COMPARATIVE PERFORMANCE

ANALYSIS OF A STEPPED IMPEDANCE

AND OPEN-CIRCUITED STUB MICROSTRIP
LOW PASS FILTER.
A Project Defence for the Award of the
Master of Engineering (M.Eng) degree
in
Electronics and Communication Engineering
BY
AKINWANDE JUBRIL
NDA/PGS/FE/M/1808/14

1
15 MARCH, 2016
TH
 PRESENTATION OUTLINE
 INTRODUCTION

 AIM & OBJECTIVES

 METHODOLOGY

 TEST & RESULTS

 CONCLUSION
 INTRODUCTION
Microwaves are Electro-Magnetic (EM) waves with frequencies ranging between 300
MHz and 300 GHz. They have short wavelengths (i.e 1mm to 100cm).

Their wavelengths are short in comparison to the average physical length of a conductor;
hence the phase and magnitude of voltage and current changes significantly over the
length of the conductor. Therefore the conductor which works as a short-circuit node in
low frequency, now behaves as a “distributed element” at high frequencies.

Modern communication systems and radar require Radio Frequency (RF) and microwave
signals for the wireless transmission of information. These systems make use of
oscillators, mixers, RF amplifiers and microwave filters to generate and process the
baseband and carrier signals.
 ABOUT MICROWAVE FILTERS
Microwave filters represent a class of electronic filters, designed to
operate in the microwave frequency spectrum. They form an
integral part of a large variety of wireless communication systems.

Due to the nature of microwave signals; microwave filters are

designed with waveguides and/or transmission line structures such
as coaxial line, microstrip line, stripline etc.
Microwave filter are employed in the
design of sub-units of communication
systems such as:
 Diplexers
 Multiplexers
 Impedance matching networks
 Time-Delay networks
 MICROWAVE FILTER SYNTHESIS
Microwave filter synthesis begin with the design of an equivalent
lumped circuit filter that satisfies the filter requirements. However
at practical microwave frequencies, the lumped circuit design has
limitations such as:

The values of the inductances and capacitances that will make up

the lumped circuit will be so small that they become very difficult
to practically realize.

At microwave frequencies the distances between filter

components are not negligible.
Thus in order to synthesize microwave filters, the lumped
elements of the filter must be converted into distributed element
realizations.
 MICROSTRIP FILTER REALIZATION
Microwave filters at the lower microwave spectrum are mostly designed using
microstrips because of its inherent light weight and low cost.
Microstrip Low pass filters are mostly realized through 2 methods:
Stepped Impedance Synthesis
Open-Circuited Stub Synthesis
Stepped Impedance Synthesis
The series and shunt reactances of the lumped circuit is synthesized as a cascade of
alternating high and low impedance lines. The length of these transmission line sections
are such that they are much shorter than the associated wavelength; so as to act as
semilumped elements. The high impedance transmission lines synthesize a series
inductor while the low impedance transmission lines synthesizes a shunt capacitors.

Open-Circuited Stub Synthesis

This synthesis method is a modification of the stepped impedance method; the low
impedance lines which synthesizes a shunt capacitor is replaced with an open-circuited
stub; while the series reactance is still synthesized as a high impedance line. Stubs are
small sections of transmission line, usually with a length of one-eight of a wavelength.
They are synthesized based on Richard’s transformation.
 STEPPED-IMPEDANCE METHOD
At low microwave frequency it can be
assumed that the electrical length of the
transmission line is short (i.e βι << 1), and
as such the transmission line can be
modelled as a T-circuit with impedances as
shown

If Zo is relatively large, then the shunt element

will become very large and will behave like an
open circuit i.e.; and hence the transmission line
behaves like a series inductance
as shown

if the characteristic impedance of the

transmission line is relatively small, i.e. ;
Hence the transmission line behaves like a
shunt capacitance.
 Open-Circuited Stub
•It is based on Richard’s transformation.
The input impedance (Zin) at any point on a lossless transmission line:
Zin = ]
If line is short circuited ZL = 0;
Zsin = Zo [ ] = j Zo tanβl (Behaves like an inductance)
If line is Open-Circuited i.e. ZL = ∞
Zoin = ]
Zoin = Zo [ ] = = -j Zo cot βl (Behaves like capacitance)
 AIM & OBJECTIVES
Aim: The purpose of this study is to present a comparative performance
analysis of a stepped impedance & Open Circuited Stub microstrip low pass
filter.

