Design of Narrow Band Filters Part 2

Similar documents
Design of Narrow Band Filters Part 1

Homework Assignment 11

Speaker: Arthur Williams Chief Scientist Telebyte Inc. Thursday November 20 th 2008 INTRODUCTION TO ACTIVE AND PASSIVE ANALOG

Filters and Tuned Amplifiers

EE221 Circuits II. Chapter 14 Frequency Response

EE-202 Exam III April 13, 2006

Lecture 14 Date:

OPERATIONAL AMPLIFIER APPLICATIONS

Sophomore Physics Laboratory (PH005/105)

EE221 Circuits II. Chapter 14 Frequency Response

Electronic Circuits EE359A

ECE 391 supplemental notes - #11. Adding a Lumped Series Element

REACTANCE. By: Enzo Paterno Date: 03/2013

Single Phase Parallel AC Circuits

Electronic Circuits EE359A

EE-202 Exam III April 13, 2015

MODULE-4 RESONANCE CIRCUITS

Exercise s = 1. cos 60 ± j sin 60 = 0.5 ± j 3/2. = s 2 + s + 1. (s + 1)(s 2 + s + 1) T(jω) = (1 + ω2 )(1 ω 2 ) 2 + ω 2 (1 + ω 2 )

Name. Section. Short Answer Questions. 1. (20 Pts) 2. (10 Pts) 3. (5 Pts) 4. (10 Pts) 5. (10 Pts) Regular Questions. 6. (25 Pts) 7.

University of Illinois at Chicago Spring ECE 412 Introduction to Filter Synthesis Homework #4 Solutions

CHAPTER 5 DC AND AC BRIDGE

ECE3050 Assignment 7

Sinusoidal Response of RLC Circuits

Solutions to Problems in Chapter 6

AC Circuits Homework Set

Network Parameters of a Two-Port Filter

Two-Port Networks Admittance Parameters CHAPTER16 THE LEARNING GOALS FOR THIS CHAPTER ARE THAT STUDENTS SHOULD BE ABLE TO:

How to measure complex impedance at high frequencies where phase measurement is unreliable.

EE221 - Practice for the Midterm Exam

Physics 142 AC Circuits Page 1. AC Circuits. I ve had a perfectly lovely evening but this wasn t it. Groucho Marx

DC and AC Impedance of Reactive Elements

RLC Circuit (3) We can then write the differential equation for charge on the capacitor. The solution of this differential equation is

LECTURE 21: Butterworh & Chebeyshev BP Filters. Part 1: Series and Parallel RLC Circuits On NOT Again

EE-202 Exam III April 10, 2008

A R E S O N A N T D U M M Y L O A D

Lecture 37: Frequency response. Context

Conventional Paper-I Part A. 1. (a) Define intrinsic wave impedance for a medium and derive the equation for intrinsic vy

AC Circuits. The Capacitor

Electronic Circuits Summary

ANTENNAS and MICROWAVES ENGINEERING (650427)

Resonant Matching Networks

Conventional Paper I-2010

Lecture 7, ATIK. Continuous-time filters 2 Discrete-time filters

Design and Realization of Fractional Low-Pass Filter of 1.5 Order Using a Single Operational Transresistance Amplifier

Operational Amplifiers

2 Signal Frequency and Impedances First Order Filter Circuits Resonant and Second Order Filter Circuits... 13

Digital Signal Processing

EE313 Fall 2013 Exam #1 (100 pts) Thursday, September 26, 2013 Name. 1) [6 pts] Convert the following time-domain circuit to the RMS Phasor Domain.

Case Study: Parallel Coupled- Line Combline Filter

CE/CS Amplifier Response at High Frequencies

Time Varying Circuit Analysis

ECE 202 Fall 2013 Final Exam

First and Second Order Circuits. Claudio Talarico, Gonzaga University Spring 2015

Lecture 11 - AC Power

Sinusoidal Steady State Analysis (AC Analysis) Part II

ON THE USE OF GEGENBAUER PROTOTYPES IN THE SYNTHESIS OF WAVEGUIDE FILTERS

H(s) = 2(s+10)(s+100) (s+1)(s+1000)

mywbut.com Lesson 16 Solution of Current in AC Parallel and Seriesparallel

To investigate further the series LCR circuit, especially around the point of minimum impedance. 1 Electricity & Electronics Constructor EEC470

