Module Unit

Topics Hrs.
No. No.

3.0
Frequency Generation &Mixer 08
3.1 One-port and two-port microwave oscillator design.
3.2 Analysis of phase noise in oscillators.
3.3 Mixers: Characteristics, Various types of Mixers: Single ended diode mixers, FET mixers,
Balanced mixers, Image reject mixers and other
types of mixers. (Refer Chapter 3)
4.0 Frequency Synthesizers 06

4.1 Direct Frequency Synthesis, Frequency Synthesis by Phase Lock, Effects of Reference
Frequency on Loop Performance,
4.2 Variable: Modulus Dividers, Down Conversion, Methods for Reducing Switching Time,|
Direct Digital Synthesis, Synthesizer Design.
4.3 Phase Noise A Model for Oscillator Phase Noise, Phase Noise
in Phase-Locked Loops, |
Effect of Frequency Division and Multiplication on Phase Noise. (Refer Chapter 4)
5.0 Electromagnetic Interference in RF circuits 08

5.1 Introduction. Natural and Nuclear Sources of EMI, EMI From Apparatus and Circuits.
Quantification of Communication System EMI
5.2 Elements of Interference, Including Antennas, Transmitters, Receivers And Propagation.
Electronic Equipmentand System EMI Concepts. Examples of EMI Coupling Modes
5.3 Equipment Emissions and Susceptibilities Types of coupling: Common-Mode
Coupling: Common Mode Coupling Mechanisms Including Field to Cable, Ground
Impedance, Ground Loop and Coupling Reduction Techniques.
Differential-Mode Coupling: Differential-Mode Coupling Mechanisms Including Field to
Cable, Cable to Cable and Coupling Reduction Techniques.

5.4 Other Coupling mechanisms: Power Supplies And Victim Amplifiers (Refer Chapter 5)

6.0 ElectromagneticCompatibility 04
6.1 The Importance Of Grounding For Achieving EMC. Grounding, Including The Reasons
(L.E, Safety, Lightning Control, EMC, Grounding Schemes (Single Point, Multi-Point And
Hybrid), Shicld Grounding And Bonding. Shielding Effectiveness, Shielding Considerations
(Reflective And Absorptive), Shielding Compromises (LE., Apertures, Gaskets, Waveguide
Beyond Cut-Off)
6.2 EMI Diagnostics And Fixes: Techniques Used In EMI Diagnostics and Fixes

6.3 EMC Specifications, Standards And Measurements. A Discussion ofThe Genesis of EMC|
Documentation Including A Historical Summary, The Rationale, And A Review of MIL-Stds,
FCC and CISPR Requirements. (Refer Chapter 6)
 Index

Module
Chapter 1 RF Filter Design ... ...1-1 to 1-74|

Module 2
Chapter 2 Amplifier Design... ...2-1 to 2-82|

Module 3
Chapter 3: Frequency Generation and Mixer.. ..3-1 to 3-46

Module 4
Chapter 4 Frequency Synthesizers.. *****°*** .4-1 to 4-24

Module 5
Chapter 5: Electromagnetic Interference in RF Circuits. ..5-1 to 5-40

Module 6
Chapter 6: Electromagnetic Compatibility. .....6-1 to 6-40
 CHAPTE

RF Filter Design

2Modale
Syllabus:
Image parameter method, Insertion loss method-
Filter transfomation and filter
Maximally flat low pass prototype, Equal ripple low pass prototype,
implementation.

1.1
Introduction..
*****************************************************************.*************.**********************************"**
***************a**************************************************"*******************************e****** -5
1.2 Overview of Basic Terms...
********************"*********************************************** ********************************* -5
1.3 Filter Design..
*********e***********°************************************************
1-7
1.3.1 Filter Design by Image Parameter Method.. *********.*****°*****************°°***°°° ..... 1-7

GO. 1.3.1 Explain synthesis of filter using image parameter methods.

OR Obtain the expression for image impedances and transfer function of two port network.
Exam. Hint: Equations 1.3.4, 1.3.15 and 1.3.16...********e******************************************************************
1-7

GQ. 1.3.2 Obtain the expression for the image impedance of n network shown in Fig. 1.3.4.

Exam Hint: Z ****************************************************************°*.*************************.********* .1-10

GQ. 1.3.3 Obtain the expression of propagation constant of r network given in Fig. 1.3.4.

ExamHint: e"= 1 +2Z,+ *************n**a**********

************ ses. ]-10

GQ. 1.3.4 Derive the expression for image impedance of T network in terms of series and shunt

am impedances shown in Fig. 1.3.6.

Exam Hint:Z= Z **************************************o**osessneo U

 RF Filter Design
RF Design (MU-8h Sem.-E&TTC) (1-2)
GQ. 1.3.5 Obtain the expression for propagation constant of t network.

**
********** ****..

.1-11
ExamHint:e"=1+2~~ '
***°**********

*************

network shown in
GQ. 1.3.65 Ompute the image impedances and propagation factor for the
Fig. 1.3.9 using ABCD parameters.

1-12
EXam Hint: coshy=AD =1-2oLC ***************°************************°******* *************°**********°

1-6

1.3.2
****************°******°***°°*****°°****** 14
Constant Kor Prototype Filter Section. ***********.***°**************

GQ. 1.3.7 What is constant K filters ? OR What is prototype section ? .**************

.
*********************°******
.1-14

****************** .. 1-14
1.4 Constant K Low Pass Fiter Design by Using Image Parameter Method. ******

********°****°°**°*°*°********** 1-14
Ga. 1.4.1 What is constant-K-ow pass filter ?...*****************a********°°******"**

1.4.1 .1-14
Analysis of Constant K Low Pass Filter.. ***************************************************°°****°°************ ***************

GQ. 1.4.22 Design a constant K low pass T section fiter to be teminated in R, 2 having cut of a, rad/sec. .1-14

1.4.2 Reactance Curve and Expression for Cut off frequency f *************************************°°****************

1.4.3 Expression for Propagation Factor.. *************.***********************************°**°* .. 1-16

1.4.4 Analysis of attenuation (a) and phase shift (B) characteristic of constant KLPF. .. 1-16
*****e*****.

1.4.5 Design Steps and Equations of Constant K-LPF.***************************************************************************** 1-18

1.5 Constant K HPF.. ************"" *******a******************************************************

******** .. 1-19

GQ. 1.5.1 Design constant K-HPF... ************************************************ *************n*********** 1-19

1.5.1 Attenuation (a) and Phase Constant (B) Characteristic of Constant K HPF...
********°*************************** 1-20

1.5.2 Summary of Design Equations of Constant K HPF. nosenseeene**"******" 1-21

**********************o****

GQ. 1.5.2 The nominal characteristic impedance of T section of LPF is Ro. Find Z at 0.9 fc..
**************************** 1-22

1.6 m-Derived Filter...........a....s**e*a*********e*********************************************a*******o****e ssee.

****************e*********. 1-22

GQ. 1.6.1 How m -

derived filters are

superior than constant K -

filter ? OR How m- derived filters

can be obtained from constant ?...
1-22
GQ. 1.6.2 How T section of m -

derived filter is obtained from constant- k- T section ?.

***********a************a********* 1-23

1.6.1 m- Derived Low Pass Filter. .

* . T-24

Analyze m - derived T and n sections of LPF.

GQ. 1.6.3 ************ ****°*
******************. *
*******°**° .
.1-24

Analysis of Attenuation (a) Versus Frequency characteristic of LPF

1.6.2 m - Derived Section ... 1-25
*******.******

Tech-Neo Publications . here Authors inspire innovation

A SACHIN SHAH Venture
 RE Design (MU-8h Sem.-E&TC) (1-3) RF Filter Design
1-26
1.6.3 Cascading of m -
Derived T-Half Sections... ****************************°*.************************************************

M-derived High Pass Filter..**********************************as**bs****************************************************** 1-28

1.6.4

GQ. 1.6.4 Draw and explain m-derived T and n section of HPF. Explain graphical representation of cut off
. . 1-28
frequency.. **********°***********°***************n*******************.******so*************************************************

**********************************u*s******** ********************************************°*********
.1-30
1.7 Impedance Matching Using Half Sections...

GQ. 1.7.1 Why half section is used for terminating section in filter. ?.****** ***************** *********°"*** **** 1-30

1-30
GQ. 1.7.2 Explain the need of using half sections.. ***********************************************************
*********** ********

... 1-30
m-Derived Half Sections for Low Pass Filter.***********************************************************°****°***********
1.7.1

M-Derived Half Sections for High P a s s F i t e r .*****************************************************°"****** .. 1-31

1.7.2 **********

Ga. 1.7.3 Write a note no teminated m-derived half section for HPF.. 1-31

1-32
1.8 Composite Filter. ******************************o******************************************************°**

Ga. 1.8.11 e******** 1-32

Draw and explain block diagram of composite filter. Write advantage of composite filters...

1-38
1.9 Filter Design by Insertion Loss Method..
Ga. 1.9.1 What is perfect filter ?. ************°*********°**** .1-38
*******************es**********. ***°*******°****

1.9.1 Characterization by Power Loss Ratio (PLR) .1-38

***********************************

Ga. 1.9.2 Explain the concept of insertion Ioss method.. ***** *************
************ 1-38

1.9.2 Types of Filter...a***uso**** e****. ***************a********************

****° 1-38

GQ. 1.9.3 Explain different types of low pass filter. ****************** ***e************.**********°****** 1-38

1.9.3 Process of Fiter Design by Insertion Loss Method. *************************************************** 1-40

GQ. 1.9.4 Explain design process of fiter by insertion loss method.. **** 1-40

1.9.4 Maximally Flat Low Pass Filter Prototype. ****** "*********** 1-41

Ga. 1.9.5 Derive expression for the values of L and C for second order low pass filter and maximally

flat low passfilter..********* *****e*************e*****« **** ************** 1-41

GQ. 1.9.6 Explain ladder circuits for maximally low pass fiter and element definition...
*********""°*** ********* 1-42

1.9.5 Equal Ripple or Chebyshev Low Pass Filter Design. ******* *******.******************
**. 1-44

GQ. 1.9.7 Explain chebysheve low pass filter synthesis process.. **************************** 1-44

Tech-Neo Publications here Authors inspire innovation A SACHIN SHAH Venture

 RF Filter Design
RF Design (MU-g Sem.-E&TC) (1-4)
*************************************************************e***************************** 1-48
1.10 Scaling of Low Pass Prototype Filters.
****************°******************* 148
GQ. 1.10.1 Explain two types of scaling. ****************************************************************

*****************************************.************************************************************* 1-49
1.11 Filter Transfomation..

********************************************************************°*************************
** l49
GQ. 1.11.11 Write a note on filter transfomation.

*°°****°° .1-49
1.11.1 Low Pass to High Pass Fiter..
***********'*********************************************************************

GQ. 1.11.2 ********************************************************************** .1-49

Explain low pass to high pass filter trans ation. *

Low Pass Filter to Band Pass Filte.

********************.**********************s************************************************* .1-52
1.11.2

GQ. 1.11.3 Explain and obtain the expression of capacitorand inductor..****************** ****°*********°°*********** 1-52

********** 1-56
1.11.3 Low Pass Filter to Band Stop Filter . *****************************************************************************

Ga. 1.11.4 Obtain the expression for inductor and capacitor of band stop filter using low pass

filter prototype.. **********e**********************************.

**************************************************
1-56

1.12 Fiter Implementation.. 1-59

ssseeeansnsns* **a .snoa...s*ssssssnnanssnssnsosanessesnnessen****************************

GO. 1.12.1 What is redundant and non-redundant filter synthesis ? *********o**** ********o9************* 1-59

1.12.1 Richard's Transformation. *********s*ssesesesnessranassese***esonsrsnsssosopnsse************ ****************osee***** 1-59

Ga. 1.12.2 How Richard's Transformation is usetul in RF fiter designing ?.

GQ. 1.12.3 Why does fiter response using Richard's Transfomation match with filter design using
lumped element ?...e***********************sass***************************s*********************.*********************.**************.*****1-60

1.12.2 Kuroda's ldentities..*********************************°°**********°e*************a*********°******.****°°************°****e ****a** 1-61

Ga. 1.12.4 Explain how Kuroda's ldentities are useful in RF fiter desig...
******a**********°°************************* **********. 1-61

Ga. 1.12.5 Prove the Kuroda's 1 identity. ****************************************************a** 1-62

1.12.3 Impedance and Admittance Inverters. ****

*************************°.*****. 1-66

1.13 Mixed Solved Problems.. ****************************************************°*********** **

***************************** 1-67

Chapter Ends. ************

************************************************e**********
************************************************************ 1-74

Tech-Neo Publication. here Authors inspire innovation

A SACHIN SHAH Venture
AF Design (MU-8" Sem.E&TC) (1-5) RFFiter Design
1.1 INTRODUCTION 5. Stop band: The range offrequencies where signals are
stopped or suppressed called as 'stop band' or

The "Short wavelength" feature of microwave region attenuation band.

has made the microwave signals to travel over long 6. Cut off frequency : The frequency which discriminate
distances under most sevenre weather conditions which the pass band and stop band is known as cut off
includes both civilian and military and wireless
frequency.
applications. Usually it is denoted as f .
As microwave frequency spectrum (1 GHz to
1007, There may be two cut off frequencies in the complete
GHz) is a finite resource, filters play vital role in order
operation of filters.
to treat this spectrum with cared and respect. They are as : f> lower cut off frequency;
Radio filters are very crucial component of different
electronic systems including radar, satellite and cellular
faupper cutofffrequency
communication.
8. Classification of filters
These are two types of filters
In these systems, RF filters differentiates between
wanted and unwanted frequencies. Classflcatlon
of filters
The design of filters include network synthesis with
which it becomes easy to transform designed filters into
microwave networks including waveguides, TEM (i) Depending on
) Depending on relation
transmission lines and dielectric resonators. So this unit between series arm frequency response
begins with the basic revision of some terms and impedances and of filters
shunt arm impedance
properties of filters.
(a) constantfiter F (a) Low Pass Filter (LPH)
H 1.2 oVERVIEW OF BASIC TERMS or prototype filter F(b) High Pass Fiter (HPF)
(b) m derivedfiters F(c) Band Pass Filter(BPF)
1. Filters: filters are electrical circuit which passes signal ( ) Band Stop Filter or
within a specified frequency range. Band Elimination Fiter
(a) Table 1.2.1 shows the classification filters with their
2 Network analysis: It is the method offinding out the
Response when network and Excitations are known. In pass band and stop band specification

network analysis each problem has got the unique Table 1.2.1

solution.
Consider a two port network Nameof fiter Pass band (Hz) Stop band(Hz)
|Low Pass O to f to o

Excitation High Pass f.tooo 1 to f

o Network Response
Energy source
Band Pass fato f Oto fa f to o

MATaFig. 1.2.1
Band Stop Oto fofa to f to fa
3. Network synthesis : It is the method of finding out the 9. Attenuation and its units:
network when Response and Excitations are known. In Attenuation is loss of power in electric network or
network synthesis there may be multiple solutions toa transmission lines. It is expressed in Nepers and
given problem. (Refer Fig. 1.2.1) decibels.
4. Pass band : The range of frequencies where signal
passes is called pass band.

Tech-Neo Publications. bere Autbors inspire ianovation A SACHIN SHAH Venture

RF Design (MU-8h Sem.-E&TC) (1-6)
RF Filter Design
10. Image impedances (Z) 1. Insertion loss
Two-port network is shown in Fig. 1.2.1(a). When a network is inserted between source and load as
Z Z Z Z3 shown in Figs. 1.2.2(a) and 1.2.2(b), the power
1o 2
TW2 delivered to the load decreases as load current

decreases.
The loss in power delivered from source to load is due
1'o o2' to insertion of network is known as insertion loss and it
is expressed in decibel or Neper.
(141)Fig. 1.2.1(a) : Two port network (1A2)Fig. 1.2.1(b)
The image impedances are denoted as Z at 1-11
terminal and Z2 at 2-2 terminal as shown in
Fig. 1.2.1(b).
When
then the
port 2 i.e. 2-2' terminal is terminated with
Z, v
impedance measured at port 1 i.e. 1-1
terminals, will be the image impedance of port 1 as (145Fig. (a) : Without network
shown in Fig. 1.2.1(c).
E.g.
Z1
Rs Network
R

(1A6Fig. b): With network

Fig. 1.2.2: Insertion loss and two port network

(1A)Fig. 1.2.1(c)
Let 1, be the current flowing through R;, when network
Z Z +(Z, 1(Z, +Z) is inserted between source and-load. The insertion loss
Z, (Z+Z is given by,
.Z 4*Z,+Z,+Z
Similarly, if port 1 is terminated with image impedance Insertion loss(a)=InNeper
of Z then the impedance measured at port 2 will be the

image impedance of port 2. i.e. Zn as shown in Insertion loss(a)=201ogo Decibel

Fig. 1.2.1(d).
The insertion loss in terms of power ratio can be
Z1 expressed as follows:
-
The P is the power delivered to load when there is no
insertion of network (as Fig. 1.2.2(a)) P2 is the power
Zn
delivered to load when there is insertion of network (as
1 2 Fig. 1.2.2(b)).
(1A4Fig. 1.2.1(d) Insertion loss ()nNeperr(N)
Za Z, + (Z, +Z) 1Z)
2(Z+Zy) Insertion loss ()=10log decibel (dB)
= Zy+Z,+Z,+Zu
2. Return loss: The part of input power that is lost due to
impedances the when
In short, image impedances are reflection at its input port is called return loss and it is
one of port is
terminated with impedance then the other

is seen at the other port. expressed in dB. RL = 20 log

impedance

.bere Authors inspire innovalion A SACHIN SHAH Venture

Tech-Neo Publieations
RF Design (MU-8 Sem.-E&TC) (1-7) RF Filter Design
1.3 FILTER DESIGN
i) Expression for image impedances and transfer
function of two port network
Filters can be
designed using active devices such as
transistors and operational amplifiers and V, and V, are the terminal voltages at port I and port 2
passive respectively. It is noted that I, is current entering into
elements such as inductors and
capacitors. port I whereas 1, is current leaving port 2. Z and Z2
However, the filters designed with passive elements are
less expensive then the filters designed with active are known as
as follows
image impedances and they are defined

devices. Image impedance at port1 when port 2 is terminated

Also passive filters operate with Z
fairly well at microwave
frequencies. Z2 Image impedance at port 2 when port 1 is terminated
Therefore this with Zr
chapter presents the design procedure of
passive filters only. Further Zn is the input impedance at port 1 when port 2
is terminated with
Filters can be designed by two methods
(i)
Zp
Similarly Z, is output
Image impedances at port 2 when input port is terminated with
Parameter Method (ii) Insertion Loss Method.
Zi.
The transmission or ABCD parameters of two port
1.3.1 Filter Design by Image network 1 can be written as
Parameter Method V, = AV2+BI, ...(1.3.1)
and 1 = CV2 + DI ..(1.3.2)
GO. 1.3.1 Explain synthesis of filter using inmage Therefore, the input impedance Zzn using Equations
parameter methods. OR (1.3.1) and (1.3.2) when port 2 is terminated with Zi2
can be obtained as
Obtain the expression for image
impedances and transfer function of V, AV+
Zn CV,+ D1, ...(1.3.3)
two port network Dividing by h on right side of Equation (1.3.3)
Exam. Hint: Equations 1.3.4, 1.3.15 and 1.3.16
This method (Image Parameter method) of filter design
can pass or stop a certain frequency band for a cascade of
two port network.

