$49.95 EXPERIMENTAL METHODS yy Ww7zol ‘ Rick Campbell, KK7B 4 Bob Larkin, W7PUA & Z "he rote A RRL Suis FBGn ‘THIS ITEM MUST BE RETURNED BY RECORDED DELIVERY ‘CAGE— | fe & od? an Se) ss Wes Hayward, W7ZOI Rick Campbell, KK7B Bob Larkin, W7PUA JBRARY. DOCUMENT SUBSRLY CENTRE ops 42 NOV 2004 Editors: Technical Illustration: Jan Carman, K5MA David Pingree, NiNAS Steve Ford, WB8IMY Dana Reed, WiLC Cover Design: Jim Talons, NOJT Sue Fagan Lamy Wolfgang, WR1B Bob Inderbitzen, NO1R Proofreaders: CD-ROM Development: Kathy Ford Dan Wolfgang Jayne Pratt Lovelace Production: Michelle Bloom, WB1ENT Paul Lappen Jodi Morin, KA1UPA (NGA REG-27762906 reumove 05 2 RODECO9 Request Ref. No. \VDxTL4432450 LOAN roan to: If no ether library incieated please return Tas entish Libraty baeument Supply Centre, Boston Spa, Yetnerny, West Yrkshira, Unted Kingdom 1523 72Q.CONTENTS Contents Preface 1 Getting Started 1.1 Experimenting, “Homebrewing,” and the Pursuit of the New 1.2 Getting Started ~ Routes for the Beginning Experimenter 1.3 Some Guidelines for the Experimenter 1. Block Diagrams 1.5 An IC Based Direct Conversion Receiver 1.6 A Regenerative Receiver 1.7 An Audio Amplifier with Discrete Transistors 1.8 A Direct Conversion Receiver Using a Diserete Component Product Detector 1.9 Power Supplies 1.10 RF Measurements, 111 A First Transmitter 1.12 A Bipolar Transistor Power Amplifier 1.13 An Queput Low Pass Filter 1.14 About the Schematics in this Book 2 Amplifier Design Basics 2.1 Modeling Simple Solid State Devices 2 Amplifier Design Basics 3 Large Signal Amplifiers | EGlirwe, cen torte cine (MMILANINAITTI 2.5 Differential Amplifiers and the Op-Amp il or ans 2.6 Undesired Amplifier Characteristics 49961449 is5 only] 2.7 Feedback Amplifiers Petunia ae : 218 Bypassing and Decoupling | goeca 0239 2.9 Power Amplifier Basics | cuss eat No, 2.10 Practical Power Ampliliers OFZ11984 2543425 LOAN 2.11 A 30.W — 7-MHz Power Amplifier It me other library incicated please return loan to ‘The Briveh Ubtary Dacument Supply Centre, Boston Sp, 43 Filters and Impedance Matching Circuits |) Wetherby, West Yorkshire, L823 780 3.1 Filter Basics 3.2 The Low Pass Filter, Design and Extension 3.3 LC Bandpass Filters 3.4 Crystal Filters 3.5 Active Filters 3.6 Impedance Matching Networks 4 Oscillators and Frequeney Synthesis 4.1 LC-Oscillator Basics 4.2 Practical Harley Circuits and Oscillator Drift Compensation 4.3 The Colpitts and Some Other scllators 4.4 Noise in Oscillators 4.5 Crystal Oscillators and VXOs 46 Voltage Controlled Oscillators 4.7 Frequency Synthesis 4.8 The Usly Weekender, MK-II, A 7-MHz VFO Transmitter, 4.9 A General Purpose VXO-Bxtending Frequeney Synthesizer5 Mixers and Frequency Multipliers 5.1 Mixer Bi 5.2 Balanced Mixer Concepts 5.3 Some Practical Mixers 5.4 Frequeney Multipliers 5.5 A VXO Transmitter Using a Digital Frequency Multiplier 6 Transmitters and Receivers 6.0 Signals and the Systems that Process Them 6.1 Receiver Fundamentals 6.2 IF Amplifiers and AGC 6.3 Large Signals in Receivers and Front End Design 6.4 Local Oscillator Systems 6.5 Receivers with Enhanced Dynamic Range 6.6 Transmitter und Transceiver Design 6.7 Frequency Shifts, Offsets and Incremental Tuning 6.8 Transmit-Receive Antenna Switching 6.9 The Lichen Transceiver: A Case Study 6.10 A Monoband SSB/CW Transceiver 6.11 A Portable DSB/CW 50 MHz Station 7 Measurement Equipment 7.0 Measurement Basics 7.1 DC Mesaurements 7.2 The Oscilloscope 7.3 RF Power Measurement 7.4 RF Power Measurement with an Oscilloscope 7.5 Measuring Frequency, Inductance, and Capacitance 7.6 Sources and Generators * 7.7 Bridges and Impedance Measurement 7.8 Spectrum Analysis, 7.9 Q Measurement of LC Resonators 7.10 Crystal Messurements 7.11 Noise and Noise Sources 7.12 Assorted Circuits ct Conversion Receivers 8.1 A Brief History 8.2 The Busic Direct Conversion Block Diagram 8.3 Peculiarities of Direct Conversion 8.4 Mixers For Direct Conversion Receivers 8.5 A Modular Direct Conversion Receiver 8.6 DC Receiver Advantages 9 Phasing Receivers and Transmitters 9.1 Block Diagrams 9.2 Introduction to the Math 9.3 From Mathematics to Practice 9.4 Sideband Suppression Design 9.5 Binaural Receivers 9.6 LO and RF Phase-Shift and In-Phase Splitter-Combiner Networks 9.7 Other Op-Amp Topologies, Polyphase Networks and DSP Phase Shifters 9.8 Intelligent Selectivity 9.9 A Next-Generation R2 Single-Signal Direct Conversion Receiver 9.10 A High Performance Phasing SSB Exciter 9.11 A Few Notes on Building Phasing Rigs 9.12 Conclusion10 DSP 11 DSP Components 10.1 The EZ-Kit Lite 10.2 A Program Shell 10.3 DSP Components 10.4 Signal Generation 10.5 Random Noise Generation 10.6 Filtering Components 10.7 DSP IF 10.8 DSP Mixing 10.9 Other DSP Components, 10.10 Discrete Fourier Transform 10.11 Automatic Noise 10.12 CW Signal Gener 10,13 SSB Signal Generation Applications in Communications 111 Program Structure 11.2 Using a DSP Device as a Controller 11.3 An Audio Generator Test Box 114 An 18-MHz Transceiver 11.5 BSP-10 2-Meter Transveiver 12 Field Operation, Portable Gear and Integrated Stations 12.1 Simple Equipment for Portable Operation 12.2 The “Unfinished,” A 7-MHz CW Transceiver 12.3 The S7C, A Simple 7-MHz Super-Heterodyne Receiver 12.4 A Dual Band QRP CW Transceiver 12.5 Weak-Signal Communications Using the DSP-10 12.6 A 28-MHz QRP Module 12.7 A General Purpose Receiver Module 12.8 Direct Conversion Transceiver for 14d-MHz SSB and CW 12.9 52-MHz Tunable IF for VHF and UHF Transceivers 12.10 Sleeping Bag Radio 12.11 14-MHz CW Receiver Contents of CD-ROM Index‘The predecessor for this book, Sotid State Design for the Radio Amateur (SSD), was first published by ARRL in early 1977. The ‘goal for that rext was to present solid state circuit design methods oa community much more familia with vacuum tube methods But, another goal was integrated into the text, that of presenting the material ia a way that would allow the reader to actually design his or her own circuits. Handbooks uf the day presented only an encyclopedic overview of solid state devices with brief ‘qalitative discussions about functionality. SSD deseritedcireuit Clements in terms of models that could be used for analysis Design consists of more than merely combining representative circuits from a catalog or handbook ‘SSD succeeded with design becoming the key word in the ttle, especially in lacer years as the world became accustomed co all, electronic equipment being predominantly solid state. What surprised many is that the book remained popular, even after many of the transistors used in the elreuits were no longer available. Experimental Methods in Radio Frequency Design (EMRED) is the sequel to SSD. with design remaining as a central theme. (Our goal is to present models and discussion that will allow the user to design equipment at both the ciscuit and the system level (Our own interests are dominated by radio frequencies, so the text, discusses problems peculiar ta radio communications equipment A final emphasis in EMAFD is experimentation. A vital part of fan experiment is measurement, We encourage the reader 10 not only build equipment. but to perform measurements on that gear as it is being built ‘The word “experiment,” often conjures memories of schoo! exercises where a teacher has assembled equipment and we. as students, go theough a prearranged set of steps to ative at a conclusion, also predetermined. Although efficient, this isa poor representation of seionce. Rather, experimental science begins ‘with new idea, An experiment 10 test the idea is then generated, the experiment is built, mousurements are made, and the results are pondered, which often results in new ideas to test. This ean all be done by one person working alone. EMRFD encourages the participating reader t© build equipment with an attitude of contiaually seeking to understand the equipment and 10 tunderstind the primitive concepts that form the basis for the equipment and the circuits contained therein. Our greatest hope is that the text will Mustrate the potential of amateur radio, and cher personal science, a6 4 training ground for the individual This texts aimed ac variety of readers: the radio amateur who designs and builds his own oguipment; college students looking fir design projects or wishing to garner practical experience with, ‘working hardware: young professionals wishing to apply their Trest engineering and physics coursework to kitchen table Projects: non-engineers wanting to dabble in a technical field; engineering managers recapturing the fun of making things {instead of people} work; and technical explorers of ull types. The firs chapter of EMRFD deals with the problems of getting started with experimentation, Numerous projects are presented, timed at assisting the experimenter in beginning investigations inelectranies, Chapters through 5 then deal with specific cireuit functions. Chapter 2 presents amplifiers while filters are discussed in Chapter3, Oseillatorsemerge in Chapter 4, including PREFACE the natural extension of frequency synthesis, Mixers, including Frequency multipliers, appear in the fifth chapter. These chapters are laced with projects that can be constructed, but they also ‘emphasize important basic concepts. Chapter 6 moves on to present communications equipment, predeminantly using super-heterodyne methous. System design considerations a included, especislly with regard to distortion and dynamic range. The chapter contains several projects including a high performance receiver. Chapter 7 deals with measurement ‘methods and includes considerable test equipment that the experimenter can build. Chapter 8 then moveson toa fundamental discussion of direct conversion. This is followed by & thorough treatment of the phasing method of SSB in Chapter 9. Chapters 10 and 1] present fundamental concepts of digital signal processing and illustrate them with projects, The book concludes, ‘with Chapter 12 featuring a variety of experimental activities of special interest to the authors. ‘A Compact Disc is included with the book, This CD contains some design software, extensive listings for DSP firmware celated to Chapters and 11. and asizeable collection of journal articles, relating to material presented in the text. The design software is ‘written for personal computer using the Microsoft Windows ‘operating system, while the journal papersare presentedin Adobe Acrobat (PDF) format. ‘This book isu personal one in that we have only writen sbout those things we have actually experienced, We specifically avoided an encyclopedic discussion of material that we fuad not ‘actually experienced through experiments, Equipment of interest tothe three of us dominates. The amateur bands up to 2 meters are ‘considered, ard are illustrated with CW and SSB gear. The book uses some mathematics where appropriate, It is, however, keptat abasic level The book contains numerous projects that are suitable for uplication. Printed cireuit hoards are not generally available for these, although boards may become available at a later time. Readers should keep an eye on the world wide web for PCB §nformation and ether matters related to the book. See http! www.arrLorginotes/8799, We generally prefer tht builders use the projects as starting points for their own designs and experiments rather than duplicating the projects presented. Acknowledgments ‘The following experimenters have contributed 10 this book through experiments, direct correspondence, encouragement ‘and by example, We gratefully acknowledge their contributions. Bil] Amidon (sk); Tom Apel. KSTRA; Leit Asbrink, SMSBSZ: Kirk Bailey, N7CCB; Dave Benson, KISWL: Byron Blanchard, NIEKY, Denton Bramwell, W7DB; Guy Brennert, K2EFB: Rod Brink, KQ6F: Kent Britain, WASVIB: Wayne Burdick, NOKR Russ Carpenter, AATQU: Dennis Criss; Bol) Culter, N7FKI: George Daughters, K6GT: John Davis, KFSEDB: Paul Decker KG7HF: Rev. George Dobbs, G3RJV: Pete Eaton, WB9FLW: Gerry Edson, WAOKNW: Bill Bvans, W3PB:George Fare, G30GQ; Johan Forser, KC7WW; Dick Frey KaXU: Barrie Gilbert; Jack Glandon, WB4RNO; Joe Glass, WB2PIS: Dr, Dave Gordon-Smith, G3UUR; Mike Greaney, K3SRZ: Linley Gum, K7HFD: Nick Hamilton, G4TXG; Mark Hansen, KITN: Markus Hansen, VETCA: Neil Heckt; Ward Helms, W7SMX; Don Hilliard \WOPW: Fred Holler, W2EKB: Robert Hnghson Pete Juliano, W6IFR; Bill Kelsey. N8ET; Ed Kessler, AA3SJ; Paul Kieiak, N2PK: Don Knotts, W7HIS: O. K. Krienke: Beb Larkin. W7SLB; John Lawson, KSIRK; Roy Lewallen, WTEL; Jon Licbentood, K7RO: Larry Liljeqvist, W782; B. F. Logan Jr, WB2NBD: Stephon Maas, WSVHJ; Chuck MacCluer, WEMQW; Jacob Makhinson, NeNWP: Ernie Manly, W7LHL; Dr. Skip Marsh WOTRQ (skh; Mike Michael, W3TS: Jim Miles, KSCX; Dave Newkirk, WOVES; Gary Oliver, WATSHI: Paul Pagel, NIFB Dave Roberts, GSKBB; Mike Reed, KD7TS; Don Reynolds, K7DBA (sk); Dr. Ulrich Rohde, KA2WEU: Dr. Dave Rutledge, KNEK: Tom Rousseau, K7PIT; Bill Sabin, WOLYH: Tom Scot, KD7DMH; Marty Singer, KTAYP: Derry Spittle, VETQK: Frod Telewski, WA7TZY: Paul Wade. WIGHZ: Al Ward, WSLUA; Dr, Fred Weiss: Jiem Wyckoil, K3BT Bob Zavrel, WISX: Bob Zulinski, WASMAM, We have certainly missed some folks in our list. Please aocept urapologies for our oversight ang our thanks for your help with About the Cover Photograph ‘The cover photograph is an experimental 2.4 GHz TC direct conversion ceesiver front-end on a gallium arsenide die. The die is litle more than one millimeter wide, and less than one mile limeter high. Gold-bond wires connect to the metal squares around the edge, The large spiral is a quadratuee hybrid coupled inductor, and the matehed inductors at tae top are in a Wilkensen the book and related experiments, Some folks have made special contributions and deserve special thanks. Colin Horrabin, G3SBI; Harold Johason, W4ZCB; and Bill Carver, W7AAZ, collectively formed the “Triad.” a group building the high performance transceiver partially described in Chapwer 6. We sincerely appreciate their willingness to share their efforts ang zesults with us. Thanks 20 to Roger Hayward, KA7EXM, for building some equipment described in the hook as well a8 helping with Tield testing of ‘numerous designs. Jef? Damm, WATMLH, deserves special thanks for his efforts. He built equipment described in SSD and provided encouragement for this version. Special thanks to Merle Cox. WIYOZ, and Jim Davey, K8RZ, for several decades of bouncing around radio ideas, huilding the second prototypes, and manning the distant station for countless experiments, Very special thanks are extended to Terry White, KTTAU. Terry did high quality PC layouts for several ofthe designs presented in the text and in earlier OST articles. He also built some equipment shown in the book und provided measurement assistance on several occasions. Special mention should be made of the efforts of the late Doug. DeMass, WLEB, As co-author of SSD, he provided interest and encouragement for this sequel, One of Doug’ greatest qualities was his intense, sincere interest in radio communications. He designed and built radio equipment, used it on the aiz, and then clearly wrote about the efforts, establishing & standard for all to follow. We missed him often through the generation of this text Finally. we want to thank our families, and especially ou wives: Churlene (Shon) Hayward, Sara Rankinen, and Janet ‘Larkin. A book requires time and intense effort that often detracts from other activities. Our “better halves” have all tolerated these moments of distraction, splitter. The passive cireuitry is similar to Fig 9.39, and the pho- tograph on page 9.43 shows this IC connected to baseband cir cuitry described in Chapter 9. Note the call signs on the dic. MAL," who was not licensed in 2001, is now K7MTL. Photo- graph by Dean MontheiAbout The Authors ‘All theee ofthe authors share a sinilar early exposure to eadio, obtaining an amateur license as teen or earlier. ‘They all started with the novice class license. Their early ham experiences expanded to become careers in science and electronics. All three are members of the IBEE Microwave Theory and Techniques Society and have published extensively in a wide variety of journals and books. All three writers contributed to all chapters of this text, but each author had a primary responsibility listed below. Wes Hayward, W7ZOI Wes received a BS in Physies from Washington State University in 1961 and an MSE trom Stanford University in 1966, He worked on electron deviee physies at Varian Associates, The Boeing Co.,.and Tektronix. He then did RF circuit desiga, first at Tektronix and then at TriQuint Semiconductor, Wes is now semi-retired. dividing his, ‘ime between writing and consulting. Wes was the primary contributor to Chapters | through 7 and large parts of 12 and ean be contacted at W7z01@ arrl.net. Rick Campbell, KK7B Rick received a BS in Physics from Seaitle Pavilie University in 1975, after wo yeurs active duty as a US Navy Radioman. He worked for years in erystal physics basic research at Bell Labs in Murray Hill, NJ before returning (o graduate sehoo! at the University of Washington, He complcted the MSEE degree in 1981 and the PhD in EE in 1984. He served on the faculty a¢ Michigan Tech University until 1996. Since 1996 he has been with the Advanced Development Group at TriQuint Semiconductor, designing microwave receiver eiseultry. Rick Mad primary responsibility for chapters &, 9, and large parts of 12, He can be contacted at KK7b@arrlsnet. Bob Larkin, W7PUA Bob received u BS in EE from the University of Washington and a MS in EE from New York University. He ‘worked for 12 years at Bell Labs in New Jersey in areas of circuit design and signal processing, Tn 1973 he and his wife Janet started Janol Labs whee « variety of radi requency products were manufactured. They moved the company so Corvallis Oregon in 1975 where it operated until being acquired by Celwave RF in 1991. He now ‘works as a consultant specializing in microwave circuits. Bob was the primary contributor to Chapters 10 and 11 fand srote a section in Chapter 12. Readess can contuet Bob at Wpua@arrlanet.CHAPTER Getting Started 1.1 EXPERIMENTING, “HOMEBREWING,” AND THE PURSUIT OF THE NEW Amateur Radio iss diverse and colorful vocation or hobby where the parteipants ‘communicate with each other through the use of radio signals, The communications. which can encompass and extend beyond the planet, are often routine and predict able, but cam at times be ethereal. The romance of communicating with the ether side of the world blends with the jay of observing acomplicated part of nature. For some of us, the wonder never disappears. ‘Although radio can be fun, our pra matic society demands more than excite ‘ment when resources are used. The virtue that most often justifies our use of the radio specirum i the growth of a profi cieit communications system that can be called upon in times of emergency. The examples of its use are numerous, But, “hany” radio is more than this. Ibis a technical avocation of diverse educa- tional potential. [thas values that go well beyond that ofa supplementary communi- colons network. ‘Most radio amateurs have an interest in the technical details of the equipment they use, Historically. this was « requirement: The only way a radio amateur could as semble an operating station was to person- ally build his or her gear. Commercial equipment was rare, and was offen pro- hibidively expensive. But today, high qual- ity “ham” gear is readily available in most ‘of the world, much of it at modest prices, Although no longer necessary. its still common for radio amateurs to build at least some of their own equipment. The reasons sre varied and as numerous as the partici- pants, A few purists consider building the equipment they use to be a non-optiensl integral par oftheir hobby in he same way that a fly fishing enthusiast would never consider fishing with a fly that he or she had not fabricated. ‘The majority take an intermediate pach, building parts of their radio stations while purchasing others. For some, building is an exercise in exafismen: ship, an opportunity to generate equipment vwith an individual imprint and personality ‘Common to all of these, amateur radio presents an opportunity that is rare among vocations, a chance for individual, unre- strained investigations in fundamental sei- cence and technology. This is @ rarity in an ‘age when mest research and design is per formed by teams of investigators within large organizations, be they universities or theenginecring arms comorations. There, the subjects chosen for investigation ae of- ten those of corporate or national interest. is increasingly rare chat a study is initiated ut of simple curiosity. Fortunately, we are not so constrained within our personal in vestigations of radio science. Consider an example, An experimentally inelined radia amateur envisions & new scheme for a receiver. It might be a beter front end circuit, a new block diagram, o & ‘way to realize some receiver functions with a computer. The experimenter can analyze the scheme, design an exansple, build a pro lotype, build and sssemble needed test equipment. measure the receiver perfor mance, compare it with predicted results and use the receiver on the air. Favh part of the investigation can interaet withthe oth- ers. All of the activity ean be done without interference trom other sources. The pro: gram will never be cancelled by the ehang- ing goals of an organization. Nor will «be rushed by the economie pressures of a cor- porate program. ‘The inspiration for experiment varies. In rare cases, the experimenter may feel that his or her work could lead to 2 new ewistin the state-of-the-art, a better receiver. But ‘mote often it will just be a casual thought that “Hey, I've never built one of these before and IL Iearn something it do.” The mosteommon is aneffort spurred bya nced) ‘ham wants a rig t0 take along on a hiking trip when ne such thing ean be purchased "No matter whar the origin, che experimenter can enjoy the knowledge that he or she is Jeuming more about the subject and about the research process In this book we encourage all levels of what has become known as radio “home brewing,” ranging from beginner projects to sophisticated multi-mode ereations. We generally emphasize simple equipment described by primitive explanations. By primitive, we intend that the discussion relute 10 the most fundamental and basic cizeult design concepts. The equipment and systems presented ure themselves basic. often without the frills, bells, and ‘whistles of commercial equipment. Some refinements willbe discussed, allowing the experimenter to add those he or she needs, This book emphasizes equipment de- sign. Our interest isin basie cizcuit fune- vions and the underlying concepts that alloy them to be understood. This book is generally NOT a collection of projects for reproduction and construction, Although some of the equipment may be directly duplicated, we would prefer to have you sdapt our results (© fit your own needs. This book i, in many ways, a sequel to aneatlier effort, Solid State Design or the Radio Amateur. That 1977 book, cco-suthored with the late Doug DeMaw, Getting Started 1.4WIPB, had goals similasto those outlined above, plus that of introducing solid-state methods to readers with experience lit ited to vacuum tube electronies, The later need hus become arguable, for virtually all of our equipment is now based upon solid-state technology ‘All of the circuits presented in this text have boon constructed, tested, and used in practical, on-the-air situations If there are exceptions where the authors have not setually built an example of what is diss ‘cussed, we will so stare inthe related tex We emphasize the traditional commu: nivations modes of CW, the original dig tal mode, and SSB phone. Building litte rigs and cadiating and receiving continu fous waves are 10 2 radio experimenter ‘much like playing scales and folk tines are to a musician. They are the first things ‘we learn, are important parts of the dail practice routine throughout life, and we neglect them at our peril. The little rigs, tnd the concepts they represent, are at the core of wireless technology. It is not ‘enough to play with them asa novice and then move on to other things: they need to be revisited over and over again at difter- tent stages of one’s vocation, each time achieving a new level of mastery until fie nally one is probing the deepest mysteries of the ar. 1.2 GETTING STARTED—ROUTES FOR THE BEGINNING EXPERIMENTER What to bui A frequent question asked by the pro- spective experimenter regards an initial project or subject for pursuit. A common Choice for a first project comes from a desire to extend the capabilities of an ex- isting station. The future experimenter al- ready has experience with on-the-air ac- tivity anda working station, Heor she then wants 19 extend that station 1o new bands, improved tansceiver performance, or fab ricale a rig offering portability. While these goals are all worthy, they can be df= Ficult. They may be conceptually impos- sible for the beginner, and impractical for he seasoned experimenter with other life commitments, A better “frst” experiment ‘may well be somiething that is mueh sim= pler. Several simple projects are offered later inthis chapteras suitable beginnings. How to build it: Another getting-started question re ¢gards the methods 1o use in building elec- tonics. There are several options, al with their assets and weaknesses. A few are discussed below. PRINTED CIRCUIT BOARDS ‘The primary construction scheme used. in modem electronics isthe printed circuit board (PCB). Here, pads or islands of metal are attached 10 an insulating mate= Fial. usually epoxy-fiberglass, Wires on the parts are pushed through holes in the ‘pour and soldered to the pads, which are interconnected by printed metal runs, thus, forming the circu A PCB begins as a fiberglass sheet wih ‘copper laminated to one er bath sides. The metal surfaces are then coated with a light sensitive “photo-resist” material. A pattern forthe circuit is optically transfeered to the surface and the unexposed material is washed aay. The board is mow placed in a solution that chemically etches some of 4.2 Chapter 1 the copper away, leaving only those regions needed to form the desired circuit. After etching, the Board is washed and Grilled, Pure copper is easily corroded, so it is common to plate boards with a tin coating, forming a more stable and solidcrable surface. Refined boards include ‘copper on both sides, and even plating on the inside of the holes. Industrial bourds ‘will often incorporate many layers. ‘Modern practice features surface mount technology, SMT, using small compo- nents without wire leads. The leads have been replaced with metalized regions on the parts that are then soldered directly to the boued. The soldering provides phy'i- cal mounting as well 2s electrical connee- tion. The SMT boards are cheaper to build and usually much more dense. SMT parts ‘can be so small that they are hard to handle without a good microscope. SMT is an interesting way to build if there is a need for really small equipment. The small size ‘of SMF circuits often results in improved high Frequency performance. Growing SMT popularity in manufac- luring means that surface mounted is the only available form for acomponent. Many parts don't exist in leaded forms. In some ‘cases they can be handled by the "Surt hoards" by Capital Advanced Technalo- ties which are found in DigiKey catalogs ‘These are small SMT boards with an inter face that will adapt to other board forms, ‘Gireuit boards have been builtin a home environment by hams for generations. The ‘reader should review the sulijectin The ARE. Heenaook wo find out more about the meth bods. A major problem with bome etched boards is the disposal of the used etchant usually asolutionof ferric chloride. Disposal practices eommon in the past are now ques- tioned in this era of enlightened recycling Although some of the projects deseribed in thistextuse etched boards. few ofthe boards were etched in our home labs BREADBOARDED CIRCUITS Breadboard, 25 applied to electronics, isa term from a time when early radio experimenters built their equipment on slabs of wood, often procured from the Kitchen, The term remains as an industry ‘wide description of a preliminary experi- mental circuit, There ure numerous modern methods that can be used to gencr- ate w one-of-ackind circuit UGLY CONSTRUCTION A particularly simple method was out- lined in an early QST paper and is now know as “Ugly Construction.”