The objectives of the project will be:

• To derive the optimal design parameters for a microstrip low pass filter within
the specified bands using the stepped impedance and open circuited stub
synthesis methods.
• Determine the performance of both synthesis methods by comparing their
frequency response with the use of the Advanced Design System (ADS)
microwave simulation tool.
• Fabricate the microstrip layout on a FR4 PCB layout, capable of carrying out
the low pass filtering function with the selected cut-off frequency.
 SCOPE OF STUDY
• Cut-off Frequency: 2GHz

• Stop-Band Edge Frequency: 4GHz

• Filter Approximation functions: Chebyshev & Butterworth LPF

• Microstrip Substrate: FR4

• Microstrip Filter Synthesis: Stepped Impedance & Open-Circuited Stub

• Method of Analysis: Advanced Design System EM simulation & Analysis

 METHODOLOGY
The design process involves 4 main steps:

1. Lumped Element Filter Design: A Low Pass

filter lumped-element prototype will be
synthesized using the Butterworth and
Chebyshev filter approximation function.

2. Conversion of Lumped to Distributed

Circuit: The obtained LC ladder circuit in step
1; is converted into an equivalent distributed
element form. Stepped impedance & open
circuited stub method will be used.

3. Optimization of Microstrip Filter Layout:

The Advanced Design System (ADS) software
will be used to simulate the performance of the
obtained microstrip layouts. Using the features
in the ADS, the designed layout is tuned in order
to obtain the optimal dimensions.

4. Fabrication of Microstrip Layout: The

microstrip layout will be fabricated by using the
 TEST & RESULTS
The following parameters will serve as basis for performance
comparison:
S(2,1) parameter plot: This describes the level of insertion loss across
the passband and stopband.
S(1,1) parameter plot: This will describe the level of return loss at the
input of the microwave filter. i.e Measures the level of impedance
matching.
Sharpness of transition between Passband and Stopband
Occurrence of spurious peaks in the stopband
Butterworth Microstrip Layout
Designs

Butterworth Open-Circuited Stub 2GHz Low Pass Filter

Butterworth Stepped-Impedance 2GHz Low Pass Filter

Chebyshev Microstrip Layout
Designs

Chebyshev Open-Circuited Stub 2GHz Low Pass Filter

Chebyshev Stepped-Impedance 2GHz Low Pass Filter

 Insertion Loss -Performance
Comparison

Chebyshev Open-circuited Stub 2GHz Filter

Chebyshev Stepped-Impedance 2GHz Filter

 Insertion Loss -Performance
Comparison

Butterworth Open-circuited Stub 2GHz Filter

Butterworth Stepped-Impedance 2GHz Filter

Insertion Loss Comparison of
Lumped vs Distributed Circuit

I. Performance in the passband is very similar to the lumped

model.
II. Open-circuited stub exhibits better stop-band
characteristics than stepped impedance & Lumped circuit
at some frequencies.
 Fabricated 2GHz Open-Circuited
Stub Low Pass Filter

Di-electric : FR4 Substrate

Approximation Function: Butterworth
Cut-off Frequency: 2GHz
Synthesis Method: Open-Circuited Stub
 Laboratory Test Setup
In Place of a Vector Network Analyser, a Microstrip Trainer’s kit
was used to test the stopband performance. The setup consists of a
VCO, a 50ohm terminated circulator & a crystal detector, for the
detection of the output microwave power in terms of voltage.

Setup Procedure
I. Output Power from the VCO between 2.4GHz – 3.63GHz is
measured i.e the input microwave power P1 entering the Low
Pass filter (LPF).

II. The Output Power P2 from the (LPF) for the frequency range
Microstrip Trainer Setup for LPF Measurement
2.4GHz-3.63GHz is measured. This achieved by inserting the
LPF in between the circulator and crystal detector as shown.