Lecture 090 Multiple Stage Frequency Response - I (1/17/02) Page 090-1

TC 412 Microwave Communications. Lecture 6 Transmission lines problems and microstrip lines

Module 4. Single-phase AC circuits. Version 2 EE IIT, Kharagpur

ECE 410 DIGITAL SIGNAL PROCESSING D. Munson University of Illinois Chapter 12

6.17 The Lossy Voice-Coil Inductance

ECE-343 Test 1: Feb 10, :00-8:00pm, Closed Book. Name : SOLUTION

Lecture 9 Time Domain vs. Frequency Domain

Refinements to Incremental Transistor Model

Physics 4B Chapter 31: Electromagnetic Oscillations and Alternating Current

8. Active Filters - 2. Electronic Circuits. Prof. Dr. Qiuting Huang Integrated Systems Laboratory

Physics 405/505 Digital Electronics Techniques. University of Arizona Spring 2006 Prof. Erich W. Varnes

Use of a Notch Filter in a Tuned Mode for LISA.

DATA SHEET. HEF4067B MSI 16-channel analogue multiplexer/demultiplexer. For a complete data sheet, please also download: INTEGRATED CIRCUITS

Handout 11: AC circuit. AC generator

Impedance and Loudspeaker Parameter Measurement

ELEC 2501 AB. Come to the PASS workshop with your mock exam complete. During the workshop you can work with other students to review your work.

Sinusoidal Steady-State Analysis

Circuit Analysis-II. Circuit Analysis-II Lecture # 5 Monday 23 rd April, 18

Ver 3537 E1.1 Analysis of Circuits (2014) E1.1 Circuit Analysis. Problem Sheet 1 (Lectures 1 & 2)

EE-202 Exam III April 6, 2017

Digital Control & Digital Filters. Lectures 21 & 22

Assessment Schedule 2015 Physics: Demonstrate understanding of electrical systems (91526)

Analog and Digital Filter Design

Problems of the XV International Physics Olympiad (Sigtuna, 1984)

Berkeley. Matching Networks. Prof. Ali M. Niknejad. U.C. Berkeley Copyright c 2016 by Ali M. Niknejad

Schedule. ECEN 301 Discussion #20 Exam 2 Review 1. Lab Due date. Title Chapters HW Due date. Date Day Class No. 10 Nov Mon 20 Exam Review.

COURSE OUTLINE. Introduction Signals and Noise Filtering Sensors: Piezoelectric Force Sensors. Sensors, Signals and Noise 1

Some of the different forms of a signal, obtained by transformations, are shown in the figure. jwt e z. jwt z e

Annexure-I. network acts as a buffer in matching the impedance of the plasma reactor to that of the RF

Frequency Dependent Aspects of Op-amps

Sinusoidal Steady State Analysis (AC Analysis) Part I

Physics 240 Fall 2005: Exam #3. Please print your name: Please list your discussion section number: Please list your discussion instructor:

Name: (Please print clearly) Student ID: CIRCLE YOUR DIVISION INSTRUCTIONS

6.1 Introduction

Impedance/Reactance Problems

I. Frequency Response of Voltage Amplifiers

Frequency Response part 2 (I&N Chap 12)

Steady State Frequency Response Using Bode Plots

Second-order filters. EE 230 second-order filters 1

Module 4. Single-phase AC Circuits

Total No. of Questions :09] [Total No. of Pages : 03

Transcription:

E.U.I.T. Telecomunicación 200, Madrid, Spain, 27.09 30.09.200 Design of Narrow Band Filters Part 2 Thomas Buch Institute of Communications Engineering University of Rostock Th. Buch, Institute of Communications Engineering, University of Rostock, Germany

E.U.I.T. Telecomunicación 200, Madrid, Spain, 27.09 30.09.200 Design of Compact Filters Filter Design, PZ-Map Loss Transformation Modified Transfer Function Determination of the normalized component values of the Continued Fractions Arrangement Inductive and capacitive coupling Denormalization Th. Buch, Institute of Communications Engineering, University of Rostock, Germany 2