® Although certainly not unique, the scheme works well and continues as a recommended method. The sctieme consist of the follow ing! T. A ground plane is established using an un-etched setup of copper clad circuit board material 2, Following the schematic for acireuit being built, grounded components are sal- dered dircetly to the ground foil with shoxt leads 3. Some non-grounded parts are sol- dered to and supported by the grounded ‘components 4, Other non-grounded components sre supported with suitable “tic down points” ‘consisting af high value resisters. ‘5. Once finished and working, the board can be mounted in a suitable bos, hidden from view if desired, where it becomes 1 permanent application of the idea. Usly construction is ilustrated in Fig 1.1. Casual circuit analysis allows’ the builder to pick the standoff resistor valucs ‘Any “high R” value resistors can be used. Usually, 1-MQ resistors work well any- ‘where within RF circuits. The typical 1/4 W resistor of any value has a stray Jead-to- lead parallel capacitance of about 0.3 to Da pF, perhaps a little more with longer leads, and a series inductance of 3 t0 5 nH\ é ‘Solder. \ == Fig 1.2—An example of * 1lMeg = .01 1 Meg (Solder), up through 150 METz or sb. High R means that resistance is high with respect tothe reactance of the inductance. Wesometimes use R values as low as 10K0. It is often surprising just how few standoft cesistors sre needed in an ugly breadboard. The greatest virtue of the ugly method is low inductance grounding, Any canstruc~ thon scheme that preserves this grounding Integrity will work as well. Picking @ method is a choice that the builder has. a place where he or she can develop the methods that work best. Integrated cireuits ean be placed on an ugly board with leads sticking up. “dead bug” style. There is litle need to glue the chipsdowa, for components and wires will eventually hold them in place. Grounded IC Teads are bent and soldered directly wo the foil Some builders prefer to maintain ICs \ith the IC label facing upward, allowing later inspeetion. They then bend all Tesds out in a “spread eagle” format We have never had a problem with ugly equipmentbeing less than robust, Many of our gly rigs have been hauled throug the mountains of the Pacifie Northwest in packs without incident. An outstanding caample, the work ofa fiend isthe W7EL Optimized ORP Transceiver arg tar has raveled around the World in suiteases and packs.? Few if uny standoff resistors were used in that rig MANHATTAN BREADBOARDING Several other construction sehemes of- fer similar grounding fidelity, including those where small pads of eireuit board material are glued ot soldered w the round foil, These pads then have compo- nents soldered to them, We have found this, massive components such as floating. aon srounded, trimmer capacitors. The spe- Cific glue type has little impact on circuit performance Vaciations oF this method have been called “Manhattan Construc tion,” and can be mixed with other bread- boarding schemes. The reader can find ‘numerous examples on the Web on sites dealing with ORP experiments. as well as in Fig 1.2 The proponents of Manhattan Construc- tion often use small round pads that are sled 10a ground foil with epoxy vr simi- Jar glue. The pads are placed So that all components are parallel to board edges and close tothe eround fil. This produces am attractive board resembling a commer: ial, PC hoard. This does mot scemte com promise performanee. With traditional ugly construction, parts can be moved about to make room for ‘mother stage. In the extreme, an entire ciccuit can be lifted and moved, a stage at fa time, to another board ‘A primacy virtue of a bread-boarding scheme is construction speed and flexitil- ity, especially important when the primary purpose of building gear is information about circuit behavior. Some folks prefer 10 rebuild a circuit after « breadboarding phase, replacing un ugly prototype with a more permanent, production-like version. These efforts ake axiditional time and rarely produce pesfor mance superior to the original bread- boards. Even looks can be deceptive when ‘one hides usly breadboards behind more attractive front panels QUASI-PRINTED BOARDS Some experimenters. prefer to build equipment that looks like x PCB, even ‘scheme for breadboarding. The installed resistor here is soldered to {ground and to a pad that connects to the rest of the circuitry. when the board is not etched in « circuit specific pattern. One method, called “checker-board,” uses doublesided circuit bbourd with one side funetioning as a ground foil. The other side consists of a ‘matrix of small iskinds of eopper. These regions are created cither by etching or manually with a back saw. Patteras oF squares on 0.1-ineh centers necommodate traditional ICs, Holes ure drilled inthe ise lands where components must reside. A large drill bit then removes growad foil sround the hole without enlarging it. No holes are required where a ground con- nection is needed. Components usually reside on the ground side of the board. See Fig 13, ‘The double sided checker-boord can also serve for breadboarding with surface mounted components, Paris then reside fon the patiern side with holes drilled to reach ground. Small leaded components can also he surface mounted ‘The checkerbourd scheme, “Manhat- tan" variants, and even double-sided printed boards have fairly high capaci- tance from pads to ground. These are often poor quality capacitors with low Q, under 100 for epoxy fiberglass boasd material and are subject to water absorption. A single sided formacis preferred for critical sections of a LC oscillator application. Getting Started 1.31.3 SOME GUIDELINES FOR THE EXPERIMENTER With Solid-State Design for the Radio Amateur came considerable interaction ‘with the rest of the amateur radio commu- nity. A frequent question we heard was How do [ get started with experiment- ing?" Or, “Y've read about and have even ‘ult some kits and published projects, but want to go further. T want to do my own. design, What is the neat step?” ‘A set of guidelines is offered in an tempt to answer Some of these questions. ‘These are not firm, well established rules, ‘but mere impressions and personal biases that we have generated, approaches that ‘work for us, They are offered without guarantee. ‘KISS: This British term is shoxt for “Keep lt Simple, Stupid.” Weoftendesign equipment that is more complicated than needed. Tt is well worth some extra time during design to evaluate every part to see if itis really needed. The function of each part should be understood and justified. ‘The cireuit should function as intended. This does not imply that designs with the minimum number of parts ate best. How ever, it is zarely justified to overdesign by adding extra components “because a prob- lem might occur.” For example. designs ‘with a profusion of ferrite beads and "Sta bility enhancing” resistors may besuspect lore: Lore, inthis ease, refers to “knowledge” that is based upon expe ences that are divorced from careful thought. A classic example in amateur ra dio regards the thermal stability of LC oscillators, Envision the amateur experi enter who built an oscillator using a tor oid, The circuit drifted when he opened the window 10 the winter weather. The nextevening he replaced the inductor with ‘one wound en a ceramic coil form, notie ing less drift when he opened the window. He coneluded that ceramic forms are bet- ler than toraids, having never considered the specific coll forms taat were used, the ‘other components in the cizcuit, or the Fact that the weather had improved. Poorly executed experiments like this often gen- erate erroneous conclusions. The zesult ing lore, although interesting, should always be questioned. ILiswlways better © «do meaningful measurements ‘Plan your projects with block dia- ‘grams: Start with small diagrams where each block is a global element, perhaps containing several stages. Expand these to show greater detail. Block diagrams will de discussed further below. Generate modular equipment: A high performance receiver, for example, should 4.4 Chapter 1 consist of several sections, each designed sothatitcan be built, ested, modified, and redesigned as needed, with minimal change to the rest of the system, Even the simplest little rig should be built «stage at atime, turned on sequentially, rested, and modified as needed. Single board trans ceiver designs are popular in the QRP arena, But realize that the ones that work well are probably the result of several re= builds, and even then, some don't work very well; others are super’. # Avoid excessive miniaturization: It takes much more time wo build small things than those where the cirouitry can expand without bound. Even when building small portable QRP transceivers, i's often worthwhile to establish the design with a larger breadboard ‘© Base projects on your own goals: Our centeal personal goal is learning through experimentation. Hence, we base projects ‘on questions that need investigation rather than what we need or want for on-the-air operation. But your gouls may be differ- ent, Ttis worthwhile to review and define them as a means of picking the best projects for you. Isolate primary goals from those that ate serendipity. Be wary of “Creeping Features.” The cerm “appliance” often describes. the lranseeivers that we purchase for ‘on-the-ait communications. Applianess, even ones that we build ourselves, are usually expected to have many features, bout these belts and whisrles can actually impede experimental progress. A single bband, single mode transceiver can be as experimentally enlightening and informa tive as a multiple mode, general coverage transceiver: © Use the literature, Peruse catalogs, data manuals, web sites, and even instruc tion manuals for circuit ideas. When a cir- cuit method is not understood, it should be studied in texts appropriate to the technol ‘ogy. It is useful to build something with the part as a way to really understand that par. ‘© While planning is necessary, don’t spend excessive time in the preliminary design phase of a project. Rather, outline preliminary ideas and goals, do initial cal- culations (on a computer only if they are really complicated). gather parts, and begin building. Enjoy the freedom that allows you to change your mind in the ‘middle of an investigation. Refined caleu- lations can occur during and after con- struction und are not just “design phase” activities. ‘#1e'snot about craftsmanship: A portion of the homebrewing community was schooled with the ides that “nice looking” citeuic construction went along with good performance, But the (wo Factors are gen- erally isolated, This is ilustrated in Fig 1.4. There is no relationship berween hav- ing a nice looking, orderly circuit board and good performance from that board. In- deed, those suddled with the chore of de- signing a printed board to perform as well sas an ugly breadboard may wonder f there ‘might be an inverse relationship! Use breadbosrding aver ground plane for communications citeuits, especially vwhen investigating new ideas. Use vector board or wine-wrap methods for slow digital circuits, buttreat fast digital citeuitsasifthey were RF fusctions. In general, build with those methods that will offer the best low inductance, grounding while allowing ci cuits tobe quickly designed, assembled, and tested, If you are concerned with acsthetic tails, build second version. Alterna tively, analiactive panel ean be used tohide ugly, hut highly functional breadboards, ‘© Build what you use, and use what you builé: Taose of us in the homebrew end of amateur radio often kid our appliance op- erator friends. suggesting thata“real ham” should build instead of just operate. Some ‘avid experimenters may take this too far; they build « rig, use it just long enough to confirm functionality, and go on to the next project, missing some exciting di coveries along the way. By using the equipment with tempered intensity, the experimenter will discover the strength and weakness of the rig, allowing the nest project to be even more successful. The same arguments might be applied to soft ware developments! © Beware of the golden screwdriver: A good friend. WA7MLH, encountered a fellow on the air whose sole method for experimentation was to adjust all of his equipment for maximum output. He did this witha favorite screweriver, whieh he ‘weated as golden, After careful tweaking fof all circuit elements that could be ac justed, he was almost always able to coax a 100-W transceiver into delivering 110 W of output. Unfortunately, what started asa good piece of equipment had become a distorted disaster. While we all tend to adjust cizeuits for“maximum smoke.” Tin~ ear circuitry should be confined to oper- ate under linear conditions. Itis important ‘that the limits be recognized and adhered to, This is especially important when building SSB gear. Alignment mens ad- jusiment to the proper, measured level.Fig 1.4—"Nice looking” ele Which may differ from maximum. ‘Always Keep notebooks for experi ments: Record those wild circu ideas that come up while you eut the lawn ox wateh TY; record important datn during experi- ments, including the temperature when you open the window: take notes on the Circuits that you build, including changes that are made during building and “tra ‘on. Date the notebook and place small dated labels inside the rigs so you can find the data when it's needed. Use bound or spital notebooks rather than loose-leat documents, for they are more permanent. Along term computer based index of note books is very useful, (© Find others with the same passion for experimenting: Although this guideline ix pretty obvious, it's also easy Tor the ex- perimenter to become isolated in hisorher own world. Builder hams are rarely iso- lated. Finding the local ones will give you a place to communicate your ideas, hear about new thoughts, and éo share junkbox parts as well as test equipment. Ask at local elubs to find out who is building. Lis- ten to the appropriate nets and attend the specialty clubs. Write to fellows who author articles of interest, especially if they live nearby. Watch the chat sessions ‘on the Tnternet or the Web, Amateur radio is about communications, so don't hesi- tate te communicate # Look toward the ordinary for expla- ations: When a design is not working as ‘well ts itshould, we look for explanations hat will explain the differences. All too often we consider the complicated an- swers, only to discover that the real swer is in the “obvious.” It is always worthwhile to return 10 fundamentals, ® Strive to build equipment that does not pollute the already abused radio spec~ trum: Make an effort to generate clean equipment, meaning that it does not emit signals at frequencies other than the in- tended ones. While most ofthis concern is \with transmitters, the ideas should also be spplied to receivers. The difficult ques- tion is “How clean is clean enough?" The onstruction does not always equate to good circuit performance. FCC has specifications for spurious emis- sions from US transmitters. These spe cations depend upan transmitter output power. Even for equipment running full power. the specifications are generally easy lo mect at HF. When power drops below 5-W output, they become even easier. Throughout this text we take the approach that even greater levels of elean- Jiness will be sought. This book includes a chapter on test equipment. One of the items featured there isa spectrum analyzer that will allow che builder to measure spec tral purty. A final “rule:” Don't let any of these rules get in the way of experimenting and building! Jr's OK if there aze things that you don't understand even if that includes the project you are about r0 build, for you ‘will understand much more when you are finished. The real goal ofthis pursuit, und of this honk is to fear by doing. The same can be said for other “rules that may ap- pear in the literature or on the web: Don't Jet them keep you feem experimenting, Getting Started 1.51.4 BLOCK DIAGRAMS Fig 1.5 shows a collection of elements that can be used in a detailed block dix ‘aram of aradio, This short listis generally extensive enough to describe the non-digital designs in this book Schematic and block diagrams serve a variety of purposes in electronics. The purpose of the block diagram isto present the fonctions and their interconnection used in a pieee of equipment. Schematic diagrams present the details. A block diagram is « useful way to plan and describe the equipment we wish to build. The block diagram will serve asthe stacting point for mathematical analysis that we may apply to the overall systera. 1 can also eniphasize the functions required tocomplete the design. This is illustrated ‘with Fig 1.6 showing a direct conversion transceiver for the 40-meter band. Several filtersaxe shown, illustrating the functions, that are important for good performance. ‘The low pass and the high pass between the mixer and audio amplifier ate simple, ‘consisting of one component each, There may be no components forthe signal split- ter, but the function remains, Fig 1.7 shows a more elaborate circuit, super-heterodyne SSB/CW transceiver for the S0-MHr band, The phasing method cant ‘also be used: such a S0-MHz transceiver is presentedin Fig 1.8. Designing any of these systems begins by forming the block dia- grams, which includes specifying cach of the blocks, Once this is done, the individual circuits can be designed, Some elements are missing inthe block diagram in the interests ‘of clarity. [Lill be useful to add block detail during circuit design Some block details may differ from the final implementation, but functions re- rmiin, For example, the splitter and phase shifting functions are often combined in ‘quadrature combiner circuits operating at RF. We sometimes show a 90-degree phase shift in one path with none in an ‘cher where actual circuitry merely main: tains a 90-degree difference These figures offer a glimpse of what test will eaver, Te design of the block ments will each be discussed in indi- chapters, Then, the blocks will be nbled in system chapters relaied to fil phasing, and digital signal processing 1.6 Chapter 1 Basic Block Diagram Elements amplifier. Provides net power gain. Mixer. Provides an output frequency that is a sum/dif of input frequencies. (AD) Oscitlator. Generates an output at \U) a single trequency. Conbiner/splitter. adds two signals or 43-] splits one into two parts while isolatin then. © _o, Inputs/outputs. Coax’, speaker,| ® 4] pe B aicropione, neapiones: \] Low Pass Resonator a Bandpa All Pass Filter (Phase Shift network) $ Fig 1.5—Common block diagram loments. weave ausio 1e/iee Fig 1.6—Block diagram of a direct conversion transceiver1.5 AN IC BASED DIRECT CONVERSION RECEIVER tion effort ‘The basis for this receiver is the NES6O2 (or NE612) integrated circuit. Originally introduced by Signeties in the late 1980s the chip is easy to use and offers good per= This reeciver design is one of the sim- plest possible that will allow CW and SSB. signals to be received, It offers perfor mance enough for oa-the-air contacts While serving as an introductory construc- super-netroayne Single-sideband ‘Transceiver ovue nop tgter chats SRS Sa eine wae Fig 1.7—Block diagram of a super-heterodyne SSB transceiver. rnput Hf 080.9 BB rower ap. SA t ete cnaioy | Jormance among very low current receiver components, The NE6O2 coatains a mixer and an oscillatar, two essential blocks needed for areceiver. The mixer in direct conversion szoziver serves lo heterodyne the incoming antenna signal directly davsn to audio. The oscillator provides mixer LO (local oscillator) injection Tor this conver= sion. The oseillator within the NEO? isu single transistor followed by a buffer am- plifier of undisclosed complexity. The NE6O2 mixer is a doubly balanced eiceuit of @ type known as the Gilbert Cell with ‘opecation outlined in a later chapter. ‘The LM386N audio amplifier follow ing the NE602 completes the receiver. The LM386N will drive a small speaker. or headphones of high or low impedance. The ideal sot of “cans” tose with this recelyer isa light weight pair of the sort used with ingging receivers or similar consumer gear. ‘The receiver is shown schematically in Fig 1.9. Our version is built using the ugly” methods outlined earlier. Ifyou use 1 pre-etched and drilled cizcuit board, ake ‘the time to study the board layout in detail ‘und trace the cireuit while studying the schematic diagram. Merely stuffing parts and soldering will provide you with no more than soldering practice ‘The signal from the antenna connector is applied to a pot that serves as a gain contro! with ourputrouted toa single tuncé cireuit using LI. a toroid inductor. This cireuitdrives ihe mixerinput at NEBO2pins 1 and 2. The load within the IC looks like fa pair of 1.3-kQ resistors trom the input pins to a virtual ground ‘The NE602oscillatorhas acollector tied to the positive power supply. The base of that transistor ig available at pin 6 while pin 7 goes to the emitter. Internal bias resistors etthe voltage and establish a cur rent of about 0.3 mA in the Colpitts oscil- lator, Feedback capacitors in our eizcuit sum between pins 6 and 7 und from pin 7 to ground. A 270-pF capacitor then ties the base fo the rest oF the tuned cireuit, A simplified version of the oscillator cizeuit is shown in Fig 1.10. This illus luates the way a simplified eircuitis used to calculate the resonant frequency. Fig 1-104 shows the complete escillator. But the 10 680-pF feedback capacitors have a series equivalent of 340 pF, as shown in part B of the figure. In going from Fis LOB to Fig 1.10C, we resolve the $0-pF variable and 10-pF fixed into 8.3 pF; the 270and 340 pF become 150 pF. We evala- ated both variable capacitors at their maxi- mum value, Fig [,10C has nothing but patallel capacitors which add directly to Getting Started 4.7L4.L2, 20 t #26 on 787-6 toroid for 6.9.7.6 MHz Fig 1.9—Direct conversion 7-MHz receiver using two integrated circuits. 10 270 340 I B) © dg) ! 4.8 Chapter 1 1.10—Simplified version of the oscillator in a NE602. See text for explanation. form Fig 1.10. A simple resonance cal- colation shows tuning to 6.9 MHz ‘Two variable capacitor (C1 and C2) are used in our oscillator. They are nearly the same value. The larger, CL, directly paral- lelsthe inductor. A detailed analysis shows that it will tune over a wide zange, the full 6.910 7. 