Disadvantage : This is simple method but unable to shape 01

it's frequency response characteristics.
) Schematic Diagram: Fig. 1.3.1 shows two port
network with its ABCD parameters.
Port 1 Port 2 (1AB)Fig. 1.3.2
Apply KVL to output side
V,-Zh = 0 ...from Fig. 1.3.2
V Zal
(1.3.4)
Solving Equations (1.3.1) and (1.3.2) for Vz 1, we get,
AV = V1- Blh
Lin o
(1A7Fig. 1.3.1: Two port network terminated with image
impedances Zy and Zu A -BL (13.5)
Dh = 1-CV,

Tech-Neo Publieations . u . Bhere Authors inspire innovation A SACHIN SHAH Venture

 RF Design (MU-8 Sem-E&TC) (1-8) RFFilter Design
Now dividing by 1, and we get
C
D (1.3.6)
D +B
Putting Equation (1.3.6) into Equation (1.3.5) ...(1.3.11)
V B-CV
AD -C+A
becomes
V,D V, = - Z , Equation (1.3.11)
AD B-CV,)
From Fig. 1.3.3,
V2 AD 2 D+B
..(1.3.12)
V,
Va V,D-B-CV) ZZC+A
AD As it is seen that from Fig. 1.3.3,
V B1+BCV, Zn Z and Z, = Z2

AD So Equation (1.3.4) and Equation (1.3.12) represents

two equations for image impedances as

v,AD AD Z
AZ+B
CZ+D and ZZC+A
Z D+B

vAD-BQ ,D-BI
AD AD 1.3.7) From Fig. 1.3.3,

By knowing the condition of reciprocity is,

AD-BC 1,
TWO
Equation (1.3.7) becomes port
V, = V,D-BI,

Similarly using Equation (1.3.5) in Equation (1.3.6)

(1.3.8)
2
IA10) Fig. 13.3
h AD BI)
AD AZ+B
Here, Z CZ+D
Al CV,+BCh
AD AD
Putting CZ+Ain above equation

CV
1-AD= AD AD ACZ+A+B
DZ+B
4 Z TDZ* D
AD C). Aj-cy,
AD cz D

..1.3.9) ADZ + AB + BCZ + AB

Z CDZ +BC+CDZ +AD
Then output impedance Z, at port 2 when port 1 is
CDZ+BC+CDZ+ADA =ADA +AB +BC +AB
terminated with Z can be obtained as
2CDZ2AB
.(1.3.10)
CD

[Negative sign indicates that current I, is leaving the port 2] Taking square root on both sides

Using Equations (1.3.8) and (1.3.9) in Equation (1.3.10), ...(1.3.13)

z DB)-VD-BI,
Z A-Cv -CV+Al If network is
becomes
symmetric A =
D, so Equation (1.3.13)

.Z

Teeh-Neo Publiceations. here Authors inspire innovation A SACHIN SHAH Venture

RF Design (MU-8" Sem-E&TC) (1-9) RF Filter Design
Similarly for Z2
V D-BCD
DZ +B V D- AB
=
Za CzZ+A
AZ+B
DA-yBCD
Putting CZ+D in above equation VA
DAZ2 +BB - VAD-V ..(1.3.15)
LD+B
Now, consider Equation (1.3.9), the current ratio is
AZ+B 1 Al-Cv,
LC+DJ*A Dividing I on both sides
ADZ+BD +BCZ +BD
2
ACZ +BC+ ACZ+ AD A-C
AC+Z BC+ACZ+ ADZ = ADZ +BD +BCZ
A-Cz,
+BD 2z AC=2 BD
AC from Fig. 1.3.3, Z Ignoring minus sign

. Za BD AB
=

NAC ...(1.3.14) From Equation (1.3.13), Z4 CD

Expressing Z in tems of Z as follows
From Equation (1.3.14), divide and multiply by AD A-C
AB
CD
Z ABD
ACD = A ABCA-1
Equations (1.3.13) and (1.3.14) are the expressions for
image impedances in terms of ABCD parameters. If = A- .Divide and multiply D
network is symmetric, then A = D and Z Z as
expected.
AvD-ABCE .1.3.16)
VD
Expression for voltage and current transfer Using Equations (1.3.15) and (1.3.16)
function of two port network terminated with
its image impedances. - AD-V) ..(1.3.16(a))

Consider Equation (1:3.8) as: Expressions for propagation constant

V2 DV -BI In Equations (1.3.15) and (1.3.16) factor
As
both sides of Equation (1.3.8), the
Dividing by V2 on
present in reciprocal form and hence can be considered
voltage ratio is as transformer turns ratio. Apart from this a
propagation factor for the network is defined as
D-B = AD-VBC .(1.3.17)
where,y a+jß
1.3.3 from Equation (1.3.17)
(Tgnoring minus sign) From Fig.
AD+BC
D-2 (WAD-BC) WAD+VBC)
NAD+BC
Using Equation (1.3.13) AD-BC = 1
(AD-BC).
D-BAB e = AD+VBC ..(1.3.18)

A SACHINSHAH Venture
Teeh-Neo Publications.. here Authors inspire innovation
RE Design (MU-8h
Sem.-E&TC) (1-10) RF Filter Design

Also, coshy t ,here, cosh 27.42+2,Z+2Z,Z]

y=vAD
=

22 +Z2
(1.3.18a)) [ 2Z, Z +Z, Z2 +27,Z,2]
GO. 1.3.2 Obtain the
expression for the image 27,+ 27 +Z
pedance of network shown
Fig. 1.3.4.
T in 4ZZ+222z, +2Z, Z2 Za +Z, Z+ 27, 2,
4Z, Z,+2Z, Z, Za+ 4Z,Z
Exam Hint: Z 4Z,Z +Z Z = 4Z, Z2

4Z, Z,
Z
Z +4Z2
Multiplying numerator and denominator by factor

22 272 z .(1.3.19)

(1A11)Fig. 13.4 : T- network Taking Z,Z, common

Terminate the given network with Z at terminal 2-2.
1 l2
A
42,Z+1)
272 V2Zi2 . Z
..(1.3.20)
1 422
(1A12)Fig. 1.3.4(a)
Z
GO. 1.3.3 Obtain the expression of propagation
1 ww constant of T network given in

Za l| 2Z2 Fig.1.3.4.
in 222
11o Exam Hint: e 1+
Zn222 lI (Z,+Z2 l| 2Z2)
--+-
(1A13)Fig. 1.3.4(b) Terminate the network with
t
Zi2 =
Z, Z1=
on both
sides as shown below and apply V
The network is symmetrical, So Z, Zo =Z2 voltage.
Is R
The image impedance of symmetrical n network is w
given by, Z
Zin Z=2Z,|I|[Z, +(2Z, Z
Vo 272
Za 22"22+Zn
Z+27, 2
(1A14)Fig. 1.3.5 : T network is correctly terminated with Z
Z (27,+Z)+2,Z2 The above network is
symmetrical t network, so by the
Za 27" =

22+2 definition of propagation constant.

27,+Z2 e

Tech-Neo Publieations.. Where Authors inspire innovation

A SACHIN SHAH Venture
RF Design (MU-8 sem.-E&TC) (1-11) RF Filter Design
By applying potential divider nule to left side of Z,/2 Z,/2
Fig. 1.3.5

(27, Z,) 1
27,Z
V 2Z+7
Z, +(2Z, 17,)J 27,
Z 2Z+7 (1415Fig. 1.3.6: Symmetrical T network

V 22, 2,
z z,+2)+2,7,J
Terminate the given symmetrical T network with

ZZ=Zo
4+22,Z,+Z,2,] Z/2 Z/2 2

27,7 1
Z
1Z27 1'o
Rearranging the terms
2
Z (1A16)Fig. 1.3.7: Symmetrical T network terminated with Z

1+27 .(1.3.21)
As network is symmetrical, image impedance Z of
Using Equations (1.3.19) and (1.3.20) such network terminated with Z2 = Zo at other port is equal

Z, Z ..(1.3.22)
to Z
Z=Zo=Z2
Use Equation (1.3.22) into Equation (1.3.21)
,7(4-z)]
Z
122 Z
Z2+2.
Z

zzz2z5+z,).,zz
+2,2
1+2Z Z 2224,222
= 122*47

ZT- +Z,Z
1+2z .(1.3.23) Z

...(1.3.24)
GQ. 1.3.4Derive the expression, for image
impedanceof T network in terms of G2.1.3.5 Obtain the expression for propagation
series and shunt arm impedances constant of t network
shownin Fig. 1.3.6
Exam Hint: e 1+
Exam Hint:Z +Z Z2
To obtain propagation constant of T network, terminate
the T network correctly as shown in Fig. 1.3.8.

24Z 14Z,

Tech-Neo Publications . . . Where Authors inspire innovation A SACHINSHAH Venture

RE Design (MU-8h Sem.-E&TC) (1-12)
Filter Design
Z,/2 2
-
s-IR Loop 2 Port 1 Port 2
Loop1
c

(IA17Fig. 1.3.8: Correctly terminated T network (IA19)Fig. 1.3.9

L
Applying KVL to loop in clockwise direction
1 m

V
c

'o -o2

(1I419)Fig. 1.3.10

The transmission ABCD parameters of Fig. 1.3.10 are

defined as
V,= AV, +Bl, .(1.3.27)
= CV2+DI, .(1.3.28)
Z
...(1.3.25) Let I 0 i.e. output part is open circuited, then
Equation (1.3.27) and (1.3.28) reduces to
V = AV2
But from Equation (1.3.24), z, = and
CV
putting Z in Equation (1.3.25). Assuming loop current as shown in Fig. 1.3.10 and
applyingKVL to loop 1-A-B-1'- 1.
1.Vizz V-1 (joL)-(jolL + ) 0

V-4(jolL)-, (joL)-1,
1*2Z*\N V-1 joL+ joL +
jwC,
= o

1 V,-,2joL+
4Z

13NE12) .(1.3.26)

the image impedances and 2 jLjoC

GQ. 1.3.6 Compute
propagation factor for the network
shown in Fig. 1.3.9 using ABCD .V
2ito'LC+1
parameters. joC
jwC
hAD / 2jo'LC+1

Tech-Neo Publieations. bere Authors inspire innovation A SACHIN SHAH Ventur

RF Design (MU-8h Sem.-E&TC) (1-13) RF Filter Design

Applying KVL to 1-2-2-1-1

1-2Lc ..(1.3.29)
V,-(joL) = 0
But
V = (j»lL) 1

foLC+1 Here I l
Where Z joL + JOC
jC
B = joL ...(1.3.35)
JwC alv,=0
V = joC
LC .(1.3.30)
As, I = lh

Using Equation (1.3.29) into Equation (1.3.30)

D v,=0 ..(1.3.36)

v, =(-'LC oC
The ABCD matrix for network given in Fig. 1.3.9 using
..(1.3.31) Equation (1.3.32), (1.3.34), (1.3.35) and (1.3.36)
Rearanging Equation (1.3.31) 1-200LC joL
1 -o'LC
1-oLC AB ..(1.3.37)
1=0 1-20Lc) Lc D joC
1-LC
V=0 (1-20 .1.3.32)
"1-oLC(13.32)
From Equation (1.3.13), Z =VC
From Equation (1.3.32), we can write as:

V, = (-20'LC .(1.3.33)
(1-20 LC. ioL
(1-o'LC) JoL

joL
Using Equation (1.3.33) into Equation (1.3.29)
(1 o LC)'

4 (joc .v,
1-26 LCT-oLCV sL1-20 LC)
jac
1-oLC
Z V1-2w 10) .(1.3.38)

C: .(1.3.34)
BD
Now, using Equation (1.3.12) Z42 =\W AC
Let V2 = 0 i.e. output part is short circuited as shown in

Fig. 1.3.11. (joL) (1)

Zg
L WE20LC(joC)
L
1- 2 10 2 1-o LC 1-o'LC)
6L) -o'LC
2 Z V1-20o' LOSC
2 1'o
1'o-
Z (1-oLOc-2oLC.13.39)
(1A20) (1421) From Equation (1.3.18(a))
Fig. 1.3.11
coshy = VAD

Tech-Neo Publications..here Authors inspire innovation A SACHINSHAH Venture

 RF Filter Design
RF Design (MU-8h Sem-E&TC) (1-14)
CONSTANT K LOW PASS
1.4
FILTER DESIGN
BY USING
coshy 1-oLC IMAGE PARAMETER
..using Equation (1.3.37) ..(1.3.40)
METHOD
Table 1.3.1: Image parametersfor T- and t-Network

What is
constant-K-low pass filter ?
T-Network T-Network GQ. 1.4.1
and t section of
Refer Fig. 1.3.6 Refer Fig. 1.3.4 and (b) presents T
Figs. 1.4.1(a) filter is
constant K low pass
filter. Constant K low pass
ABCD parameters ABCD parameters which passes certain
band of
defined as a filter
attenuate all
off frequency and
A=1+2Z A=l+27 frequencies upto cut
other frequencies above the
cut off frequency as shown
B Z
in Fig. 1.4.1(c).
B Z+4Z of series arm impedances and shunt
Also the product 2
Z2 R,. It is simplest
C- impedances is equal to Z
=
arm

Z
D 1*2Z type of filter.

D1*2Z U2 U2 Z
Imageimpedance: Image impedance
2Z2 222

(1A22) (a)T section (1A23)(b) T section

Propagation Constant:
Propagation constant: ***

-1+2
*

=1+2Z Stop
Pass
Dand band
*****ww**venwvvenn

*********************

1.3.2 Constant K or Prototype fHz)

Filter Section (1A24)(c) LPF attenuation Vs frequency characteristics
-----

Fig. 1.4.1: Constant K low pass filter

GQ.1.3.7 What is constant K filters ? OR What
is prototype section ?
1.4.1 Analysis of Constant K Low
In constant K filter, the product of series arm Pass Filter
npedances Z, and shunt arm impedances Z, satisfy the
Ga. 1.4.2 Design a constant K low pass T
ndition as Z Z, = Rwhere R, is a nominal
section filter to be
haracteristic impedance and do not depend on frequency. terminated in R
2 having cut of o, rad/sec.
Types of constant K filter (1) Constant K Low Pass
Filter (2) Constant K High Pass Filter Constant K-low pass T section filter is shown in
Fig. 14.1(a). In low pass filter, series arm
consist of inductor and shunt arm
impedances
consist ofimpedance
capacitor.
Tech-Neo Publirations. Where Authors inspire innovalion "A SACHIN SHAI Venture
 RF Filter Design
RF Design(MU-9 Sem-E&TC) (1-15)
Expression of nominal characteristic In Equation (1.4.5) Z is characteristic impedance of T
section of low pass filter. The reactance curve characteristic
impedance Ro
of T section low pass filter is shown in Fig. 1.4.2.
From Fig. 1.4.1(a),
Total series arm impedance

.(1.4.1)
X oL
2r
Total shunt arm impedance

ZioC .(1.4.2)

characteristics of
DC (1A25)Fig. 14.2: Reactance Vs frequency
constant K LPF
Multiplying Equations (1.4.1) and (1.4.2)

2,joa) Putting the value of Z Z and from Equations

C
(1.4.1), (1.4.2) and (1.4.4) respectively, into Equation (1.4.5)
.-1 becomes:

i s being real magnitude]

R, .(1.4.3)

a1.4.2 Reactance Curve and

Expression for Cut off
=R 1 oLC
frequency f
From Fig. 1.4.1(a),
Z = joL
Ro (14.6)

X, = oL
2
Z LC
4
. o LC .(1.4.7)

joL joL joC Zor Ro

4
jC where LC
-oLC 1.4.4)
4

But

...(1.4.8)
Z 7=Zg=\
Taking Z, Z, common
From Equation (1.4.8) it is clear that
G)Z R,foro = 0
.(14.5) V

Teeh-NeoPublications. Where Authorsinspire innovation SACHIN SHAH Venture

 RF Filter Design
RF Design (MU-8h Sem.-E&TC) (1-16)
i) . L C real. :1+5 2

cä) 1 Zq =R,\1-is imaginary.

Gv) Z becomes characteristic impedance of pass band =1- 20'LC +V fo (0j1C+4)
whena Z is real
) Z becomes characteristic impedance of stop band 1 2 + oLojc-4)
when
oL >1 as ZT becomes imaginary.
(vi) It cans
that the cut off frequency will be at particular
.LC
condition when oLC1 1 4(LC
i.e.

Hence constant K LPF has a cut off frequency

expressed as.

.(1.4.9)
LC or fLC
+LC
It is also clear fromEquation (1.4.8) that, when f < f
ZTis real in pass band. When f> f, Zis imaginary in
the stop band. 1-22L V 1
So the cut off frequency of constant K LPF can be
expressed in Equation (1.4.9).
-1 (1.4.11)
section.Z, - Rg .(1.4.10)
a 14.4 Analysis of attenuation («)
and phase shift (8)
a 1.4.3 Expression for Propagation
characteristic of constant
Factor K LPF
From Equation (1.3.26), he, propagation factor for
T-section constant K-LPFis given as FromEquation(1.4.11), coshy=1-4
4
Putting

With reference to Fig. 14.1(a), here, Z, =joL and Z, = k coshy= 1-4 LC

2
Putting the values of Z and Z, into Equation (1.3.26) 1-LC sincey a+jB
or cosh ( +jB) =
=

jOL
1+2 joC
OR cosh a cos B+ jsinh a sin B= 1 - 2 ..(1.4.12)

joL
Equating real and imaginary part of Equation (14.12),
joL) we get

Cjoc joc cosha cos B = 1-LC

..(1.4,13)

Tech-Neo Publications. here Authors inspire innovation

A SACHINSHAH Ventur
RF Design (MU-8 Sem.-E&TC) (1-17) RFFilter Design
sinh a sin B = 0 ..(1.4.14) ie. sin
Since imaginary part is zero. In order to fulfill Equation
(1.4.14) either a =0 or B = nm, n is being integer. Taking square root on both sides
For pass band
The attenuation 'a' is zero in pass band i.e. a = 0 sin
From Equation (1.4.12)
=sin
cos B=1- 2
i.e B 2 sin" radian ..(1.4.17)
Since cosh a = 1 .(1.4.15)
The value of cos ß varies from + 1 and - 1 F o r stop band
In stop band attenuation 'a' is not equal to zero (a # 0)
Hence cosB= 1-0LC
2 as from Equation (1.4.15) B =nm.
From Equation (1.4.13), for n = 1,

-cosh a =l-29
L2 o i.e. o= 0is the lower cut off
Since CoS Tt - 1 ...(1.4.18)
frequency of filter operation.
This proves that lower cut off frequency of filter is a =cosh 1 Neper ..(1.4.19)
Zero.
Now, cos ß= - 1 gives higher cut off frequency of Also -cosh a =1-2 cosh
filter
From Equation (1.4.18),
. 1-LC =
1 1-2cosh =1-20
-2 which gives
2 -2 cosh a 20
oLC =4
2 cosh a =

LC
Taking square root on both sides
VLC Biving i, cosh a=
Now in Equation (1.4.15) substituting@
= cost"Neper
cosB= 1L1_20
d = 2 cosh"()

B = coSs
radian ..(1.4.16)
2 2cosh .(1.4.20)
The attenuation (o) and phase shift (B) characteristics
Also, cos 1-2sin=1-=12
of LPF have been shown in Fig. 1.4.3.