Idea of the Procedure E.U.I.T. Telecomunicación 200, Madrid, Spain, 27.09 30.09.200 Idea of the Procedure Design method looked at till now permits only the realization by reactive four-terminal networks. (no losses!) Therefore loss transformation by moving of the jω - axis to the left. s-data s -data Realization of a reactive four-terminal network for s -data. The circuit then carries out the actual data of the s-plane with losses. jω' jω u, u' Th. Buch, Institute of Communications Engineering, University of Rostock, Germany 6

Idea of the Procedure E.U.I.T. Telecomunicación 200, Madrid, Spain, 27.09 30.09.200 Loss transformation Formulation: All inner resonant circuits have the same quality factor and with that the same attenuation. δ 0 The st circle has the attenuation δ > δ 0 Consideration of the internal resistance of the source. The n-th circle has the attenuation δ n > δ 0 Consideration of the input resistance of the following amplifier stage. Th. Buch, Institute of Communications Engineering, University of Rostock, Germany 7

Idea of the Procedure E.U.I.T. Telecomunicación 200, Madrid, Spain, 27.09 30.09.200 Loss transformation 2 s + δ 0 s Input impedance of a re- jω' jω Z e (s) Z e (s ) active four-terminal net- δ 0 work X e (s ) δ 0 must be lower then the smallest absolute value of the real parts of the poles! u, u' Th. Buch, Institute of Communications Engineering, University of Rostock, Germany 8

Continued Fractions Arrangement E.U.I.T. Telecomunicación 200, Madrid, Spain, 27.09 30.09.200 Continued Fractions Arrangement With that becomes the continued fraction decomposition of the input impedance for the coupled bandpass filters: Z e (s ) R /δ = (δ δ 0 ) + s + s + s +... x,2 2 x2,3 2 x 2 n,n (δ n δ 0 ) + s Th. Buch, Institute of Communications Engineering, University of Rostock, Germany 9

Continued Fractions Arrangement E.U.I.T. Telecomunicación 200, Madrid, Spain, 27.09 30.09.200 Simplification Simplification: Z e (s ) = R δ (δ δ 0 ) + s R δ + s δ + R x,2 2 s R x 2,2 δ x 2 2,3 + s δ x 2 2,3 R x 2,2 x2 3,4 +... Th. Buch, Institute of Communications Engineering, University of Rostock, Germany 0

Continued Fractions Arrangement E.U.I.T. Telecomunicación 200, Madrid, Spain, 27.09 30.09.200 Resulting Circuit R ( δ δ 0 ) δ R δ R x δ 2 2 2 x23 n odd δ R 2 x2 δ x R x x 2 23 2 2 2 34 n even Th. Buch, Institute of Communications Engineering, University of Rostock, Germany

Continued Fractions Arrangement E.U.I.T. Telecomunicación 200, Madrid, Spain, 27.09 30.09.200 Remark The input impedance results from the loss transformation for a completed equivalent LP-Reactant from the input impedance of the doubly-terminated lossy bandpass circuit. The normalized components can be won directly by continued fraction decomposition of the fractional rational impedance function Z e (s). It is important to say that only transfer functions without zeros can be realized by coupled resonant circuits. Th. Buch, Institute of Communications Engineering, University of Rostock, Germany 2

Further Design Method E.U.I.T. Telecomunicación 200, Madrid, Spain, 27.09 30.09.200 Further Design Method Transformation of the given BP-DTS into the normalized LP-DTS and determination of the PZ-data. 2 Loss transformation with the aim of generating the PZ -data of the s -plan. 3 Building the transmission function from the PZ -data and outline of the reactive four-terminal network. 4 Continued fraction stripping down of the input impedance. 5 Denormalizing and determination of the components of the coupled bandpass filter. Th. Buch, Institute of Communications Engineering, University of Rostock, Germany 7

An Example shoes the further Design Method Example: Three section bandpass filter with Butterworth approximation. Data: f m = 200 khz B = 4 khz a d = 3 db j Ω' jω Ω PZ-Data of the normalized LP: u 0 u u, u' u k Ω k 0 0,5 0,8660254 δ 0 Th. Buch, Institute of Communications Engineering, University of Rostock, Germany 9

Loss Transformation Loss transformation: Condition: δ 0 < min{ u k } = 0,5 Selected: δ 0 = 0,3 δ 0 = Q 0 Conclusions: Q 0 = δ 0 = Q 0 = 66,6 f m δ 0 B Th. Buch, Institute of Communications Engineering, University of Rostock, Germany 20