5-MHz spain, C2 is"paded dawn" with a 10-pP series capacitor. C2 has a value ranging from 5 to 50 pF. The series capacitor then generates # composite C ranging from 3.3 t0 8.3 pF, a 5-pF differ. ence. Add capacitance in parallel with C2 tocreate even greater banulspread (resolu tion or low tuning rate). Al fixed capacitors should ideally be INDO ceramic types, readily available from ‘major mail ordcr sources. But, don’t hesi- tate to try other caps if you have them in your junk box. The werst that will happen is that the receiver will drift more than desired, New parts are easily substituted later. ‘These capacitor variations are doubly significant. First, you ean adapt a tuned circuit to work with whatever you have on hand. For example, common 365-p8 AM broadcast capacitors can be used in both positions with appropriate padding. Sec fond, the use of two capacitors is a very practical means for building simple recciv- ers while avoiding the mechanical com= plexity of adial mechanism. We have used double cap tuning for transceivers in other parts ofthe book. Adapt the circuit to wht you have available “The mince input network at LL that in= jects antenna signals into the NE6O2 uses ‘an inductor identical to that in the oscilla. tor, tuned with a mica compression trim- ter capacitor. Any variable can be used here. Ifa 365-pF panel mounted cap is used, the 270-F capacitor could be re- duced in value. I the only available vari- able capacitor is much smaller than 180 PF, you may have © resize LI, or add or subtract net eapaeitance a Dit 10 hit reso- hance. The inductance ean be reduced by spreading or removing turas, or increased by compressing turns. Both cisouits ace very tolerant of such changes. ‘Onee the mixer has been wired. most of the receiver is finished. The LM386 is a low power part with no heat sink required. ‘This receiver draws only 7 mA when siz~ nalsaze low, with more current with louder signals, A simple 5-V power supply works well. A 6-V battery pack will run the receiver for extended periods, ‘The NE6O2 mixer features excellent LO to BF isolation. This means that there is litle LO energy appearing at the mixer RF por, and henee, the rocviver antenna terminal “The presence of such encrgy can lead to a eommon problem of “tunable humm” withFig 1.11—Direet conversion recel some direct conversion receivers. ‘Tho seeciver also has problems, Some, the audio images, a intrinsic to all simple direct conversion receivers. This is the price, hut also the thal! of such # design. ‘Thesolectivity is lacking, This ean be rem- died with audio filters that can be placed in the receiver. Examples of audio filters are Found elsewhere in this book, These filters would go between the mixer and the audio amplifies. It is easy to add such things a bresdboarded receiver, but more difficult with a printed board. “The greatest periormancedefieiency isthe sembly. poor strong signal handfing capability ofthe receiver. Although helped s bit by placing the only gain conteal inthe ancenna lead, the problem is intrinsic tthe NE6O2 mixer. The basie Gilbert Cell is capable of much more, but only when biased to draw considerably more current, The current is kept low in the NEGO? by design, fori is intended for bat tery powered consumer equipment and not hom gear. Strong, high performance direct ‘conversion receivers are described later in the book. Initial wen-on and adjustment is straight forward. Apply power initially with a 1.6 A REGENERATIVE RECEIVER There was a time when simple vacuum tube regenerative circuits were the only receivers available to the radio amsteur Even when super-heterodynes. became possible, the regenerative design remained 2g the entry level radio, Regenerative receivers have become popular again, but they now generally use semicondsctors, Much of this popularity has been fueled by the work of Charles Kitchin, NITEV 58 People now build re- ‘generative receivers for the sheer joy of listening to a receiver that is extremely simple, yet is capable of receiving signals from all over the world. The radio offered here tunes from 5.5 10 16 MHz. covering three amateur bands, 7, 10.1, and L4 MHe, as well as international short-wave broad- ‘casts at 6, 7, 9.5, 12, 13.5, and 15 MHZ ‘The core of regenerative receiver is the detector, Fig 1.12 shows a IFET ver- sion of a classic regenerative detector us- ing a “tickler coil.” Signals from the enna or a preceding radio frequency amplifier are applied to the tuned citeuit, producing a voltage at the FET gate. This produces FET drain currents that vary at the RF rate, The RF drain current flows in the tickler coil which couples energy hack tothe original coil through inductive tans- former action. Irenough energy is coupled ‘back, the eireuit oscillates. Even when the coupling is weaker, insufficient for oseil- lation, the eireuit can have very high gain. ‘This makes the weakest signal large within, the detector cireuit. The presence of any large signal in a “square-taw” device like a JFET will produce detection, which 100-2 resistor in the power supply line The resistor serves as a fuse if yeu have done something drastically wrong. Insert ing the headphones when the output capacitor is uncharged will produce an audible pop. ftir audio seems to be work: ing, turs the receiver off, emove theextra resistor, and startagain. Attach an antenna, advance the gain control and tune C1. Sig- nals should be heard. Adjust the froneend tuned eireuit for maximum signal. If you have a ealibrated signal generator you ean inject signal and see if the operation is at the right frequency, If you have a general coverage receiver available, yowcan attach the antenna of this receiver 10 that of the general coverage receiver where you will beable tohear the LO signal. If'an antenna is not available, you ean throw 20 or 30 {eel of wire out on the floor. While this is rot going v9 compete with a good outdoor antenna, i¢ will provide signals in abun~ dance to listen to and confirm receiver ‘operation, ‘The receiver in Fig 1-11 was built for the 40-meter band. Ifyou wantto try a diferent band, all that is required is to change the two inductors. Increasing the 1.16-UH in- ductor 4.5 tH will drop the receiverright inco the 80 meter band. A band switching version would be practical ‘The first popular receivers of this sort appeared in the USA in a OST paper by WASRNC~! Variations of a similar sort were generated and published in Europe by George Dobbs, G3RIV. George used a double tuned circuit in the front end to improve signal handling properties means that audio also appears witbia the cireuit. It need only be coupled out and applied to headphones or an audio ampli- Fier to complete the receiver. ‘Our receiver uses some slightly unusual circuits that simplify the design. The de- tector is based upom a little appreciated variation of a traditional Hartley oscilla. tor, a variant withou transformer action. Instead, to series inductors, L1 and L2, serve as the traditional “tank,” or resoma- lor. Toroids were used. although Q is not critical and traditional cylindrical eoils Will also work, Indeed, low Q radio fre- ‘queney chokes offer opportunity to the ex- perimenter. The detector, Q2, uses a junction field cffect transistor. While we used 2 2NS854, the detector worked well withany N-chan- Getting Started 1.9‘Tuning nel depletion mode FET we could find in ‘our junk box, This included the U309, B10, 2N4416, 2N3819, and MPF-102, a8 ‘well as some even more obscure parts. We couldn't find a FET that would aot work. ‘Use what you have !The complete receiver schematic isshownin Fig 1.13, anda front panel photograph appears in Fig 1.14. We wound our own I-mH choke for L3 Fig 1.128 classic regenerative detector. using large ferrite bead. A I-mH or 2.5 mH RFC will work well in this posi- tion. A_L-K resistor even functioned in placeof 3. although the regeneration con- trol was not as smooth as it was with an inductor. ‘The mechanical complications of a dial mechanism axe avoided by tuning the te ceiver with two variable capacitors, C2and C3, each with elarge knob. C2 is a“*band- set” while C3 isa higher resolution “band- spread” tuning, an action resulting from the series and parallel fixed expacitors around C3, Regeneration iscontrollad with ‘nother 365-pF variable capacitor. None ofthe variable capacitor values aretetzibly critical. Ifyou find others at a flea market orbamfest. ou can adapt the cirenittouse thom. That's part of the charm of a person- alized regenerative receiver: it applies positive feedback to your imagination, This eireuit uses am RF amplifier. QU ‘The gain isnot really needed, or even de- sired. However, the amplifier provides a relatively stable driving impedance for the detector, and isa convenient way of vary~ ing the strength of the Signals arsiving st the detector. The RF amplifies is preceded by a 5th order low pass and 3rd order high pass filters. The high pass rejects signals from the AM broadcast band that could overload thereceiver. The low pass alten ates FM and TV broadeast signals that could inter-modulate in the RF amplifier ‘or detector, producing distortion Within the receiver tuning range. Audio gain is provided by Q3 driving Lz 20t #22 768-6 Lat St #22 130-6 13: 1 mH, 30t #28 FR43-6301 2,3, 14,5: Le: 2) 1,3,45 1 365 pF see text 12t #28, 730-6 #26 TS0-! 2n3904, 262222, ete. 02: 25454, see text. Di,2: 1Nd152, or any si sw. Fig 1-13—A regenerative 4.10 Chapter 1 elver tuning from 5.6 to 16 MHz. See text for discussion of parts and construction.Fig 1.14—Front panel view of the regenerative receiver. Fine Regen Ul, acommon LM386N output amplifier This will drive either low impedance Walkman” type phones or a small speaker. Walkman isa Sony trademark. Q4 isan active decoupling filter that provides hum-free de to the detector. Although the receiver af Fig 1.13 is shown with a 12-V power supply, it worked well with volt .es as low as 6, Typical current is 20 mA 12V, A Signal generator with frequency counter is useful during initial experi ments with the receiver. However, many ‘builders may noc have them available. Fig LAS shows a suitable substitute, a crystal oscillator that willl eperste anywhere ‘within the receiver range. Numerous ines- pensive crystals are available from the popular mail order sources that will pro- Vide a starting point. For example, 9 10-MH7 crystal available for under $1 will mark the 10.1-MEz amateur and the 9.5 to 10-MHz SW brosdeast bands, ‘The receiver can be built in any of many forms. A metal front panel is a must. af fording shielding between cireuitry and Fig 1.16—Alternative regenerative detector. the operator's hands. However, the est of the receiver could be as simple as a block ‘of wood found in the garage, Our receiver ‘was built “ugly” with scraps of cireuit board material, One scrap will suffice, although our receiver used three, an indie tor of earlier experiments, Other bread boards will work as well, but a printed cic- cuit board should mever be used for a regenerative receiver. Even if dozens are te be built, such as in a club effort, the project should emphasize open ended. flexible breadboarding to encourage ex perimentation. Some experimentation may be required to set up the regeneration, Increasing L2 by a umn or decreasing RI will both increase regeneration, However, toomuch inductance at L2 or too litle resisiance at RI will produce sueh robust feedback that regeneration cannot be stopped or easily controlled. Operation of this, or any regenerative receiver isa multiple contcol effort. Begin with the regeneration control, CA, at min mum capacitance, unmeslied, and set the vo ning controlsat halt. Setthe RF gain for maximum gain, +12 V on the ampli- fier, with the audio gain in the middle and attach an antenna. Tuning C2 may produce a signal, Now slowly advance the regen= ration, adding C at C4. It is normal for background noise to increase with « mild “plop” occurring in the hewphomtes a the detector begins n oscillate. I'the detector becomes overloaded. reduce the RF gain control, Tune the receiver untilan AM sig nal is found. Then reduce regeneration until the “squeals” subside, CW ane! $SB fre best received with the regeneration well advanced. While the receiver works best with an outside antenna, it will func tion with as litte as a few feet of wire tacked co the wall. The signal generator of Fig 1.15 requires no more than a to foot piece of wire on its output, somewhere in the same room as the receiver. There are numerous interactions. be- teen controls, features that offer chal- lenge and intrigue for the experimenter who takes the time to enjoy them. Numer- ‘us cireuit retinements are available to the coxperimenter who wishes 10 continue the quest. The experimenter will discover a ‘areat deal from his or her efforts in operat- ing this receiver, The availability of very high gain through positive feedback ean be used to great advantage. But operation ccan be a greater challenge than found with su more advanced receiver, A more reeent experiment used a differ- cent regenerative detector, shown in Fig 1.16. This circuit eliminates one of the vVariuble capacitors used in the other cir- cuit, replacing it with a pair of potentiom- cers, This circuit was featured in a recent issue of SPRAT by George Dobbs, G3RIV. although the circuit seems to he the brain child of GI3XZM.? Performance ofthe two circuits is similar Getting Started 1.111.7 AN AUDIO AMPLIFIER WITH DISCRETE TRANSISTORS cur literature is ich with older + using high impedance head- nies, These designs are often very bat- fry eificient, a vital performance virtue portable or emergency equipment. But zh impedance headphones that can be used with the more efficient designs have necome rare. The answer to this dilemma is @ simple audio samplifier that will drive low impedance headphones while main: ‘ining reasonable elliciency One solution (© the problem is one of ‘many integrated circuits. Throughout the book weused the LMAS6 orop-ampstodsive headphones of the Sony “Walkman” vati ty. An alternative circuit is shown in Fi 1.17. This amplifier uses commonly avail able diserete transistors. The version af the circuit that we built used leaded pasts, but could just as well be built with SMT compo- ‘nents. QI functions 2s a gain stage. The 2.2 KO collector loud (R8) with 100-0 degeneration (R4) produce QI bias eurrent ff 2 mA for an approximate voltage gain oF 20.Q2 functionsas a floating voltage source thar establishes bias for complememary ‘emitier-follower output transistors Q3 and Qs. Negative feedback through R3 reduces {gain and establishes overall bias. This ci 1.42 Chapter 1 Fig 1.17—Simple audio amplifier using discrete components. cuit is similar 19 many of the simpler inte However, the diserete solution is available sated circuits, This eircuit functions well when an IC is not. Al of the transistors in ‘with power supplies from 5 10 15 V. this circuit are very inexpensive and usu- ‘Aa [C is usually the preferred solution, — ally found in the experimenter’s junk-box1.8 A DIRECT CONVERSION RECEIVER USING A DISCRETE COMPONENT PRODUCT DETECTOR The ditect conversion receiver de. scribed earlier used » NE-602 integrated cirouit to fulfill hath che deveetion and the local oscillator factions. Discrete (non- integrated) componentscan also be usedin these applications. The receiver shown in Fig 1. 18usesadifferential amplifiers the product detector, This design, shown for ‘operation inthe 40-meter hand, has been built with both traditional leaded compo- sents and with surface mounted teclanol= ogy (SMT) parts and appears in Kig 1.19. ‘QI functions as 4 local oscillator. Volt- age control is used with any of several eom- ‘montuning diodes. The Colpitts circuit uses small powder iron toreids for both leaded and SMTcomponents. Cis a combination ‘of NPO capacitors, selected during construc- tion to resonate at the desired frequencies. With the parts shown, the receiver tunes over about a S0-KH range in the 40-meter band. The range may be expanded by paral- Jeling additional varactor diodes, inereas- ing the value ofthe 82-pF blocking expaci tor, decreasing the value of the 2.2-k0 resistor in series with the tuning contro, ar combinations of these measures. ‘The oscillator is buttered with Qa common-emitter amplifier with emitter de- generation. This eizcut, using negative fee: buck, usesaform found throughout the hook, ‘one where an add component reduces gain to improve pertormanes. The ouput drives the mising product detector consisting of 3 and Q4, An RF signal is extracted feom the antenna through a zain eontro, low pass file tered, and applied tothe base of QS where it iswmplifieciand converted toxeurrem' source Feeding Q3and Q4, The mixercollectors are bypassed tor RF. ‘The detector ourput feeds adifferential sig nalte aLM3¥6 audio amplifier. De-coupling became important with this stage, owing t the internal resistance found with a normal 9V battery. An uncomfortable “howling” os cillation disappeared with high decoupling ‘capacitance for the audio amplifier. ‘sur Top views . ) Fig 1.18—Direct conversion receiver using ee, TE Blatte cotter and detector commoonets ano wan Ingrid eete sr uae for he alo ce] me eo empiier snd for wolage eplton 7, Butloul io us discrete component: Tis an recive suhati tor consteion ik pote either leaded or StIT components. Getting Started 4.131.9 POWER SUPPLIES Among the many tools needed by the cireuit experimenter, beginning or sea soned, isa power supply. Indeed, several are always useful. Batteries serve well for simple, low current applications, How: ever, the more useful power supply ex- tuacts energy from the power mains. That ac voltage is applied to a transformer. is rectified, filtered with a large capacitor, snd regulated with transistors and/or inte- ‘rated citeuits. ‘Two major design questions are pre- sented to the beginner: What transformer should be selected and how large should the filter capacitor be? Fig 1.20 shows an example 12-V, 05-A design we use 10 address these questions, Transformers are rated for RMS output voltage with a load. The peak voliage will be higher by a factor of 114, so a 12.6-V transformer will bave a peak out putof 17.8 V. The transformer current ral- ing should equal or exczed the maximum desired de current, so a 0.5-A transformer is adequate for this application. This is shown in part A of Fig 1.20. A switch and protective slow-blow fuse is added to the transformer primary. ‘A bridge rectifier using four diades is added to he eirevit 1 generate a de ourput. ‘The bridge is preferred over circuits with just two diodes, fora center tapped trans Tormer is then not required. Bridge recti- fier diodes should have an average current rating above the maximum power supply current. 1-A diodes would be fine for this application. Some waveforms ate shown in Fig 1.21 ‘The “before filtering” voltage isthe result of rectification forthe circuit a Fig 1.204. Tho “V-cap” trace shows the voltage across the capacitor when itis added to the circuit, Fig 1.208. The significant decal is the ripple, or variation in unregulated out put voltage occurring at the filter capaci (or. Fig 1.22 shows ripple for twodifterent capacitor values when the load current is OA. ‘A suituble regulator is the popular 7812, This three terminal regulator IC will pro vide the desired output with a dvopou of about 2.5 V. Dropout is the minimum volt age difference between the regulated ute putand the higher unregulated input. With 4 2.5-¥ dropout, the unregulated input most be 14.5 V ox more over the entire eycle. Fig 1.22 shows that « 2000-UF ca: pacitor will be adequate, but 500 UF will not, If we define AV as the difference be- tween the peak rectified voltage and the ‘minimum unregulated value, 17 ~ 14.5 = Jas the output current, and tas the fime for @ half eyele (0083 second for 60 Hz). the minimum capacitor value in 4.44 Chapter 1 2 rectifier Be Rectifier + Filter cap Re cireust Rc Rectitier, careuse wilter cap. Fig 1.20—Fundamental power supply. Part A shows the transformer and rectifier, B adds the critical output Filter capacitor, while C uses a 12-V regulator IC. Fig 1.21— Wave- forms for a simple . fon foo, | power supply. The "betore filtering shows the raw testified signal without any’ i capacitor, The “ cap” shows the voltage across the filter capacitor attached to the rectifier when loaded to a modest (oe current. o Fig 1.22—Wave- n forms showing the voltage across titer capacitors of we i ‘wo values vihen loaded with 0.5 A. ‘See text ws . eet __.. | discussion.unregulated Input Regulated output Fig 1.23—Extending ‘the output current ‘capability of a Fegulator with a “wrap-around” PNP wansistor. Fig 1.24—Practical dual output power supply featuring the LM-317 regulator. 1.10 RF POWER MEASUREMENTS [Before one can do uny meaningful ex- periments with transmitters, you must be thle to measure RF power. A basie scheme ‘ordoing this is shown in Fig 1.28, The RF «applied to the 30-£1 termination through cable, Iris necessary that a well ‘defined impedance be available to absorb the transmitter power. The load must be capable of dissipating that power in the form of heat. So if the transmitter is ca pable of delivering, for example, 100 W. ‘he 50-62 load resistor must be capable of dissipating this power. The lad must be 2 resistor that really appears as « resistor to Peak a Detector Voltmeter Farads is given by [eat av For thisexample, Eq 1.1 predicts amin mum C of 1700 uF. A practical value of 2500 LF would be 2 good ehoice. ‘Tae complete circuit with the regulator is shown in Fig 1.20C. Extra eapacitors placed close to the regulator IC. serve to stabilize the IC. Phe user should check di sheets for the IC that he or she uses lo evaluate stability. The 1-k@ bleeder resistor consumes Tittle current, but gui antees thal the supply turns off soon after ‘the switeh is opened. ‘The 0.5-A rating of the 7812 becomes a problem when more current is needed. Fig 1.23 shows a circuit that will extend the output current rating by adding a power transistor. QU now carries most ofthe cure rent with the split being determined by the ratio of RIKI. Tae dropout for the total circuit is now that of the IC plus 2 little more than a volt for the diode/transistor and RI and R2. Fig 1.24 shows a supply using a LM317. ‘This is 2 programmable voltage pari that can supply outpuis from 1.2 up t9 37 ¥, set with two resistors, for an urput current of L5 A. The power supply we built, used extensively for developing many of ihe eir- cuits in this book, was variable voltage ancl also included a 12-V regulator as a second output. An 18-V transformer was used, for wwe wanted regulated outputs up to 20 V. ‘Many other regulators are found in ven- dor cataloys. many with considerably igher ‘output currents and lower dropouts. The ex- Perimenter is encouraged to build his own circuits using them, Switching made regukt- tors offer interesting performance virtues with equally interesting challenges. (Eq 1.) P) never Getting Started 1.15the radio frequency applica to it, This ‘means that the usual power resisters sold by vendors, even if capable of dissipating 1100 W. will not be suitable. They are us ally built as a “wire wound” part, making them highly inductive for RF. It is some- times possible to tune them, an interesting avenue for the advanced experimenter: Suitable $0-0 terminations, or*“dummy fouds” ean be built with parallel combina- tions of 2-W carbon resistors, or similar 2 for 3-W metal oxide power resistors such as those manufactured by Yaego or Xieon Some of these are used in power attenus- (ors described in Chapter 7. The RF power dissipated in the resistor will develop a corresponding RF voltage. ‘That is rectified with a simple diode detec- tor, providing a signal acres the capacitor equaling the peak RF voltage, less 0.7 V for the diode turn-on voltage. “The powermeter is completed with x suit- able de volt meter. It can be as simple as a O-LmA. current meter anda resister. aFET voltmeter, oreven a digital voltmeter Fig 1.26 shows a dual range power meter, Essentially it is a pair of power merers sharing a single meter movement. ‘The higher power part of the ciecuit stats with e-4-W load built from two parallel 100-2, 2-W resistors. These ean be car- bon or metal film resistors. IF 2-W resis- tors aze not available, four parallel 200-0 1-W parts will work as well. The resulting RF vollage is rectified with a silicon switehing diode, This should be a L00-V part suchas the IN4148, IN4152, or simi- lardiode. The velimeter part ofthe circuit sa 20-k@Qresistor driving a0-1 mA meter Assumeatansmitteris attaches and keyed con to produce an indication of 0.6 mA. This represents a peak of 12 V. tor the meter ma: tiplier is the 20-40 resistor. The resulting power is then calculated from the formula aiven withthe figure, 1613 mW, er L6 W The S0-mW input to the power meter uses a Single 51-9, YW, resistor wich more sensitive IN34A recifier diode. The meter multiplier is now just 1.5 kQ. An approximate calibration curve is shown in Fig 1.27. The finished meter is shown in Fig 1.28, Other schemes suilable for RF power measurement include terminated oscillo- scopes, microwave power meters (usually ‘using calorimeter measurement methods.) specirum analyzers, and wideband loga- rithmic iniegeated cireuits. Some of these will be covered in a later chapter. ‘Olten we wish toexaminean RF voltage 10 sec ifa circuit is “alive.” und perhaps to adjust it. The classic method for doing this used an RE probe with a high impedance, usually vacuum tube or FET voltmeter ‘The method is still yery useful, especially 4.16 Chapter 1 P(milliwatts)= 10 (¥ + 0. ( curve} (Use Calibration (2 W scale) 4 watt a Input ont sae Fig 1.26—Dual ie. range power 2 Q t W input uses “a 2 i the formula to ealculate power = = inmilivetts, ‘The S0.mW range 50 nm uses the curve of Input imam | oy Fig 1.23. on srubuBene Fig 1.27—Callbration curve for the 50 mW range of the previous power meter. Fig 1.28—The front panel of the dual- range GRP power moter. To High z Voltneter Fig 1.29—RF probe suitable for uso with a VTVM, FET voltmeter, or even @ DVI, Resistors marked with * are standott r have little Impact on circuit operation. when instrumentation is limited. Fig 1.29 shows « very simple RF probe. The photo in Fig 1.30 shows an open breadboard ver- sion: its the sort af eircuit chat one builds when a measurement must be done imme diately. A long lasting version of the same circuit might better be built inside a cylin- Ger at the end of the coaxial cable. The probe may require calibration. This is best done with one of the other power tors used for probe construction and meters and a small transmitter or similar RF source, The transmitter is attached the power meter and the ouput is mea- sured. The corresponding RF voltage is noted and the RF probe is attached to the powcr meter $0-0 resistor. producing & result that can be compared Fig L.3f shows nhiah impedance de vole meter suitable for use with this probe. [Lis also a good starting measurement tool forFig 1.20—Close up view of an RF built on a strip of PC board material. ‘The probe is a capacitor load. use in the lab. For gonoral utility. iis useful tohave the 5.1-MA2eesisto at the tip end of probe that is inserted intoacirevit for mea ssurcments, This allows the de to be mea- Fig 1.31—Simple high impedance voltmeter for measuring de voltages in circuits. It can be used with the AF probe of Fig 1.29 and Fig 1.30. a TA] eso SP a 1 ai sured without upsetting signals that may be present in the circuit, This circuit can be calibrated with a fresh 1.5-V battery: vary 1.11 A FIRST TRANSMITTER ‘This section describes the design of a simple transmitter suicable as a first rig. a project for someone who has never buill transmitter It uses robust circuits with few adjustments roquired during construction. I-can he built with nothing more than a volt meter, power meter. and power sup- ply. We used an oscilloscope and a spec tram analyzer during the tig design phase and those results are presented. However, that equipment is not necessary for con: sirnction, The crystal controlled 2-W 40-meter transmitter is built with bread- board methods cather than with a printed circuit, “The circuit, shown in Fig 1.32. begins ‘with QI functioning as a erystal controlled oscillator, Our crystal had a marked tre- {queney of 7045 KHz. This was the speci fied frequency for operation with a 32-pF load capacitance. This Colpittscircuituses 1 pair of series 390-pF feedback cupaci- tors, The equivalent 195 pF parallels the crystal. Because this eapacitance is much larger than the specified 32 pF. the opetat- Ing frequeney will be less than the marked 145 kHz Ifyou want the frequency to be exact, a small trimmer capacitor ean be placed in series with the crystal. We will eventually do thisas a method of obtaining some slight ning, but don't bother with chistefinement in the beginning. The com- plexity of erystals Is discussed im later ehapters. The oseillator is built on the end of sorap of circuit board material. The erystal was held on the board with # piece of double sided foam tape (Tess, 67601). The oscillator worked right off with several V peak-to-peak observed at both the hase and the emitter with an oseilloscape and 1X probe. The RE probe described earlier could also be used, The oscillatar Func: ‘oned well with supply voltages as low as 2.5 V. A quick check with a reeciver confirmed the frequency +12¥ De 100 3 Fig 1.32—Crystal controlled oscillator that Ie the start of the Beginner's ‘ransmitter the 6.2-KO resistor if needed, We will have more to say about RF power measurement in Chapter 7. ‘The oscillator és followed by a butter amplifier. A buffer is an amplifier that allows power 10 be estracted from an ‘oscillator. rather stage, without adversely disturbing it, An ideal buffer often has a high input impedance so it can be attached without extracting any power. The best dnuffers have good reverse isolation, mean- ing that any signal present at che output is Dreavily attenuated at the input ‘The first buifer ried was an emitter fob ower, common choice to follow a crystal oscillator, Performance was poor. While the loading was light. Une output was highly distorted. This problem behavior is dis cussed in detail in Chapter 2. The desizn ‘was changed to the degenerated common emitter amplifier shown in ig 133. We obtain the butter input from the oscillator base instead oF the more common emitter for the waveform is cleaner, simewave-ike, at that point ‘The buifer is added to the crystal ascil- lator by soldering the requited parts to the board orto other components. The board is not installed in a box at this time. Rather it's loose where it is easiest 10 build and measure. We ean tuck solder small loud resistors or coax connectors co he board to facilitate experimentation. The busfer output transformer has a 4:1 ‘turns ratio. The primary, the 12-tum wind- ing on a FB43-2401 ferrite bead, or a T97-13 toroid, which is virtually identi ‘eal, has un inductance of about SOuH. This Getting Started 4.47has a7-MHz reuetance of 2.3-k0. The load ‘on the output is tunsformed from $0 Q up by the square of the tums ratio 10 $00.2. the approximate impedance presented (© the collector of Q2, The inductive reac- tance is much higher, so it does not impact the cireuit operation. ‘The ouipot is not tuned, allowing it 10 function Well over a wide frequency range ‘We measured the power from the 3mm ‘ouiput link on T! by attaching a small length fof coas eable that ran tothe S04rW input of the power meter described cater. The aul pat was +10 dBm. 10 mW, with R1=2702, and was up {0 +15 4Bm with RI of 150. Recall that the power meter has a 50-2 impedance We want more than 10 enV from our transmitter and willeventually add « power amplifier ro reach an ouiputat two W. That ‘amplifier will require modest drive of 200 6 300 mW. We could obtain more power by biasing the second stage for higher gain fand output. A more conservative and stable, free iromself-oseillation, approach adds a thitd stage ‘The evolving design isshown in Fig 1.34 with « class C amplifier for Q3. We want this third stage to provide a power gain of 16 and pick anotter 2N3904, With an F, mote than ten times the operating fre- quency, gain will be good. The 2N3904 also has a beta that holds up well at high currents, a useful characteristic for apower amplificr. While we wanted class C opers- tion in the 3° stage, stability was deemed vital, so the eireult is degenerated with « 10-0 emitter resistor and a 100-0 load is placed at the base. Class C operation is assured. Q3 current disappears when RF +12 pe 2aKt 1 s1e 1003 1223 | Fig 1.93—evetving tranemiter | 13) TE | Shematte showing the addition of a Fig 1.34—A Class C driver amplifier, 03, 1s added to the transmitter. 