2ai
Tech-Neo Publications .ww.Where Authors inspire innovation A SACHIN SHAH Venture
RF Design (MU-8h Sem.-E&TC) (1-18)
RF FilterDesign

a (Nep) Pass CT Ro ..(1.4.22)

Stop
B(rad) band band
Table 1.4.l:: Design parameters of Constant-K LPF
B 2sin (Hz)/B Parameters Value
a - 2c0osh
Ro
Slow attenuation rate
a 0
nVLC
f(GHz)
Ra
(1A26)Fig. 1.4.3:a and Bcharacteristics of constant K-LPF Tfe
C
a 1.4.5 Design Steps and Equations Tt Roe
of Constant K-LPF
ZT
Selection of elements like nominal characteristic
impedance and cut off frequency, inductance capacitance
plays very important while designing the filter. Ro
These are expressed as follows

(Equation (14.3) )
Ex. 1.4.1

f TLC (Equation (1.4.9)) Design constant K LPF for the following specifications
Cut off frequency f = 2.5 kHz ; Nominal characteristic

Tf impedance R, = 700 2

) Find image impedance Z and Z

LC Gi) Also find the frequency at which this filter produces
attenuation of 19.1 dB.
.LRrc
The inductance and capacitance are expressed using
ii) Find phase constant ß at pass band and stop or
attenuation band.
above two Equations as follows, Soln.
asR-y Given: f 2.5 kHz =2500 Hz, R= 700 2

Attenuation, a' =
19.1 dB =
6RG= 2.199 Neper
Ro . R LC
To find the values of L and C
L =R VLC The design equation of inductor and capacitor for
L yLC constant K LPF are
expressed in Equations (L4.21) and
(14.22).
So using Equation (1.4.21)
.L T 14.21)
700
L = T t T X 2500 89.126 mH
Similarly for C:
Using Equation (1.4.22)
RV C
Tt
R fe Tx 700 x 2500
C R 7t
C = 181.89 nF
Ro

Teeh-Neo Publications. here Authors inspire innovalion A SACHIN SHAH Venture

RF Design (MU-8" Sem.-E&TC) (1-19) RF Filter Design
() The image impedance ZqT of T section and Z of n
Ex. 1.4.2
section of LPF is given as:
AT section constant K-LPF has series inductance 80 mH
and shunt capacitance 0.022 uF.
ZR1-(E). (i) Determine the cut off frequency ii) nominal
R characteristic impedance 'R'. Also design an equivalent

V-CE
But in this case, f > f. therefore the image impedance
T section.

Soln.
ZT is imaginary. Given: L=80 mH and C = 0.022 F

(ii) Attenuation in stop band of constant K LPF is given

)The cut offfrequency f, is given as
as

a = 2 cosh Neper
Where f is frequency at which this filter produces
attenuation 19.1 dB or 2.199 Neper TV80 x 10x0.022 x 10
From above Equation.
f 7.587 kHz
Gi) The nominal characteristic impedance is expressed
as

2500 coah 80 x10

f
f =
=2500 x 1.66785
4.169 kHz
R 0.022 x 10
R 1.906 kHz
(ii) The phase constant ß in pass band is given as
The required T and t section of constant K LPF shown
B 2 sinE in Fig. P. 1.4.2(a) and (b).

2 sin(4170 U2 40mH U2 40mH L 80mH

25 radians.

In pass band As f>f, ß in pass band is also C/2 0.011uF

C 0.022uF c/2=0.011uF=
imaginary.
In stop band: The phase constant ß in stop band is
given as p=T radians. (1A29)(a) (1A30) (b)
(iv) Desigm of constant K-LPF Fig. P. 14.2: Constant K-LPF with T and n section
The T and n section constant K LPF along with its
elements is shown in Fig. P. 1.4.1. H1.5 CONSTANT K HPF
U2 U2 44.56mH

89.126mH =44.56mHA GO. 1.5.1 Design constant K-HPF

2 C 181.89nF
Fig. 1.5.1 presents T and n sections of constant K-HPF.
2C 2C
(IAZT(a)
L 89.126mH
3 2
CI2 90.94nF C/2 90.94nF

(1A31(a) T section (1A32)(b) t section

(1A28) (b) Fig. 1.5.1: Constant K-HPF
Fig. P. 14.1: T-type section of constant K-LPF T section

Tech-Nee Publications.. here Authors inspire innovation A SACHIN SHAH Venture

 RF Filter Design
RE Design (MU-8 Sem.-E&TC) (1-20)

In T and T sections, a series impedance and shunt

ZinZon=
impedance are given as

ZiCC ..(1.5.1) v ...fort section..(1.5.7)

Z = jo»L ..(1.5.2) represent the image
Equations (1.5.6) and (1.5.7)
The image impedance which is equal to characteristic section of HPE.
impedance of T and T
impedance of T section HPF is given as
Attenuation (a) and Phase
1.5.1
Constant (6) Characteristic

Using Equations (1.5.1) and (1.5.2) of Constant K HPF

44 jcjol ==R ..(1.5.3)
For T section, the propagation constant yis given as:
Where in Equation (1.5.3) R, is nominal characteristic
(1.5.8)
impedance. cosh Y T2z
By using Equations (1.5.1) and (1.5.3), ZT of HPF can
be expressed as
By using Equations (1.5.1) and (1.5.2)
1/joC
1+2 1+ 2jol.
ZT LC
1o2jo
(1.5.9)

Z ...For T section.(1.5.4)
120LC
But y 2+jB ...(1.5.10)
From above equation, two conclusions can be drawn: Comparing real and imaginary parts in Equations
(1) If 4 LC > 1, Z is real and filter operate in pass (1.5.9) and (1.5.10)
band.
sinh a sin ß = 0 ..(1.5.11)
(2) If 4 wLC < 1, ZT is imaginary and filter operate in
or cosh a cos B= 1- .(1.5.12)
stop band. C
Hence the cut offfrequency is expressed as To fulfill Equation (1.5.12), either o = 0 or B = nt
4 oLC = 1
where n is integer.
1 Pass band: Attenuation
.Oc2 LC constant B exist with cos ß =1 -
a =0 in pass band. Whereas
when a =0 in
phase

..(1.5.5) 2 LC
fc4TLC Equation (1.5.12).
The value of cos ß varies from+ 1 to - 1.
Also Z7 R
cos B = 1-
2 wLC
..utilizing O 4LC

...(1.5.6)
1-20c 2 0LC
1
0 2LC
For n section, image impedance Z Znis given as:
=

R OxLC*
=
ZonZT
This is higher cut off frequency.
Tech-Neo Publications...hereAuthors inspire innovation
d SACHINSHAH Venture
RF Design (MU-8" Sem-E&TC) (1-21) RF Filter Design

For cos B= - 1, from Equation (1.5.12) 1

cosh O =

20LC
COS T=-1|
1
-1 20LC
cosha = -

2
20LC
20LC
cosh a = - 1+2-

LC
Hence = cosh| .(1.5.17)
oLc =
and
stop band a is given by Equation (1.5.17)
In
4LCC phase constant is given as B =
. 2 LC .(1.5.13) Stop Pass em.ww.ienweedaennwrihoranoodiooeeetaorni
a(N) band band
From Equation (1.5.13) gives a value of o when cos B1
B-1 acosh 2 B= (rad)
Therefore, the lower cut off frequency is expressed
aS

p- 2sin
Oc2LC 1.5.14) **** *m*.gaanrme**********;*

t=0
******

1- .(1.5.15)
fc 4T LC
(1A33)Fig. 1.5.2: a and ß characteristics of HPF
Also from B
cos
1-2LC
=

a 1.5.2 Summary of Design

2 C2
Equations of Constant K HPF
cosB =1- .4LC
Nominal characteristics impedance "R' and cut off
. B = cos
frequency fe if both are being known, then the values of L
and C of filter can be easily obtained by using following
Again cos B = 1-2sin' =1 expression.
Ro (Using Equation (1.4.3))
-2si- 7-2%
(Using Equation (1.5.15))
2
fe4Tt LC
Solving above two equations the expressions for L and
2 C obtained as

RLC
sin R,LiLC
sia R VLC =L
4Tt RVLC = L
B 2sin' .(1.5.16) 4T

F o r stop band L Ro
4tfc ..(1.5.18)
As B=m in stop band, Equation (1.5.12) reduces to

Tech-Neo Publications.. here Authors inspire innovation A SACHINSHAH Venture

 RF Filter Design
RF Design (MU-8 Sem.-E&TC) (1-22
of constant K
-

filter are
Similarly for C Two disadvantages
as

follows
Ro (a) does not increase in stop band
(1) The attenuation

C VLC 47 LC so it is not sharp.

Ro 41t Ro There is wide variation
in the value of
(2)
characteristic impedance in pass band from
C 4T Ric ..(1.5.19)
desired value.
Table 1.5.1: Design parameters of Constant- K HPF
(3) Non constant image impedance.
Parameters Value The m -
derived filter removes these two disadvantages
Ro of constant K or prototype filters.
Here the m -
derived filter is designed from constant K
filter which has same image impedance throughout
4T VLC
pass band and stop band.
L
R Also in, m -
derived. filter, it ispossible to get fast
47T fc attenuation rise in stop band and beyond the cut off

C frequency. Figs. 1.6.1 (a) and 1.6.1 (b) shows prototype

and m - derived T and T sections.
4T R fc

ZT
4 o'Lc R
222

Ro (1A35)
(1434)
mz,/2 mZ/2
: GQ. 1.5.2 The nominal characteristic impedance
of T section of LPF is R Find Z at
09 fc (1A36): T section of constant m - derived

Given: 0.9

Using Equation (1.4.10), Z

=
R m m
1 (f
Ro (1A37):T Section of prototype and m-derived filter
Zn 1-(0.9) 2.3 Ro Fig. 1.6.1:Tand n sections of prototype and m - derived filters
Zi 2.3 R
1 . 6 M-DERIVED FILTER Here, Z and ZZ
Using Equation (1.3.24),
: GO. 1.6.1 How m derived filters are superior:
For constant K T
than constant k filter ? OR How -

section,
m - derived filters can be obtained
from constant ?
ZT ..(1.6.1)

Tech-Neo Publications. here Authors ispire innovation A SACHINSHAH Venture

RF Design (MU-8" Sem.-E&TC) (1-23) RF Filter Design

Form- derived T section

..(1.6.2)
272
By equating above two equations,

mz (1A39)(a) : t Section of constant K filter

Squaring on both sides,

22
Z,7 4 +m Z Z m

mZ,z a-m)+z,z (1A40)(b) : n Section m - derived filter

m -derived filters.
- z m
Fig. 1.6.3 :Constant K and

For T section of constant - K filter,

term should be positive to calculate Z Using Equation (1.3.20)

physically. i.e. 0 < m< 1.
Z, is an impedance which is consist of two series

impedances as Z, and both of these

142
For T section of m - derived filter

impedances can be calculated if i s positive. Here, Z = Z and Z,m

i.e. 0<m<1.
Thus in m - derived T section

Z = mZ, .(1.6.3)

amd - )z,+ 1+4

..(1.6.4) m

Fig. 1.6.2 presents the m -

derived T section obtained
-

Equating above two equations,

from constant K filter by having modified series and shunt
impedances shown in Equations (1.6.3) and (1.6.4).
mZ/2 mZ/2
132

m 1+4Z
(1A38 Fig. 1.6.2: m -
derived T section filter
-

GQ. 1.6.2 How n section of m-derived filter is

obtained from constant k

section P

The T section of constant k filter n section of m

derived filter is shown in Figs. 1.6.3(a) and 1.6.3(b).
. m 4m

Tech-Neo Publieations . u . here Authors insgpire innovation A SACHINSHAH Venture

 RFRF Design (MU-8h Sem.-E&TC) RF Filter Design
(1-24)
Figs. 1.6.6(a) and 1.6.6(b) presents general m derived T
4ZZ, +m Z, Z = 4Z Z,+Z Zm and T filter.
mZ
mZ,/2 mz,/2

m
4m
4m 4m

Z Z2
Z2m 4m
(1A42(a): T-section
(1A43(b) t section

T and n section of LPF

Fig. 1.6.6:General m derived
Z, Z
the shunt
particular frequency
-m)+A 4m
For Fig.
branch of T section
1.6.6,
or
at a

series branch of T section will be in

Multiplyingg and dividing by 4m2

t o Numerator and resonance making short circuit to the input of T - section
- m) and open circuit for n section, therefore causing infinite
Denominaor attenuation.
-4m With reference to Equations (1.4.1) and (1.4.2)
ZZ2 (1-m)
Z 42, m Z joL and Z,o c
m+ m (1 m))
At resonance frequency,
m 4m
1-
moC
..(1.6.5)
4
mZ,+Tm -m) LC
From Equation (1.6.5), it is clear that, Z, is an
impedance of n section of m derived filter which is a-m)Lc
consisted of two parallel impedances as m Z, and

1-m TTLC(1-m)
Thus, T section of m - derived filter is shown in
Fig. 1.6.5.
mLC 1-m
mZ m-derived Tsection:
For low pass filtler,

1-r
m
m
1-m

(1A41)Fig. 1.6.5 : Filter General m

-
derived n section -
a 1.6.1 m -
Derived Low Pass Filter 1-m

Analyze m derived T and n sections

GQ. 1.6.3
of LPF

here Authors inspire innovation A SACHIN SHAH Venture

Tech-Neo Publications ..
RF Design (MU-8" Sem.-E&TC) (1-25) RF Filter Design

1.6.2 Analysis of Attenuation (a)

m Vi-4/-1- Versus Frequency
The m- derived T section of LPF is shown in characteristic of LPF
Fig. 1.6.7(a. m - Derived Section

The propagation constant of m - derived LPF is

expressed

mC

(1a43a)Fig. 1.6.7(a) : m- derived T section LPF

For low pass m derived filter, by referring Fig. 1.6.7(a)
m-derived n 8ection Z = jolLm
Similarly, for m - derived n section, inductance and
and mjoL
jøCm
4m j o L

capacitance in series branch make a resonant circuit.

joLm
1joL (1-m)
4m
j»Cm
m L = 4m Lm
(1-m) o,c 4m +joL (1- m) jøCm
jCmx 4m
4
(1-m) LC joLm joCm 4m_
4m+jwLCm (1 - m)
, - m ) LC
ion-mLC
41m-wLCm (1 - mn)

TVLC(1-m -ommLC
4m-oLCm (1 - m)

LCmLC V1-m) o'm LC

For low pass filter 41 oLC(1-m|
4
TVLC - (20wm/@)
:
1-(1 m)' (o/ o
Where GThen,
mL
1-(w/o
142 1-1-m)( a
From above equation, e'becomes real if m is 0< m< 1

( c m and e > 1 for »> o. Hence stop band starts at

@ = 0 , like constant K - section.

For = where = * e becomes infinity,

(1a430)Fig. 1.6.7(b) : m -
derived 7 section low pass filter
producing infinite attenuation.

Teeh-Neo Publieations. here Authors inspire innovation A SACHIN SHAH Venture

 RF Filter Desiq0n
RF Design (MU-8h Sem-E&TC) (1-26)
, mZ/2 mZ,/2 mZ,/2
From . 1-m it is clear that o > O. SoO mZ/2

infinite attenuation
produces after the cut off frequency
a s shown in Fig. 1.6.8. Steep of a decreases after
0>is not desirable. This can be solved by
cascading constant K section to give composite filter. (1A45)(a)

a (Neper) mz

m-derived ***** 222.

attenuation **
*****

wwww
ConstantK
attenuation
www.w
211-m)2
4m
***************** *g*****

(1A46)(b)

Fig. 1.6.9:m-derivedn section

**********
. .

(1A4Fig. 1.6.8: Attenuation Vs frequency characteristics

is
comparison of m - derived and constant K fiter The image impedance of Fig. 1.6.9b), expressed as

a 1.6.3 Cascading of m - Derived

T-Half Sections

From Fig. 1.6.8, it is clear that a very sharp cut off

frequency response received. However its attenuation
decreases for o > o As it is always expected that to
have infinite attenuation as o, then m - derived But 2T
section can be cascaded with constant K - section to

provide the composite attenuation response as shown in . Zn ZT

Fig. 1.6.8.
In order to match the image impedance of m - derived- Z
T section with image impedance of constant
Using Equation (1.4.8), ZqT =
Ro\ and
K (independent of m) section. Them - derivedIT
section was designed. However, still there is problem Using Equations (1.6.3) and (1.6.4),
of non constant image impedance.
z mZ,
The image impedance of n section will depend on m
and this can be used to get an optimum matching
section. (mZ , | ( m ) ,

Figs. 1.6.9(a) and (b) shows an infinite cascade of m -

derived T- sections. It is considered that the n section

is a piece of infinite cascade of m - derived T sections.

Tech-Neo Publications . u here Authors inspire innovation

A SACHINSHAH Venture
 RF Filter Design
RF Design (MU-8 Sem-E&TC) (1-27)

z ,mZ,
R1- Pass
band m1
Stop
band

-7,4 4

m 0.6
R- m 0.3

Z, Z, + Z1-m)
4
Z .(1.6.6)
(147)Fig. 1.6.10: Z Vs frequency in pass band of LPF for
different values of 'm'

Using Equation (1.4.3) i.e. Z, Z, ==R,as Z, =jol | Ex. 1.6.1

Design m -
derived T section of LPF having cut off
5 KHz and nominal characteristic impedance of
frequency
times
600 2. The frequency of infinite attenuation is 1.25
4C the cut off frequency.

-LC4LR Soln.
4C Given: f. =5 kHz, R = 600 2,
W,
1.25 5 kHz 6.25 kHz
-0L C.4:L 1.25 xf, = x =

4C 4C
(1) Design of L and C

For low pass filter, the cut off frequency

So Equation (1.6.6) becomes andR

Hence, using design Equations are found as :

4 LL = 600
x x 10 38.19 mH.
Z RRo1-() TI Ttx 5x10
C
1
TR fTx 600x5x 10
C 106.10nF
R,V1-(o/o
For m-derived LPF,
Z (-n)(0/o R, (1.6.7)
V1-(o/ o m

As Equation (1.6.7) is dependent on value of 'm' the

chosen value of 'm will minimize the variation of Zm m = Sx10
6.25 x 1 0 ) 0.6
=0.6
=

in pass band of filter. The actual values m - derived T - section of LPF, are
Fig. 1.6.10 shows the variation of Z with respect to given as follows
frequency for various values of 'm'. It is found that mL
2
0.6 38.19x
2
10-- 11459 mH
m=0.6, gives the bestresultfor Z
mC = 0.6 x 106.10 x 10"= 63.66 nF

Tech-Neo Publieations.hereAuthors inspire innovation SACHIN SHAH Venture

 RF Filter Desi
(1-28)
RF Design (MU-8" Sem.-E&T .(1.6.9)
m 4m Z
(mL (1-(0.6Ix38.19 x10 and (1.6.9), Z1 and Z, are
the
(1.6.8)
In Equations shunt branch of
of
of the
series and the
L = 10.184 mH impedances
constant K-HPF.
T section HPF ie
11.459mH m=11.459mH The shunt
arm of m-derived
at the frequency of its seriees
impedance
offers zero
consisted of series combination of 1
resonance (as Z, is
E 10.1m
and C).
frequency f. infinite
denoted as
mC 63.66nF At this resonant inductive reactance X.
and
attenuation is produced

becomes equal to
capacitive
reactance
(X
Fig. P. 1.6.1: complete designedm- derived T section of LPF
Hence,
a 1.6.4 M-derived High Pass Filteer
m
4 . C
GQ. 1.6.4 Draw and explain m-derived T and n 1-
m
Solving above Equation
section of HPF Explain graphical
representation of cut off frequency ..(1.6.10)
4LC
The m-derived T and n section of HPF are shown in
-m
Fig. 1.6.11(a) and (b). 2 VLC
mz
mZ/2 mz,/2 -m ..(1.6.11)
Or f 4mLC
But cut off frequency of constant K-HPF, is expressed
(4m22
1- m as

f 41LC Hz
(1A48(a) (1A49)(b)
OR c 2LC
Fig. 1.6.11: Tand n section of m-derived HPPF
Therefore using w Equation (1.6.10),
The shunt branch of T-section of HPF, is consisted of

series combination of capacitance

-4m c
1-m2 or f = (V1-m).f
Thus m for m-derived HPF is obtained as
ie ( z,)andindutanee
The shunt impedance Z is expressed as m

mj+
joc ..(1.6.12)

For T-section of Fig. 1.6.12(b), it is observed series

= mZ .(1.6.8)
branch that is consisted of parallel LC circuit which
Likewise, total seríes impedance Zi is given as produces infinite impedance at it's resonant frequeney
of parallel LC circuit.