PZ-Data If the quality factor Q 0 (e.g. the filter coil) is predefined, then the bandwidth B cannot be chosen freely. It is valid: B f m δ 0 Q 0 PZ data of the loss transformed LP: u k Ω k 0,7 0 0,2 0,8660254 Th. Buch, Institute of Communications Engineering, University of Rostock, Germany 2

Transmission function H B (s K ) = (s + 0,7)(s 2 + 0,4s + 0,2 2 + 0,8660254 2 ) K = s 3 +,s 2 +,07s + 0,553 H B (s ) = K N(s ) K has to be chosen so, that HB (s ) s =jω = HB (jω ) H B (jω') < Ω' Th. Buch, Institute of Communications Engineering, University of Rostock, Germany 22

Design of the reactive four-terminal network H B (s ) H B ( s ) = H E (s ) H E ( s ) H E (s ) H E ( s ) = N(s ) N( s ) K 2 N(s ) N( s ) Determination of the zeros of H E (s ) H E ( s ), since the poles are known. H E (s ) H E ( s ) = s 6 0,93s 4 + 0,077s 2 + 0,305809 K 2 s 6 0,93s 4 + 0,077s 2 + 0,305809 Th. Buch, Institute of Communications Engineering, University of Rostock, Germany 23

Zeros of H E (s ) H E ( s ) Approach: K 2 = 0, Zeros of H E (s ) H E ( s ): = numerical solution! u E Ω E +0,7077 +0,829324 +0,7077-0,829324-0,7077 +0,829324-0,7077-0,829324 +0,646687 0-0,646687 0 H E (p') j Ω' H E (-p') u' Th. Buch, Institute of Communications Engineering, University of Rostock, Germany 24

Remark If K is set too greatly, so that H E (jω ) >,the zeros of H E (s ) and H E ( s ) do not let themselves divide. The zeros then lie on the imaginary axis. Th. Buch, Institute of Communications Engineering, University of Rostock, Germany 25

Continued Fractions Arrangement of Z e (s) It gives up: H E (s ) = s 3 + 0,88084s 2 + 0,8529096s + 0,453644 s 3 +,s 2 +,07s + 0,553 From this the input resistor Z e (s ) is determined. Z e (s ) = r + H E(s ) H E (s ) = 2 s 3 +,98084ss 2 +,9229096s +,006644 0,2959s 2 + 0,270904s + 0,0993586 r Th. Buch, Institute of Communications Engineering, University of Rostock, Germany 26

Continued fraction stripping down of Z e (s )/r Z e (s ) = +9,257945s + r 0,256687s + 0,227433s + 0,3396 From this the normalized LP gives up: R σ σ 0 σ ( ) R σ 2 x 2 2 x 23 R σ R σ 2 x 2 R σ x 2 2 2 x 23 ( σ σ ) 3 0 Th. Buch, Institute of Communications Engineering, University of Rostock, Germany 27

System of Equations The system of equations arises: R δ (δ δ 0 ) = r () R δ = 9,257945 r (2) δ R x 2,2 R x 2,2 δ x 2 2,3 = 0,256687/r (3) = 0,227433 r (4) R x,2 2 δ x2,3 2 (δ 3 δ 0 ) = 0,3396 r (5) One gets from (??): R δ r = 9,257945 This expression is contained in all other equations. = Recursive solution of the system of equations. Th. Buch, Institute of Communications Engineering, University of Rostock, Germany 28

Results Result: δ = 0,4095795 > δ 0! x,2 2 = 0,508098 x,2 = 0,7280559 x2,3 2 = 0,45336369 x 2,3 = 0,673322485 δ 3 =,29060986 > δ 0! Use of standard coils. Specification of an identical inductance for all circles. Selected: L = 0, mh Th. Buch, Institute of Communications Engineering, University of Rostock, Germany 29

Measure of the coupling It is valid: x,2 2 = ωm 2 M,2 2 δ δ 0 = M 2 ωm 2,2 R R 0 B 2 L 2 M,2 = L x,2 General: M k,k+ = L x k,k+ A bandpass filter with an inductive coupling results. Th. Buch, Institute of Communications Engineering, University of Rostock, Germany 30