4.48 Chapter 1 drive is removed from the amplifier, The desired driver output power is ‘A W. This can be realized by properly loading the stage. We must present a re sistive load to the collector given by Wee —Y, DP, (Bq. 1.2) where V.: is the supply, Ve is the emitter voltage, and Ry is the load resistance in Ohms, (Vaq-¥Q) is about 11 V for this ex ample, so the equation predicts a desired oad of about 150 Q. An Lenetwork, L1 fand the 200-pP capacitor, is designed te transform a 50-0 loud to “look Like’ 200 @ at the collector. An RF choke pro- Yides collector bias for the transistor While mnable components could have been used in the L-network to get the op- simum output, we elected t0 use fixed vale nies. We measured LL und set the value to that desired, We then used a 5% value for the 200-pF capacitor. Variahle elements are only needed in highcr Q situations, or ‘where itis not possible t9 find tight toler ance components. Power output could be measured with the 4-W position of the watt meter. We used an alternative approach here. A 31-0 's-W resistor was tack soldered into the sireuit at the outpur point shown in Fig [34 and the outpus voltage was measured with an oscilloscope and 10X probe. The (Q3 output was 123 mW. 7 V peak-to-peak atthe load, with R1=270 82 in the butter. Changing RI 10 150.0 increased output 10 314 mW. The DC current, 43 mA, was de> lermined by measuring the voltage drop across the 10-0 decoupling resistor, The calculated efficiency is then 62%. good for fan amplifier which coatains resistors in both the emitter and collector. The 2N3904 at Q3 is operating well within ratings, Gen- erally, « TO-92 plastic tansistor like the 2N3904 can dissipate a quarter of a watt for extended times. or hall a watt for the shomterintermittemt periodsencountered in 8 CW transmitter. This “cule of thumb” can be strotelied with hect-sinking, or eas ily violated in thermally isolated settings. Owing 10 the good efficioney. the dissipa- tion is only 200 miW in Q3. (03 power output varied smoothly from very low levels up to the maximum 314 mW as V,_ was adjusted from 5 t0 12 V. ‘This is generally u useful method for ex: amining stability, We will eventually adda “drive control” 10 the citeuit. Before continuing werced toaxldress the issue of spectral purity. Some observed waveforms have departec froma sinewave, This means that these waveforms areharmonie-rich. This transmitter uses a crystal oscillator operating at the output frequency. The only signals that should be present anywhere within the transmitter fare at 7 MHz of harmonics at 14, 21, 28, ... MHz. The only filtering needed is 2 Jow pass filter at che transmitter output While the L-network that makes 4 50-02 load appear as 200 02 at the Q3 collector has a low pass characteristic, it has only {wo components and is not very effective asa filter. If the driver amplificr is going to be used by itself as a transmitter, ane cother low pass filter should be added to the output, There is, however, Tittle value in auding a better low pass filter after the driver if itis to be used only 10 driveanotherstage which will also he ore- 1.12 A BIPOLAR TRANSISTOR POWER AMPLIFIER ‘The project now stars to get exciting as wwe begin to experiment with higher ouput powers, The transistor we have seleeted for 2.2-W power amplifier (PA) is 2 2NS32L ‘This is 2 NPN device in a TO-39 ease with collector dissipation of 10 W in an infinite head sink, or 1 W in free aie, 50-V break- downs, the ability to switch acurrent of 2A, ‘and a S0-MHz Fr, all for less than $1. The low Fy restricts the device to the lower bands. hut it also means that high frequency stability will not be an issue. The 2-W PA schematic is presented in Fig 1.35. ‘The first detail we must consider with the PA is heat sink. Our intention was to increase power by about 10 JB to the ? to 3-W level. Ifefficiency turns out tobe 50%, se will have a collector dissipation that is the same as the RF output. The transistor can’t support this power without a heat sink. We had 2 Thermalloy 2215 in the unk box which should be more than ad equate. The transistor was mounted in the sink which was then bolted to a PC board scrap. Holes through the board made he leads available for soldering. Be care fol to avoid any short circuits that are not tended. The transistor case is attached to the collector terminal in most TO-39 pack- aged deviees, I's always difficult ro estimate heat sink sizes. While one can do thermodynamic calculations, it’s generally adequate with ‘small transmitters 10 experimentally creat the problem, Touch the heat sink often dur- ing initial measurements. IF 1's 0 hat 10 tovch, he heat sink isnot large enough. We always seem to err in the conservative area ‘with more heat sink chan is needed ‘The formula presented in Eq 1.2 shows that a 25-9 load resistanee presented to the collector will support the desired ourput. A simple pi-aetwork was designed. The net: work Q was kept low, but was picked 10 generate anetwork with standard, and junk box available, capacitors. A matching net ‘work design is presented in Chapter 8 A33-V Zener diode is atached from the calleeior to ground. The collector voleage will never reach these levels wich normal CClass-C operation, so the diode is uanspas- tent except for the sometimes substantial apacitance tha it adds to the collector eir- cuit. But, the diode conducts ifthe output loud disappears, und provents collector breakdown that might otherwise destray the ating harmonic distortion. Spectrum ans- lyzer_ measurements showed spurious driver outputs at -27, ~30, 43, and 49 Be for the second through fifth ha monies when the driver was delivering ‘ull output. Theharmonic suppression was, actually worse at lower output levels. The term de refers co dB down with respect to the carrier transistor. Care was taken 1 keep the emit- ter lead short when the amplifier was buile for even small amounts of inductance can alter the performance, This is nor always bad. ‘Transmitter testing afvays begins by ale taching « 50-82 load to the output. This can he a power meter or a resistor ofthe proper rating. The PA should never be run without a load, ‘The frst PA we built forthisproject used the simplified circuit of Fig 1.36. This cir- cuit suffered from instabilities whieh be- ‘came clear as We varied the drive from the earlier part of the transmitter. At one point, the RF output and the collector cureent both changed abruptly. The oscillascope showed frequencies well below the desired ? MHz. Changing the collector RF choke feom the original 15 UH toa smaller 2.7-2H molded choke moved the Frequency up, but the in stabilicy was stil present. However, chang- Sng the base circuittoone witha lower drive Smpedance completely solved the problem. “The output power and collector currentnow vary smoothly as the drive is varied, The base transformer ig a 2:1 turns ratio step- down that now drives the base from a | -TuH yare ans2572 Sav 72, 5 bifilar 12=12t#22, 750-8, space over half core. Q5+2N5321 with Heat sink feose 750p, Mica ens #22, FB: +01 pK +O aot Fig 1.35—A 2 W power amplifier. Fig 1.36—Earlier simplified PA design which suffered with stability problems, See text for discussion. Getting Started 1.1912.5. source impedance. The 33-0 base resistor absorbs some drive nd tends 10 st Dilize the amplifier. Changing this resistor is one of the experimental “hooks” avail able to thc esperimenter fighting instability ‘The 2-W amplifier is installed in the teansmitter. An autpue power of 2.25 W resulls ftom a drive of just over 100 mW Increasing the drive produces higher out- When the 2-W amplifier drive is adjusted for 2.25-W output, the measured cfficieney was 47%. A spectrom analysis showed 24 and 3°/harmonies st ~36 UBe and 47 dBe. Addition of un outboard low pass filter removed al spurious responses to better than —75 dBc “The outbound low pas filters shown in jg, 1.37. This is a T!-order Chebyshey design with a 7.5-MEL ripple eutof! ire quency and a ripple of 07 dB. The rather obscure ripple was picked to fit standard value capacitors that were on hand. The inner capacitors are parallel combinations ‘of 680.and 180 pF. The measured insertion loss for the filter was 0.11 UB at 7 MHz, The filter was built into & small aluminum box, Fig 1.38, as an outboard appendage so iteould be used for other projects. Also, tho performance is superior when the shielding around the filier is absolute. Tf tho same filter was built into the transmit ter. theze is a greater chance that ground curgents and radiation could provide paths for signals to leak around the filter. ‘This extreme fillering is probably re dundant. A much simpler filter could be built into the transmitter, near the output put But once the output gets much beyond 3 W, QS begins to heat. Although a higher power tas observed with che oseile Joscope when the Key was first pressed, the power decreases over a period of a Jew seconds before stabilizing. We inves- tigated this by looking at the collector waveform at differing drive levels. When riven to 2.25-W output, the collector voll- age varied between 3 and 23 V. As drive .13 AN OUTPUT LOW PASS FILTER ‘coax connector, for adequate harmonic at renuation, Chapter 3 provides detail. Practical Details ‘The modules built so farare mere seraps of circuit board material sitting on a bene with short pieces of wire to tie them. 10 ether, They need to be refined and puck aged to create a transmitter that we ean put fon the air. An almost complete sehematic of the transmitter is shown in Fig 1.39 ‘The first refinement is a keying ciceuit, This function is performed by Q4. a PNP switching infogrator. This is afavor- ine keying scheme of ours, allowing a grounded key to control the positive sup- ply to a transmitter stage. Keying in the positive supply allows the grounded pares of the circuit to remain grounded without over being disturbed by keying. Q4 serves the additional function of shaping the key~ ing. When the key is pressed, current begins to flow in the 3.9-KQ2 resistor. The current flows from Q4 base whieh “ties” totum Q4 on. As the Q4 collector voltage begins co increase, the change is coupled back tothe base through the capacitor. The Fig 1.37—Low pass filter far use with the experimental transmitter. 41.20 Chapter 1 increases. the hottom of the collectorswing drops toward zero. But at this point the amplifier is fully loaded. Further excur sions are not consistent with simple class C operation. More drive will eause bigher current with litte increases i auiput, al Towing efficiency to decrease. This causes the heating, Changing both the matching neiwork and drive power is needed for higher output. positive going signal opposes the current extracted by the 3.9-K0 resistor. Hence, the ealleetor does aot swith immediately toa high state, Rather, itramps upward at an approximately steady cate until Q4 be~ comes saturated. Forcing the stage to tra ‘onsmootaly over couple oF milliseconds restricts the bandwiuth ofthe modulation related tofumning the cavier on. That band width will extend a few hundred Hz on ther Side of the carier. Beyond that, no clicks will be heard in a good recciver. ‘A power output contol is added to the emitter of Q2. Owing 1 the class C nature ‘ofthe following amplifiers, the outputco: two will allow the sramsmiter to run from the maximum cusput down 10 virwally nothing. The contol is a screwdriver ad- jisted pot mounted an the board A variable capacitor, Cl, s added tothe erystal oscillator. The capattor used in our teansmittr tuned feom 5 t0 80 pF and pro- vided a tuning range of 3 to 4 KHz. Use whatever is it your junkhox. While cer tainly nora substitute for a VFO, it allows the user to dedge some interference. A “spat” switch, $2, allows the oseillator co function withoot placing 2 signal onthe air. Finally, « transmit-receive system is scdded. This function is performed with « multipole toggle switeh, a simple but ad Fig 1,98—Inside view of the 7-eloment low pase fiker bull to {g0 with the beginner's rig. The filter is also used with other equipment.VEO Freq. control, ete2e, TIH6L 1,22, 93203904 PRs 20#22, 750-6, space over half core. with Heat sink ga=203906 2b #26, 3t link #22, FR43-2402 5 bifilar turns #22, 402 95=2N5322 wt Tee r? Receiver 51 + 5 To antenna I or Tuner. Fig 1.39—A nearly complete schematic of the transmitter. This version combines the PA with the earlier stages, adds shaped keying, power output adjust, T/R switching, and VXO action. ‘equate solution. SIA applies the +12 V supply 10 the oscillator during transmit periods. The supply i always available 10 }3 and QS and does not need 10 be switched. The keying circuit, Q4, controls the supply reaching Q2. S1B switches the antenna from the receiver to the transmit: ter. The miniature toggle switch at SI is suitable for powers up through a few watts More refined T/R methods are presented From Receiver +12V Audio out bo Hey 300 ~ ta alt RE sic aay 220% —y ra gu mL pt Headphones 2 10K od 06=2N3906 Fig 1.40—Sidetone oscillator for the transmitter. This circuit i code practice oscillator. elsewhere in the book. iis transmitters tobe used witha high quality modem receiver with a wide AGC range, @ two pole switch i al chat is needed at S1-The user can then listen to the trans imitter inthe receiver as the key is actuated. ‘The more common scenario places this transmitter witha simple direct conversion receiver such as that described earlier in this chapter. [twill then be impossible to tum tbe sainin that receiver down far enough 1 pre vent overload. An answer to the problem is presented in Fig 1.40 where asidetone oxcil- lator is added to the system. A 555-timer imtcgrated circuit functions as the square wane oscillator which is keyed on and off with QS. Q5 base current routes through 2 10-K8 resisior attached to the key in Fig, 1,39. R2 must be adjusted for the head phones used sth the transmitter. The head Phones are disconnected from the reciver using transmit intervals, atached only 10 the sidetone oscillator. Two phone jacks are included on the transmiter. A short cable ten routes the receiver andio output from the receiver (0 the transmitter where it is switchod, This scheme does not prevent the receiver from heing overloaded, but guaran- tees that you don’t have to listen when it happens. The receiver won'tbe damaged by Getting Started 4.21Fig 1.41—Overall view of the complete transmitter ‘onsiruction. Fig 1.42—Outside view of the Beginner Station, At left is the beginner's direct conversion receiver with the transmitter at the right. the overload. A third poke is needed on the switch for this refinement. Three pole double throw toyele switches are umususl so te used one with four poles. ‘The complete transmitter is packaged in ‘a standard box as shown in Fig Lf. This fone measured 2x 3.5 x 6 inches, although whatever is available will work, Alterna tively, you can build your own box. The ‘outside of the box can he fined t© be as attractive s you would ike it be, consis- tent with personal tastes. The variable c&- pacitor, C1, the spotting switch .S2, and the TIR switch are located on tae front panel as shown on the right hand side of Fig 142. The key jack and a headphone jack are also located on the front. The rear ppane] contains power receptacles, a jack for the angio input from the receiver, and coaxial connectors for the antenna and a ceable tothe receiverinput. The box we pur chased for the transmitter bad gray paint ‘on it, Unfortunately, it had nearly as much paint oa the inside us was on the outside, Fig 148—Tho inside view of tho transmitter shows the capacitor and TR switch fixie paint was removed where campos mounted tothe front panel with power and coaxial connectore on the rear. The left as Tena” where cone board contains the first three stages while the Fight board contains tha 2-W power "ents ere grounded tothe case, Details o amplifier. A hest sink is under thet board. A small board under the T/R switen ‘He internal construction appear in Fig Gontains the sidetone osc 148. 41.22 Chapter 11.14 ABOUT THE SCHEMATICS IN THIS BOOK The schematic diagrams used in this book differ slightly ftom other ARRL pub- lieations in that we use slightly different conventions, Not all details are presented inal schematics. Capacitors are in microfarads iPelectro- Iytic or if they have decimal values less than (. Values greater than unity ate in pi ‘cofarad if they are not electrolytic. Elec tuolytie caps always have a voltage rating fareater than the Voc oF Vg value used REFERENCES 1, W, Hayward and D. DeMaxt, Solid Stave Design for the Radto Amateur, ARRL, 1977. 2. R, Hayward und W. Hayward, “The Ugly Weekender,” QST. Aug. 1981, pp 18-21-See also G. Grammer, Understanding Amateur in the circuit with 25 V being typical. In some applications we will use C values in AP, which stands for nanofarad. 1000 pF Uae. RF transformers are specitied by turns ratio rather than impedance ratio, Often lis datas presented within the schematic diagram rather than a8 part of a caption. ‘The same holds for inductance values. We sitive to load the schematie with as much information as possible. Radio, ARRL, 24 Buition, 1976, p 144. 3. R Lewallen, “An Optimized QRP Transceiver,” OST, Aug, 1980, pp 14-1 4. J. Dillon, “The Neophyte Receiver.” QST, Feb, [988, p 14-18. 5.C Kitehin,“A Simmple Regenerative Radio We generally label schematics with the parts that we used. Bur that does not mean ‘that this is what you might want touse. An example is our frequent use the IN4IS2 silicon switching diode. In all cases, vir= tually all of these can be replaced by the more common 1N4148 or INB14. When. there is a question about such details, look the part up and see if the parts you have on hrand are similar. Then try the substitu tion for Beginners.” OST, Sep, 2000, p 61. 6.C. Kitchin, “Aa Ultra Simple VHF Re ceiver for 6 Meters." QS7, Dec, 1997, p29. 7. G, Dobbs, “A Stable Regenerative Receiver,” SPRAT, Issue 105, Dec, 2000, pil Getting Started 1.23Amplifier Design Basics 2.1 MODELING SIMPLE SOLID STATE DEVICES Small signal amplifiers are used in a receiver 1o bring weak signals up to the point that hey canbe heard inheadphones. Large signal amplifiers in transmitters ere ate even Targer signals that, when applied toan antenna, propagate to be heard by the receivers, Clearly, the amplifier function {central to all that we do as eadio experi enters, Before we get into the details of the amplifier circuits, weexamine devives that can amplify. preliminary look at diodes soon evolves into « discussion of bipolar and field cifect wansistors. But, prior to that, we examine the modeling process. Even the simplest electronicdeviews ean be very complicated in their overall bbchavior, especially ifall power levels and all frequencies arc considered. Such a complete description can be overwhelm: ing. Indeed, such a camplete device pie ture would conceptually bury the salient behavior that the designer may seek whem Fig 2.1—Forward biased junction dod. he or she uses a device, What is needed is, something simpler, a modet with enough complication to be useful in practical applications, hut with no extra frills We use models for even the simplest of parts, A resistor for example, is madeled as an ideal element, a past thet ol (Ohm's Law. with no other characteristics The real resistor is more complicated even the smallest surface mounted part has capacitance and inductance. Wire leads only make the effects larger. The Land C alter circuit bohavior, hut can be described by more elaborate models, The Junction Diode ‘The first device we model in detail isthe junction diode. The diode is a deviee thet ‘has polarity dependant properties, Specifi- cally. if we insert an ideal diode in a func~ tioning de circuit that carries a current, the cizeuit will be unchanged by the presence of the diode if tne polarity is for “forward Dias." But, current flow will cease if the diodeis reverse biesed. The schematic dis- gram of Fig 2.1 illustrates a forward bi ased diadc defined by this beliavior. Re= versing the diode leads eliminates eusvent flow in the circuit, ‘The current in the cizcuit of Fig 2.1 is shown ia Fig 2.2, a curve called an LV charaeteristie. The current is that flowing through the diode and the voltage is that across the diode, Fig 2.2 plots a current that is completely determined by elements external (o the diode. This particular part is culled an “ideal” diode, A real world digde departs from the jdleal. Firs, a slight voltage drop appears across the Forward biased diode, Current remains very small until that level is exceeded. Secon. the flow of dinde cur- rent causes a slight additional voltage drop. A refined model with these charse~ teristics is shown in Fig 2.3. The model becomes an ideal diode. # 0.6-V battery and a diode resistor, Ry that i the ratio of 4 small increase in applied voltage, AV. and the resulting small change in current AL. We sometimes refer 10 the threshold (0.6 Vin the figure) asa diade affser volr- ‘age. The offset will vary with diede type Silicon junction switching and rectifier diodes usually have an offset of 0.6 10 0.7 V. Germanium und hot-cartier silicon diodes will have lower values, while some compound semiconductor parts have Fig 2.2—IV Characteristics for an ideal of perfect diode. The curve shows | for any possibio V that might be applied to the Ideal diode. Amplifier Design Basics 2.1mw) Curent v Diode Bias, Votes Fig 2.3-1V characteristic for a rotined diode model. thresholds exceeding one voll ‘The model of Fig 2.3 is more accurate than the ideal diode, but is still less than perfect in some situations. A much better diode representation is 2 mathematical model where current is given by an equa- ion ats 8 = Is wun Bq 2 Tete" Es ‘where I, is called the saturation current in amperes, q is the charge on an electron, js Bolizman’s constant, and T is the diode temperature in degrees Kelvin. The see- ‘ond, approximate form is common. This mode], known merely as the diode equa son, is Mustrated in Fig 24 for the ease of 100 K (near room temperature) and 1 3x10 A. a value that we inferred from measurements for the popular IN4148/ IN4I52 series of parts, Changing Is generates new offset values. The diode equation is also significant because it briginates as a description evolving from bbasie physics. Physics based models are generally preferred because they follow from fundamentals, even though they may not bo as intuitive More refined diede models will include reverse breakdown, igh frequency parameters (inductance and capacitance.) and even carrier lifetime, No matter what methods we use to apalyze 2 circuit, the results ofthe analysis will only be as good as the models. SMALL SIGNAL DIODE MODEL The antenna signals that our reveivers amplify are often in the microvolt region for less. We ask how the diode would 2.2 Chapter 2 Fig 2.4—1V characte follows the dio behave if one microvelt was applied to it ‘The current flowing in the diode, Eq 2.1, would be essentially zero if a microvolt was applied directly. But, the diode might hhave a much different response if the diode already had a bias current flowing. Fig 2.5 shows part of a diode 1V curve. ‘The point corresponding to 5 mA DC cur rent flow is marked with a cangent Tine The slope of this line defines a resistance, 4 change i current for an applied change in voltage that occurs when a small signal is applied to the biased diode. The diode has a resistance of about 5. when the current is 5 mA, generally represented by 26 Fq22 ‘The factor 26 mY (or ,026 V} comes from differentiation of Eq 2.1 and isa very ‘common parameterin semiconductorelec~ tronies: xr ATS a6 Eq 23 ’ 4 [A small signal diede model is n0 more than a simple resistor. We will make extensive use of small signal models as we The Bipolar Transistor ‘The bipolar transistors athzee terminal device. If we use the same equipment that ‘we used to examine diodes, we might eon: clude that the bipolar transistor is just « pair of diodes in onc package, attached as shown in Fig 2.6. This is an incomplete yet useful model Let's place this model in a test circuit, shown in Fig 2.7. A variable voltage lerstic for a common Junction diode. This squation. source with a large base resistor is used, allowing us te control base current. A posi- tye voltage is applied to the collector, reverse biasing the collector-base junction. ‘The two-diode made] would prediet zere- collector current, But, collector current does flow in proportion to the current in the base. This is transistor action, The ratio of collector ty hase eurrent is usually si Fig 2.5—Small signal model for @ junetion diode represents it as 3 resistor with the slope shown. See text. NPN ) ? e Fig 2.6—Apparent model of a bipolar(a) : aN ‘On a5 1 en b a. Fig 2.7—The circult we used to bias a bipolar transistor tor active operation. See text nified by the greck Ieter beta, B. A typical value might be 100, ‘The simplified model on the right side of Fig 2.7 i clearly in error. The “collee- tor" diode is reverse biased by V,., yet considerable current flows against the diode arrow. A better model is shown in Fig 2.8A where the original diode pair is supplemented by a current source propor: ional to the eutrent in the base-emitter diode, The model in Fig 2.8B is the model ‘we use for evaluation of biasing circuits. It rncglects the collector-hase diode snd re fines the base-emitter diode SMALL SIGNAL BIPOLAR TRANSISTOR MODEL ‘What happens with the bipolar transis- torfor small signals? How do we model it? The methods used with the diode are ex panded to describe the transistor, as shown in Fig 29, In Fig 2.9A, the input diode is replaced for small signals