzin, “Fossil and recent biofilms: a
natural history o f life on Earth”, Kluwer Academic Publishers, Dordrecht, Boston,
2003.
15. M.S. Good, C.F. Wend, L.J. Bond, J.S. McLean, P.D. Panetta, S. Ahmed, S.L.
Crawford, D.S. Daly,
“An estimate of biofilm properties using and acoustic
microscope”, IEEE transactions on ultrasonics, ferroelectrics, and frequency
control, vol. 53, no. 9, pp. 1637-1648, September 2006.
Reproduced with permission of the copyright owner. Further reproduction prohibited without permission.
16. S.F. Chang, Y.H. Jeng, J.L. Chen, “Dual-band step-impedance bandpass filter for
multimode wireless LANs”, Electronics Letters, vol. 40, no. 1, pp. 38 - 39,
January 2004.
17. T.L. Tsung, O.T.C. Chen, “A piezoelectric SH-leaky SAW biosensor for antibody
detection”, IEEE International Workshop on Biomedical Circuits and Systems,
Sl/5- 9-12, December 2004.
Reproduced with permission of the copyright owner. Further reproduction prohibited without permission.
Документ
Категория
Без категории
Просмотров
0
Размер файла
2 289 Кб
Теги
sdewsdweddes
1/--страниц
Пожаловаться на содержимое документа