1- 4m
m
jOC At this resonant frequency infinite

produced so, transmission through filter is zero.

attenuation s

Tech-Neo Publications Where Authors inspire innovation

A SACHIN SHAH Venture
 RF Design(MU-8 Sem.-E&TC) (1-29) RFFilter Design
Also at resonating frequency inductive reactance ) With reference to analysis of m-derived HPF
Xbecomes equal to capacitive reactance Xc

Thus,
4mL = m
1-m C
m

-m
m
V-4x 10
2VLC .(1.6.13) m = 0.436

Or f .(1.6.14) using Equation f. mTGand R, =V

4T VLC 4TV
From Equations (1.6.13) and (1.6.14) it is observed that The series arm capacitance and shunt am inductance

the expression for f. for both T and n section of m- are calculated as follows
derived HPF is same.
Therefore, the value m of T section is applicable for t CTRaf. 4 xTx 4x 10 x 600
section. = 0.0334 F
The circuit resentation of T and n section of m- Ro
and L 4Tf
600 = 11.94 mH
derived HPF is shown below in Fig. 1.6.11(c) and (d).
47x4 x 10
In T-section of m-derived HPF, the values of series
and shunt arm elements are calculated as

2C2X =0.151 F
11.94
0.436 27.39 mH
z (4mc-m
4m 4X0.456 x0033 x 10
1-m 1-(0.436)
(1A50 C) = 0.071 uF

In m-derived n section of HPF, the values of series

and shunt arm elements are as calculated as follows
mZ 2 2x11.94x 10 54.78 mH
0.436
mZ
0.033x0.43610 0.076 uF
m m

(IA51 d) 1-
s 4x0456 x 11.94 x 10
1-(0.436)
Fig. 1.6.11
25.704 mH
Ex. 1.6.2
2m 2
m
Design T and n section of m-derived HPF having cut off
frequency 4kHz, design impedance 600 and infinite
0.151 F 0.151uF
attenuation at 3.6 kHz.
27.39 mH
Soln.
Given fe= 4 kHz 4mc=0.071F
Ro = 600 2

f 3.6 kHz (IA52(a)

Fig. P. 1.6.2(Contd...)

Tech-Neo Publications ...Where Authors inspire innovation A SACHIN.SHAH Venture

 RF Filter Design
(1-30) Half
Sections for
R F Design (MU-8 Sem.-E&TC) m-Derived

a 1.7.1
Low P a s s Filter
= 0.076uF

terminating
half sections is
filter with
A low pass
25.704mH shown in Fig.
1.7.2(a).
mZ/2
mZ,/2
(4mL m
-
1-m 54.78mH 2Z2
54.78mH Low m
22 pass R
filter
R (T-ype)
(1A53Kb)
Fig. P. 1.6.2: Complex T-section T and
msections ofm-derived -
HPF
4T
1.7 IMPEDANCE MATCHING sections for T-types LPF
Terminating half
(IA55Fig. 1.7.2(a) :
uSING HALF SECTIONS and
Z, =
joL
With reference to theory of LPF,

Ga. 1.7.1 Why half section is used for j

terminating section in filter P values of Z and Z2, actual components
By using above
Ga. 1.7.2 Explain the need of using half in the terminating half sections with 'm'
can be found

sections
out as
i) Since a filter is being consisted of cascading
of sections, it is necessary that the impedance
at each
of number
. ntie-je()
junction is to be correctdy matched.
sections do not match
2
(i) The image impedances of T or t m m

with each other.

Section.
(iii) Therefore T section cannot be connected to n
of t and
Civ) When it is required to design a filter consisting Hence terminating half sections with actual elements
T sections, a half section should be used for impedance
are shown in Fig. 1.7.20b).
matching between T and n Sections.

(v) Such half impedance matching half

section is m
constructed either from T section or t Section.

Half Section has two different image impedances at mC/2 mC/2

(vi) Low
port I and port 2. R pass
fiter
Fig. 1.7.1 shows constant K half section

z/2
w- T GT

Z1 2 Z224 (1A56) Fig. 1.7.2(b): Terminating half sections for T-types LPF

(ASA) Fig. 1.7.1: Half section used to match Z, to Z

Tech-Neo Publications. here Authors inspire innovation

SACHIN SHAH Venture
 (1-31) RFFilter Design
RF Design (MU-8 Sem.-E&TC)
20 2C 44.33nF
1.7.2 M-Derived Half Sections for m m

High Pass Filter

2Um 2/m15.9mH
GQ. 1.7.3 Write a note no terminated m-
derived half section for HPE 24.93nF c
With reference to Fig. 1.7.1, instead of considering T-
T
type LPF, now consider T-4ype high pass filter. (1458Ka) (b)
For high pass filter, m-derived half section of HPF
Fig. P. 1.7.1: Terminating
Z and Z =joL 2 2x4.77x 103 = 15.9 mH
Then the actual components in terminating half 0.6
sections with 'm' can be found out as follows 2C 2x 0.0135 X1U--44.33 nF
0.6

(2m.c (2x0x0.0133x 10 =

1-(0.6)
24.93 nF

22 2io-jo( m Ex. 1.7.2

7 kHz and
(1 m2 Design aLPF filter with cut off frequency

e
-
-T section and
R 500 SQ which will consist of constant
m-derived two terminating halfsections. (Use m =0.6).
The terminating half section, with actual components
are as shown in Fig. 1.7.3. Soln.
Given f 7 kHz, Ro 5002
2C terminating
Required filter = Constant T section and two
half sections
2/m
Low
2Um 1. Design of T section of constant K low pass filter.
.R pass R
2 filter
4TVLCndR, =/
The actual components are calculated as:

ZT Ro 500
for T-types
L TTX7x 10 22.73 mH
(1A57)Fig. 1.7.3 : Terminating half sections
m-

derived high pass filter

C E = 90.94 nF
Ex. 1.7.1
nf Ro
Thus T section of constant K LPF is shown
4.77 mH and
pass filter with L
=
A constant K-high Fig. P. 1.7.2(a).
C 0.0133 uF operates with cut off frequency of 10 kHz.
The nominal characteristic impedance of filter is
600 2. In LI2= 11.365mH L/2 22=11.365mH
section with
order to cascade, determine terminating half
m 0.6 without disturbing cut off frequency.
C 90.94nF
Soln.:
L = 4.77 mH, C 0.0133 uF
Given
R = 600o2
f= 10kHz, (a)T section constant k LPF
The terminating m derived half sections of high pass (1A60)Fig. P. 1.7.2: Composite LPF
filter is shown in Fig. P. 1.7.1.
A SACHINSHAH Venture
Tech-Neo Publications. here Authors inspire innovation
 RF Filter Design
RF Design (MU-9h Sem.-E&TC) (1-32)

2. Design two terminating m-derived half sections.

Fig. P. 1.7.26) shows two terminating half section.
m/2
m/2

(1A61)Fig. P. 1.7.2 (b): m-derived terminating two half section

Thus, the actual components are calculated as follows

0.6x22.73 x 10 6.82 mH
2

m 06x90,94 x10 = 27.28 nF

2

(mL- ( x 2.73x 10 =12.12 mH

The complete composite LPF is shown in Fig. P. 1.7.2 (c).

6.82mH 11.365mH 11.365mHH 6.82mH

Rs 27.28nF 27.28nF
R
90.94nF 5002
12.12mH 12.12mH

Teminating half Constant K-T section of LPF Terminating half

section using section using
m=0.6 m =0.6

(1A83C) Composite low pass filter

Fig. P: 1.7.2d) represents with combining elements.

18.18mH 18.18mH

27.28nF 27.28nF:
90.94nF
12.12mH 12.12mH

(IABFig. P. 1.7.2 (d): Design ofcomplete composite low pass filer

1 . 8 COMPOSITE FILTER

GQ.1.8.1 Draw and explain block diagram of composite filter. Wirite advantage of compositefilters

The attenuation Vs frequency characteristic of prototype (constant-K) filter in stop band does not provide sharp cut off
and increases slowly from zero.

Tech-Neo Publications bene Autbors inspire innovation .A SACHINSHAH Venture

 RF Filter Design
RF Design(MU-g Sem.-E&TC) (1-33)
Whereas in m-derived filter attenuation provides sharp cut off but achieves very high value at a frequency for very close

to cut off frequency.

a
(Neper Pass Stop
band band Compositeflter

m-derived

Constant k

(AGAFig. 1.8.1a): Attenuation Vs frequency characteristicsofcomposite frilter

However from Fig. 1.8.1(a) it is observed that, in stop band of in low pass filter, attenuation reduces drastically after t

and reduces drastically before f. in stop band of high pass filte.

These drawbacks can be overcome by connecting number of filters including prototype filters and m-derived filters with
is termed
terminating half sections. Such combinations of different filters together along with terminating half sections
as 'composite-filter'.
Fig. 1.8.1b) shows the block diagram of composite filter.

Oned One or Terminating

Terminating more more half section
Ro halfsection
for impedance
Constant
K filter
m-derived
and
for R
impedance
matching forhigh filters for matching
Cutof sharp cutoff

4T
(IA6 Fig. 1.8.1(b): Composite filter with four stages

the conmposite filter, one or more constant K flters provide high attenuation into stop band at specifñc cut off
In
frequency and hence make possible transition from pass band to stop band at specified frequency.
The m-derived filters provide on infinite and sharp attenuation in stop band at a frequency close to cut offfrequency
c) Further, terminating half sections atthe source and load end of the composite flter provides the impedance matching
between source and load impedance 'R,' and intemal image impedance Z of constant K and m-derived filters.

Ex. 1.8.1

Explain attenuation (a) Vs frequency characteristics of composite filter.

Soln.
Fig. 1.8.1) presents attenuation Vs frequency characteristics of constant K filter, m-derived filter and composite
filler. Each individual filter including constant-K and m-derived does not meet expected attenuation Vs frequency
characteristic. But the combination of constant K and m-derived filters form composite filter which provides desirable
attenuation Vs frequency characteristics from Fig. 1.8.1(a). It is visualized that the attenuation rises very fast with frequency
in stop band (G, to f) and falls only marginally with frequency.
The composite filter is designed from three parameters
) Cut off frequency (ü) impedance (ii) infinite attenuation frequency 0 or m.

Tech-Neo Publications.. here Authors inspire innovation A SACHIN SHAH Venture

 RF Filter Design
RF Design (MU-9 Sem.-E&TC) (1-34)

Ex. 1.8.2
Design low pass composite filter with following specifications:
() Cut off frequency 2 MHz
i) Characteristic impedance 75 2
(ii) Provides infinite attenuation at 2.05 MHz.

Soln.
Given: f 2MHz
R 75Q
f. = 2.05 MHz

It is assumed that the design of composite low pass filter consist of all T sections.
1. Design at T section of constant K low pass filter
For constant - K LPF

LCnd and R=1/

L2 5.96uH L/2 5.96uH

Using above twoequations, actual components are as follows:

L 75 = 11.93 uH
Ttte x 2x 10 C 2.122nF

= 2.122 nF
RTX2x Tx2x 10'x 75
K LPF
(1A66)Fig. P. 1.8.2(a): Tsection of constant
-

The T section of constant K-LPF is shown in Fig. P. 1.8.2(a).

2. Design of T section of m-derived LPF: The frequency of infinite attenuation i.e. f =2.05 MHz.

m V-E
. m -1-(205-0219
The components of series and shunt branch of T section of m=1.306uH =1.306uH
m-derived filter are calculated as follows
mC 464.71pF
. 0219x 11.93x10 = 1.306 mH
2

(L- (1-(0.219x11.93 x10=12.96 H

4m 4x0.219
mL-129h
mC = 0.219x 2.122 x 10 =464.71 pF
The m-derived T section of LPF is shown in Fig. P. 1.8.2(b). (1467)Fig. P. 1.8.2(b): T section of m-derived LPF

3. Design of m-derived half sections:

The half sections are designed for m = 0.6.

The series and shunt arms of terminating half sections are calculated as follows

m 0.6)x(11.93 x10)-3,579 uH 2

mC (0.6) x(2.122x 10) 636.6 pF

Tech-Neo Publieations . w u . Where Authos inspire ianovation A SACHINSHAH Veature

 (1-35) RFFilter Design
RF Design (MU-8" Sem.EATC)
( 2m
L- ( 2x0.6
x I1.93 x10 =6.362 uH
The terminating m-derived half sections are shown in Fig. P. 1.8.2c).

mL/2 = 3.579uH mL/2 3.579uH

To
To
Source
-6362u4 2m Load
Filter
m-6366pF Filter mC 636.6pF
O

(1A68)Fig. P. 1.8.2(c) : Matching sections

4. The complete composite filter is shown in Fig. P. 1.8.2(d).

3.579uH 5.96uH 5.96uH 1.306uH 1.306uH 3.579uH

j464.71pF: 6.362uH
Rs
6.362uH R
752 2.122nF: 752
12.96uH 636.6pF
:636.6pF

T section of section of Matching

Matching
half section constant K m-derived half section
LPF LPF

(IA69Fig. P. 1.8.2(d) : Low pass filter of composite

Ex. 1.8.3
for the following specifications.
Design a composite low pass filter by image parameter method
f= 50 2, f. 50 MHz, f 52 MHz
= =

Soln.
f 50 MHz
Given =

50 2
f = 52 MHz

It is assumed that all sections are T type.

1. Design of Tsection of constant KLPF.

For low pass filter

and R=

follows LI2 =0.159uH L/2 0.159uH

Using above two Equations, the actual components
are as
-
L x 50100.318 uH
50x
TfTx 50x 10 C 127.32pF
C 1 = 127. 32 pF
R x 50x 10°x 50
constant K-LPF
Design of T section of constant K-LPF is shown in Fig. P. 1.8.3(a). (1470)Fig. P. 1.8.3(a) : T-section of
A SACHINSHAH Venture
Tech-Neo Publications. here Authors inspire innovalion
 AF Filter Des
RF Design (MU-8" Sem.-E&TC) (1-36)
is f. = 52 MHz.
2. attenuation
Design of T section of m-derived LPF. The frequency ofinfinite

The
m
-y-)- v-.7
components of series and shunt branch of T section ofm-derived filter
are
calculated as tollows

m0.0429uH 0.0429,H
m027x0.318 x10 = 0.0429 uH

(L-( 4x0.27 |.0.318x 10'=0.272 uH

mC 34.37pF

mC =
0.27 x 127.32 x 10 =
34.37 pF 027,0.27pH
The m-derived T section of LPF is shown in Fig. P. 1.8.3(6).
(IA7)Fig. P. 1.8.3b):
T-section ofm-derived LPp
3. Design of m-derived half section for matching
The half sections are
designed for m = 0.6
m
2 0.6x0.318x10 95.4 nH

mL0.6x127.32
2
x10 38.196 pF

(L- (1-(0.6x0-318
The half sections fot
x 10= 169.6 nM
matching are shown in Fig. P. 1.8.3(¢).
mL/2 95.4nH mL/2 95.4nH

L= 169.6nH
(1-L169.6H Loae
Filter
Filter
.TT*38.19pF m=38.19pF
(1A72Fig. P. 18.3(e) : m-derived half section for matching impedances

4. Design of composite low pass filter is shown in Fig. P, 1.8.3(d).

95.4nH 0.159uH 0.159uH, 0.0429uH 0.0429uH 95.4nH

169.6nH 34.37pF 169.6nH

127.32pF
38.19pF 0.27uH R
38.19pF
Half section T section T section of
Half section
for constant K m-derived for
Matching LPF LPF
Matching
(IA73Fig. P. 1.8.3(d): Composite LPF step-wise design of composite filter

Tech-Neo Publications. here Authors inspire innovation

SACHIN SHAH Ventun
 RF Filter Design
(1-37)
RF Design (MU-g Sem-E&TC)
Ex. 1.8.4 n = -(-020
Design a composite high pass filter by
the image parameter and shunt branch of m
The components of series
method with the following specifications derived Tsections
as follow
are calculated
=50 MHz, f. =48 MHz. x21.22x 10
R, 75 2, f
= 2 51.57 pF
0.280
Soln.:
50 MHz, f. 48 MHz L 119.36 XI -426.28 nH
R =7592, f, = =
Given: 0.280

Design of T section ofconstant K high pass filter: 4x0.280 12

.
For constant K high pass filter (1- m C1-020 x21.22 x 10

= 25.78 pF
R and
fan yLC Thus, m derived T section of
HPF is shown in

. The expressions for L and C components are Fig. P. 1.8.4(b).

obtained as follows: 2C
m
151.57pF 151.57pF

L 4=119.36nH H
426.28nH

C4TR.f 21.22 pF
4mc
m
25.78pF
HPF is
Hence a prototype i.e. T section of constant K -

shown in Fig. P. 1.8.4(a) m-derived T section of HPF

(IA74) Fig. P. 1.8.4(b) :
2C 4244pF 2C 42.44pF with m = 0.6. The
3. Designof terminating half section
shunt branch of half sections
components of series and
SL 119.36nH
arecalculated as follows:
2C 2x2122
0.6
x10 -70.7 pF
filter
(IA73) Fig. P. 1.8.4 (a) T-section of constant high pass
K
=
2x119.36x
0.6
10 - 397.86 nH
filter.
2. Design of m-derived T section and high pass
The value of m for HPF is
(C-0 X 21.22x
m
39.78 pF
Thus m - derived half sections are shown in Fig. P. 1.8.4(c).

70.7pF 2C 70.7pF
m

397.86nH Fiter iter 2397.86nH

m
Load
Ro
Source
+2C38.re T1-m
(1A75) (1A76)

Fig. P. 1.84 (c) m-derived half sections for impedance matching

here Authors inspire innovation A SACHINSHAH Venture

Teeh-Neo Publications.
 RF Filter Design
RF Design (MU-8" Sem.-E&TC) (1-38)
shown in Fig. P. 1.8.4(d).
4. Design of composite high pass filter is
151.57pF 151.57pF 70.7pF
70.7pF 42.44pF 42.44pF

397.86nH 119.36nH426.28nH 3 397.86nH

Load
Ro
Source 39.78pF
39.78pF 25.78pF
m-derived Terminating
Terminating Constant K T section half section
half section T section of HPF
for m = 0.6
for m = 0.6
fHPF

(d)Composite high pass filter

1.8.4: Step wise design of composite high pass filter
Fig. P.