Application of the capacitive coupling M R, 2 R 2 C R 2 R 2 C L L C L C C 2 L ω 2 m = L C C = C + C 2 C 2 C 2 + C 2 As C 2 << C 2 Approximation: C C + C 2 For the inner circles: Approximation: C C k,k + C k + C k,k+ Th. Buch, Institute of Communications Engineering, University of Rostock, Germany 3

Calculation of the coupling capacities (Philippow: Taschenbuch der Elektrotechnik, Bd. II, S. 580, Bild 7.45) Conversion: It is valid: C k,k+ = C ω m M k,k+ C (ω 2 m M k,k+ C) 2 With that: ω 2 m M k,k+ C = ω 2 m x k,k+ L C = x k,k+ C k,k+ = C x k,k+ ( x k,k+ ) 2 Th. Buch, Institute of Communications Engineering, University of Rostock, Germany 32

Denormalizing Numerical values for the example: C = = 6,332574 nf ωm L 2 C = C C,2 = 6,2422682 nf C 2 = C C,2 C 2,3 = 6,56984 nf C 3 = C C 2,3 = 6,2472898 nf C,2 = 90,3057 pf C 2,3 = 85,284 pf Th. Buch, Institute of Communications Engineering, University of Rostock, Germany 33

Circuit Diagram with Source and Load (C-coupling) I 0 C C 2 3 R L L L R a i C Q = 66,67 Q = 66,67 Q = 66,67 C 2 C 3 U a All circles get the same swinging Q factor. The greater attenuations for the circles and 3 are realized by the input resistor of the source and the load resistor (input resistor of the following amplifier). Required quality factor of the st resonant circuit: Q = / δ The inner resonant circuits have a quality factor: Q 0 > Q Th. Buch, Institute of Communications Engineering, University of Rostock, Germany 34

Calculation of source and load resistance It is valid: R 0p = Q 0 ω m L R p = Q ω m L } R p = R 0p R i R i = R 0p R p R 0p R p R i = Q 0 ω m L Q ω m L Q 0 Q = ω m L ω m L (Q 0 Q ) Q 0 Q = ω m L δ 0 δ It is valid correspondingly: R a = δ 0 = δ ω m L = 6,34395 kω (δ 3 δ 0 ) ω m L = 57,3393 kω (δ δ 0 ) Th. Buch, Institute of Communications Engineering, University of Rostock, Germany 35

Testing the designed Filter Testing the ready filter circuit with the help of a network analyzer software, e. g. PSpice Design-Center (PSpice for Windows) Circuit Diagram for PSpice: C 4 2 3 4 5 C R4 L R C L2 R2 L C R C 2 3 3 3 R 4 G Th. Buch, Institute of Communications Engineering, University of Rostock, Germany 36

Electrical devices Resonant impedances of the oscillating circuits: R p = ω 0 L Q = 2π 2 0 5 0 4 66,67 = 4π 0 66,67 R p = 20,944 kω L = L 2 = L 3 = 0, mh R = R 2 = R 3 = 20,94 kω R 4 = 57,34 kω R 5 = 6,344 kω C = 6,242 nf C 2 = 6,57 nf C 3 = 6,24 nf C 4 = 90,2 pf C 5 = 85,28 pf Th. Buch, Institute of Communications Engineering, University of Rostock, Germany 37

PSpice Plot Th. Buch, Institute of Communications Engineering, University of Rostock, Germany 38

Application Application of a compact filter in the IF amplifier of the HiFi tuner ReVox A76. It becomes a 8-stage Gaussian filter with a linear phase response (constant group delay!) used. Th. Buch, Institute of Communications Engineering, University of Rostock, Germany 39

HiFi tuner ReVox A76 Th. Buch, Institute of Communications Engineering, University of Rostock, Germany 40

Summery E.U.I.T. Telecomunicación 200, Madrid, Spain, 27.09 30.09.200 Summery Consideration of the losses by left shift of the jω-axis. With the new PZ data the transmission function of a reactive four-terminal network is calculated. From the continued fraction expansion of the input impedance the normalized network elements can be dimensioned. From the elements the couple factors and resonant circuit losses can determined. Calculation of the normalized devices. At use of customary coils the capacitive coupling of the circles is more favorable. Th. Buch, Institute of Communications Engineering, University of Rostock, Germany 46