III. The Insertion Loss Ratio is calculated as: ILR = dbm

Laboratory Test Result for 2GHz Butterworth LPF
Direct With LPF
VCO 10logP2/P
S/N f(GHz) (Volts) V1(Volts) P1(mW) V2(Volts) P2(mW) P2/P1 1 0
1 2.45 3.28 0.8 10.5 0.487 4.2 0.4 -3.9794
2 2.50 3.81 0.83 11.3 0.436 3.49 0.30885 -5.10253
-2
3 2.55 4.31 0.824 11.1 0.394 2.6 0.234234 -6.3035
4 2.60 4.78 0.788 10.2 0.373 2.5 0.245098 -6.1066
5 2.65 5.5 0.76 9.6 0.347 2.3 0.239583 -6.20543 -4
7 2.75 6.9 0.72 8.6 0.32 1.9 0.22093 -6.55745
-5.1

Insertion Loss (dbm)

8 2.80 7.72 0.73 8.9 0.297 1.72 0.193258 -7.13862
-6
9 2.85 8.52 0.73 8.8 0.277 1.67 0.189773 -7.21766
10 2.90 9.48 0.676 7.6 0.257 1.39 0.182895 -7.37799
11 2.95 10.51 0.632 6.7 0.266 1.48 0.220896 -6.55813 -7.14
-8
12 3.00 11.62 0.586 5.8 0.258 1.39 0.239655 -6.20413
13 3.05 12.8 0.616 6.4 0.205 0.98 0.153125 -8.14954
-10 -10.21 -10.21
14 3.10 13.95 0.574 5.6 0.16 0.6 0.107143 -9.70037
15 3.15 15.19 0.549 5.1 0.135 0.5 0.098039 -10.086
16 3.20 16.4 0.492 4.2 0.121 0.4 0.095238 -10.2119 -12
17 3.25 17.6 0.488 4.2 0.125 0.45 0.107143 -9.70037
18 3.30 18.82 0.522 4.7 0.14 0.5 0.106383 -9.73128
19 3.35 20.09 0.636 6.9 0.169 0.7 0.101449 -9.93751 -14
20 3.40 21.41 0.754 9.3 0.198 0.89 0.095699 -10.1909 2.3 2.4 2.5 2.6 2.7 2.8 2.9 3 3.1 3.2 3.3 3.4 3.5 3.6 3.7
21 3.45 22.9 0.706 8.4 0.19 0.8 0.095238 -10.2119 Frequency (GHz)
22 3.50 24.62 0.708 8.4 0.193 0.8 0.095238 -10.2119
23 3.55 26.38 0.784 10.05 0.2 0.89 0.088557 -10.5278
24 3.60 28.41 0.852 11.9 0.174 0.7 0.058824 -12.3045
25 3.63 30 0.592 5.9 0.175 0.74 0.125424 -12.4471
Comparison of EM Simulation & Laboratory Test
Actual Test Result with Microstrip
EM SIMULATION 2GHz Butterworth LPF Trainer
0

-2

-4

-5.1

Insertion Loss (dbm)

-6

-7.14
-8

-10 -10.21 -10.21

-12

-14
2.3 2.4 2.5 2.6 2.7 2.8 2.9 3 3.1 3.2 3.3 3.4 3.5 3.6 3.7
Frequency (GHz)
A comparison of both tests shows marginal
average difference of less than 5% between the EM
simulation & the Actual test. As a result we can
conclude that the EM simulation results mimic the
actual performance fairly and can be used as the
basis for analysis.
 LIMITATIONS OF RESEARCH
• An actual field test could not be conducted due to the non-availability of a Vector Network
Analyser (VNA), which is required to measure scattering parameters of the fabricated microstrip
filter. The performance analysis was predominantly based on the result of EM simulation using
the ADS EM simulation software.

• The FR4 dielectric was used for this study, and as a result the order of filter that could be feasibly
fabricated was limited and the FR4 is a lossy substrate that exhibits poor performance with
increase in frequency.

• The width of the transmission lines were between 0.5mm and 1mm; this requires the use of
specialised tools in order to fabricate the layout on a plane FR4 board. Unavailability of such
specialised tool, limited the fabrication to the use of manual methods, and only the Butterworth
Open-Circuited Stub filter could be fabricated due to limitations of this fabrication technique.
 CONCLUSION
I. The Stepped impedance filter exhibited marginally better return loss in the
passband, compared to the open circuited stub filter. i.e the stepped
impedance filter presents better impedance-matching in the passband.

II. The Open-circuited stub approximates a shunt capacitor more accurately than
a low impedance line, as used in the stepped-impedance synthesis.

III. Overall, the open-circuited stub filter approximates the lumped circuit model
more accurately, in the passband its performance was very close to the
lumped circuit model, its stopband characteristics surpassed the design
specification; while the stepped impedance filter failed to meet the stopband
specification for the 2GHz low pass filter.
THANK YOU