1.9 FILTER DESIGN BY Pnc ...(1.9.1)

INSERTION LOSS METHOD Pload 1-IT(0)r
is matched with source
When load impedance
impedance, then PLR = IS
Go. 1.9.1 What is perfect filter ?
**
The insertion loss (L) in dB is expressed as
The characteristies of perfectly designed filters are IL = 10 log PLR ..(1.9.2)
) Zero insertion loss in pass band Due to causal properties of passive network, IT (»)r is
(ii) Infinite attenuation in stop band even function of w it can be expressed as polynomial in
an
(ii) And a linear phase response in pass band in order
to avoid signal distortion in passband. M (w) ...(1.9.3)
T(o)F = M()+N (b
However, practically it won't be possible to have such
filter which will meet all the characteristics at a time. Where M and N are real polynomials in o.
So compromise must be made depending upon the Substituting Equation (1.9.3) into Equation (1.9.1)
requirements. Though design of filter by the image
power loss ratio, PLR can be expressed as:
parameter methods is simple it has restricted freedomn
to design the filter so that it will meet above PLR
M(0
characteristics.
The filter design by insertion loss method uses network
M(o)+N (o)
synthesis technique so that it provides high degree
control to shape the characteristics in pass band and M(0)+N (0)-M (0)
M (0)+N (0)
stop band.
M(+N(o)
1.9.1 Characterization by Power N(w)
Loss Ratio (PLR PLR1+M(o)

.(1.9.4)
N(o)
GQ. 1.9.2 Explain the concept of insertion loss Thus, for filter to be physically realizable its power loss
method ratio must be the form in Equation (1.9.4)

In filter design by insertion loss methods, power loss 1.9.2 Types of Filter
ratio (PLR) or insertion loss are defined as

Power available from source GQ. 1.9.3 Explain different types of low pass
PLR Power delivered to load filter.

Tech-Neo Publications.. here Authors inspire innovation

A SACHIN SHAH Venture
RF Design(MU-8 Sem-E&TC) |-39) RF Filter Design

Based on Equation (1.9.4), there are four types of low 2. Equal Ripple or Chebyshev filter:
pass filter. A chebyshev polynomial is used to represent the

Maximally flat Butterworth or Binomial filer insertion loss at N order low pass filter as
It can satisfy requirement for sharpest cut off region.
This type of filter provides the flattest possible
passband response for a given filter order. P 1+K T, ..(1.9.8)

For low pass filter, PLa is expressed as, Where Ty x) is the chebyshev polynomial. Generally,
2N
Nis chosen to be odd integer when source and load
an
Px1+K' .(1.9.5)
impedance are matched. This type of filter provide the

sharpest cut off, with ripples in the pass band. The

In Equation (1.9.5), N is order of filter, o, is cut off
magnitude ofripple is 1+ K" as shown in Fig. 1.9.1(b).
frequency. The PiR versus frequency characteristics is
shown in Fig. 1.9.1(a). PLR
In pass band from o = 0 to o = 0, at the band edge,
1+K
PLR1+ K', o =0.
PLR Flat wnipunwrew

Pass O.5 1.0 Stop o/@c

************
band band

1/2 Pass Stop (1A79Fig. 1.9.1(b) : PLR Vs frequency characteristics of

band band
equiripple LPF

**** wwfa. w w Since Ty(x) varies from - 1 to +1 for Ix l 1. Thus, K

(1A78)Fig. 1.9.1(a) PLz VS frequency characteristics decides the amplitude of ripple.
of maximally flat LPF
For large X Tu ()= 2x", for o>> o, the
If K = l then PuR = 2 for w =0 which is 3 dB
insertion loss from Equation (1.9.8) is expressed as:
frequency. 2N
1.9.9)
For > o,, the attenuation increase monotonically
with frequency. For very large 0 i.e. 0 >> o and with From Fig. 1.9.1b), insertion loss increases fastly in
K= 1, then stop band as compared to binomial. Also insertion loss.
24
Pa K -( (1.9.6) of chebyshev filter is greater than binomial filter
for any frequency where >> @c-
From above the insertion loss IL is defined as:

L = 10 log (PR) ..(1.9.7) For N 3, log1o () = 12.0 dB

Which increases at the rate of 20 N dB / decade. The N=5 log -24.1d

first (2 N - 1) derivatives of Equation (1.9.6) are zero
It is also concluded that both maximally flat and equal
at = 0.
ripple responses both have monotonically increasing
This type of filter meet the need of minimum insertion
attenuation in stop band.
loss by providing flattest pass band response.

Tech-Neo Publications...Where Authors inspire innovation .A SACHIN SHAH Venture

 RFFilter Design
(1-40)
R F Design (MU-8" Sem.-E&TO)
"faster" roll off than
band. It has
3.
Elliptic function band and stop
response in both pass
filter has equiripple
This type of low pass 1.9. 1(c).
chebyshev filter.
It is shown in Fig. and minimum attenuation
maximally flat and attenuation Anax in the pass band
the maximum
of low pass filter is specified by
This type

Amin in the stop Band.

PLR
Amin

Better Adequate
Amax cut off rate attenuation

1.0 2.0

**

characteristics
function low pass filter
(1A80)Fig. 1.9.1(c) : Elliptic

4. Linear phase Low pass filter linear phase response in pass

band.
distortion in pass band, it is necessary to have
signal sharp cut off
In order to avoid
systems. At the cost of filter
for communication
Such linear phase filters are used as multiplexing filters
can be achieved with the following
response
characteristics a linear phase

wo)= Ao +P
of filter and P is constant.
Where o(w) is phase of voltage transfer function
Group delay t, is expressed as
..(1.9.10)
do
do -1+P(/2N+1»)|
From Equation (1.9.10), group delay for linear phase
filter is maximally flat function.

Loss Method
1.9.3 Process of Filter Design by Insertion
GQ. 1.9.4 Explain design process of filter by insertion loss method.
-----*.- --****~********
The Fig. 1.9.2 shows the block diagram of process of filter design by insertion loss method.

Filter
Low pass Scaling
prototype & Implementation
specification design conversion

(AB)Fig. 1.9.2: Block diagram of process ofiter design by insertion loss methods
The general procedure for designing a filter using the insertion loss method can be summarized in three steps:
1 Filter specifications: The first step include cut offfrequency stop band attenuation, insertion loss in pass band and
frequency response of filter in pass band etc.
2. Design of low pass prototype or constant K circuit : In 'this step the design of low pass prototype include
R=12 and o, = 1 rad/sec. Filtertables are used for this step.
3. Scale and conversion: Finally the filter is scaled to the proper impedance level and if desired to a high pass, band pass
or band stop topology.

Tech-Neo Publications...WhereAuthorsinopire innovation A SACHIN SHAH Venture

 RF Filter Design
RF Design (MU-8h Sem.-E&TC) (1-41)

1.9.4 Maximally Flat Low Pass PLR

Filter Prototype
-+152+
GO. 1.9.5 Derive expression for the values of L
PLR
and C for second order low passfilter Z2--2+1
and maximally flat low pass filter Z Z+, +Z,+1
Fig. 1.9.3 shows low pass filter prototype for N=2. FLa +Z+Z+1-2,Z+Z, +Z-
L Zi n +2n +Z+1
,+2,+1
PLR 2Z+ 22n
(Zi +1) (2+)
in
22+2
GA8DFig. 1.93: Low pass filter prototype for N =2 PLR Z+1
2Z+Zi ...(1.9.12)

From Fig. 1.9.3, From Equation (1.9.11)

+Z RI-Rioc) R1)
Inputimpedance Z= joL+ R
1+R'C) 1+o'RC
i.e. imaginary part is zero
Zin Since ZR+0j
r
Power loss ratio from Equation (1.9.1)
Z+ Zin (1+0RC) ...(1.9.13)

PLR 1-
OCR
oL 1+oR'
Now Z = jol+
R
So, we have,
jac oCR
R Pa .R')a1.RC)J
joC
joL+1+RiøC 4R (R+2R+ 1 +R'Co+ol'+o'L'cR
JOC
-2 LCR)
R
joL+1+RjøC 1+R I(1-R)+(R C+L'-2LCR)of
Multiplying 2d term by complex conjugate of. +LCRo .(1.9.14)
(1+ Rjoc) to (1 -Rjoc) to numeratorand denominator
If low pass filter of Equation (1.9.5) is second order
R(1-Rjoc) filter, N 2 , and is normalized (Z, 12 and a@, = 1), then
Z joL + PLR Ofsuch filter is expressed as:
(1+Rjoc)(1-RjoC)
R(1-RjoC) PLR 1+.. K=1 .(1.9.15)
joL+-oRC As at = O, IL = 10 log (1 + K)=3 dB

R(1-Rjo 3 = 10 log (1 + K)
jolL+ 14+0 RC) 0.3= log (1 +K)
.(1.9.11)

Tech-Neo Publieations. here Authors inspire innovation A SACHINSHAH Venture

 AF Filter Design
(1-42) flat low pass
filta
RF Design (MU-8" Sem.-E&TC) values for
maximally

Element for the

resistance
Taking antilog on both sides Generator in Fig. 1.9,4(a)
ladder
network shown
for the
10 1+K Generator
resistance

shown in Fig. 1.9.4(b)

1.999 =K* ladder network
0.99=K? 1 L2 9 2
R, 90 ----
K = 0.9976 1 Generator TUn.

Thus K 1 C3 9N 1 Load
impedance
resistance

From Equation (1.9.15) PLR = 1 at o = 0, substituting C 91

this in Equation (1.9.14), gives
I = 1 R a-R)N in
shunt element
Ladder circuit
begins with
R (1A83Ma):
4R L 93
L 91

9N
. R R =1

R 1
series element
From Equation (1.9.15), coefficient of w is 0, equating begins with
circuit
(LA84(b): ladder
flat low pass filter
this coefficient in Equation (1.9.14), with R = 1, 1.9.4: Ladder
circuits for maximally
Fig.
R (RC+L-2LCR)=0 & 2sin )I

C+L-2 LC =0 K= 1,2,...N ..(1.9.15a))

(C-L) =0 Thus C=L Inductance for series inductors

Similarly the coefficient of o in Equation (1.9.15) is 1, 8,k=1 to N) =

Capacitance for shunt capacitors
is shunt capacitor
equating this to Equation (1.9.14), with R = 1 gives
SLoad resistance if g a

EN»1 \Load conductance if gy is a series inductor

c) Sample calculations are shown below, and (calculator
L = 4 mode should be on radian mode).
Thus L = C=y2 For N = 1, number of elements =2
For higher order of N filter, this method of finding the . From Equation (1.9.15(a))
values of L and C are impractical.
.
81 2 2 sin
Ga 1.9.6 Explain ladder circuits for maximally 82 1
low pass filter and element definition
For N 2, number of elements =3.
The normalized low pass filter with two ladder

prototypes
impedance
are In these two types of ladder source
designed.
is 1 2 and cut frequency o = 1. The elements
81 2 sin ]-14142
values are numbered from g, at the generator impedancee to
with N elements.
2s J-14142
8N1 at load impedance offilter and 83 1
However the element values for ladder type circuits are Table 1.9.1 shows element ('L' and C') values for
calculated as follows and are shown in Table 1.9.1 for normalized (g= 1 and o, 1)
maximally flat (Butterworth)
=

N= 1 to 10. Elements are alternate between series and shunt low pass filter prototype.
connections.

Tech-Neo Publications... here Authon inspire innovtion

A SACHIN SHAH Venture
 RF Filter Design
RF Design (MU-8h sem.-E&TC) (1-43)
Table 1.9.1: Element values for maximally flat low-pass filter prototypes (e= 1, o, =1, N=1 to l0

N & 6 8 810 B11

12.0000 1.0000
2 1.4142 1.4142 1.0000
3 1.0000 2.0000 1.0000 1.000

0.7654 1.8478 1.8478 0.7654 1.0000

0.6180 1.6180 2.0000 1.6180 0.6180 1.0000

6 0.5176 1.4142 1.9318 1.9318 14142 1.5176 1.000

7 0.4450 1.2470 1.8019 2.0000 1.8019 1.2470 04450 1.0000
8 0.3902 1.1111 1.6629 1.9615 1.9615 1.6629 1.11110.3902 1.0000
9 0.3473 1.0000 1.5321 1.8794 2.0000 1.8794 1.5321 1.0000 0.3473 1.0000
10 0.3129 0.9080 1.4142 1.7820 1.9754 1.9754 1.7820 1.4142 0.9080 0.3129 1.0000

The Figs. 1.9.4 (a) and(b) are dualof each otherandboth willproduce samefrequencycharacteristics
Taking log on both sides
Ex. 1.9.1
Design a maximally flat low pass filter with cut off log 30.62 = 2N log

frequency of 2 GHz, impedance of 50 2 and at least 15 dB

1.4860 2Nlog
insertion loss at 3 GHz.
(a) Find the order 'N' of filter. 1.4860
0.1760= 2N
(b) Find unscaled values of the filter reactive elements.
. 2N = 8.4431

Soln. N = 4.2215
Given:f=3 GHz, f =2 GHz IL 15 dB N = 5
(a) The order of filter 'N" is calculated as:
Thus required number of elements for maximally flat
LdB) = 10 log Pr! low pass filter are 5.
15 = 10 log PLR!
Thus, the order offilter N =5.
1.5 =
log IPL (b) Unscaled values of reactive elements of filter
are obtained either from Table 1.9.1 or
Taking antilog on both sides
calculated as follows:
10 P
PLR 31.62 8 2sin
Power loss ratio with K = 1 is : for N = 5 and K=1 to 5
2N

PR1+K(
81 2 sin o=0.6180=C
2N
31.62 1 & 2 sin|1=1.6180 =L

31.62-1- ( 8, 2 sin 2-C,

2N
30.62 = 8 2 sin10 1.6180 =L4

Teeh-Neo Publieations bere Authors inspire innovalion A SACHI SHAH Venture

 RF Filter Design
RF Design (MU-8 Sem.-E&TC) (1-44)
Taking log on both sides,
& 2 sin =0.6180 =C Log 99 = 2N log (1.375)
and & 1.9956
. 2N 0.1383
The filter circuits with five reactive elements are shown
in Figs. P. 1.9.1(a) and P. 1.9.1b). 2N = 14.42

1.6180H 1.6180H N = 7.2147 8

2 92 4 94 Thus N = 8
98 The required number of elements to design maximally
C 91 3 93 Cs 95
0.6180F
2F 0.6180F flat low pass filter are N= 8,

(1A85Fig. P. 1.9.1(a) : Ladder circuit begins with shunt element 21.9.55 Equal Ripple or Chebyshev
L 91 93 L5 95 Low Pass Filter Design

G 90 502
Ga. 1.9.7 Explain chebysheve low pass filter
C4 94 96-13
synthesis process.
The power loss ratio "PLR' for an equal ripple low pass
(1A86Fig. P. 1.9.1(b): Ladder circuit begins with series element filter with cut-off frequency o =
1, from Equation (1.9.1) is
expressed as
Ex. 1.9.2
A maximally flat low pass filter is to be designed with a cut
PLR1+kT (o) ..(1.9.16)
Where 1+k'=ripple amplitude in pass band.
off frequency of 8 GHz and minimum attenuation of 20 dB
at 11 GHz. How many filter elements are required ? But Chebyshev polynomials have the property that:
JO for N odd
Soln.: Ty(0) 1for Neven
Given f = 8 GHz
Also, the chebyshev filter provides a sharp transition
L = 20 dB
from pass band to stop band. So, in this case Chebyshev
f = 11 GHz polynomials are used to find insertion loss mathematically
In order to design maximally flat low pass filter,
as follows:
required number of elements 'N' are calculated as follows
ILI
L (dB) = 10 log PLR
VI+eTy ()
20dB 1 0 log PLR
where N= 1,2,3... (1.9.17)
2 logPLa In above Equation (1.9.17), Eis constant, o' normalized
Taking antilog on both sides,
frequency and Ty (0) is chebyshev polynomial of degree
10 PR N.
PLR = 100
Where, T,(o) = o'
Power loss with K = 1
2N Ta(o) 20"-1 =

PR 1+K| T (w) 40"-3 =

100 =1+( K=1 T (0) =80-8o+1

The insertion loss of chebyshev filter is defined as:
- ( IL =-20 log1o

99 - ()
Teeh-Neo Publications..Where Authorsinspire innovation SACHIN SHAH Venture
RF Design (MU-8h Sem.-E&TC) (1-45) RFFilter Design
The design steps for equal ripple or chebyshev low
= -
20 log1o pass filter are described as follows:
For chebyshev low pass filter
IL =
10log1o1+ T .(1.9.18)
...(1.9.21)
Where, 1 ifN is odd Number
BN+1 ...(1.9.22)
G, = 10 log (e + 1)
coshifNis even number
G
10 log (e +1) ..(1.9.23)

Taking antilog 4a-) ay. k= 2,3... N

and Bk .(1.9.24)
b-1) Bk- 1)
'

1010= e+ where,
G
E = 100 - 1 = 101 xG - 1 ..(1.9.19) ..(1.9.25)
Where G, is ripple amplitude in dB.
X = sinh2N ...(1.9.26)
The order N' or number of required elements of
chebyshev low pass filter is calculated as a =sin ] (1.9.27)

N =
cosh 10-1 b x+sin' ( (1.9.28)
.(1.9.20)
cosh
The sample calculation of values of elements for
chebyshev low pass filter are shown below and based on

Where L is required insertion loss in dB at frequency ' ' . these calculations Table 1.9.2 and 1.9.3 are given for
reference.

Table 1.9.2: Element values for Equal-Ripple Low-pass filter prototypes g, = 1, ac= 1, N=1 to 10, 0.5 dB ripple

0.5 dB Ripple
N 2 54 55 59 S10 B11
1 0.6986 1.0000

2 1.4029 0.7071 1.9841

3 1.5963 1.0967 1.5963 1.0000

4 1.6703|1.1926 2.3661 0.8419 1.9841

5 1.7058 1.2296 2.5408 1.2296 1.7058 1.0000

6 1.7254 1.2479 2.6064 1.3137 22.4758 0.8696 1.9841
7 1.7372 1.2583 2.6381 1.3444 2.6381 1.2583 1.7372 1.0000
8 1.7451 1.2647 2.6564 1.3590 2.6964 1.3389 2.5093 0.8796 1.9841

991.75041.2690 22.6678 1.3673 2.7239 1.36732.6678 1.2690 1.7504 1.0000

10 1.7543. |1.2721 26754 1.37252.7392 1.3806 2.7231 1.3485 2.5239 0.8842 1.9841

Tech-Neo Publications .. here Authors inspire innovalion A SACHIN SHAH Venture

 RF Filter Design
(1-46)
RF Design (MU-8 sSem.-E&TC)
1, N=1 to 10, 3.0 dB ripple
Table 1.9.3: Element values for Equal-Ripple Low-passfilter prototypes g,
=
1,oc =

3.0 dB Ripple
&10
N
1.9953 1.0000
2 3.1013 0.5339 5.8095
3 3.3487 0.7117 3.3487 1.0000
4 3.43890.7483 4.3471 0.5920 5.8095
5 3.4817 0.7618 4.5381 0.7618 3.4817 1.0000
3.5045 0.7685 4.6061 0.7929 4.4641 0.6033 5.8095
6
3.5182 0.7723 4.6386 | 0.8039 4.6386 0.7723 3.5182 | 1.0000
7
8 3.5277 0.7745 4.6575 0.8089 4.6990 0.8018 4.4990 0.6073 5.8095

0.8118 4.66920.7760 3.5340 1.0000

93.5340 0.7760 4.6692 0.8118 4.7272

10 3.5284 0.7771 4.6768 0.8136 4.7425 0.8164 4.7260 0.8051 4.5142

0.6091 5.8095
For N 2, 8=1, o, =1, G, =0.5 dB ripple amplitude
From Equation (1.9.27), for k = 1, Using Equation (1.9.24),
To calculate a 4 xa Xa
S2
b, xS1
sin 0.7071
Now using Equation (1.9.27),
Using Equation (1.9.23)
2x0.7071 sin
X = 0.7071

So to calculate X, use Equation (1.9.25) and Equation

Using Equation (1.9.28),
(1.9.26)
b X+sin
tncoh(1737 b 1.0160 +1=2.016
Using trigonometry: 4x0.7071 x0.7071
coth x tanh = 1 2.016x 1.4029
82 0.7071
.coth tan h
Using Equation (1.9.22)
. coth =
0.5 Now g3 = cosh
tanh 17.37
coth = 34.7495 = cosh 3.54
= In [34.7495] = 3.5481 83 1.4198
Using quation (1.9.26), Tables 1.9.2 and 1.9.3 exist for designing equal ripple
low pass filters with normalized source impedance and cut
X sinh22=1.0080o
off frequency o= 1. These elements can be used for
2x0.7071 1.4029 designing ladder circuits shown in Fig. 1.9.4(a) and 1.9.4(b).
1.0080
81 1.4029

Tech-Neo Publications. .Where Authos inspire innovation A SACHIN SHAH Venture

 RF Filter Design
RF Design(MU-g" Sem.E&TC) (1-47)
coth )
Ex. 1.9.3 =5.8476
3
tan h 17.31
Design 3-dB equi-ripple low pass filter with a cut-off
frequency of 2 GHz, 1 Q impedance level and at least 15 dB E n [5.8476] = 1.7660
insertion loss at 3 GHz.
X = sinh 1.7660 0.2986
Soln. 6

Given G 3 dB
f 2 GHz b (0.2986) sin(
+

G &o 502 = 0.0891+ (0.8660)

L 15 dB
= 0.0891 +0.7499
f 3 GHz
b = 0.839
Design of equal ripple low pass filter
G, 3
b (0.2986) + sin?
e = 100-1=100 - 1= 10-1
e = 0.9952 1
= 0.0891 + (0.8660)
= 0.0891 +0.7499
) The order of filter is calculated as follows:
b,= 0.8390
10
cosh1 0 - 1
1
N b =(0.2986 +sin'
cosh( 0.0891+0

cosh.
101x151 b, 0.0891
V 10013
N
The 'g' values for the filter can now be found by
cosh ...As N is odd
Eo 84=1
N= 2.4913
(ii) The element values are S1 2x0. 3.3489
0.2986
&o= 50
As N = 3,
82 4x0.5XL-0.7118
3.3489 x 0.839
... as Nis odd
83
4x1x0.5 3.3489
Using Equations (1.9.23) to (1.9.28), x 0.8390
.7118

Thus, a =sin =05 The 'g' values are listed in Table 1.9.2 for N =3.
The designed equal ripple low pass filter is shown in
sin=1 Fig. P. 1.9.3.