with a resistance. The resistance is exuetly that used with the earlier diode, 26/1 where Tis now the DC current in’ milliamperes for the base-emitter diode. The current amplify ing properties that we discovered eatlier ace preserved forsmall signals, so the small Signal collector current remains at Bx, We use a lower case “i” co emphasize the small signal levels, ‘An alternative small signal model is shown in Fig 2.98, Hore the zesistanee in series with the base has been replaced with cone in the emitter. This resistance, termed re is given by 26 BAT Eq 24 where I, is now the emitter current in mil liamperes. The collector current exceeds that in the hase by B, and the emitter cur- rent is the sum of the collector and base values, so the de emitter current is greater than the base value by (B+), Accordingly, the emitter resistor of Fig 2.9B is smaller than the resistor of Fig 2.9A by (B+ 1). Both models are equally valid, although that Fig 2.8—A currant sourea is added to the dode pair to form a representative model, The diode is often ignored as in 8. fa) Fig 2.9 Evolution of a small signal trar using x, is more common, Common emit: ter small signal amplifier input resistance Ry = (ri) Fa25 A trations viewpoint emphasizes the bipolar transistor as a current controted device wih B representing curent gain But bec can vary considerably fora given transistor ye suggesting that he ampli fer gain may aie for diferem wane tos. ohich # notte, A preferred smal signal mode is shown in 9g 2.9C, where the partis viewed asa vtiage driven com ponent. The output curren source is now Specified by a tamsconductance. ig = 8m °¥o Eq 26 ‘The transconductance, Bis given by (ma) 1 6 Bn Eq27 While 6 may vary among transistors, gy is well defined by emitter current Another festure of the model is illus- trated by a simple amplifier design, shown in Fig 2.108, An NPN transistor is biased with & base resistor attached to a positive supply. A load resistor, R., is placed in the collector, The base resistor is adjusted uniil the emitter currentis | mA. Thesmall signal model shown in Fig 2.10B is used for analysis With T mA emitter current, the trans- conductance is gq=1/26. Signal current is then v2. This current produces an out- put voltage because it flows in R., result- ing in & voltage gain of g,,*R,, Which is G eRe, Eq 28 Knowing biasing details. voltage gain ean be predicted “by inspection” as a ‘resistor ratio, independent of beta. Current ain, or Bis still of signifiesnce, for it will alter the signal current that flows. when drive voltages are applied, which defines input impedance. Note that we have said nothing about transistor ype. Our discussion has consid- ‘ered the NPN, but has sai litte else of specific nature. This is not an oversimpli- fication. Much of the utility of the bipolar Iransistor eesults from properties. that Amplifier Design Basics 2.3Fig 2.10—The simple amplifier at A is analyzed with the small signal model at 8. depend primarily upon the standing emit tee cucreat, BIPOLAR TRANSISTOR BIASING Acourate transistor currentis vital toany design, becuuse current determines small ial properties. The power dissipation, the power output capabilities. the distor tion, and even frequency dependence are also detcrmined by bias current and valt- age. Biasing methods will be evaluated with the model of Fig 2.8B, where the base-emitter junction becomes an ideal diode with a 0.6-V batiery. Colleetor cur- rent is then Bx, ‘The fisst bias example we consider is shat shown in Fig 2.11. A I-kQ2 load resis- torappears in the collector, while the base is biased [rom the 12-V supply through a 100-KO resistor, The model assumes an ‘offset of 0.6 V, so the base current is 11-4 V across 100 KO, or 114 WA. If cransistor +100, the collector current is 11.4 mA. But, the 1-KO collector resistor produces tun IR drop of 11.4 Y, leaving 3 collector voltage of only 06 V. Repeating the calculation with slightly higher B predictsa negative collector volt- ‘age, impnssible without a negative supply. Recall that earlier models included a eollector-base diode that prevented the collector from being moze than a diode Arop below the base, Whenever the collec- tor voltage equals or drops below that of the base, for an NPN, the transistor is said te be saturated. ‘The scheme of Fig 2.11 is,at best, «poor bias method. Slight changes in beta yield great uncertainty. Biasing is improved with negative feedback, with one form shown in Fig 2.12. The 100-k2 resistor is biased from the collector rather than the 12-V supply. An intitive examination shows that this is an improved method, even before we “crunch” any numbers. If the beta changes 10 drive the transistor toward saturation, the current through RL 2.4 Chapter 2 will decrease from the reduced collector voltage. A lover than nominal beta will ‘cause collector voltage to climb, forcing more base current to flow: Application of the model and some algebra provides gencrsl equation forFig 2 Vo Rit Vey “BPR, vee Me BRR, Eq 29 {An even beticr bius scheme is shown in Fig 2.134, where the base is driven from av Re Ri i 100% — a Fig 2.17—A simple amplifier used for bias analy Vee = +12 ¥ Re RL 1K 100K EH —y Fig 2.12—Improved bias Is obtained ‘rom the collector. the positive supply through a voltage divider. R, and R. The equivalent circuit tor the divider is shown in Fig 2.13B. The base voltage with the transistor tempo- rarily removed is found from divider R. Ve = Vee RRS RoR. Bq 2.10 ‘where the pre indicates thatthe base is ‘open circuited, and absent from the cale lation, The emitter voltage is below the base by the 0.6-V offset, placing the emit- ter voltage at 1.45 V. The emitter current is then determined by the 330-9 emitter resistor as 4.39 mA, The collector current is almost the same as that in the emitter, and the drop across the collector load puts Vat 761 V) “This. analysis, although close, is in crtor, Buse vurrent flow produces an IR drop inthe biasing resistor chain. This decreases the base voltage below the value shown inFig?2,13by about0.25 V. There are two solutions to this problem. One would replace Ry and Ry with a “stiffer” voltage divider. Valuos of 3.3 kO and 6802 would ‘work well, but atthe price of greater power consumption, The other slternativeis amore careful analysis. If this is performed, the cemiller current is given by ‘The I, value for the components in Fig 2.13 is 3.759 mA. PNP biasing is identical co that of the NPN, except that the voltages are mes sured with regard to the positive power supply, which may or may not be “ground.” See Fig 2.14. Fig 2.15shows a natural refinement the biasing scheme, Here another resistor is added, anormal parcofadecouplingseheme. ‘The added resistor provides negative feed- back like that used earlier in Fig 2.12. This, incombination withthe feedback from Ry of Fig 2.13 further stabilizes bis. ‘A scheme useful for biasing an NPN transistor with a directly grounded emitter is shown in Fig 2.16, A PNP transistor emitter senses the de collector voltage sid ‘compares it with the PNP base at a refer- cence, V,. established with voltage divider Ry and R), The reference dividers usually designed to put most of the power supply ‘on the NPN collector. The 0.1-UF capaci tor stabilizes the negative feedback bias, loop. With the values shown, the bias is defined byb.o5 + 8) PNP biased to the same nga we established with the Fig 2.19—Evolulion of base bias from a voltage divider. NPN example. y, = Nec“: veo ew R 4, Re Ry R-depl 90 06 eg2a2 Re R1 ~ Transistor Although the bipolar tansistor is our work horse, various forms of field effect transistor, ot FET. ae close in popularity Among FETS, one of the most commen is the junction Variant, the JFET. A JFET is much like vacuum tube triodes of the past and is easily biased and used in amplifier applications. FETs, including the JFET, generally lack the uniformity and predict= ability of a bipolar transistor. IFETS tend to be low noise devices. Not only is the noise figure low, but the low frequency flicker, or “I/F” noise is small. This com. bination makes the JFET especially useful for low noise oscillators. Fig 2.17 presents the test setup that al- lows us to meusure, and then model the JFET. The example is an N-channel Depletion mode JFET. A drain power sup- ply, +Vyy 18 applied. The gate voltage is then varied while examining the current that flows. Fig 2.18 is a resulting plot of rain current vs gate-to-source voltage ‘with constant drain voltage, The gate volt- age is negative for most of the curve. The re can be no more than 0.6 V positive for the gate of a JFET is actually x diode junction, The metal oxide silicon field ef ‘ect transistor, MOSFET. has similar prop erties, but uses an insulating gate. There is hen no diode clamping action, Once gate-to-source voltage drops oan adequate level, drain current goes to zero and the FET is said to be in “pinch-of1” The pinch-off voltage, the gate-source V where current drops to (ar neazly 0) zero, Vec=12 R2, 6.8K negative feedback to the biasing an emitter resi veo RL 203906 FSult - + atpue Input Fig 2.18—A “wrap-around” PNP biases an NEN with grounded emitter. The 0.1-uF capacitor stabilizes bias and is ‘the dominant element in the bias loop. is at -3 V for the example of Fig 2.18. These data arc typical for the popular 310 JEET. A drain voltage higher than the magnitude of the pinch-off is usually required to ensure lincar operation. This is Fig 2.17—Test setup used to evaluate a FET. ‘often called operation in the saturation region. Saturation is just the opposite con- dligion in a FET from sanarier in a bipo- lar transistor. Fig 2.19 shows the usual source resistor method used for biasingan N-Channel JET ava current below Lj, The current flow= ing through the resistor establishes 4 positive source voltage. As current increases, the source voltage increases, ‘causing the gate-to-source voltage differ ence to become more negative. This is the action needed to decrease current, eventue ally stabilizing the bias. The action of an external source R is a form of negative feedback, just as we used with an emitter resistor inthe case of a bipolar transistor. Fig 2.19 includes some JFET equations SMALL SIGNAL JFET MODEL Fig 2.18 showed s complete curve, escribing large and small signal behavior as well as JFET biasing. The simplified small signal model is shown in Fig 2.20. Here an open gale terminal accepts an input voltage. That sigmal them controls an Amplifier Design Basics 2.5aw) ™ A FET amotter © eo Buse FET equa psa Frnt aoe 7 siete Tiga [a i eae tet lige Fe estan I GRP ae) Fig 2.18Drain Current ve Source-to-Gate Voltage for J310 {ype Junetion Field Etfect Transistor. l,,-35 mA and V,=-3 V. Vpis the voltage where drain current g&s to 200. lz, fs the dain current when the gate and source are both at the samo potential. Fig 2.19 -JFET bias circuit and equations. The lot circuits @ amplifier, while that on tho right is the bias Equivalent. Pick a desired drain current, Ig (must be less than toga) and tse the miedo equation to find‘ne requires source {alstor. The resulting source voltage is given by Ohm's Law output current source related to the input by a transconductance, ui, With gts { , Me ve Ut y) bees Forexample, if we biased the FET fora gate voltage equaling half ofthe pinch-off value, with Tdss=35 mA and Vp=-3 V. the smal signal transconductance is 0.0117 Having examined busie device models and biasing, we now evaluate some basic amplifier designs, first with the bipolar junction transistor (BJT) and then the junction field effect transistor GFET). We begin with a single stage audio design, Fig 2.21. The cixcuit that we might builds presented in Fig 2.21a, while wbius- ‘ng relaied partis showa in Fig 2.21b. The vollage divider, 10 kQ and 3.3 KA, creates aan equivalent source of 2.481 V at the base. This decreases hy 0.6 V in moving through the transistor to produce an emitter voltage of 1.881 V. Tho emitter current is then 1.881 mA. With beta=100, base current is 19 HA, well belaw the 752 HA in the voltage divider, The collector voltage is. then 2.6 Chapter 2 2 AMPLIFIER DESIGN BASICS 10-1.881=8.119 V. The collector 1o-emitter voltage, Ve. is 6.238 and power dissipation 1s the product ofthis voltage with the stand- ing cosrent, 11.73 mW. ‘Small signal transistor characteristics are established by emitter current, The resulting small signal model is that in Fig 2.2le, The |-KO2 emitter resistor has dis- appeared from the circuit for it is well bypassed by the 100-uF capacitor. The small signal ris 26/1,(mA)=13.82.0. The input resistance looking into the base is almost 14 k= 1,x(B+1), ‘The input source isa |-mV voltage gen- erator in series with a resistance of 1 KO, which might represent a previous stage. ‘The source is AC coupled to the base Fig 2.20—Simplified small signal JFET model. for “amps per volt.” From the equations in Fig 2.19, we see that the DC drain current is then 8.75 mA, which is realized with a source R of 171 Q. The low frequency in pur resistanee is essentially infinite through a 10-HF capacitor which has L-kHL reactance of 16.0. Being very small compared with the amplifier input or the souree, it may be neglected for a I-kHz analysis, The same argument may be made for the output capacitor. The result is the small signal citouit of Fig 2.214, The power supply is missing in the small sig nal models where Vc is replaced by ground: the supply is fixed and does not change with audio signal current. so itis effectively 4 signal ground. We characterized the BIT by a trans conductance, 2,7-0.0724 ampivolt, Also, we noglect any effect related to the base bias divider on the small signal model, The I-mV input signal is voltageFig 2.21—Single transistor audio amplifier design. See text for details. divided between the 1-k@2 source resis tance and the 1.39-kQ input resistance ‘The base input voltage becomes 0.582 mV to produce a collector signal current of i=BnXVp.=0.0421 mA, This current flows through a resistance of 333 Q. the paralle] ceguivalont of the 500-0 load and the 1-kO collector resistance. The output voltage is then 0.0421 mAx333, or 14.02 mV for a lrcuit voltage gain of 24.1, Note that this is also exactly the ratio Ri Gy Keg 2.14 where the load is the otal impedance seen by the collector. ‘The form of this equation is especially intuitive, emphasizing the role of r, 3s 4 degeneration resistance. If we placed 10-Q resistor in series with the 100-uF emitter bypass capacitor, the net emitter resistance would be 10++ 13.8223.8 9 and the voltage gain would become 14. Tae role of emitter current is clout: Inereasing standing emitler current causes 1, t0 decrease, increasing voltage gain. Emitler degeneration is a common feedback scheme, We have tated the bipolar transistor as voltage controlled device. Beta was indi rectly used in the calculation, but only © settransistorinput resistance. This. in turn estublished the fraction of the I-mV input voltage that appeared at the base. ‘There is a counter intuitive nature tothe modeling presented in Fig 2.21D. The schematic shows the inputs tiedto ground through r,. che 13.8-0 resistor. whielt ‘would severely attenuate the signal, How= ever, the current source representing the transistor alsoattached to the inputnode, fand that current moves in unison with the input voltage. This yields the results outlined We calculated a voltage gain. The gains of geeater interest are power ratios, One of Smierest 10 the RF designer is, simply, ower gat, the output power divided by Input power, The output power is cale lated (For Fig 2.21) as V2/R where R is the '500-{2 load and V isthe 14.02-mV output ‘Output power is then 3.93 « 107 W. The input poweris the base voltage (0.582 mV) across the transistor input R of 14 KO, or 2.435 x 10-10 W. The power gain is the ratio of the uso powers, 161d relationship, this beconies 32.1 dB. This is high but reasonable for a single transistor. for this amplitier operates at low frequen. cies, Such gain from a single tansistor at redio frequencies is more uifficul Power gain is fundamental but is not always the gain we measure, We usually measure Iransducer power gain, espe- cially when working with RF circuits Transducer gain is output power delivered toa load vs the maximum power avail- table from the input generator. We have already ealeulated ouput power. The available power from the source is the povier that would be delivered to 8 termi- nation that wus impedance matched to the generator. The generator was a I-mV ‘open cireuit source behind a 1-KE2 resis tor, so the load that would allow the mai mum available power to be extracted would be a 1-KO resistor. The available input becomes 0.5 mY across | k&, or 2.5 % 10-10 watts, leaving a transducer gain of 1572, of 32.0 JB, This is newly as high as the power gain, ‘The gain difference is a consequence of the input impedance mis- mateh, We will have move to say about ‘gains and dB later in this chapter. A common practice converis a voltuge gain to decibel form with the familiar 20*Log(G,), 27.6 dB for this example. This is nor u correct result, for the source impedance is not the same as the load impedance. The decibel eonstruct is one that should only be applied to power ratios. It works with voltage ratios only ‘when the related resistances are equal. In the amplifier we analyzed, the input ‘was applied 10 the base while the emitter was grounded through « large bypass capacitor. Hence, the input was applied between the base and the emitter. The out- put was extracted from the collecter-cmit- ler port, This i a common-emitter (CE) configuration, for the emitter is common to input and output. A common-collector (CC) amplifier is shown in Fig 2.22. ‘The complete amplifier citeuit is shown in Fig 2.22A, while the small signal ver- sion is in Fig 2.22B. The open citevit de base voltage is 5 V, so the emitter bias current is 4.4 mA, leading to 1.25.91 © The follower of Fig 2.22B is driven from. a 1-kQ source impedance. Lis terminated ina pair of 1-KQ2 resistors in parallel. The input resistance of a follower is given by Rea (BHI) (e+ RL) Bars while te output impedance is R Rom = Goat Eq 26 ‘The voltage gain for the emitter follower Bq 2.17 Substituting r, into these equations shows that the follower has 4 gain oF 0.988, essentially 1, aceounting for the cireuit zname. Seiting to 100, the input resistance is $1 kO while the output resistanceis 15.8 2° The input resistance and the voliage gain both grow if the follower is lightly Toaded. The output resistance decreases as the source impedance drops. It is very common to de-vouple a fol- lower to & preceding amplifiers this is iilustracod in Fig 2.23, Amplifier Design Basics 2.74 4 wo B won Fig 2.22—Common collector amplifier, also known as an emitter follower. Fig 2.24—Common Base Amplifier with small-signal equivalent ‘The third basic amplifier configuration is tac common base (CB) amplifier of Fig 2.24 ‘The input resistance for the common base (CB) amplifier is 1 Rr Ey 218 ‘The current gain for the CB amplifier is sven by the parameter a, 2.8 Chapter 2 fi Bet which is normally very close to unity. We essentially assume thatthe current injected into the CB amphitier appears at the out- put. The voltage gain is then Gy =a Ry Fq2.19 Eq 2.20 The voltage gain for the CB amplifier ean be very large. However, this is some what synthetic, forthe input impedance is tosually very low. making the amplifier Lifficull to drive. The eommon applic ‘ions use a current source to drive the CB amplifier realized by placing an extra r= sistance in series with the input The CB umplitier has the useful prop= erty that it offers excellent reverse isolie tion, Thats, he input impedance of a CB. amplifier s not affected by anything that bappensto the output circuit. The example shown in Fig 2.24 is biased to a current of tubout 0.8 mA. producing an input resis lange of 32.12 The equations for the small signal prop- erties ofthe various amplifiers aredetived in Inrroduetion io RF Design! and are dis- cussed in The Arr of Electronics. The CC amplifier had a low ourput impedance. Nothing was said about the ‘conimon emitter and common base ampli Fer ourput resistance, Both are essentially infinive for the simple models considered where the BJT is modeled as an “ideal current source, “Most of the amplifier umalysis we have done is based upon simple models, oncs that have but one or wo parameters. Beta has only minor impact on cireuit perfor ‘mane, The dominant element in all of the models is ry, the emitter resistance. This parameter is directly related te current, & parameter under the control of the cireuit designer. This would suggest that all bipolar ansistors are moze alike than they ace differont and that the only majar differ fences are in the frequency capability and size. This is generally an uccurate view of the small-signal bipolar transistor. Smail-Signal FET Amplifiers ‘The field effect transistor fumilies are similar 10 the BIT: as three terminal devices, they can he configured into tarec different forms. Fig 2.25 shows the come mon source, common gate, and common drain (or source follower) coniigurations Tor an N Channel JFET. There are many similarities between BIT and JFET circvits, The common gate FET amplitier (CG) has a low input impedance with « high output impedance. ‘The topology offers excellent reverse isolation, The follower (CD) has a low ‘output impedance with a very high input impedance, IFET bias current is controlled by the designer, just as it was with the BIT. Resistor values may, however, have 10 be deviee specific, picked fora given FET to establish performance. Within a given JET type, for example, 23:1 variation inVad vad Out Fig 2.27—Current gain vs Frequency for abut. 4g. is commen. similar variation exists with pinchotf voltage, Tae combination of these two variables might lead one to feel that it Would be nearly impossible 1© design with FETs, Fortunately. it'saot that bad, for the variations are related to each other. That is, a given JFET in a family With a high T,, Will also tend to have & pincboff with'a more negative value, pro: ducing Jess variation in gq, the dominant small signal characteristic, ‘There is good eason forthe similarities between FET and BIT amplifiers. Many of the properties result from feedback that is added toa circuit by the configuration. For example, the follower has the load in series with the current source. The voltage developed across the load then generates a signal that contributes to the control of the cutteat generator. The JFET has an additional property not predicted by the preceding model, the j switeh section illustrated in Fig 2.26. The JFET functions here as a series SPST switch, Am input ae signal is applied to the ET channel (the source-drain path) und is routed to the output when the control vollage is positive with regard to the ehan- nel. The channel is the current path be ‘ween source and drain, The channel is bi- sed above ground by the voltage divider: ‘The switch is open circuited if the control vollage is more negative with regard «© the channel than the FET pinchoff volt age. The switching FET may be modeled 44s. vollage controlled variable sesistor in this application, Lowest R occurs when the control voltage is at or ahove the channel, The gate resistoris usually large, allowing the control to be several volts higherthan the channel, Although the gate diode is then forward biased, current is small and of litle consequence Virtually all FET types funetion well as switches, Enhancement mode MOSFETs offer the advantage of no gate diode (© complicale the circuit. GaAs MOSFETs fare useful in very high speed switching applications where they may be used for ‘microwave signal contcol. IFETS and Vdd Out S Vecontol i we Fig 2.26—A JFET operating as a series switch, MOSFETs are useful audio switehes in ‘many applications. ‘The FETs may be used as vollage vari able resistors. Ax such, they can function in gain control circuits. High Frequency Effects Little hs been sail about the effets of high fequency. Yer. mach of our intrest ss taioexperimenters isin the perforiance of cransistreicus at requencies ell yond the rage of our simple model, “The firs thing tht happens tothe BIT as frequency increases is that B decreases overthe deandaudio values, This sshown inthe curve of Fig 227 of B vs frequency The low frequency B is shown as By. The frequency where Bdropstoa vslucofunity isealled the curent gain bandwidth prog tet or more offen, just as F,. Dropping to a frequeney of F/2 wil podace B=?. The frequency where B begins to depart from Bs called the “beta culo. ‘The role off of eure gain with fre- quency is modeleé with an acted base Capacitor, Fig 228, The otherelements are generally unchanged, s0 the complete rll tf may be sttbuted to the capacitor across the input. The cirait shown in Fig 2.8 is called the hybriden model ‘AL low frequencies an output signal from a iransisior is either in phase (0 depres) oroutof piss (180 degrees) with the input signal. These simple phase rls tionships no longer hold above the 8 out off where the mathematics change, (aking on a (formally) complex character, A typical BIT isthe 2N3903, This NPN transistor has a typical F, of about 300 MHr anda low frequency Blof 100, This places the B cutoff at about 3 ME. This Alvice will have some phase shift ctfcts stall requencies within the HP specita and higher Amplifier Design Basics2.3 LARGE SIGNAL AMPLIFIERS ur previous small signal viewpoint is now expanded. We will examine over- driven receiver circuits only intended for small signals. A more common large signal amplifier is a transmitter stage, a cireuil intended to funtion at high levels, Distortion isa consequence of large sis- nal operation, Distortion in an amplifier ‘merely means that the output is Something ifferent than a replica ofthe input. A d torting circuit driven by a sinewave will have non-sinewave outputs when viewed in the time domain, experimentally with an oscilloscope. In the frequency domain, the distortion appears as harmonics. A dis- torting circuit driven by two or more signals may contain outputs that are the result of intermodulation, frequencies that are sums and differences of input fre: quency multiples, The BJT model of greatest popularity is fan extension of the diode equation, Eq 2.21 y= Inge! where Ing is called the emitter sesuration cumtent. V is the voltage on the base emitterdinde. The other parameters are the same as appeared with the diode equation in Section 2.1. This is part of the model known collectively as the Ehers-Moll equations, The non-linear exponential behavior is intrinsic to the bipolar transis lor, Detailed use of this model takes us well outside the realm of this text, but is highly recommended for those with such interests.* Many large signal properties of ampli fiers are extensions of simple circuit analysis, Although the details are always buried within refined models, much can be discerned from eureiul analysis with- ‘out analytic complenity. Some examples Will he used to illustraée this, Fig 2.29 shows a simple audio ampli- fier driven with a | KHz signal behind 3 Fig 2.29—A simple aucio amplifier ‘examined for large signal performance. 