,
=
sin 05 R 90 12 L2 92
Now,
e =ineothT73) C91 9 R=1
tanh
ann 17.37 = 0.17101 .tan x cot 6 = 1
L
(1A87Fig. P. 1.9.3: Equal ripple low passfilter

Tech-Neo Publications . here Authors inspire innovation '

A SACHIN SHAH Venture

 RF Design (MU-8 Sem.-E&TC) (1-48) RF Filter Design
=

1 . 1 0 SCALING OF LOW PASS The L, C and R are the values of original prototype
PROTOTYPE FILTERS filter
2. Frequency scaling
GQ. 1.10.1 Explain two types of scaling.
Since the cut off frequency w, of prototype filler is 1,
The designs of low pass filter
prototype by insertion so to change it from unity to different low pass cut off
loss method have a normalized source impedance of frequency values, frequency of filter is scaled by the
Rs= 12 and cut off frequency of o,= 1.
However it is
factorwhich is achieved as
possible to scales and transform low pass
prototype filter design into low pass, high pass band
pass and stop band filters for any impedance ...(1.10.5)
(Rs =
R)
value and for any cut off frequency value.
By applying Equation (1.10.5), the inductive and
There are two types of scaling: capacitive reactance are found as
1. Impedance scaling
2. Frequency scaling

L = oL

1. Impedance scaling
L= L
It is seen that, with reference to Equation (1.9.22),
equal ripple filters have non unity, load resistance value L= L
.(1.10.6))
for an even 'N' order of filter. Whereas in
prototype
filter design, source and load resistances are unity.
Since filter is linear circuit, it is possible to multiply all
impedances including terminating resistances of
prototype filter design by some factor without .C
disturbing transfer function of the filter. .c
Thus, input and output impedances will change. Let
prime denote impedance scaled quantities. If required
load and source impedances are equal to Ro.

Then, .(1.10.7)
X = R,XL
By applying both
impedance and frequency scaling,
L'@ = R, oL Equations (1.10.1) to (1.10.7) becomes as
. L =
R,L ...(1.10.1)
.(1.10.8)

...(1.10.9)
R
R= R ...(1:10.10)
Therefore C = ..(1.10.2)
Ro R= R,R ..(1.10.11)
R R1 =R ...(1.10.3) Fig. 1.10.1 shows the scaling for low pass filter
prototype response for @= 1.
R= R R;= R, RL .(1.10.4)

Tech-Neo Publications . here Authors inspire innovation

A SACHIN SHAH Venture
RF Design (MU-8 sem.-E&TC) (1-49) RF Filter Design

PLR The order of filter is 1.

Now,

a sin

Thus, &= (05

n| coth17.37)J
1 1 E = 3.5481

3.5481
(1A8 a) Low pass filter prototype for @. = 1 X sinh2
X = 2.8625
PLR
2x1
81 2.8625
81 0.6986

Now,
R8 50 x0.6986
2 Tx 100 x 10

(1A8 b) Frequency scaling for low pass filter prototype

L= 5.69 nH
R=R s 0 a
Fig. 1.10.1 Frequency scalingfor low pass filters
L
Ex. 1.10.1 w
A chebyshev low pass filter is to be designed in such way s RL
that it should maximum ripple level 0.5 dB in its pass band.
The filter must operate in the frequency range of upto
100 MHz and attenuate the signal at 400 MHz by 5 dB. The (1A90)Fig. P. 1.10.1
load and source resistance of 50 S2 each.

Soln. M 1.11 FILTER TRANSFORMATION

Given G, = 0.5 dB

f = 100 MHz GQ.1.11.1 Write a note on filter transformation

f 400 MHz As discussed previously filter transformation is used to
IL = 5 dB design new filters including low pass, high pass, band pass
R= Rs = 502 and band stop filters.

A chebyshev low pass is designed as follows: The following are the filter transformation techniques.

The order offiltered is calculated as follows

1.11.1 Low Pass to High Pass
cosh Filter
N
cosh (4) GQ.1.11.2 Explain low pass to high pass ilter
10015
cosh transformation.
N
VioTR
cosh (4)
A high pass filter is designed by transforming low pass
filterprototype. This is shown in Fig. 1.11.1.
2.1622
N
cosh
cosh O.12201
(4)

Tech-Neo Publications ..Where Authors inspire innovation A SACHINSHAH Venture

 RF Design (MU-8h sem.-E&TC) (1-50) RFFilter Design

R, Lx .(1.11.6)
HP
R ...(1.11.7)
Cand L are capacitor and inductor of high pass
filter respectively.
C
(IA9TKa) (1A92 b) HE ---IF
RsRo
Fig. 1.11.1:Low pas to high pass transformation
E L4 R
An ideal low pass filter passes all signals upto
normalized cut off frequency with almost zero insertion
loss in pass band and attenuates signals above cut off (1A93)Fig. 1.11.2: High pass filter obtained from low pass
frequency. prototype
On the other hand, high pass filter passes all signals
above the cut off frequency o, and attenuate the signals Ex. 1.11.1
those are below the cut off frequency Design a high pass chebyshev filter with pass band ripple
Therefore, following frequency transformation will be magnitude not more than 0.01 dB with cut off frequency 100
used to transform a low pass filter to a high pass filter. MHz. The filter exhibit at least 5 dB attenuation at 400
MHz. The load and source resistances are 75 2 each. Also
...(1.11.1) design scaled low pass chebyshev filter.
1. Thus, applying Equation (1.11.1), inductors wil
replace the shunt capacitor of low pass filter and
Soln.:
capacitors will be connected in series in place of
Given G, = 0.01 dB

IL = 5 dB
inductors.
f 100 MHz
X = jo L = - i k f = 400 MHz

R R = 75 2
(1.11.2) Design of chebyshev low pass filter*
jo Ck
) The order of filter is calculated as:
jBx jo Ck =
-j Cx
cosh
HP ...(1.11.3) V100
jo L N =
cosh (4)
Equation (1.11.2) indicates that series inductor Lx is
replaced by capacitor CK cosh 2.1622
0.0023
N = 4.1158
Thus, L .(1.11.4) cosh (4) 2.063
N = 1.99 2
2. Similarly, shunt capacitors C is replaced by inductors
But in order to meet symmetrical characteristics 75 2
L, thus Equation (1.11.3) reduces to each side order of filter can
be increased to N=3 (odd)

..(1.11.5) So we consider order of filter

N = 3
By applying impedance scaling using Equation
(1.11.4), Equation (1.1 1.5) becomes,

Tech-Neo Publications m s here Aulhors inspire innovation

A SACHIN SHAH Venture
 RF FilterDesign
RE Design(MU-8h sSem.-E&TC) (1-51)
82 0.9664 =C
Element values: For K =
1, K= 2 and K = 3
(i)
and 8s
4 x a Xa
b X B2
i.e. a= sin 2N 4x 1x0.5 = L3
83 3.315 x 0.9664
a sin =0.5 83 0.62425

a sin 1 (vii) By applying scaling rule as follows:

and a sin =0.5

L=L - L 5x0.62425
2Tx 100x 10

ii) e n coth17.37)J L=L 74.51 nH

C 0.9664
tanh 17.3 = s.7570 C R, 2tx 100x 10° x 75
using tan 0 x cot 6 = 1 C2 20.50 pF
in
cot 6 = (vii) The low pass chebyshev filter is shown
tan
Fig. P. 1.11.1.

coth 9.0117.37 O.01 1737.00

R Rs L = 74.51nH Ly 74.51nH
tan 17.37,
E = In [1737.00] = 7.4599 7.5
C2 RL 756n
iv) X = sinh 20.5 pF

X = sinh 1.601 (1A93())Fig-P. 1.11.1: Scaled low pass chebyshev filter

ix) To design highpass chebyshev filter, with

(v)
impedance scaling as:

b (1.6019)+sin() C =c=
= 2.5660+ (0.8660) R,
= 2.5660+0.7499
2Tx 75x 100 x 10° x0.62425
b= 3.315
C = C = 33.99 = 34 pF
b (1.6019 +sin?( R
and L = 75
2.5660+(0.8660) = 2.5660 +0.7499 27t x 100x 10" x 0.9664
C2
b2= 3.315 HP
. L= 123.5 mH
(vi) Thus, g1 x)The resulting chebyshev high pass filter is shown in
Fig. P. 1.11.1(a).
2x0.3
1.6019 0.62425 L
=

4b-18x-1
(a1 RRs c34 pF c34 pF
ie. Ex
752
4xaXa2 R RL
82 b,X 81 123.5nH 3 752

82
4x0.5x 1
3.315 x 0.62425
(1AB4)Fig. P. 1.11.1(a) :Scaled chebyshev high pass filter

Tech-Neo Publications here Authors inspire innovalion A SACHIN SHAH Venture

 RF Design (MU-8 Sem.-E&TC) (1-52) RF Filter Design
1.11.2 Low Pass Filter to Band Pass and 4 .(1.11.10)
Filter
Also from Fig. 1.11.3(a), shunt capacitor of low pass
GQ. 1.11.3 Explain and obtain the expression of filter prototype transforms into parallel combination of
capacitorand inductor. inductor as Land capacitor as C in band pass
A low pass filter
filter. These elements are obtained as
prototype can be transferred into band
pass filter as shown in Figs. 1.11.3(a) and (b). BP
L, 1.11.11)
RgR C
and .(1.11.12)
v 2
BP 2-0
Further, these elements are scaled by as required by
source and load impedances expressed as follows
(1A95 a): Band pass filter circuit
C
- .(1.11.13)
R, L
and .(1.11.14)

O 1
**********************-**
@2
.***********************
L R,(-0,) ..(1.11.15)
(1A96) (1A97)
(b) Low pass filter prototype to Bandpass filter C
and
Fig. 1.11.3: Band pass filter ...(1.11.16)
R2-o)
The perfect low pass filter passes all
frequency signals Ex. 1.11.2
with zero insertion loss in pass band upto normalized
cut off
frequency o 1 and completely stops signals
=
Design a bandpass chebyshev
filter with maximum ripple
those have higher frequency than level 0.5 dB with N 3. The centre frequency is 1 GHz and
=

In band pass filter, all signals with frequencies between bandwidth is 10% and impedance is 50 2.
and 02 passes and stops passing outside this Soin.:
frequency range. Given: N=3, fo= 1 GHz,
Hence the following frequency substitution transform a
2-@1
low pass filter prototype to a band pass filter. i.e. 10% 0.1,
= =
G, =0.5 dB

( (I1.8) (1.11.8) ) To design bandpass filter, first equal ripple low pass filter prototype is designed as follows:
Where, in Equation (1.l1.8), . o,=V o x 2
1. Above frequency substitution replaces the series ie. asin Dz]|
inductor of low pass filter prototype to inductor BP
L
and capacitor C which are connected in series as
shown in Fig. 1.11.3(a).
sin()o5
These elements are obtained as follows sin(
and s i n ( - 0 5
..(1.11.9)

Tech-Neo Publications .. here Aulhors inspire innovation

SACHIN SHAH Venture
RF Design(MU-9 Sem-E&TC) (1-53) RF Fiter Design

ie.e Thus, the element values of chebyshev equal ripple low

(i) =
n| coth 17.37 pass filter prototype of Figs. 1.9.4(a) and 1.9.4(b) are shown
below
tanh( 17.37 =002877
81 1.5963 =L,
coth 17.37 34.7495 8 1.09681 =C
B 1.5963 =Ly
. Using tan e.cot = 1
= in [ 34.7495] 841.00 =R,
Above values are also verified from Table 1.9.3.
E 3.5481
elements of bandpass filter are found
X = sinh (vii) The scaled
ii) follows
X =sin h
R,x x , xL
= 0.62642

0.1
(iv) x'+ sinN ...Putting

0.1
b (0.62642) +sin() 50x 2T x 1x 10'x 1.5963

(0.392440)+ (0.8660 C 0.199 pF

0.39240+0.74995
R,XL
b= 1.14235

-
b =
(0.62642+ sin
= 0.39240+(0.8660)
= 0.39240+0.74995 .Putting 0.1
b=1.1423
50x 1.5963
27x 1x 10
() 0.1
2x0.5 1.5963 50x1.5963
81 0.62642 * 27x 1x 10x0.1
4 xaX L
BP
127.02 mH
(vi)
Now, L R, -)
4xaXa
4x0.5x1 R,(-0
&2 1.1423x1.5963 x Cx
821.09681 50x 0.1
4x a Xa 2mx1x 10x 1.09681

.Putting 0.1
4x1 x0.5
83 1.1423x1.09681 BP
L 0.726nH
83 = 1.5963

R(o -o)
Teeh-Neo Publieations . here Authors inspire innovalion A SACHIN SHAH Venture
 RF Design (MU-8h Sem-E&TC) (1-54) RF Filter Design

R,O 01)

1.09681 .Putting =0.1

50 x0.1 x 2 x 1x 10
c2 34.91 pF
Now, R,XL
BP R,XLg
L

50x1.5963
O.1 x 2tx 1x 10 =0.1
L 127.02 mH
C
R,x ,X O,XL3
0.1
S0 x 2 Tx 1x 10x 1.5963
BP
0.199 pF
R,XRL 502
(vii) The resulting band pass filter is showm in Fig. P. 1.11.2.
Rs R50 LBP 127.02mH
c0.199pF C0.199pF
L127.02mH
0.726mH2 c=34.19pF 502

(1A98)Fig. P. 1.11.2: Scaled equal ripple band pass filter

Ex. 1.11.3 ) Design of bandpass Chebyshev filter
Design band pass chebyshev filter that exhibits with N 3,
no more them 0.01 dB in pass band. It passes signals
ripple
40 x 10°x 10x 10°
from 10 MHz to 40 MHz with zero insertion loss. Load and
f 20x 10' Hz = 20 MHz
source resistance are of 752 each.
i) Design, equal ripple chebyshev low pass filter
Soln. prototype for N = 3
Given = 40 MHz
f= 10 MHz sin sin2NJ
G, = 0.01 dB, N=3
a sin =05

Tech-Neo Publications ..here Authors inspire innovation

A SACHIN SHAH Venture
 RF Design(MU-9h Sem-E&TC) (1-55) RFFilter Design
4 sin(=1 820.9664 =C2
83 0.62425 =Ly

a sin '84 R=1.00

(vi) Further, the scaled elements of equal ripple band
ii) n coth pass filter are calculated as follows:

0.01 02-@1
tanh 17.37 =5.7570

R,xoXL
Using tan 8 cot 6= 1 27tx 10°(40 10)
coth137 0.01 1737.00 75x (2 x 10°x 20) x0.62425

E i n [ 1737.00] = 7.4599 = 7.5 C 254.95 pF 255 pF

X sinhN R,XL
L 27tx 10° (,-f)
x sin)-1.6019 75 x 0.62425

21x 10° (40- 10)

(v) x+ sin L= BP 248.38 nH

b, = (1.6019)+ sin (tw3)

= 2.5660+(0.8660) L
BP R, (-0)
= 2.5660+0.7499
oxC
75 x 27x 10°(40 10)
b = 3.315
(2tx 20 x 10°) x0.9664
ba = (1.6019) + sin? . L = 926.37 nH
b= 2.5660+(0.8660 (ix) Thus the scaled elements of chebyshev bandpass
b2 = 2.5660 +0.7499
filters are:
b2 = 3.315
=
L 248.38 nH

(vi) S1 Ca =
C=255 pF
2x0.5 926.37 nH
81 1.6019 0.62425
C
4xa-Xa C Ro(-O)
-X8-1 0.9664

82 4x0.5x 1 75 x 2Tx 10" ( 40-10)

3.315 x 0.62425
C 68.35 pF
= 0.9664

4x1x0.5 68.35 pF,

83 3.315 x 0.9664 R =R = R," =RLx 75 =75 2
83 0.62425 The resulting chebyshev equal ripple band pass filter is
shown in Fig. P. 1.11.3.
(vi) Thus, the element values of equal ripple chebyshev
low pass filter are as follows:

81 0.62425 =L

Teeb-Neo Publieations. here Authors inspire innovation SACHIN SHAH Venture

 RF Fiter Design
RF Design(MU-8 Sem.-E&TC)_ (1-56)
R 752 c 255pF La 248.38mH c68.35pF
o-

L248.38mH RL752
926.37mHSLP
c68.35pF
filter
(1A9)Fig. P. 1.11.3: Chebyshev equal ripple band pass
a1.11.3 Low Pass Filter to Band Stop Filter

GO. 1.114 Obtain the expression for inductor and capacitor of band stop filter using low pass fiiter

prototype.
By transforming low pass filter prototype as shown in Fig. 1.11.4(a), a band pass filter can be obtained.
A perfect ideal low pass filter prototype passes all the signals upto cut off frequency = 1 in pass band and stops all

frequency signals having frequency greater than cut off frequency.

The stop band filter attenuate and signals having frequency range of o, and o, and passes the signal out side this

frequency range.
Therefore, the frequency response of stop band filter is completely opposite to that of band pass filter. Thus, the
following frequency substitution transform then low pass filter to stop band filter.

.(1.11.17)

www.weveereeviparoro

02
-facdoise iu ***** *ddnrh*******

(14100) Fig. 1.1144a) : Transformation of low pass filter into stop band filter

This frequency substitution replaces the series inductor of low pass filter prototype with parallel combination of
inductor as L and capacitor C
BS
The series combination of capacitor Cq and the inductorL, are place in the placed of shunt capacitor of low pass filter
prototype
The element values of band stop filter are as follows

...(1.11.18)

BS

(o-0)4 ...(1.11.19))
(02-0) CG .(1.11.20)

1
.(1.11.21)
(o-0,)G
Tech-Neo Publications.. Phere Authors inspire innovation A SACHINSHAH Yenture
RF Design (MU-8" Sem-E&TO) (1-57) RFFilterDesign
Further these elements sealed as follows
R0-)L .(1.11.22)
(

..(1.11.23)
R, (02-0)L
...(1.11.24)
R,
R ..(1.11.25)
(o-0,)CG
The resulting band stop filter is shown in Fig. 1.11.4(b).

LS
RsRo
-
HH
R
v

(IA102Fig. 1.114b): Equal ripple chebyshev stopband fiter

The filter transformation of low pass to high pass, band pass and stop band filter is summarised in Table 1.11.1.

Table 1.11.1: Low pass filter prototype

Fiter transformation without scaling electrical circuit elementrealized afterfrequency mapping

Sr. Components Low passHighpass frequency Bandpass Bandstop
No. prototype mapping0

Series circuit
component
9k 1
(2-0

(1A103) (14104) (1A105)

(1A106)

2. Shunt circuit
E
component
1
2-)C
T (2-)

(1A107) (1A108)

(1A109)

(1A110)

ADOve elements are unscaled

Tech-Neo Publieations . . h e r e Authors inspire innovation A SACHIN SHAH Venture

 RF Design (MU-8 Sem.-E&TC) (1-58) RF Filter Design

Ex. 1.11.4 82 2=C,

Design maximally flat band-stop filter with N = 3. It stops
signal with frequency ranges from 10 MHz to 40 MHz and (ii) Now the elements of band stop filters are calculated
passes the rest of frequencies. Assume load and source
resistances are of 752 each. asfollows
BS 0-00,) L, xR,
Soln. L
Given N=3 75
LS X 10 (40- 10) x
f= 10 MHz , f^= 40 MHz
(2 x 20 x10°)
) Design of maximally flat band stop filter L 0.895 uH
fo = V 10x10°x40x 10 R, (0 - ) L,
=C
*s

= 20 MHz
75 x 2Tx 10° (40-10)
(i) First we will design maximaly flat low pass filter
= 70.73 pF
design for N=3 as follows
R
& 2sin (-) C
75
8o &=l 27T x 10° ( 40 10) x2
8 2 2 sin L= 0.198 uH 0.2uH
-0) C
2 sin)-2 R
8 2 sin 2 x10°(40-10)x2
75x (2 x 20 x 10°)
The element values of maximally flat low pass filter are
= 318.3 pF
Bo&Rs =RL=1
81 1 Ly
iv) The resulting Band-stop filter is shown in Fig. P. 1.11.4.

LS=0.895uH LS 0.895uH
Rs R
752

v S=0.2H c=70.73pF RL 752Q

318.3pF

(1A111)Fig. P. 1.114: Scaled maximally Mat bandstop filter

Tech-Neo Publieations. here Authors inspire innovation A SACHINSHAH Yenture

RF Design (MU-9 Sem-E&TC) (1-59) RF Filter Design

1.12 FILTER IMPLEMENTATION Whereas input impedance of open circuited stub shown

inFig. 1.12.1(b) is given as

(0)
GQ. 1.12.1 What is redundant and non- Zi -jZ,cot (B/) ..(1.12.2)
redundant filter synthesis ? However, the reactance of inductor and capacitor are

exprssed as
The filters designed by using lumped elements, though
..(1.12.3)
they are simple to design, works well at only lower
frequency. ...(1.12.4)
c
The working of such filters rises two problemsat
But from Equations (1.12.2), (1.12.3) and (1.12.4) it is
higherfrequency seen that impedance of stub and lumped elements such
()The lumped elements such as inductors and
as inductor and capacitor are different function with
capacitors generally have only limited range of
respect to frequency.
values and are difficult to use at microwave
frequencies i.e. use frequencies greater them 500 Thus, Z Z and Z+Z (1.12.5)
MHz. In order to satisfy Equation (1.12.5), let's us assum

(i) Secondly the gap between filter components can that at frequency ' c
not be neglected at microwave frequency. In order z Z and Zjn +Z¢
to solve the above problems, Richard's The frequency c is the frequency at which the
transformation is used to replace inductors and functions of lumped element L and C) and
capacitors by transmission line. transmissionline stubs (Z, B) are equal.
Whereas Kuroda's identities are used to separate filter Thus, replacing @ to Oc in Equations (1.12.3) and
elements by using transmission line sections without
(1.12.4)
affecting filter response. This type of design is called joL = j Z ,tan BI
redundant filter synthesis.
In non-redundant filter synthesis, microwave filters can and = jZ, tan (1.12.6)
P
be designed by using both Richard's transformation
and Kuroda's identties in oder to improve the filter Similarly, = -jZ,cotB
response. The non-redundant filter synthesis does not
include inductors and capacitors. -i2, c o , ) (1.12.7)

1.12.1 Richard's Transformation Where V, phase velocity of signal propagating on the

line. Let's set the length of transmission line stub to

; Ga. 1.12.2 How Richard's Transformation is
, where
usefulin RFfilter designing?
**

The input impedance of short circuited stub shown in (1.12.8)

Fig. 1.12.1(a) is given as: Thus Equation (1.12.7) reduces to
Z jZ, tan (B ..(1.12.1)
jagL = jz,tan
Where s is denoted for short circuited stub.

Using Equation (1.12.8)

jocL = j Z ,tan xRx
8xB

(1A112 a) (1A113(b) jocL = jZ tan(A.x4

Fig. 1.12.1a): Short circuited stub and open cireuited stub
Teeh-Neo Publications....here Authors inspire innovation A SACHIN SHAH Venture
 RF Filter Design
AF Design (MU-8 Sem-E&TC) (1-60)
Above two results are Richard's Transformation
joL =jZ,tan () results.
elements inductors
jaL = jz, ..(1.12.9) It is summarized that, the lumped
can be replaced by short
and capacitors of filter design
Similarly for Equations (1.12.6) and (1.12.7) transmission stub
circuited and open circuited
as shown in Figs. 1.12.2(a)
and (b).
respectively
-iZg co
GO. 1.12.3 Why does filter response using
match with
-j,cox Richard's Transformation
Fiterdesign using lumped element ?
-j2,ca(Ax The Taylor series approximation for tan o and cot o,
when o is small, is expressed as:
.(1.12.10)
tan o= and cot
Thus from Equations (1.12.9) and (1.12.10) it can be .(1.12.11)
for<«l,
concluded that an inductor can be replaced by short Where is in radian.
cireuited transmission stub of length 1 The Richard's Transformation of shorted stub at
frequency o' has input impedance expressed as :
frequency with characteristic impedance
Zo L. It is shownin Fig. 1.12.2(a).
jZtan(p) ..(1.12.12)

S.C. But A
Z c (Short
circuited)
in jZ,tan
I= aJ8
(1A114a) (1A115b)
Fig. 1.12.2: Short circuited stub
=
jZ%lan. 46
Likewise a capacitor can be replaced with an open Thus using Pe
circuited stub with same characteristic impedance of
capacitor C' i.e. Zo = Thus, the characteristic

impedance of open circuited transmission stub is z jZo tan

Zo=Z and it is shown in Fig. 1.12.206). The When <<a,Z can be approximated as

length of transmission open circuited stub is I at = jo-L) tan c

frequency jo,L )
Using Equation (1.12.11)
O.C.
Z drcut)
(Open z =
joLwheno<<o, .1.12.13)
8 Comparing this with impedance of inductor,
Z joL .(1.12.14)
(IA116 Open circuited (IA117Kd) Richard's Thus, from Equations (1.12.13) and
transmission stub Transformation Stub (1.12.14)
Fig. 1.12.2 z =
Z, where1
i.e. Relatively close to 1.
Tech-Neo Publications. bere Autbors inspire innovation
A SACHIN SHAH Venture
RF Design (MU-8" Sem-E&TC) (1-61) RFFilter Design
It is found that Richard's Transformation of shorted 1.12.2 Kuroda's Identities
stub has nearly same input impedance to that of
inductor for all frequencies less than o, (i.e. all
Ga. 1.124 Explain how Kuroda's ldentities are
frequencies less than o of low pass filter pass band)
usefal in RF filter design
Similarly Richard's transformation of open circuited
stub has an input impedance as Kuroda's identities are very useful for implementing
Richard's transformation. These identities provide an
() equivalent circuit for two port network, where they
have closely the same S-parameters, Z-parameters, Y-
parameters and ABCD or transmission parameters.
It means that Kuroda's identities replace a two port
...UsingEquation (1.12.11)| network with its equivalent circuit without changing
Z (H) it's scattering matrix.
When <<O .(1.12.15) Kuroda's identities are useful to :
When Equation (1.12.15) compared with input Separate transmission line stubs physically.
impedance ofcapacitor impedance as: (i) Replace series stubs into shunt stub or vice versa.
(ii) Provide characteristics into more impedance
.(1.12.16)
realizable ones.
From Equations (1.12.15) and (1.12.16), it is found that
(iv) Faciliate the design of distributed elements.
both Equations have approximately same impedance
Fig. 1.12.3 presents four Kuroda's identities which are
for all frequencies @ < < o, in pass band of low pass
used for transformation of networks. These transformed
filter.
network use Unit Elements (UE) having a length of a
Thus, the filter designed using Richard's
Transformation has same frequency response as filter transmission line stub a cut off frequency with
designed using lumped elements.
specified characteristic impedance.

n Z1
U
U.E. U.E
Z1 uE nz

(1A118) (1A119) (1A120) (1A121)

(a) (b)

1:n2
n:1
H
2 nz

(1A12) (1A123) (1A124) (1A125)

(c) (d)

Fig.112.3:FourKuroda'sldentities=14
Tech-Neo Publications ...Where Authors inspire innovation SACHIN SHAH Venture
 RF Filter Design
RF Design (MU-8 Sem.-E&TC)
-***n*^***--
(1-62)

GO. 1.12.5 Prove the Kuroda's 1st identity

The Kuroda's 1s identity is

(1A126) (a) (1A127) (b)

Z n-1
O.C.
Shunt
stub
Unit element Unit element

(IA120)(©) (1A129) (d)

Fig. 1.12.4 : Equivalent two part networks using Kuroda's identity

(1) The above two identities can be redraw by using

The ABCD parameters of Fig. 1.12.5(a) are expressed
as
Richard's transformation ie. capacitor is replaced by
open circuited shunt stub and inductor is replaced by

short circuited series stub and shown in Fig. 1.12.4.

A V,=0 BTv,=0 ..(1.12.17)
The equivalent of two port networks shown in
C = D =

Fig. 1.12.4 can be shown equivalent to each other by

A voltage source of V, is connected at port 1 and port 2
matching ABCD parameter matrices of both sides.
is short circuited. Let V is incident voltage at portl,
Here ABCD matrix of transmission line of length '
then at port 2, it will be Vine. Since reflection
with characteristic impedance Z is obtained as
coefficient of short circuited terminal is- 1, so
follows: It is shown in Fig. 1.12.5(a). reflected voltage at this port will be 180° out of phase.
-2y
Hence reflected voltage arriving at port 2 is -Vine
Thus V= Vin-Vine
V2 = 0
V Z.Y V2

Unit element
and I
(IA130)Fig. 1.12.5(a): Unit element of ' transmission length line

Teeh-Neo Publications . here Authors inspire innovation ASACHIN SHAH Venture

RF Design (MU-8" Sem.-E&TC) (1-63) RFFiterDesign
From Equation (1.12.17), Thus, transmission matrix of transmission line having
length T' is obtained as follows:
B 21-*) cosh (y Z sinh(y)1
cosh (y
Ze-e")
2e As transmission line is assumed as to be lossless, thus

Z sinh (YD Y jB bence above matrix becomes:

Using cosh jx = cos x and sinh jx =j sin x, tanh jx =j
and D tan x, coth jx = -j cotx

-uV,=0 :osh Z sin B!

cos B
D
2 in iz, Sin 7
D +e) le sin
L Z cos B
1

D + 1 jZ tan BI
=
cos ZtanB 1
D cosh (Y4)
Now substituting tan Bl and using
Now, port 2 is open cireuited and voltage source V, is
still connected at port 1. When Vi, is incident voltage at cos 8-
port 1, then Vin e is arriving at port 2. As reflection
coefficient of open cireuit port is + 1, reflected voltage .(1.12.18)
at port 2is equal to incident voltage Vine
Thus, Now, the ABCD parameters of open circuited shunt
V,= Vin +Vi e stub as
shown in first circuit and using Equation
= Vin (1 +e") (1.12.2), has characteristic impedance-jZ, cot B
As ABCD matrix of Fig. 1.12.5(6) is obtained as
V =2V
follows:
-1-*) Here Vis connected to port 1 and port 2 is short
circuited.
and

From Equation (1.12.17),

V2
A
VL-2"
and C (1A13())Fig. 1.12.5(b)
1,=0 2Ze" ThusL=- and V, =0 V
Using Equation (1.12.17).
2Ze
and B - 02
sinh (y)

Tech-Neo Publications . here Authors inspire innovation

ASACHIN SBAH Venture
 AF Filter Design
(1-64)
RF Design(MU-8 Sem-E&TC)_
and
D v,=0 .(1.12.21)
while V, is still
Now, consider port 2 is open circuited Now, ABCD parameters of short circuited series stub
connected at port 1.
as obtained as follows whose impedancetan B
Thus, V, = V, and I, =
(using Equation 1.12.2)
Using Equation (1.12.17)

A V=0 = 1 m-3
4 2

(2)Fig. 1.12.5(c)
C
V L=0 As shown in Fig. 1.12.5(c), V, is connected at port
and port 2 is short circuited.

Thus I=-h and V, =I Z volt

Using Equation (1.12.17)
Substituting 2 = tan BI

B = Z2

le
and -
0 1
2 is open circuited and V is still
..(1.12.19) Similarly port
connected to port 1. So that I, = 0, V, = V1 and I =0.

From left hand side of Fig. 1.12.4, these two elements Thus, using Equation (1.12.17),
are cascade. Thus using Equation (1.12.18) and
(1.12.19). A Vl0
nd c VlL,-0 = 0
1 j2Z
..(1.12.22)
As these two elements are cascaded and using
j2Z Equations (1.12.21) and (1.12.22).

VI-ia() 1 - | fA B7 1
n
..(1.12.20)) 1

Similarly from second part of Fig. 1.12.4, the ABCD

matrix of unit element is obtained as by considering

1
transmission line of length "'.
L 0
:1a-
...(1.12.23)

Tech-Neo Publications . here Authorm inspire innovation A SACHIN.SHAH Venturet

 RF Filter Design
RF Design(MU-g Sem-E&TC) (1-65)
b = 0.839
By using n=I+7 the Equation (1.12.23)
1 j2Z, (0.2986+sin(
..(1.12.24) = 0.0891 + (0.8660)
( ) 1-#2 b , = 0.8390
So from Equations (1.12.20) and (1.12.24), Kuroda's 2a 2x0. 3.3489
g = 0.2986
first identity is proved. (v

Bz
4x0.5 X=0.7118
Ex. 1.12.1 3.3489 x 0.839
Design a low pass f+lter to fabricate using microstrip lines 4x1x0.5 3489
for following specifications:
83 0.7118 +0.8390
4 The normalized equal ripple low pass filter prototype
GHz, N=3, R,=R,=50,
elementvalues are:
Equal ripple =3dB
81 3.3489 =L
Soln.: 82 0.7118 = C
Given f=4 GHz, N=3 (Order offilter),
R = 12
R 50 2, G, =3 dB
83 3.3489 =L
T o design equal ripple low pass filter prototype
following design equations are used: Fig. P. 1.12.1 presents normalized equal ripple low pass
filter prototype with lumped element.

1Le. a - sin I L3.3489 Lg 3.3489

2NJ Ro 12
o-
= sin(=05
C2 =0.7118 12
si(

and , sin(05s (IA131)Fig. P.1.12.1:Low pass prototype with lumped element

(2) Apply Richard's Transformation to Fig. P. 1.12.1 ie.

(ü) ln coth1737)J replace inductor by short circuit series stub and replace
capacitor by open circuit shunt stub.
tanh17-37=0.17101
Z0 3.3489 Z 3.3489
coth(3
17.37 0.17101
5.8476
coth 17.37, =
12

I n [5.8476] =1.7660
I N8 at o=1
X
= 2N)sinh
Z 07118B 1.405
X =sinh o=0.20R6
(IA132)Fig. P. 1.12.1(a): Low passfilter prototype with

iv) Richard's transformation to replace inductor with short

circuited series stub and capacitor with open circuited shunt
stub
b, 0.2986+sin(
= 0.0891+ (0.8660)

Tech-Neo Publications ..Where Authorsinspire innovation SACHIN SHAH Venture

 RF Design (MU-8" Sem.-E&T (1-66) RF Filter Design
(3) Since Fig. P. 1.12.1(a) cannot be fabricated so unit 1 Z 4.3500 Z 4.350
elements are added. After adding unit elements at ends
of filter.

o 3.3489
Z0 3.3489 Upitelement
1 1.405
Z 1.299 Z Z 1.299
12 (IA13Fig. P. 1.12.1(e): Appling second Kuroda's identity

Unit elenment Z1 (5) Apply frequency and impedance scaling. All

Z1.405 normalized characteristic impedances multiplied by
(1A133)Fig. P. 1.12.10b) : Adding unit element at ends of filter 50 2 and stub lengths are selected as / = at
50 Zo 50x 4.350 Z0 50x 4.350
(4) In order to convert Fig. P. 1.12.1(b) into open circuited w
217.5 217.5
shunt stub, we apply Kuroda's second identity to
convert series stub into shunt stub. 500
From Fig. P. 1.12.1b) and Kuroda's 2nd identity.
Z= 3.3489 Z 1.299 x 50
Z 1.299 x 50 Z 1.405 x 50
= 70.252
64.92
64.92
(IA135Fig. P. 1.12.1(d) : After impedance and frequency scaling
Now,
(6) Microstrip fabrication of Fig. P. 1.12.1(d).
n 1+}=1+3.3489
=

502 502
n .299 217.52 217.52
From nd column Kuroda's
of identity
(Refer Fig. 1.12.3)
64.92 70.32 64.92
Value of capacitor
(IA136
Fig. 1.12.1(e) : Microstrip fabrication of final filter.

1.299 x1 1.12.3 Impedance and Admittance

But it has to act Inverters
Open circuited shunt stub so the suceptance of shunt
It is always easy to fabricate a filter either with series
stub =(n Z) = 1.299
or shunt stub of transmission line.
The characteristic impedance of unit element is nZ Along with Kuroda's identity, impedance (K) or
= 1.299x 3.3489
admittance () inverters are also used for conversion of
4.350 series to shunt stub or vice versa.
Second element is already open circuited shunt stub so Impedance (K) or admittance ) inverters are used for
no need to convert. band pass or band stop filters with small (< 10%)
Now for third element i.e. short circuit series stub, this bandwidths.
will be converted into open circuited shunt stub as above.
Thus Fig. P. 1.12.1(¢) presents the low pass filters
prototype using only open circuited shunt stub.
Teeh-Neo Publications . here Authors inspire innovation A SACHIN SHAH Venture
 RF Design(MU-8h Sem.-E&TC) (1-67) RF Filter Design

1.13 MIXED SOLVED PROBLEMS 3. The element in series and shunt arms of m-derived
T section of low pass filter are given as:
Ex. 1.13.1
0.274 x318.30 x10=43.60 nH
Design composite low pass filter by image parameter
method with R,=5 02 f,= 50 MHz and f. = 52 MHz mC = 0.274 x 127.32 x 10
Soln. mC = 34.88 p.F.
Given R=5 5 02 f, = 50 MHz and f=52 MHz
To design composite low pass filter following design
equations are used.
(L- (1-(0.214L \x 318.30x 10
4x0.274

268.61 nH
1. Design of constant -KT section oflow passfilter
The m-derived T-section of low pass filter is shown in
L 50 Fig. P. 1.13.1(a).
TX 50 x 10
m43.60nH 43.60nH
L = 318.30 nH thus= 159.15 nH

C = mC 34.88pF
50x 10"x50
C = 127.32 pF
The constant K-T section low pass filter is shown in
( L 268.61nH
4m

Fig. P. 1.13.1.