2.40 Chapter 2 1-40 impedance, We observe an output voltage at the collector. The de base volt- age is approximately M the power supply, sothe emitter iat about 1.8 V. Theemitter current isthen 1.8 mn, producing wde col- lector bias voltage of §.2 V. The emitter current leads to a small sigaal r, value of about 14, Voltage goin is 70 with the I-kQ collector load. The input resistance will be a litle over | kQ if Bis 100. This ‘means thatthe hase signal voltage is just lover hal the generator value. From the bias and small signal analysis, sie predier that an input of 20:m¥” peak at the generator will produce a bit over 1 mV atthe base. The voltage gain of 70, applied to this will give a peak collector signal of 0.7 Vora peak-to-peak value of LV. The 8.2-V zero signal collector value will then move berween 7.5 V and 8.9 V. This is still a long way from the +10-V supply or the 2.5-V base where saturation would be approached. We ‘would expect a sinewave input to generate a sinewave output Fig 2.30 shows waveforms for three rive levels: 02 V, 0.1 V, and 0.5 ¥ peak The sinusoidal output is very close to the values we estimated. However, the other two cases are severely distorted. The O.1-¥ drive case, five times stronger than the initial 20-m¥ input, is enough to cause the output to reach the 10-¥ positive power supply, causing collector current to drop to zero. The other part ofthe cyele is still well behaved with approximately sinusoidal outputs The most severely distorted ouput results from the largest input signal, 0.3 V peak, also shown in Fig 2.30, At the posi- tive extreme, the transistor is cutoff with current having vanished. At the other end, the transistor current is Well beyond the ‘rigs value. The collector has dropped below the base voltage and the transistor is saturated for the bottom, voltage-flat parts of the curve, Simple models predict much of the nonlinear hehavior, without formal analy’ sis. The base-collector diode prevents eol- lector voltages more than a diode-drop below the base. But, the collector current generator is capable of increasing “as needed” to supply larger currents, but only of the prescribed polasity. The larger drive examples would sound very distorted if this audio umplitier was part of a receiver. ‘The next example is a familiar emitter follower that might be on the output of aan oscillator, A follower has a low output impedance, and should, we season, be capable of delivering power 1 a low lmpedance soch as 8 mixer. But this Fig 2.31—Emitier follower to drive a ‘50-01 load. This circuit Is not biased to = veoh) No output load Fig 2.30—Output wavetorms for the simple amplifier at several drive levels,reasoning is flawed ‘The emitter follower circuits shown in Fig 231. A pair of 3.50 resistors bias the base at half the 10-V’ povwer supply and the emitter is biased with a L-kGQresis~ tor. [a4 mA. setting r, to 5.9 . The follower is driven from 200-0 source resistance for an outpot resistance of 79.0. Tr this circuit was going to be used to drive a 50-0 filter, the 50-0 resistance would be realized by adding series 43-0 resistor to the output “This follower circuit is being driven by a signal source witha peakaupliude of V. ‘The input impedance is well above the 200+ ariving source, so vwally al f he wail able generator signal is present atthe base. The modeling pracess is applied to capacitors with the same importance that dtis to wansiscors. capacitoraccurmilates charge through current flys, never allow ing the voltage across the cupacitor to instantaneously change. The capacitor could conceptually be replaced by a bat tery. In no-signal conditions the 4.4-mA. transistor current flows in the L-KE2 bias resistor with zero eusrent in the 50-2 lose. Applying a positive going signal to the base merely totas the transistor on harder. AAs the base voltage increases from the 5-V no-signal level to 5.5 V, the emitter will follow from 4.4 V1 49 V. We now have +05 V on the output load, oreing un output current of 10 mA to flow. The cur rent in the I-kQ bias resistor has increased to4.9 mA, so the oral transistor current is 149 mA. ‘A negative-going base signal produces complications. A small negative base deive of 0.1 ¥ 19 4.9 V would drop the emitter 0 4.3 V, which drops the outputco-0.1 ¥, The current in the 50-0 load becomes -2 mA. 7.045 u¢as) « veer) AY f IN UT Tk Te Bias TTT ST / Tine Fig 2.82—Follower waveforms. 7008 = uceas) + Cent) us 7 aus Hine Fig 2.33—Follower output waveforms after Increasing the standing blas current. With the emitter voltage at 4.3 V, we still have 4.3 mA flowing in the 1-EO resistor ‘The transisior current has now dropped 16 2.3 mA, Because itis still postive, the ran- Sistor i sill controlling the output and the follower continues to follow. But what happens when the drive reaches the full negative value of -0.5 V2 Ifthe lineae, small signal model applied, the hase would drop to 4.5 V, leaving the emitter at 3.9 V with the output at 0.5 V, producing a current inthe load of -10 mA. ‘Bucthe current flowing in the bias resistor ‘would still be 3.9 mA, implying that the ‘ransistor curront would be -6.1 mA. This, is not possible! The transistor can supply current via the model current generator. ‘but chat current cunnot he negative. Fig 2.32 presents the waveforms. The negative going excursion is clipped at the point when the cransistor emitter current ‘drops to zero, leaving all ourput current to Flow in tho L-kO resistor ‘This simple circuit has ilustrated the difference between stall signal and large signal models, Currents of either polarity are allowed in a small signal model. The large signal behavior i restricted co that dictated by the model. inthis case Kimited to the positive current flow predicted by the Ebers-Moll equation, ‘Tae low small signal ourpuc impedance ‘of a follower was a consequence oF tive feedback. The load in series with the ‘output creates a voltage that is applied to the transistor in opposition to the signal driving it. Ff we allow the follower to "run ‘outof current,” the transistor iscutoff with zero cusrent flow. The low output imped- lance is no longer present during that part ‘of the eycle when trunsisior current flow haus ceased Fig 2.33 shows the output after the design was modified. The emitter bias resistor was changed from 1 k® to 330.2, increasing the emitter bias current to 12.6 ‘mA. This is larger than the needed 10 mA, s0 the output remains clean. But, even a slight increase in drive could allow the distortion co resurn. The ubimate refinement might bea complementary out- put such as is found with many audio amplifiers, The next example considered is LO-MHz Class A amplifier intended to develop afew milliwatts of output power ‘The cireuit is in Fig 2.34. The base is biased from a 10-¥ supply through 2 velt- age divider of 10 kQ and 3.3 KO. produe- ing a DC emitter voltage of 1.64 V. The 200-Q emitter resistor sets an emitter eur- rentof 8.2 mA, yielding small signal r, of 3.20. The 50-0 output load sets the small signal voltage gain at 16 A common approximation sets high Amplifier Design Basics 2.11frequency Bat FYE, placing B at 30, This ‘els input resistance ofabout 100 ©, whieh predicts that about 2/3 of the open circuit input voltage will appear at the base. An input signal of 10 mV peak produces ahout 6.7 mVon the base, Applying the small signal voltage gain, the output will be 105 mV peak. Perhaps of greater interest, the load current for this output is 2 mA peak. ‘The transistor collector current varies fromthe quiescent (no-signal) value of 8.2 mA up 10 10.2 mA and down 10 6.2 mA. While small signal charaeteristies are pre- served, the output current is already becoming a sizable fraction of the DC bias ‘A characteristic found with the present circuit that we did not see in earlier ampli- fiers results from the use of a collector RF choke, The induetor has the properties of a constant current source. AS a de current is established in the inductor, the action of the inductor “tries” to maintain that value This allows the collector voltage toexceed ‘co Which never aceurred when a collec (or resistor supplied bias eurcent. This is stiown in plots which follow We now: increase the input deive to $0. sm peak, Thisisa five times increase over the 10-mV case, se we expect a similar jerease in both the output voltage and ‘current if small signal conditions are pre- served. Measorements und computer Simulations both confirm this general behavior, although the output signals depart considerably from sinusoids, Out ppat voltage across the load is about 0.5 V peak. Collector current drops almast t0 2210 at one point inthe cycle but reaches a ‘maximum of about 19 ma, about twice the bias value. Distortion is severe. The amplitior with 0.5-V drives current limited, forthe current drops to zero al one point in the drive eycle. However. the vole tage excursions are still small, The output power with a 50-0 lead is about 2.5 mW. Consider changes in load resistance seen by the collector. If se maintain drive at 0.5 V peak, the collector signal current 2.42 Chapter 2 Fig 2.35—The class ‘A amplifier is ‘moditied with output impedance transformation for higher output power. as stgcon) sine Fig 2.36—50-0 termination on the class A amplifier. NY i 50 aY drive ine pi-pet mateh vith 1K at col Tove us ine Fig 2.97—Collector (upper) and output load (lower) voltages with the pi network output eireutiry, ‘will be the same, Output voltage can, how ever, inerease as Ry, grows. To obtain the maximum power output, we wish to pick & load that sllows the collector voltage to drop nearly to the base value (saturation) while going an equal distance above V.. at the opposite part of the cyele, This voltage excursion should oceur as the current var- is from vwice the bias value dawn 0 zero. ‘The load resistance that allows this isBq 2.22 where Ik, is the de bias value. A more familiar form expresses the load in terms of a desired output power, (Woe = Ve) 2R, RL Eq 2.23 where Ry is the load resistance in Obins, ‘Vecis the power supply. Vj is the DC base 2.4 GAIN, POWER, DB AND IMPEDANCE MATCHING Audio and other low frequency smplifi- ers are easily analyzed with the low fre- quency models used for biasing. Buc most ft our interest is in higher frequencies where measurement difficulties petsist. These encourage us to cousider power instead of the voltages and currents that dominate the view of the circuit theorist ‘This emphasis is an integral part of RF design and forms the basis for this section. "The emphasis on power measurement goes back inearly methods. Power atradio, miemo- ‘wave, and even optical frequencies was mea sured using « Bolometer. The Bolometer is based upon temperature measurements. A resistive load is embedded in a thermally wwell-insulated chamber. The application of RF power causes a temperature increase, which can be detected with a thermometer. But, the same increase in teruperature can be produced with application of direct current Measurement of thedlirectcurrentandrelated voltage then provide a very fundamental voltage, und Poy: is the output in Wats ‘This form applies to Class B and C ampli- ices as well as the class A amplifier under tisoussion, Application of Bq 2.22 predicts load resistance of just over 1000 Q tor mexi- ‘mum output, Changing the load £0 1 kO in tho circuit produces a 10-MHz output of IL V peak-to-peak corresponding to a power of about 16 mW. Even larger resis lunge would have produced voltage limit ing, so this is close to optimum. More often than not, 1000 82 is not the impedance thot the designer wishes to use as 4 termination for the amplifier just termination of the RF power. The other reason we are concerned with poser is that its power and not voltage or ‘current that is more fundamental. Power is the race that energy is transferred. whether itbe arate of dissipation, such a the power that becomes heat in a resistor. or the rate that energy may pass through 2 surface, such as the rate that a radio or light wave passes through a plane. That plane could well be the capture area of an antenna. The unit for power is the Watt (W), or Joules per second, We are more familiar with it ‘being the product of current and voltage. ‘Aa amplifier application is presented in Fig 2.38 consisting of voltage source with telated source resistance, the ampli fier, and an output load. While 50 2 is ‘common for both the source and load, this is certainly not necessary. But, if power is to be measured, we must have some resis: ance, for a voltage across an open citeuit provides no power, designed, Rather, he or she wishes to men sure the amplifier output with 50-0 instru mentation and perhaps drive other circuits swith a 50-0 impedance, The solution is found in Fig 2.38 where an impedance wansforming x-network is inserted be- ‘oveen the 50-2 load and the collector, This network makes the terminating “look like” 1040 © at 10 MH, It also has low pass filtering characteristics, attenuating energy at 20 MF, 30 MHz, and higher tnmonie frequencies. Fig 2.36 shows the collector ‘waveform when the 50-0 load is connected directly o the collector, The waveforms. ter matching are shown in Fig 2.37. Consider the simple circuit of Fig 2.39 consisting of a voliage source. V, and a source resistance, Ry. We will terminate this in a load R, Ohms Law provides the net current, while voltage divider setion aves the voltage across the load, yielding the power Rs pw Vv WS Fig 2.38—A voltage with a source resistance A delivers power to a load R. resource fran vegen, acca ( @ AT Y ite Ingadance Matching Reload | ie gos af & jo 7 in | 2 oaslt | | 3 | cw) ws: Normale Ld Rosotance impedances. load resistance equal Fig 2.40—Power delivered to the load is maximum when the ‘hat of the source. Amplifier Design Basics 2.VOR (eo) Eq 2.24 A plot of powers Ris given in Fig 2.40 where we have normalized the curve, The maximum power is shown as | and the normalized resistance, defined as r= R/R, is 1 when power is maximums, This is the famifiar result chat the maximum power transfer occurs when the load resistance, R, equals that of the source, R.. We then say thatthe source is marched to the load. Tn the general case, the source impedance ccan have a reactive part. Then, maximum power transfer occurs when the load is a complex impedance with the same resis tanee as that in the source impedance. When a generator voltage and. the related source resistance are specified, the power exiracted when the generator is ter ‘minated in a matched load is called the available power, for it is the maximum jpovier that is available from that generator. ‘The amplifier of Fig 2.38 has an input resistance, Rig, and an ouput resistance, Ryu. The rest of the amplifier is modeled with a controlled current generator. The amplifier willbe matched a the input when R=Rjq. The output is matched with @ load R,=Rout- Picking these source and load resistances will produce this perfectly matched amplifier. While it sounds easy enough, it can be very complicated in a practical RF application. In a practical Aamplifice R,, sill depend upon the loed, R,,, while Rig, Will dependion Rs, Evente: ally stability becomes « dominating issue. Cireuits thatare unconditionally stable ean eventually be matched perfectly at beth input and output. Source und. load resistances are not changed directly as a means of achieving matched conditions. Rather, a 50-0 gen: trator might be applied to an impedance transforming network that presents a dif- ferent impedance to the amplifier input. ‘These networks are discussed in greater detail in Chapter 3. We always are interested in the "gat of an amplifier. This usually means power gain, which is the ratio of two power lev els, With a known source voltage, V, and source resistance, Rs, and the modcled input resistance Ryy (item Fig 2.38), we can calculate the input power. Output power can also be calculated when the amplifier is well modeled. Knowing the powers, the power gain is Eq 2.25 ‘The maximum possible gain is that 14° Chapter 2 occurring when both input and output are matched. ‘The power gain of Eq?.25israrely mea- sured directly. Instead, we more often measure or calculate rransducer gain, first ‘mentioned in Section 2.2, Transducer gain Pour Gr- Te Pay Bq 2.26 whore Pyycis the power delivered to the oud and Pay is the power avaitable from the source. Power gain und transducer gain ‘re equal ina perfectly matched amplifier. ‘A variant of transducer gain is the ayer son power gain obtained when a transmis sion line is broken, and an amplifier is inserted, This occurs when both Rs and Ry are identical, usually 50 £2 The Decibel, or dB. Gain ean be expressed as a numeric ratio, but is more often specified in deci bls, given by Ey 227 where P, and P, are nwo different powers. If ‘an amplitierhss a5 25WV output and is being riven by 8 generator with an available power of I mW, the power ratio Pyy/P ay is 5. for a transducer power gain of 7 dB. The dB construct was nol invented o con fase the prospective designer. Rather, itis a natural consequence of the mathematics ‘Output power is calculated from an input powerand a numeric gain by using ealtipl cation, It is also calculated from a Ontoseore Attenvator s0 ohe Termination Fig 2.45—Using a return loss bridge with an amplifier ‘1 aL be measured. The bridge is Festapen ei» 4% aaa cuited at the “X" port, and the detector response is noted. Then, a $0-Q termina tor is placed on the “X" port. A large decrease in detector response should be noticed. This response isa measure of how well the RLB is funetioning and is called the bridge dfrecrivity, An amplifier (power on) snow attached to the "X” port through coaxial cable, and a terminator is attached to the amplifier output, The detector response will be lower than the lovel present with the *X” port open cir- cuited by a ratio called the return Joss, a AB value. The Step attenuator inthe detec: tor can be adjusted to attenuate the refer- ence to better meusure return loss Return loss is related to T throu R.L. =~ 20-Logr Bq 2.30 ‘The inverse form is 2.5 DIFFERENTIAL AMPLIFIERS AND THE OP-AMP ‘The differential amplifier, or diff-amp, is the foundation for most silicon analog integrated circuits in-use today, making it a very important topology. Here we inves- tigate differential samphifier fundamentals and examine a major derivative of it, the operational amplifier, or op-amp. Following the name, the differential amplifier is acisouit intended to amplify 2 difference, The differential amplifier has ‘eo input termes, The output, whieh can bbeberween two collectors or from just one, 2.46 Chapter 2 is then proportional to the veltuge (or cur- rent) difference between the inputs. The basic differential amplifier using NPN bi- polar transistors is presented in Fig 2.46. We start with two identical transistors biased at the same de hase voltage. The ‘two emitters are attached and returned to ‘ground through a common resistance, as in Fig 246A. Two identical collector resistors are attached, biased from a com- ‘mon supply. This eireuit ean have signals applied in two ways. If the ewo bases are While we have illustrated the RLB with oscilloscope detection, 2 30-0 power meter or spectrum analyzer is preferred Both are described in Chapter 7. These are 50-92 instruments, so they da not requ the external terminator so vital to the oscilloscope. The ‘scope suffers from two problems that compromise this applica Gon. First, its a wideband instrument, so noise limits the sensitivity, making it dif ficult to see the weak signals that are readily seen in a spectrum analyzer. Se fond, many of the terminations thet we might measure are nacrow bandwidth loads. As such, they will produce high ath loss at one frequency. but not at the harmonics. The usual signal generator is harmonic ich, The harmonics are revolved ‘and, henes. ignored in a spectrum analyzer ‘measurement driven together. the composite cireuit ‘would beliave as one transiscor, The 19 collector signals would then be identical. This operation is called common-mode drive or excitation, The laege emitter resistor becomes s degeneration element, causing the common-mode gain to be low The other diff-amp drive is the differen- sial-mode, where one base is driven in one direction while the other is driven by an ‘opposite polarity. Assunse that Ql and Q2 are biased with ade hase voltage of 5. Thevoltage a the common emitter is then 4.4 Total current will be 4.4 mA for an emitter resistor of 1 KO. Ifthe two transistors are identical, cach will be biased to an emitter current of 2.2 ma, We now apply «differ ential signal causing Vj) 10 increase by 10 mV while Vio drops by an equal 10 my, The emitier voltage remains essen tially constant, Vi, decreases while Vex ‘noreasos by an amount related to the gan. ‘A useful property of this circuit is that total eutrent does not change with difier- ‘ential drive Fig 2-46, part B shows the circuit varia~ tion found most often in integrated cireuits ‘whore the emitter resistor is replaced by a third transistor, Set Vig (02 volts and pick the Q3 emitter resistor for the same 44 ‘mA. This leavesbias conditions forQ] and Q2 as they were, although the common mode gain is even lower. QBs constant current source, acireuit that acts as if the bias for QI and Q2 came ftom a very large negative power supply wwith sm equally large resistor. The effect of this topology is to force the sum of the currents in QU ancl Q2 to remain constant. ‘This has iwo important consequences. Fig 2.46—Ditterential Amplifiers. First, a differential amplifier is very easy to decouple, With constant rol current, signals are not injected onto the Ve, power supply, vory important when the diff-amp is one of many Such circuits within an TC. The other consequence of the constant curtent source is that drive applied ro just ‘one input will result in differential ovrput signals. This is shown ia the amplifier of Fig 2.47, The ew collector voltages have ‘equal amplitudes and are ouof phase with each other, ‘Although differential amplifiers are abundant in integrated circuits, they are also useful and practical in discrete form. Fig 248 shows a diff-amp with readily available parts that might be used to pro- vide belunced local oscillator drive to a mixer without (ransformers. This cireut is +V¥ec > -Vee Fig 2.49—Schematic diagram for an ‘operational amplifier. oe Fig 2.47—Ditferential Amplifier that converts a single ended signal into a differential one having two outputs with a . second order IMD occurs at frequencies i IP 30h PBouti2)=4 Fig 2.61—A cascade of two amplifiers, ‘ach well specified for gain and output Intercept. The composite intercept is ‘easily calculated. An extension of this allows an entire system to be analyzed for IMD. 6B Hybrid Combiner 25:50 ohm ot [orn cc [oss ‘Spectrum Analyzer oo 3.48 Hybrid Combiner/Spitter | Fig 2.62—Test setup for measuring IMD. A low pase filter Sometimes follows the hybrid. (f+ fy) and (f,-f). These distortion fre- ‘quencies are usually far removed from the inputs. Hence. they ean be removed with a filter following the amplifier. ‘This is not possible with third order products very close to the Trequencies causing the distortion, The lest amplifier was found 10 have a second order output intercept of +44 dBm, Second order intercepts are generally numerically higher than the third order ones. although the second order distortion does not drop as quickly. Second onlor IMD can be a majorditticulty in wide band designs, such as general coverage receiv fers or spectrum analyzers, Ir is interesting to compare the 1 dB ‘comiprossion power with output intercepts. ‘Our test amplifier hid Pyy(=1 dB)=#16.5 dBm and TP3,,,.=#20 dBm, a difference of 13.5 GB. Differences of 13 to 16 dB are common for amplifiers with bipolar tran- sistors. Smaller values (7 to 10 dB) are ‘more common with silicon JFETsund with GaAsFETs. The difference is nov intended to be a Figure-of-merit, Indeed, simaller numbers indicate that @ devieo can be ‘operated closer to it's intercept. Typically any of the devices we commonly use for amplifiers cannot operate at powers as, high as thetr output intercepts, ‘A test set used to measure 2nd and 3rd onder intercepts is show in Fig 2.62. The Koy to the scheme is the hybrid contbiner that adds the output of two signal genera. ons while preserving impedance match and isolating the two generators. A 6-dB hybrid is the prefereed scheme owing t© the excellent isolation afforded. But a 3 AB hybrid ean be substituted if good qual ity Signal generators are used. A 6-UB hybrid is a nctwork with un output that is 6 GB lowerper tone than each input. Note that the 6-dB hybrid has the sumo sche- matic as a return loss bridge. Hence, one instrumenc ean be used to measure imped- ance match and to isolate signal sources. Every home lab needs at least one hy’id combiner, The intercept formalization is generally restricted to circuits with constant, or nearly constant, bias current. A Class AB ‘or B amplifier where cursent grows with applied drive is not generally described by an intercept. Rather. it is characterized with a simple TMD ratio, usually at full power outpat Further information on distortion and noise is found in dntroduction to Radio Fre: queney Design. The reuder is ulso referred {Bill Sabia’s presentation inthe 1995 (and later] ARRE HerlbooK” concerning distor tion, including that of 2nd order IMD. Amplifier Design Basics 2.232.7 FEEDBACK AMPLIFIERS ‘A circuit form appearing often in this book is the Feedback amplifier. This is a cizeult with wo forms of negative feed- back with (uswally! a single transistor to obtain wide bandwidth, well controlled in, and well contolled, stable inpuc and ‘Output resistances, Several of chese amnpli- fiers can be caseaded 10 form a high gain circuit that i both stable and predictable. The small-signal schematic forthe feod- back amplifier is shown in Fig 2.63 without Dias components or power supply details. The design begins with « NPN tan sistor biased to a stable de current. Gain is reduced with emitier degeneration. ineress- ing input resistance while decreasing gain Additional feedback is then added with parallel feedback resistor, Rj, between the Fig 2.64—A practical feedback Components marked with "B" are predaminantly for biasing. The 50-2. output termination is transformed to 200 0 at the collector. A typical trans- former is 10 bifilar turns of #28 on a FT-37-43 ferrite torold. The Inductance ‘ot one of the two windings should have 2 reactance of around 250.0 at the lowest frequency of operation. 