I2 159.15nH U2 = 159.15nH (1A138)Fig. P. 1.13.1(a): m-derive low pass filter

Design of matching sections
The -matching sections are designed for m = 0.6 for
C 127.32pF
designing terminating half sections.
The series and shunt arms of terminating half sections
(1A137)Fig. P. 1.13.1: Constant K-T section of LPF areas follows:
mL 0.6) x 318.30 x 10 = 95.49 nH
2. Design ofm-derived T-section LPF as follows:
The frequency of infinite attenuation is f. = 52 MHz 0.6 x 127.32 x 10
38.19 pF

m--()-y-9 1-0.6Lx 318.30x 10

m = 0.274
= 169.76 nH 170 nH

The terminating m-derived half sections are shown in Fig. P. 1.13.1(b).

95.49nH m=95.49nH

To L=170nH filter
To Flr 17 To
Load
source
m38.1pF 2 38.1pF

(1A139) (14140)
Fig. P. 1.13.1(b) : m-derived matching sections

A SACHINSHAH Venture
Tech-Neo Publications. "here Authors inspire innovation
 RE Design (MU-8h Sem.-E&TC) (1-68)
RF Filter Design
5. Design of composite filter

The composite filter is shown in Fig. P. 1.13.1(c).

159.15nH 159.15nH 43.60nH 43.60nH, 95.9nH
95.94nH

Rs 34.88pF 170nH
502
170nH
127.32pF: R 502
Vs 38.1pF 268.61nH 38.1pF
Terminating T section m-derived Terminating9
half section constant K T section halfat msection
= 0.6
at m =0.6 section

(IA141)Fig. P. 1.13.1(c)

The simplified composite filter is shown in Fig. P. 1.13.1(d).

225nH 202nH 139nH

Ro 170nH 34pF 170nH

127pF
502
38.1 268nH
38pF

(1A142) Fig.P. 1.13.1(d)

Ex.1.13.2 2C 42.4pF 2C=42.4pF

A composite high pass filter is to be designed with
R, 75 2, f = 50 MHz and f. = 48 MHz.
L 119.3nH
Soln.:
Given Ro= 752 f50 MHz
and f 48 MHz (1A149)Fig. P. 1.13.2: Tsection of constant K high pass filter

A composite high pass is designed by using following 2. Design of m-derived high pass filter.
design equations.
1. Design of constant - K-T section of high pass filter. m

75
L
4Tx 50x 10
19.36 nH Thus the eleme
-y-( = 0.28

of series of shunt arms of m-derived

C
T section of high pass filter are given as follows:
4 7 T f,R 4Tx 50 x 10° x 75 2x21.2x10- = 151.4 pF
0.28
21.2 pF
2C = 42.4 pF L 119.36x10 426.2 nH
m 0.28
Thus T-section of constant K of High pass filter is
(1-m*C (-028.x21.2 x10
shown in Fig. P. 1.13.2.
= 25.76 pF

Tech-Neo Publications.WhereAuthors inspire ianovation

A SACHINSHAH Venture
 RF Design(MU-8 Sem.-E&TC) (1-69) RFFilter Design
Thus, m-derived high pass filter of T section is shown 3. Design ofm-derived matching sections with m =0.6.
in Fig. P. 1.13.2(a).
The elements of series and shunt arms of matching
2C sections are obtained as follows
151.4pF 2151.4pF
= 70.66 pF

Um 426.2nH 2x119.36X 1--397.8 nH

0.6
1 m
mC25.76pF
(20c
1-m (-2-0.6x21.2x
x (0.6)
x21.2x 10 10!12

(A14Fig. P. 1.13.2a): m-derived high pass filter (T-types) = 39.8 pF

Thus, m-derived terminating half sections are shown in P.

Fig. 1.13.2(6).
20 151.4pF 2C 151.4pF
m

397.8nh 397.8nH

+c39.8pF m

ZOT OTtm
(14145)Fig. P. 1.13.2(b) :m-derived matching section
Hence the composite high pass filter (T-section) is shown in Fig. P. 1.13.2(0)
2C/m 2C 2c 2C/m 2C/m 2C/m
70.66pF 42.4pF 42.4pF 151.4pF 151.4pF 70.66pF

2Um Um 2Um
397.8nH 426.2nH3 397.6nH

39.8pF 119.3nH 25.76pF 39.8pF

1-m |1Amc
m
Matching Constant K m-derived Matching
section T section T section section

(1A146Fig. P. 1.13.2(¢): Complete composite high pass filter

The simplified composite filter is shown in Fig. P. 1.13,2(d). Here 2C

capacitors and 2C are in series, so simplified
value is
2C 2C

2C
m
1514x 424-33.12
151.4 +42.4
pF
33.12 pF
m+2C

Tech-Neo Publieations.. here Authors inspire innovalion A SACHIN SHAH Venture

 RF Filter Design
RF Design (MU-8 Sem.-E&TC) (1-7

25pF 33.12pF 48.17pF

R
750 397.8nH 119.3nH 426nH397.8nH R 750

25.76pF 39.8pF
39.6pF

(1A147) Fig.P.1.13.2/d) Simplified composite high passfilter (T-section)

:

Ex. 1.13.3 2sin io=0.618

Design a low pass maximally flat filter with pass band range
0 to 3 GHz and an attenuation 20 dB at 5 GHz. The & 2 sin0=1.618
characteristic impedance is 752.
Soln.: 8 2sin 10)=2
Given: f=3 GHz, f 5 MHz, PLR=20 dB g42 sin 1.618

) Design of maximally flat low pass filter

Filter order can be calculate as
2 si(10)-06018
ii) The unscaled elements of maximally flat low pass
PLR10logto1+K filterare as follows:
As N=5, first element is capacitor
20
10log10 1+ .K=1
BoBR, =R;
2 log1 ) 81 0.618 = C,

Taking antilog on both sides 82 1.618 =L2

2N
8 2 C
2N 84 1.618 = L

99 = 8s 0.618 = Cs
Taking log on both sides
iv) Applying the impedance and frequency scaling the
loB099 2Nlog elements of maximally flat low pass filter are:
1.9956 =2N (0.2218)
C Ra 7 5 x0.618
21x3x 10 0.437 pF
2N 1.9956
0.2218
2N = 8.9972
RB275x1.618
1096.45nH
2Tx3x
N = 4.498 5
B 2 1.41 pF
Here order of filter is 5. R, 75 x 2n x 3x 10
(i) For maximally lat low pass filter, element values L4 R, 75x 1.618_.= 6.44 nH
are calculated using Equation 2mx3x 10
&s 0.618
B=2 sin 2N R , 75 x21 x3x 10
BoB=1 0.437 pF
RL = R=Rs = 75 2

Teeh-Neo Publications.. Where Authors inspire innovation

A SACHIN SHAH Venture
 AF Design (MU-8h Sem.-E&TC) (1-71) RF Filter Design
(v)The low pass filter maximally flat with scaled elements is shown in Fig. P. 1.13.3.

L6.43nH L4=6.44nH
750
C 0.437pFC 1.41pF Cs 0.437pF R;=752

(14148) Fig. P. 1.13.3 : Maximally flat low pass filter with scaled element

Ex.1.13.4 X = 0.1775

Design high pass filter with 5 lumped elements with a 3 dB b, =x +sin

iv)
equal ripple response cut off frequency of 1 GHz and
impedance of S0 2. What is resulting attenuation at 0.6 GHz b,= (0.1775)+sin
=
()
0.0315 +(0.5877)* = 0.0315 +0.3453
Soln. b = 0.3768

Given N=5, f= 1 GHz,

f=0.6 GHz, G,=3 dB
b (0.1775)+ sin?(
= 0.0315+ (0.9510) = 0.0315 +0.9044
) To design high pass Chebyshev equal ripple filter, b2=0.9359
first low pass equal ripple prototype is designed as
follows: b,(0.1775+sin' (
For, N= 5, 0.0315+ (0.9510) = 0.0315 +0.9044
bg.0.9359
sin I
b = (0.1775)+sin?(:
asin1=0.3090 0.0315+ (0.5877)
=
=0.0315 +0.3453
b=0.3768
az sin = 0.8090
a3 sin
(v) 81= 2a 2x0.3090
0.1775
81 3.4816
4 ax 4xaxa
a4 = s i n 0 ) =0.8090 (vi)
Dx-1 Bx-1
sin (=0.3090 g =
XO.3090x 0.809
0.3768 x 3.4816
e G,
In coth1737)
(i) 82 0.7621
4x ax a3
Now, tanh1737 = 0.1710 S3 4x0.8090x1
0.9359 x0.7621
83 4.5375
coth
17.37) =
0.1710=5.8479 4x aX a4
e = In [5.8479]= 1.766 84

(ii) X = sinh2N) B4
4x1x0.8090
0.9359 x 4.5375
X = sinh766 B4 0.7620

Tech-Neo Publieations.. here Authors inspire innovation

A SACHIN SHAH Venture
 RF Filter Design
RF Design (MU-8h Sem.-E&TC) (1-72)

8s
4 x aX as Ex. 1.13.5
A band stop three lumped element filter has 0.5 dB equal
centered at 3 GHz and
4x0.8090 x 0.3090 ripple response a band width of 10%
0.3768 x 0.7620 filter of 3 lumped
impedance of 752. Design a band stop
& 3.4825 elements.
(vii) Thus element values of low pass prototype equal Soln.
ripple filter are: Given N=3, Ro 752, G, =0.5,
81 3.4816 =L
f 3 GHz and (2 -0.1
82 0.7621 = C2

8s4.5375 =Ly ) In order to design band stop filter with 0.5 equal
ripple, first low pass filter prototype is designed by
84 0.7620 =C4
85 3.4825 =L_ using following equations:
Ro = R=Rs = 1 Design of low pass filter prototype:
(viii) Now converting low pass prototype into equal =sin2N
ripple high pass filter as follows:
By applying impedance and frequency scaling the sin=05
elements of high pass filter are obtained as follows:
1 sin(-1
Ro 50 x2x1x10'x3.4816
= 0.914 pF sin )=05
L Rg 50
C 2t x 1x 10x0.76211
LHP 10.44 nH First, tanh 0.5
17.37 0.02877
HP
3 R,oL 50 x 2nx 1 x 10 x 4.5375 coth7.37=34.7495
C=0.7015 pF using cot 6 tan 0=1
50 E 3.5481
L C2 21x 1x10x0.7620
,HP i) X = sinhN
= 10.44 nH
L
C Ros 50x27x 1x 10x 3.4825 X =
sinh( =0.62642
HP=0.914 pF (iv) b X+sin'(
ix) The equal ripple chebyshev high pass filter is shown
in Fig. P. 1.13.4.
b, (0.62642ý+ sin' ()
= 0.39240+ (0.8660
HP
RsR 502 C C3 C = 0.31240 +0.74995

0.914pF 0.7015pF 0.914pF b 1.1423

RL 502
[ 10.44nH 10.44nH b = (0.62642)+ sin?

= 0.39240+ (0.8660
(1A149) Fig. P. 1.13.4
ba1.1423
Tech-Neo Publications . here Authors inspire innovalion
ASACHIN SHAH Yenture
RF Design(MU-8h Sem-E&TC) (1-73) RF Filter Design

BS 5X0.1 x 1.5963
() 21n x 3x 10

81
2x0.5
0.62642 1.5963 = 0.635 nH

(vi) 8 4xaX RoC0-0) L, o,x Rl(0-0) x L

b-1 X &%-1
82
4 xax a 27tx 3x 10 x 75 x 0.5 x 1.5963

c= 0.88 pF
82
4x0.5x 1
1.1423 x 1.5963
82 1.09681
c 0.1x 1.09681
3
4XaXa 75 x 2 x 3x 10

4x1x0.5
C= 0.077 pF
83 RX
1.1423 x 1.09681 Ro
(0-0,) C2 (-o,) xo,xC
83 1.5963
75
Thus element values of low pass filter prototype with 0.1 x 27t x 3x 10x 1.09681
0.5 equal ripple filter arc
L 36.27 nH
8 1.5963 =L
BS (-0,)La75 x0.1x1.5963
82= 1.09681 =C2 2Tx 3x 10
83 1.5963 =L BS =
nH
0.635

&41.00=R.=Ro
C (0)Ca
Ro(0-) L R,
(vi) =

C C1
(vii) The resulting band stop filter is shown in Fig. P.1.13.5.
LS =0.635nH L 0.635nH
R 752

36.27nH C8S=0.88pF RL
v c=0.88pF 750
-0.07pF

(1A150)Fig. P. 1.13.5: Band stop filter with equal ripple 0.5 dB

(1) To design maximally flat low pass filter of 3 order

Ex. 1.13.6
using series stubs following design equations are used.
Design a low pass third order maximally fiat filter using For maximally flat low pass filter the element values
series stub. The cut off frequency is 6 GHz and
only are as follows,
impedance is 50 2.
& 2 sin K)E
ZN
Soln.
f=6 GHz, Ro= 50 Bo 841
Given N=3,
A SACHIN SHAH Veture
Tech-Neo Publications8. . here Autbors inspire innovation
 RF Filter Design
R F Design (MU-8h Sem.-E&TC) (1-74)
(4) Since Fig. P. 1.13.6(a) can not be fabricated without
2 sin=1 unit elements. So unit elements are added at both ends,

8 2 sin -2 Zo1
8 2 sin Ro 1
1
(2) Thus element of maximally flat low pass filter arc as

follows
1
&o &4R =R, =

81 1=C, (1A159) Fig. P. 1.13.6(b) : Richard

transformation

82 2L2 To convert open circuited shunt stub, to series stub

(5)
83 1=C Kuroda's 1s identity is used two times.

Thus maximally flat low pass filter is shown in Here

Fig. P. 1.13.6. Z 1
L22
Z2 1
v C1 Cg1 12 117
n = 2
identify two times
(6) Apply first Kuroda's
1A151) Fig. P. 1.13.6: Maximally lat low pass filter
(3) Applying Richard's transformation to Fig. P. 1.13.6
i.e. replace inductor by short circuit series stub and Ro
1 1/2 1/2 10
replace capacitor by open circuited shunt stub. It is
shown in Fig. P. 1.13.6(a).

Ro Fig.P. 1.13.6(c): Applying Kuroda's

2 (7) Scaling to 50 2

v 1
502 250 | 1002 2

252 252 500

1
(1A152) Fig. P. 1.13.6(a) : Applying Richard's transformation
Fig. P. 1.13.6(d)

Chapter Ends..
O00
RF Design (MU-8" Sem-E&TC) (2-69) Amplifier Design
The center and radius of input and
circles are
output stability Soln.
The test is expressed as:
For output stability circle
The centeris 1-IS
IS-AS+Is, S
CL S-As*
IS- AI For device A

0.4525°-(0.1174-50 )(0.341170) ) K-A test

(0.45)-(0.117) A =S S22-Siz S21
CL (0.4831-28)
0.1888
0.483128 A= (0.34 -170°) (0.452-25°)-(0.06 70°) (4.3Z80°)
0.1888
CL = 2.55 L28° A =0.665 L-1.7°
The radius of output stability circle is 1-1S-1s,+1ar
K
(0.0670) (4.3 80°) 2IS2 Sa1
R IS-1A (045)-(0.117) K = 1.36> 1
Since A < l and K > 1, device A is unconditionally
0.258150
R 0.1888 stable.

RL 0.258 1.37
0.1888 i) -test
For input stability circle 1-(0.34
NOW(O.452-25) (0.665 2-1.7) (0.34 2 170°) 1
The centeris
1.94

Cs (S-AS
1S,- 1A
Since u>1 and 1Al <l
So the device A is unconditionally stable.
C. 0,342-170)-(0.1172-50°) (0.45/25°)
(0.34) - (0.117)
For device B
C=5843-174°)* ) K-Atest
0.1019
A S S2-S21 Si2
Cs 0.3843174
0.1019
3,77 L174° A = 1.366 2 - 14.5°

Al = 1.36>1
The radius for input stability circle is
-IS
R IS S, +|S\ S,l
O.06 270 4.3280) 0.283
(0.34) -(0.117) Since 1Al >1, device is potentially stable. No need to
R = 5 8150° 0.258 determine K test.
Rs .1019 0.1019 F o r device C

Ex. 2.12.8 ) K-Atest

devices
Use test to determine. Which of the following are

unconditionally stable and of those which has the greatest

A
S Sn-Su S2
A = 0.393< 164.95
stability IAl =
0.39 <1
Device S S12 S
0.34-170 0.06 270°|4.3 280° 0.45 -25° K = 1-ISF-ISf+1af
A 2IS12 S21
70° 5.0 290° 0.51 260 K = 1.24>1
8 3 0.75 2-60° 0.2
60° 2.4 50 0.70 2-65 Since IAl<1, K>1
device unconditionally stable.
C 0.65 L-140° 0.04 2
ech-Neo Publications here Authors inspire innovation ASACHIN SHAH Venture
RF Design (MU-9h Sem.-E&TC (2-70) Amplifier Design
(i) For H-test

1-1S
S S,,A+S S,l
1.056

Since A <1, K>1 and u>

So the device C is unconditionally stable.
Above u test summrise the results in table as,

Device S S2 S S Stability
0.34 2-170° 0.06 70°|4.3 280 0.45 2-25°|1.194 unconditionally Stability
B 0.752-60° 0.270° 5.090° 0.5 290° Potentially Stability
C 0.65 2-140° 0.04260° |2.42 50° 2.4 2501.056 unconditionally Stability|
From above values of it can be said that greater the values of u indicates that the device has greater
degree of stability
So the device A has best stability
Ex. 2.12.9 Ex. 2.12.10
Prove that for S12 = 0 (unilateral device), u test implies that Design an amplifier for maximum gain at 5.0 GHz with a
IS,<1 and I S< Ifor unconditional stability GaAs FET having following S-parameters (Z% 502 );
S 0.65 4-140°, S21 = 2.4 2 50°, S =0.04 L60°,
Soln.: S 0.70 2 - 65°. Design matching sections using open
The test is expressed as
circuited shunt stubs
1-IS
IS-sal+IS1IS2 Sid Soln.
IfS12= 0, as unilateral then A reduces to Given GaAs FET transistor
A = S1S2n-S,2S21 S 0.65 2-140°
A = SSn . . .
S21 2.4 2 50°
S120

1-1IS S20.04 260

So IS2-Sa S22 0.70 2-65°

1-1S () Stability test

IS-S S,SS
First we will determine the stability of transistor using
Putting S S1S and A=S,S22
K and A test
1-IS
IS-Is, S A
SS22 - S1S21
Since the denominator is positive due to mod sign, so A = (0.65 2-140)° (0.70 2-65°)
s
positive, then the numerator must be also positive thus -

(0.04 260°) (2.4 2 50°

IS<1. Then the above u reduces to
A = 0.393 2 165

Sso lAl = 0.393

So IS< 1 Now K = S-S,+1A

Hence it is proved that u implies IS,<1 and IS,< for
2 ISS2

unconditionally stable devices K = 0.65)-(0.70)+(0.3932

2 1(0.04 60°) (2.4 L 50°)N
Tech-Neo Publications bere Authors inspire innovation
A SACHIN SHAH Venture