2.24 Chapter 2 collector and base. This is much like the resistor between an op-amp output and the inverting input which reduces guin and decreases input resistance. Several additional eincuits are presemed showing practical forms of the feedback amplifice. That in Fig 2.64 shows a com- plete cireuit, The hase is hiased with a resistive divider from the collector, How- ever, much of the resistor is bypassed, Teaving only Ry active for actual signal feedback. Emitter degeneration is. ac coupled (9 the emitter. The resistor Re dominates the degeneration since Rp is normally much sialler tha the emitter A B a RFC a out ena vec Fig 2.55—A variation of the feedback amplifier with a 50-9 output teeming at the collector. bias resistor. Components that are pre dominantly used for biasing are marked with “B.” This amplifier would normally be terminated in 50.2 at both the input and output. The transformer has the effect of ‘making the 50-0 load “look like" a larger oad value. R,=20002t0 the collector. This isacommon and useful value for many HF applications Fig 2.68 uiffers from Fig 2.64 in two places. First, the collector is biased through an RFC instcael of a transformer. The collector circuit then “sees” 50 @ when that load is connected. Second, the cmiticr degeneration is in sesies with the bins, instead of the earlier parallel connec: tion, Either scheme works well. although the parallel configuration affords experi mental fleaibility with isolacion between setting degeneration and biasing. Ampli fiers without an ouiput translormerare not constrained by degraded transformer pers formance and often offer flat gain to sev= eral GHr, ‘The variation of Fig 2.66 may well be the most general. Ituses an arbitrary tans- former to mateh the collector. Biasing is taditional and does not interact with the feedback. Feedback is obtained directly from the output tap inthe circuit of Fig 2.67, While this scheme is common, itis less desirable than the ethers, far the transformer is part of the feedback loop. This could lead to instabilities. Normally. the parallel feed back tends ta stabilize the amplifiers. The ‘equations and curves presented below per- tain to cizeuits with feedback taken directly from the collect ‘The circuit of Fig 2.68 has several fea- vee mr Ree e 2 out RES 3 moH 2 HE ze Fig 2.56—This form uses an arbitrary transformer. Feedback Is isolated fram bias components. Fig 2.67—A feedback amplifier with feedback irom the output transformer lap. This is common, but can produce unetable resultstures, Two transistors are used, exe with 18 Separate crniter biasing resistor. How- ‘ever, ac coupling causes the pair to oper ate as a single device with degeneration sel by R,, The parallel feedback resistor R,. is bath a signal feedbuck element and part of the bias divider. This constrains the valucs slightly. Finally. an arbitrary out- pur load can be presented tothe composite collector through a R-type matching network. This provides some low pass filtering, but constrains the amplifier bandit Design Procedure Design begins by picking a bias current, usually dictated by output power and [MD requirements, Next the ourput load imped- ‘ance presented 10 the eollector (or drain) is chosen. A value of 200 0 is probably the migst conimon, for it affords good gain with reasonable current. With that load, the output power will be restricted to around 200 mW in 12-vall s gressively lower impedanees will allow higher output power. Most feedback am= plifiers end up being designed for 50-2 input resistance. ‘The emir deger tion and feedback resistors are chosen nest, A reasonable input and output impedance mateh occurs with RR, = RyRy, Eq 240 where Ry is the parallel feedback and Reis the net degencration resistance, r+Ry Here Ry is the external deyeneration, tnd 1, is the current dependant value, 2e/l,(mA). For example, an amplifier driven by 50 £2 and terminated in 200 & might use 10-@ external degeneration and 10-mA current for Re = 12.7 Ohms. R,= 7879 would produce R,, Ry, with Rj, and R, being the input and ‘ourpit resistances far source and lead Rg and Ry, A practical choice would be R= $20.0. a standand value "There is sil a wide range of values that can be used for degeneration and feedback. ‘Tae final choice is made on the basis of desired gain, which ean be determined by the ‘equations presented in Fig 2.69. The choice is eased by example daa in Table 22. While the data in the table is for one current, 20mA, it will provide an initial estimate ‘The equations of Fig 2.69 appear long sand messy, but are easily programmed for a calculator or computer Fig 2.70 shows the gain obtained when Table 2.2 Simulated Gain vs Degeneration and Feedback Resistors for 2N3904 blasad with ,=20 mA whore -=1.3 0. Gain was calculated at 14 TiHz, so B=300/14=21. Resistors were picked as standard values and to provide an input return loss better than 10 dB. The first example is the amplifier described in the previous section. Load Redegen Feedback Gain gn 62 13K 20308 ssa she 245 0B 47° 27k 239.08 5eQ 2ke 22.308 882 16k 20.7 0B jon 9102 16303 wa 7300181 B wea 5600 126.08 ia 4300 103.08 229 3502 7768 son 272 Bn 20.008 59a gana 18208 470 5600 16908 56Q 4700 15608 Bea 300 © 14.168 won 2702 10.7 eB i2n 2200 8308 ea 1500 3408 en Gain vs Dugeneration hen Matched Degeneration Resist Fig 2.68—Foodback amplifier with two parallel transistors. ampli Fig 2.70—Gain Vs net dager 3 ig matched. This evaluation occurred at 14 MHz with 2 2N3904 biased to 20 mA with a 50-01 source and 200-0 load, ration stance when the 0 tog] 4 RE Ry Re = (148) (Rp BRe Bi) aye, PR Ry] [Bevieg-2eRcesR,|seiRe [[1 +8) Ree Ry) Res (Ry +Re+ BR, +HR,_)R,+ BR, Ry +RyR,| | IG +B) R, +R, +8 R,] fas (r+) }Re+R ERs] Fig 2.69—Transducer Gain Gin dB, Input resistance, Rix, and Output resistance, R,, both in Ohms for a feedback amplifier. The analysis. fa the tot Bis the current gain anc operating frequency, both in MHz ‘approximated 2 scalar value, restricted to the case where parallel feedback is obtained from the collector. Ry Is the parallel feedback and R, ‘omittor degeneration (soe text.) A, and R, are the source and load resistances, and are arbitrary for this analys “dF where F; is the current gain-bandwidth product and F the Amplifier Design Basics 2.25Fig 2.71—Galn Vs degeneration for fixed feadback of 1.3 KO. ay Sse gs Hn aaa Fig 2.72—Input resistance Vs ‘dogoneration for fixed feedback sistance. Fig 2.73—Output resistance Vs dogensration for a fixed 1.3-k2 toedback rosistanc: Supa Ye Soee RS 6 ot) npr ie Leo (LSE 6 Ob) Fig 2.74—Output the source resistence. Ey 2.40 is applied, forcing 4 reasonable input and output impedance match. Tris common ro build an amplifier only to then find that the gain must be changed alittle, The effect of changing the emitter resistor is presentedin Fig 2.71 fora fixed 213 kM. The same 14-MHz, 20-ma, bins case is assumed, Fig 2.72 and Fig 2.73 shoo the related effect on terminal resis tunees. A chnrweteristie of Feedback amplifiers (sometimes useful, sometimes frustrating) is that they are partially transparent. The ‘input resistance becomes steong function of the load while the output resistance depends upon the source. This is illus- trated in Fig 2.74 and Pig 2.75. Again, a L3-KQ fecdback R and 6-0 extemal degeneration arc used. The amplifier ‘eansparoney is partially “fixed” with the addition of an altenuator at the amplifier ‘output, especially useful when the ampli Fier must interface with filters and switching-mode mixers. Pads must be added with care, for they will decrease overall guin, available output power and ‘output intercept 2.26 Chapter 2 Feedback extends the bandwidth of wansformer terminated amplifiers. Fi 2.76 shows gain vs F for the example am plifier with « 2N3904 at 20 mA. 6-0 de. generation and 1.3-K02 Ry, 50-82 source ‘and 200-62 load, There is less than a 3-dB variation over the HF spectrum, and the amp is usable up to 50 MH, even with a modest 2N391}. Higher F,tansistors can produce much greaicr bandwidth, espe- tially when configured for low or modest gain without any transformers that might compromise frequency response. While se usually think in terms of buikk ing feedhack amplifiers with bipolar tran- sistors, they are just as tenable with FETs. Fig 2.77 shows a JFET version of the amplifier, This circuit uses no degeneri- tion resistor. The FET is self-biased with a hypassed source resistor, andthe biased FET ransconduetance is caleulsted using ‘equations presented curler. Having this vilue, we can then ask “what current (Fit 4 bipolar transistor would produce the same transconductance?” Finding that value, we then use the same equations for analysis that were applied to the bipolar, Fig 2.69. ] Fig 2.76—Feedback tends to flatten frequency respon: ‘dramatic with lower gain amplifiers. Fig 2.77—A feedback amplifier using a FET. See text for design details. Feedback amplifier noise figure is usu ally greater than that from the same tan sistor without feedback, Noise available from the feedback resistors is injected into the circuit, A feedback amplifier was builtusing a 28C1252 transistor (Fj=2 GHz) ‘with degeneration and feedback resistors of 5.1 Q and 1.8 KQ. Noise figure was measured with an HPS970B test set for differing standing currents. The noise figure was 1.8 dB in the HE spectrum for 10 mA, increasing (0 3.3 dB with 63 A. Noise figure tor the 2N3904 example amplifier featured in this section (20 mA, 6 Dand 1.3 2, 200-82 load was measured ul 6 dB. Fig 2.78 shows a feedback amplifier ‘with two transistors in @ Darlington con figuration, This circuit is typical of sev eral popular silicon monolithic integrated circuit amplifiers that are presently ayail- able. Those components within the dotted line are part af the IC. Qt and Q2 usually have F, ubove § GHz, so the amplifiers offer useful performance 10 2 GHz and beyond with gain from If)tonearly 2048. Fig 2.78—Feedback amplifier with Darlington connection of transistors. ‘These amplifiers are specified by their dis tribulor fora voltage on the output pin with 1 specified curvent allowing the user to pick R, foran available V.., For example, the Miniciceuits MAR-2 's specified for 25 mA at 3 V. Hence, for 2 12-V power Supply, 280 would be needed for R: This IC should not be used without a drop- ping sesistor. The power dissipation in the resistor should be checked, It's only 175 mW in this example, sow 4-W resistor ‘would suis Fig 2.79 presents another two diserete transistor feedback amplifier. This is a butter amplifier designed by W7EL. This cirouitis similarto MAR citeuits parts, but uses transformer ourput coupling for even Fig 2.79—Feedback amplifier, the design of W7EL, often used as an ‘oscillator butter. ‘greater available gain. The input resistor Should be driven from a souree at DC ‘ground, Bandwidth depends on the output transformer with severe distortion pose sible at low frequencies if it does not have adequate reactance, A typical 7-MHz application uses 2 20-tur primary on T= 37-43 torvid with @ 5-turn output link A common base amplifier with tran former owiput coupling is shown in Fig 2.80, This circuit uses no feedback other than the 47-0 degeneration, This is pre- sented as an evolutionary step toward a feedback amplifier, but itis very useful as shown, The coruman base topology fea tmges excellent reverse isolation, amd, as such, itis an excellent VFO buifer. The amplifiers biased to about 4 mA collector current, 50 has an input resistance atthe emitter of 6.5 Q. Adding a series 47-0 resistor creates a reasonable input match to a 50-2 souree, The power gain will be determined by the ratio of turns on the ‘output auto-transformer. An interesting variation of this circuit is presented in Fig 2.81, The 47-2 input resistor has been teplaced by a single cum link through the transformer core. The operation is easily understood if we think of driving the input with a eurrent source, The low inpu: impedance at the emitter has no impact on the current flowing. Essentially the same current flows in the ccolleetor (recall that the current gain of a common base amplifier is unity), but it now flows inthe high impedance multiple turn transformer windings. This allows the cireuit to provide power gain. We now “sample” the collector current with a winding, creating a voltage across the winding. The nev “voltage” is place series with the low emitter input imped- ww wes ene 201 on eo n-turns 3 x1s50 } "7 + $ 510 Res=50 Fig 2.80—Commen base amplifier with an input re See text. Norion of Anzac. Fig 2.61—A transformer feedback amplifier designed by D. Amplifier Design Basics 2.27Fig 282—A modified Input impedance. tance Wo create a 50-02 input termination, However, this is done without any resis- lots, so the woise figure is not compro- ‘mised. This amplifier is the brainchild of David Norton of Anzac. ‘The Fig 2.81 amplifier will be matched it m?—m-1 Eq 241 to produce a transducer power gain of 20Log(m) dB. For example, if m=3, 0 is ‘then 5, and the power gain is 9.5 dB. The transformers for these amplifiers are often wound on a binocular-type balun core. A um through such a ferrite core is counted assingle pass of wire through both holes. Polarity is vital to construction of the ‘wansformer. If wound wrong, the input impedance will be negative, almost guar- aantced to create oscillation. Tn amplifiers of this kind that we have built, we meu- sured excellent input impedance mate jedback increases (25-dB return loss) over 4 5 to 100 MHz range with noise figure under 2 dB. This amplifier, however, suifers from a major problem; the terminal impedances depend strongly on the termination at the other port. Tho circuit is worse than resistive feedback amplifiers in this regard. Transformers can be further applied t0 extend performance of amplitiers. Fig 2.82 shows a generally traditional feed= back amplifier that is modified by passing the input lead through the transformer core to alter input impedance. This topology is curly work of Rohde” Fig 2.83 shows a FET amplitier (small signal circuit only) using an input trans- former, A twpped transformer feeds signal tw both the FET source and the gate, The winding driving the source sees a Tow Impedance, so adjustment of turas ratio can censure 4 perfeet match, The gate winding even though there isno signal current flow 2.8 BYPASSING AND DECOUPLING Our amplifier designs fave included grounded points that Were not really at ‘ground. Ruther. those points are “signal grounded” through bypass capacitors, Obtaining an effective bypass can be difficult and is often the roure to design difficulty ‘The problem is parasitic inductance. Although we label and model parts as “capacitors.” a more complete model is ceded. The better model isa series LRC, shown in Fig 2.88. Capacitance is close «© 2.28 Chapter 2 the marked value while inductance is a stoall value that grows with component lead length, Resistance isa loss term, usu- ally controlled by the Q of the parasitic inductor. All components show some inductance, including a wire. Even leadless SMT component will display inductance commensurate with the dimen sions. A wire has an induetanee of about 1 nH per mm of length, Bypass capacitor characteristies can be measured ia the home Jab with the test Fig 2.83—Small signal circuit of a transformer type feedback ampiifior using a JFET. This circuit supr ment. With V.=12, I (O1P3=+42 dBm, Gi bandwidth exegeds 50 MHz. ing, provides the gate voltage needed for {gain and low noise performance. Design doails are given in Introduction to Radio Frequency Design, p 216.9 Bill Carver WTAAZ, has built practical versions of this, amplifier. See QS7, May, 1996." with fur ther discusvion in Chapter 6. “Transformer feedback amplifier design is a subject that continues 10 produce design activity. The reader can find more information starting with papers by Trask! and Koren, Fig 2.84 shows an example of a feed back amplifier. soup of Fig 2.86, Fig 2.87 shows a test fixture with an installed 470-pF leaded ‘capacitor. The fixture is used witha signal ‘generator and spectrum analyzer to evalu fc capucliors, Relatively long capacitor leads were requited 10 interface to the BNC connectors, even though the capacitor itself was sinall The signal generator was tuned over its range while examining the spectrum analyzcr response, which was ‘minimum atthe series resonant frequency. Parasitic inductance i calculated from thissheen 500m, $21 Ret. 00008 Fig 2.85—Model for a bypa capacitor. ‘Signal Generator | a)» j Y on =e) {70 pF Leads or Leads capacitor. Fig 2.96—Test set for home lab measurement of a bypass stop 3000-600 Mia, Fig 2.88—Network analyzor moasurement of 470-pF shunt capacitors. Both SMT and leaded paris are studied Fig 2.87—Test fixture for measuring self resonant frequency of capacitors. frequeney, The C value was measured with alow frequency LC meter, Measurement gear is discussed in Chapter 7, ‘The measured 470-pF capacitor is mod sled as 485 pin series with an ineuetance of 7.7 nH, The L is larger than we would see With shorter leads. A 0.25-ineh 470-pF ceramic disk capacitor with cero lead length will show 9 typical inductance closer to 3 nH. Tae measured capacitor Q ‘was 28 at self- resonance of 82 MHz but is higher at lower frequency. Data from a similar measurement, but with a network analyzer is shown in Eig 2.88, Two 470-pF capacitors are mea sured, one surface mounted and the other ‘leaded part with D.]-inch leads, Fig 2.89 shows two calculated plots for the 470-pF capacitor. The one on the let is 1 Smith Chart showing the behavior vs, re~ quency, while that on the right is a plot of fomponent reactance vs. feequeney. Reve to Be Fig 2.89—Impedance of a 470-pF bypass c tance dominates, keeping the data on the edge of the Smith Chart for the Q is moder ate at 28, Bypassing is “perfect” at only one froquoncy. that of series resonance, An ideal (no inductance) capacitor would have a capacitive reactance of about 2 9 at 150 ‘MHz. The actual 150-MHz. value is induc tive with magnitude of about 5. ‘Traditional lore tells us that the band- ‘width for bypassing ean be extended by paralleling & capacitor that works well at fone frequeney with anotler to necommo- date a different part of the spectrum. Hence, paralleling the 470 pF with a OL-UF capueiter should extend the bypassing to lower frequencies. The eal- culations are shown in the plots of Fig 2,90 The results are terrible! While the low frequency bypassing is indeed improved, a high impedance response is created at 63 MHz, This complicated behavior is again the result of inductance, apacitor. Each capacitor was assumed to have a series induetance of 7 nH. A parallel reso- nance is approximately formed between the L of the larger capacitor and the C of the smaller. The Smith Chart plot shows us that the impedance is ncatly 50 9 at 63, MHz, Impedance would be even higher swith greater capacitor Q. This behavior is 2 dramatic example of lore that is generally wrong! Bypassing can be improved by parallel- ing. However, the capacitors should be approximately identical. Fig 2.91 shows the result of paralleling two capacitors of about the same value. They differ slightly at 390 and 560 pF, creating « hint of reso- nnance. This appears as 2 small “burble” in the reactance plot and a tiny loop on the ‘Smith Chart, These anomalies disappear a the C values become equal. Generally, paralleling is the scheme that produces the best bypassing. The ideal solution is t0 Amplifier Design Basics 2.29place a chip cap om each side of a printed eitcuit run oF wire at @ point that isto be bypassed ‘Additional capacitors were measured. O1-uF disk (leaded, 50-V, 0.2-inch diam- eter) was resonant at 20 MHz in the test fix tute shown, indicating au induetance of 6.5 aH. The Q was $7. Two different 0.1-uF leaded capssitors were investigated, Both dra identical capacitance even though one ‘wus larger than the other. The inductance was about 4.3 nH with Q=5 for both ‘Matched capacitor pairs form an effee- tive bypass over a reasonable frequene: range. Two of the .O1-uF disks have a reactance magnitude less than 5 €2 from 2 (0 265 MHz. A paic of the 0.1-uF capaci tors was even better, producing the same bypassing impedance from 0.2 co 318 MHz, The 0.1-uF capacitors are chip com- ponents with atiached wire leads, Even better results ean be obtained with multi layer ceramic chip eapacitors. Construe tion with multiple layers creates an Integrated! paralleling. We have measured some 0.2-4F parts with an inductance of 2 all. The multilayer components are more expeusive than the monolithic 0.1-aF parts investigated. Some applications (e.2., IF amplifiers) require effective bypassing at even lower frequencies. Modern tantalum electrolytic capacitors are surprisingly. elfective through the RF spectrum while offering hhigh enough C to be useful at audio ‘The parts should be evaluated for evitieal applications We have discussed the problem of bypassing, but have neglected the related problem of decoupling. The bypass capacitor usually serves a dual role, first creating the low impedance needed to gen- crate a “signal” ground, It also becomes part of a decoupling low pass filter thar passes de while attenuating signals. The ailenuation must function in both direc tions, suppressing information inthe power supply that might reach an ampli- fier while keeping amplifier signals from reaching the power supply’ ‘A low pass filter is formed with alter- nating seties and patallel component con- nections. A parallel bypass is followed by 4 setics impedanee, ideally 4 resistor Additional shunt elements can then be ‘added, although this must be done with cure. An inductor between shunt capaci- tors should have high induetanee. It will resonate with the shunt capacitors (© ene ate high impedances just like those that ceamte from parasitic L inthe bypasses. This, imuakes it desirable to have an inductance ‘hac is high enough that any resonance is below the band of interest. But series inductors have their own problems: they 2.30 Chapter 2 2 Paral Bose ‘capsctors, 4709 and “Diu each th 7 nH 7 Toe a Fig 2.90—Tho classic technique of paralleling bypass capacitors of two values, here 470 pF and .01 pF. This Is a ieribie bypass! ‘See text. Two Paral pass Capectors of amost al Vatu 0% 500 seach ah 7m Sens Tntance 0 Fig 2.91—Paralleling bypass capacitors of nearly the same value. This results in Improved bypassing without complicating resonances. Fig 2.92—Two different resistor values parallel a decoupling choke. The lower, 20-0 value is more atfactive, Sea text.hhave parasitic capacitance that ereate their ‘own selferesonance, A couple of available RF ehokes were casured (now as series elernents) with the equipment described eatlier. A2.7-Ht ‘molded choke was parallel resonant at 20 MHz, indicating a parallel capacitance of 0.24 pF. The Q at 20 MHz was 52. A 15-uH molded choke was parallel resonant at 47 MHz, yielding a parallel C of 0.79 PE. This part had a Q of 44 at 8 MHz. Large inductors can be fabricated from seties connections of smaller ones. The best wideband performance will result only when all inductors in a chain have about the same value, The reasons for this (and the mathematics that describe the be- havior) are identical with those for paral- Jeling identical capacitors, Low inductor Q is often useful, whic encourages us to use inductors with ferite cores, Inductors using the FairRite (Amidon) -43 material have Q in the 410 10regien in the HF spectrum. One can also create Tow Q circuits by paralleling a seties L of modest Q with a resistor Fig 2.92 shows « decoupling network tnd the resulting impedance when viewed fzom the “bypass” end, The 15-yH RFC resonates with a 0.L-uF capacitor 16 destroy the bypass effect just above 0.1 ‘MHz. A low value parallel resistor fixes the problem. ‘A major reason for careful wideband bypassing and decoupling isthe potential for amplifier oscillation. Instability that allows oscillations is usually suppressed by Jow impedance terminations. The base and collector (or gate and drain) should both “see” low impedances to ensure star bility. Buc chat must be true at all frequen- cies where the device can produce gain. It isnever enough to merely consider the op crating frequency for the amplifier. A par- 2.9 POWER AMPLIFIER BASICS ‘The remainder ofthis chapter deals with power amplifiers, 4 subject dear to the radio experimenter, The earliest tinker among us cut our teeth on attempts to extract more power from the already stressed amplifier devices of the day. We all recall stories of 6L6 receiving vacuum tubes being coaxed into providing high output power by immersion in an oil bath. ‘The rest of us have tried to extract power from transistors, only to see the device dis- appear “in smoke.” Experience ofthis sort is a “right of passage” for all RF experi- ‘menters; don't miss it! Glasses of Amplifier Operation “Many of the amplifiers considered so far have been "Class A.” The cats of opera tion of an amplifier is determined by the fraction of dtive cycle, of duty cycle ‘where conduction occurs. The Class A amplifier conducts fr 1008 ofthe eyee. Iris charneterized by constant supply eur- rent regardless ofthe stength ofthe div ing signal, Most ofthe amplifiers we use for RF applications and many audio cir cuits in rteiversoporate in Class A. A Class B amplifier conducts for 50 % of the eyele, which is 180 degrees if we examine the circuit with regard toa dcv- ing sinewave, A Class B amplifier draws 10 DC current when no input signal is ap- lied, But current begins to flow with any Input, growing with the input strength A Class B amplifier can display good envelope linearity, meaning that the out put amplitude ar the drive frequency cchunges linearly with the input signal. The total absence of current flow for half of the drive cycle will ereate harmonies of the signal drive ‘A Class C amplifier one that conducts for less than half of a eycle. No current flows without drive. Applicationof asmall drive produces no output and no current flow. Only after a threshold is reached does the device begin to conduct and pro- vide outpat. A bipolar transistor with no source of bias for the buse typically oper- ates in Class C. ‘The large-signal models discussed ear Hier are suitable for the analysis of all amplifier lasses, Smull-signal models are generally reserved for Class A amplifiers The mest common power ampliticr class is across between Class A and B, the ‘Class AB aruplifier that conducts for more than half ofeach eycle. A Class AB ampli- fier at low drive levels is ndistinguishable from a Class A design, However, increas- ing drive produces greater collector (or drain) current and greater output. Amplifier class letter designators were augmented with 2 numeric subscript. A ‘vacuum tube Class ABI amplifier wus one ‘operating in AB, but with no grid current flowing. In the absence of grids, the num- bers have disappeared. While wide bandwidth Class A and Class B amplifiers ere common, mast cir- allel resonance can be a disaster, When the tulimate bypassing is not possible, nega tive feedback that enhances wideband sta bility is often used Capacitors also appear in circuits as blocking clements. A blocking capacitor, yor example, appears between stages, cre- ating ® near short circuit for signals while accommodating different de voltages on the two sides, A blocking capacitor is not as critical as a bypass, for the impedances ‘on either side will usually be higher than that of the black: Emitter bypassing is often a critical application. As we have seen, afew Ohms fof emitter degeneration can drastically alter amplifier performance. A parallel resonant emitter bypass could be a pro- found difficulty while a seties resonant ‘one can be especially effective. Clearly, detailed modeling isthe answer to compo- nent selection. cuits operating in Class C and higher are tuned at the output, The tuning accom plishes twe things. First, itullows different terminations o exist for different frequen cies. For example, aresistive load could be presented atthe drive frequeney while pre senting short circuit at some or all harmonics, The second consequence of tuning is that reactive loads ean be created and presented to the amplifier collector ar drain, This then provides independent con {ol of current and voltage waveforms While not as common as A, B, and C, (Class D and E amples ure of increasing. interest. The Class D eireuit is a balanced (owo transistor) switching format where the input is driven hard enough to produce quate wave collector waveforms, Class E amplifiers usually use asingle device with output tuning that allows high current 10 flow in the device only when the impressed voltage is low ‘Class A and AB amplifiers ate eapable ‘of good envelope Linearity, so they are the most common formats used in the output of SSB amplifiers. Class B and, predomi nantly, Class C amplifiers re used for CW and FM applications, but lack the enve lope linearity needed forSSB. Recent work ‘with a4" method of SSB mnay change that, allowing distorting amplifiers to be used in SSB service. Efficiency varies considerably between amplifier class, The Class A emplitier ean reach a collector efficiency of 50%, but no higher, with much lower values being 2.31 Amplifier Design BasicsT = 10 bifilar s12v eS Pout Vs Pin at 5, 10, 20, 30, 50 MHz e+ 30 isu 6 REC g eo Tt Bis . S10 2N3866 a a 3.9 6 Pa atinput, dBm turns ¥T-37-43 Fig 2.93—Class AB amplifier chain. typical. Class AB amplifiers are capable of higher efficiency, although the wideband circuits popular in HF transceivers typi- cally offer only 30% a1 full power, A Class C amplifier is capable of efficiencies ap- prouching 100% as the conduction cycle becomes small, with common values of 50 to 75%. Both Class D and E are capable of 90% and higher efficiency. ‘An engineering text treating power amplifier details is Krauss, Bostian, and Rash's Solid Stare Radio Engincering."® ‘A landmark paper targeted to the home experimenter was that presented by a group from Cal Tech in QS7 for May and Tune, 1997.0 A Two-Stage General Purpose Class AB The cireuit of Fig 2.93 operates in Class AB with on output of halla watt in the HF spectrum. This circuit was originally built 2.32 Chapter 2 power chain. as a general purpose gain block for CW twansmitiers, Total current is about 80 mA, ‘ith no RF drive, reaching 200 ma or more When drive is increased with most of the increase aceurring in the second stage. Pig 2.94 shows Poy Vs. Pq at 5, 10, 20,30, and 50 MH for this amplifier when operating with 2 12-V supply. The measurements were done with a signal generator and a spectrum analyzer. Low frequeney gain is hhighat 35 dB, dropping to 28 dB at 50 Mz. Low frequeney oulput power is over halt 2 ‘watt, with overa quarter of a watt available ac 30 MH, However, goin is severely com= pressed at this level. Higher output power is available with impedance matching. ‘A heat Sink isused on the output transis tor, for dissipation becomes ‘high with large drive. The dissipation in the 23904 ig 350 mW, safe forkeyed (low duty eycle) ‘OW applications, but marginal for SSB or digital modes "The third order intermodulation distor- tion was measured at 14 MHz, With an Fig 2.94—Gain compression charactevatios fo he ample output of +10 dBm per tone, the output intercept was +32 dBm, Increasing drive Yor +20 dBm per tone ovtput (00 mW/ tone or 400 mW PEP) yielded a higher value of IP3oute+35 dBm, This is expected, or total current is now higher at 180 ma. ‘The power supply for the input stage is normally Keyed when used for CW trans- mission, The bias for the output stage is derived from the same supply resulting in ‘a typicul buckwave 70 dB below full out- put. “Backwave” is the residual signal present froma CW transmitter during key- up period, ‘This design, although lacking in effi- ciency, is otherwise very useful and has been used in over # dozen transmiters oF transceivers in our stations, Tt can be driven by a crystal oscillator on any HF ‘band o form an effective QRP transmitter. Preceding it witha feedback amplitier pro- duces @ DSB of SSB chai suitable for CORP use, or as a driver for u five want PA.2.10 PRACTICAL POWER AMPLIFIERS This section presents several design imples for power amplifiers, A wo watt bipolte power amplifier was presented in Chapter I with the “Beginnee’s Transmit- ter.” Some simple power meter circuits were also included. A CW-QRP Rig Amplifier A familiar RF power amplifier encoun- tered by the experimenter is that used with «low power (QRP) transmitter. The popu- lar design provides about 1-5-W oviput from a 12-V supply. The load resistance the collector would “Like ro see” is then Eq. 2.42 Evaluation yields Rp=48, so close 10 50:Q that no impedance matching network is required at the output. Only a low pass filter is required to attenuate the strong harmonies that are often created by the it. The umplifier circuit is shown ig 2.98. The 7-MHlz design illustrates the design ideas, which are frequency invariant. ‘The amplitice input isto be driven from 8 50-0 source, While not required, it pro- motes convenient measurement. The builder can then test and adjust the dsiver stages alone, with the earlier transmitter stages, and without the complications of the output amphifier. This amplifier will usually require adrive power of 20 to 100 mW, depending upon the transistor type used in the amplificr. The 50-0 drive is transformed downward to “look like” a 12.5-Q source si the base. This transfor mation provides the high base current required for efficient operation. The 18-2 base resistor serves as a wideband load for the input driver, even duting the part of the drive cycle when the base is reverse biased, Decreasing this resistance can improve stability al the price of gain, ‘Base matching occurs with TI a simple transmission line transformer consisting of abifilar winding ona ferrite core. These transformers are diseussed in the fl chapter. Other impedance transformation circuits can also be used, including tuned, L, .or Tee networks. The stage that must drive this will probably be loaded with 4 higher impedance, perhaps 200 9. Another bifilar ansformer could be used. for a single ferrite transformer with a 4:1 tums ratio could make the transition from 200 to 12.5 Qin one step. tis important that the base drive be pro- vided by a low impedance source. A higher source resistance might supply the needed base current, bt then develop high voltage uring the negative part of the drive cycle This could lead to emitter base breakdown, 4 phenomenon that creates cransmitted noise and a slow performance degradation in the oumput transistor, Emitter-base breakdown is easily observed with a wideband oscilloscope. A low driving impedance also helps stability ‘A small heat sink is needed for a TO-39 transistor such as the 2N3866 or 2N3353. A clip-on heat-sink will suffice. ‘The wan: sistor can even be soldered into « hole in a circuit board. Ifthe latter method is used, the hole must be isolated from cireuit ground with extra capucitance absorbed into the design. ‘The amplifier includes extra compo- nents that are not alssays needed. One is the familiar Zener diode at the collector ‘This should have a breakdown value of about 3 times V.. but less than the transis tor breakdown. The diode’s purpose is to load the amplifier if it loses an output ter mination. The diode conducts only if the collector voltage becomes too high, thus a rT ote bina = wou Ba af be Ow J ogee GB ¢ ome Fig 2.95—Typical output amplifier in @ Q IAP transmitter. saving the more expensive output transis tot from damage. The typical Zener diode Will have a relatively high ‘even before breakdown, requiring input C in the low pass filter be reduced in value The virtue of this diode is open to debate, Itis often seen in amateur applica tions, especially with transistors not intended for Clays C RF applications. Its ‘not so common incommereisl applications ‘using transistors intended for RF, The pro- tection Tunetion is easily studied with a high-speed oscilloscope ‘An RF choke routes bias to the collec tor, An extea inductor is placed in series ‘vith the supply. providing a series imped: ance for decoupling. A resistor then paral Tels the decoupling choke, as discussed in anearlicrsection, An optimum decoupling RFC uses large lossy ferrite beads, AT-MH? series tuned eireuit is formed by the 30-pF, 10-uH combination, The back-to-back diodes provide a short cit- cuit for large RF signals, generating a con- venient electronic T/R system. This scheme, and similar TIR methods are dis- cussed in Chapter 6. A Tow ripple Chebyshev Tow pass filter with a cutoff frequency of about 7.5 MHz is recommended. Details appear in Chap- ter3. The capacitance atthe transistor end ‘of the filter should be reduced to account for Zener diode capacitance and the 50 pF related (0 the T/R. No component values. are shown for this example, ‘The ideal transmitierdesign will include variable RF drive, Besides being useful for communications, it isa very useful experi ‘mental tool. ‘Amplifier adjustment consists of noth ing more than varying the drive power while watching the output to a 50-02 lod Amplifier operation without a load should bbe avoided. The output power should change smoothly with drive, with any Jumps suggesting instability Tis interesting to monitor efficiency while drive is varied. Drive is adjusted, ‘output power is measured. poxer supply ‘current is nated, input power iscalculated, and the resulting efficiency is calculated. Efficiency is usually low when the output is considerably less than the design level but increases with drive. It will often be possible to drive the amplifier wo an output {greater than 1.5 W. usually at the price of efficiency. If you are interested in higher omtput, the output network shoul be re-designed accordingly Its usefal to examine amplifier perfor- mance with a variety of loads, This is eas ily done sith a transmatch, The dummy Amplifier Design Basics 2.33Waveforms of a Class-C Amplifior In an effort to gamer intuition about the voltages in Class-C ampilfiors, a low power experiment was per- formed with the circuit of Fig A. A signal generator provided base drive to the 2N3804 amplifier. The coleo- tor was biased at 5 V through a 4.5-uH high @ inductor, A variable capacitor allowed the inductor to be tuned to the drive frequency, or be detuned for an inductive collector termination. A Zener diode could be added to the circuit. Test points are available at the transistor base and collector, allowing the voltages to be monitored with a high speed oscilloscope, @ Tektronix 7704A in this case. "The fiest case examined was the reference for the oxperiment with resuits shown in Fig B, The low FF drive baraly excites the base, but tums the transistor on al the peaks. The resulting current is a short spike, but stil produces @ very clean collector waveform, just reaching soncow Fig AWAF Drive Is applied to the base ve Bence ‘of 2 BJT while the un-torminated 200 9 0 collector is biased through a tuned 200 9 no circuit, The data table relates results to tee aa operating conditions. Fig B—Low drive produc the upper trace. The lower tr Inall cases, the vertical sensitivity is shown for each trace, ‘and the 0-V'line is marked at the left of the trace Fig C—Increased drive produces sovere clipping In the base voltage and an 18-V poak collector signal. 2.34 Chapter 2 Fig 2.96—Schematic for a 10-W output Class ¢ amplifier. ‘The input auto- transformer might consist of 3 turns through a binocular type balun transformer core. A ‘Thomson 2861968 would be a good transistor choice, but try other parts as well. See text. Jou isplaced atthe ransmatch output, and the collector valtage is observed with an oscilloscope and 10X, 10-MO probe. The ourput power will be 1.5 W when the transmatch is properly adjusted. However, ‘output pawer will drop considerably as the transmatch is “tweaked.” The collector voltage will undergo major changes during this adjustment, with voltages sometimes going well heyond the expected 24-¥ value observed when oper= ating in the usual class C moxie with &Fig D—Operation with an inductive load allows the collector voltage to ring up to over 40 V on positive peaks. zer0 al the boltom of the oscillation. Tha positive collecior peak easily reaches twice the supply value. Just a hint of base conduction can be seen at the peak of the base waveform, The conduction must be occur- fing only over a stall fraction of the applied waveform, for the base spends most of the cycle below 0.6 V. The Zener diode is disconnected for the first experiments. ‘The RF drive is now increased to 30 mW, mare than we would normally use with this small transistor. Tho base vollage exceads 1-V peak, which causes the collector voltage to drop to zero, The base voltage tries” to stay on for more than half of the cycle, evi- ence of charge storage, a phenomencn intrinsic to the BJT. But when the base does stop conducting, the collector voltage “rings up" to 18 V, well beyond the 5-V supply. These results are in Fig C. Base voltage ringing at higher frequency is evident. The collector resonance of the last example is eliminated by detuning the capacitor to a low value. The collector now sees a predominantly inductive impad- lance, resulting in the over 40-V poak signal of Fig D. Fig E—The Zener diode is attached, effectively protecting the transistor from excose voltage. Note the change in vertical scale. The transistor ie probably on the verge of damage at this point. Note also that the base voltage has changed, having been altered by the stressed collector. The amplifier has no rasistive load other than that represented by the unloaded resonator Q and provides ‘no oulput power. The collector could be loaded by adding a resistor across the inducior, which would reduce the collector voltage. Even with loading, an inductive component in the collector impedance will allow high voltages to be generated, The final experiment connects the Zener diode producing the waveforms shown in Fig E. The collector voltage is now clipped at the 24-V breakciown of the Zener diode. The base conduction duly cycla is still high, a result of the high drive and charge storage. But the transistor is now saved from damage. These experiments illustrate the effects of an induc- live collector termination, Zener diode protection, and variabie drive. The experiments could be extended with other devices, more aggressive applied stress, and loading that would allow DC collector current to increase ‘proper” termination. It is not unusual to see the amplifier go into oscillations dur- ing the severe mismatch that bappens with this transmatch experiment, The oscille tions sbould not be destructive at this power level, so long as the transistor bas a modest heat sink and is protected against excessive collector voltage. This a good lea to monitor the heat sink temperature by touch is good enough) uring these experiments. A current limited power sup. ply is always useful, if not vital. during experiments of this sor Consider placing a pad between che transmitter and the transmateh. Ifwe use, for example, # L-4B pad, the worst-case return loss would be Wwiee the attenuation, or 2 dB. The corresponding wors VSWR is 8.7:1 (see Eq 4.6.) If the ampli- fier can now withstand all possible adjust= ments of the transmatch, we say that the amplifier can withstand an 8.7:1 VSWR at all angles. The pad is. of course, removed after the test A 10-W CW Amplifier While the 1.5-W amplifier is ideal for the seasoned QRP operator, others may want abit more power. Outputs of 101020 ‘Ware interesting. A few dB gain can make ‘aig difference in results while still spozt- jing und practical for portable operation. Amplifier Design Basics 2.35‘There are numerous inexpensive bipolar transistors that will provide this power in cluding many not normally used for: RE. (One should Took for devices specified for ‘ peak current that exceeds twice the an- ticipated level (1.5 to 2 A for this case), collector breakdown voltages well above the expocted level (24 V here), and an F, at Jeast 3 t0 5 times the expected operating ‘Frequency. Powor dissipation should equal br exceed te planned output. A suggested 10. amplifier is shown in Fig 2.96. The input resistance is expected to be lower than for the 1-W amplifier, so we rive the circuit from a lower impedance source. This can be an auto-transformer, as shown in Fig 2.96, or 03:1 of 4:1 turns ratio classic transformer, Binocular type ferrite balun cores are excellent in this application, noting thateach turn now con- sists of one full pass through both holes in ‘he core. Other wideband transformer con- figurations are listed in the wansformer discussion in the Filter chapter. The input can also be driven from a low QL-C.C Tee network like that used in the output, designed for an impedance of a few Ohms. ‘4 LOAW output calls for a resistance of 7.2.0 presented to the collector when Vee=12. (See Eq 2.42) This amplifier uses tuned circuitry in the form an L-C-C type Tee network. This particular topology is excellent in that component velues are usually practical, nerwork Q can be kept low for low loss, and once designed, the network is easily “tweaked” for slightly \ifferentimpedances. A good design value for Qis 2 10 3. The network between the ddowed lines in Fig 2.96 is used for imped- ance transformation while the filter atienu- ates harmonics, ‘The normal Tee network is modified slightly; a fixed capacitor witha reactance magnitude near the load resistance value is placed at the collector. This kills high frequency gain, helping to ensure VAF stability. Silver mies capucitorsare a good choice for network capacitors with cerant- ies for bypass and blocking elements ‘A suiuble est loud is six paralleled 300-2, 2-W resistors, The drive is increased slowly while monitoring the RF output and collector current. The output ‘Tee network capacitors are tuned for maxi- ‘mum output at each power level. An oscil- lascope is especially useful during such experiments, allowing easy observation of oscillations. should they occur, More often than not, oscillations will occur at low frequencies, so a wideband ‘scope is not mandaiory, This amplifier will prob- ably use no moze than YW of drive, so the builder may wish 10 add a pad if the driv: ing transmitter delivers more than this, ‘The amplifier is set up for Class-C 2.36 Chapter 2 aw fam) + Input: 0.5-0.7 watt 7 une Fig 2.87—High efficiency amplifier after W7EL, T1=34urn primary, 1-turn secondary, #30 wire, on Falt-Rite 2843002402 Balun core. Count Gne turn on a balun coro as a pass through both holes. L1=0.71 yH= 13 ton T44.6; L2= 1.05 uM = 19 on 737-6, L3=15 mH molded RFC. is a GE 042C9 plastic power transistor IRFS11 TRF510 or Fig 2.98—Simple of ip HEXFET linear amplifier for GRP *e oR ‘s Py $y (TE) if ar SK mz 1 lev 42 only 21 uaz 10u network shown <01 180-200 ne Fig 2.99-Dual band Direct Coupled HEXFET Amplifier after W7EL. This circuit ‘operates at 14 and 21 MHz. Lt is 7 turns on a T97-6 and is the inductor for an LiNetwork at 21 MHz. The 1N5367 Zener diodes protecting the FET drain add about 140 DF to the circuit and are a vital part of the network. The band-switch adds me Inductance for a 14-MHz L-Network, Both impedance transforming networks followed by low pass filters. RI, § KO, ent current in the IRF511, while R2, 5 kA, sels the quiescent current driver power supply is less than +12 and Is varied to establish output power.‘operation, although it could be modified for class AB linear operation with litle other change required. Linear biasing is, discussed below. An Enhanced Efficiency Amplifier An interesting and subse amplifier from Roy Lewallen, W7EL, is presented in Fig 2.97. Dubbed the "Brickett intended to follow a 1.5-W output, 7 Mz QRP transceiver. ‘This amplifier used an unusual iransis~ tor, a GE Da2C9. The available drive is aenuated with a 3-dB pad, which was needed for stability. The original W7BL pplication used a 6-dB. pad. The ampli- fier contains the usbal Zener protection diode, but now with 8 75-V breakdown. A peak collector voltage of 65 was measured With this eireuit, even with V=120 V. The iret transforming the 50-0 load 4 Tower value atthe collector isa simple LLnetwork. The resistance presente tothe collector is higher than expected, and is inductive, allowing the high RF voltages The net results a collector efficiency of 8546 or ereater with un output of 7109 W. ‘What began as a Class C design probably nov operates in Class E. The measure- ments have becn repeated and confirmed with several versions of the circuit, all showing high efficiency ‘The adjustment procedure was sinilar to that presented for the 10-W design However, Roy kept increasing drive while aujusting the ouzpat network for inereased power and efficiency. “The TYR scrics-tuned circuit is attached to the collector. Although the networks present an impedance less than 50 910 the Feceiver, the mismatch is nota problem at 7M. HEXFET Amplifiers Power FETs became popular in the late 1970s, While some ritanufacturers intro- duced devices specified for RF, the mar- kket was dominated by switching applica tions. A major supplier is International Rectifier with line of devices called HEXFETs, ‘The HEXFETs are available as both N and P channel enhancement mede parts with a gate threshold around 4 V. The transconductance ofthe typical N-channel device is very high, often rivaling that of 3 bipolar power transistor at comparable cur- rents, While the input gate is a very bigh impedance st DC, high eapacitance at all three terminals limits high frequency gain. HEXFETs are often high voltage devices, allowing a wide variety of supply voltages. Fig 2.98 shows an RF ampliticr using an IRFS11 or the [RFS10, preferred for higher breakdown, Either past has a low ‘gn” resistance of 0.6 0, important for efficiency. This circuit is set up for an ‘output of about © W from @ 12-V supply A 21 turns ratio transtormer generates 12-R drain load, This class AB circuit will function in either CW orlinear SSB appli- ‘cations. The bias should be adjusted fara quiescent current of 100 mA or more for SSB while lower levels are suitable for CW. The output transformer is a bifilar ‘winding on a ferrite core an is suitable for any of the HF bands, We have used this circuit up through 14 MHz, The FET should reside on a modest heat sink, The HEXFET amplifier uses 2 10-Q ‘gate resistor to preserve HP stability. A ferrite bead should not be substituted for the resistor.'8 ‘An interesting discet-coupled amplifier appears in Fig 2.99. This circuit, another creation of WTEL, uses » de coupled IRFS1L to generate an output of 5 W at either I4-or 21 MHz. with a measured effi clency of about 75% Higher Powers HEXFETs offer an inexpensive and interesting route to higher power. We have builesingle band CW amplifiers for output powers from 10 10 50 W on many of the HP bands. The inexpensive IRFS30 HEXFET js an excellent choice for the ‘bands up cheough lé MHz, A 30-W 7-MHz CW amplifier is described later. ‘The IRFP440 and IRFP4S0 have been used in high efficiency CW amplifiers dis cussed later. These parts should also offer interesting opportunities for the experi menter. Although more expensive than HEXFETs, some ‘vendors build parts especially for RF power applications. A search of the web can yield numerous data ‘with suggested experiments. See, for ex- ‘ample, an interesting paper by K4XU and the related Web site of Advanced Power Technology at www.udvancedpower. SSB Amplifiers The bipolar and FET amplifiers pre- seated ean be adapted for linear operation as shown in Fig 2.100, Bipolur transistor base bius should come from a voltage source, Ifthe more typical curtent source is used, the DC current cannot easily increuse with RF drive as is needed for Class AB opezation. A yoltuge source bias uses a diode as a shunt “regulator,” Fig 2-100A. The diode is biased with a zesistor from the same supply that powers the amplifier. The silicon diode isin inti= mate thermal communications with the ‘output transistor, Some designs us a stud mounted diode bolted to the PA uansister hheat sink. Others attach the diode wo the ‘wansistor with epoxy. ‘The BIT amplifier is usually bissed at the quiescent level recommended by the transistor manufacturer. A 10-W part sight use an idling collector eurrent of 20, to M0 mA. A larger current should flow through R-bias with the diode serving as a shunt regulator. Tneveesing the resistor current increases the standing current ia the amplitier, one of the handles available to the experimenter for improved IMD pectormance from the amplifier. DT waa Fig 2.100—Bi Jing schemes for linear amplifier operation of (A) bipolar transistors ‘and (B) power FETs. The base AFC used with the BJT can have small r for the input impedance is low. The diode is bypassed with 2 500-uF electrolytic Capacitor. The base resistor may or may not be needed. R-bias in (A) should have ‘moderate dissipation, for the current may be high. 2.37 Amplifier Design BasicsFig 2.1008 shows FET biasing for SSB. ‘This is generally simpler than with a BIT, for bias curtent is low, The FET bias is controlled with small transistors, 1g TIR switching problems. As with Dipolar wansistor amplifiers, the FET cir- cuits present a compromise between eff ciency and linearity. Amplifier IMD can be reduced with higher standing currents, although the heat sink requirements grow. Amplifier biusing methods are dis- cussed in more detail in the text by Dye aang Granberg. Included are schemes for temperature compensation. Push-pull operation is common with both FET and bipolar linear amplifiers ‘There are several advantages to this. Fist, wo devices are used instead of one, spreading the thermal load over a larger region, Second, transformer coupling bberween device inputs will prevent Inrge reverse voltages on bipolar base-emitter junctions. One forward biased junct serves to clamp the reverse veltage on the other device, Finally, the balanced opera ton will reduce even order harmonic and intermodulation distortion. Negative feedback is often used with Class AB amplifiers, usually in the form ‘of an ac coupled resistor between base and collector, or gate and drain. Feedback sta- bilizes gain over frequency. The negative feedback is applied individually tw each levice ina push-pull pair. Negative feed buck is sometimes extracted fram a wind: ing in an output transtonmer or bias ele~ ment in a pust pull pai, Push pull bipolar transistors ate essen: tially ia parallel tor biasing. For this roa- son. and to help maintain RF balance, RF power bipolar transistors are often sold in matched pairs. This has beeome so com mon that the price penalty is minimal ‘The ease of FET biasing includes push pull amplifiers, which is illustrated in the practical circuit shown in Fig 2.101. This SSB linear amplifier, the work of AA3X (now K3BT), uses a pair of IRFSIIs in a push pall cireuit to develop an output of 30 W PEP, The circuit uses a solid ferrite block forthe output transformer. Fig 2102 shows a sketch forthe oulput transformer, 73, Separate bias lines set up 2 quiescent current for each PET. A DVM measuri total current duriag bias adjustment allows the two currents to be sct equal 10 each ther, While matched transistors might be Fig 2.101 ‘of AASX, is capable of up to 30-W output I par of IRFS11s. This circuit, the creation it with Vddo28 V on the lower HF bands. Reduced output and gain aro avallablo at 14 and oven 21 MH2. Input transformer T1 i 12 tfilar tens 426 on a FT50-43 ferrite toroid. 72 is 12 bifilar urns of #22 on a ‘slack of two FT37-80 toroids. This amplifier was originally in QST, Hints and Kinks, {for January, 1993, page 50.” See reference and text for practical details. Fig 2.102—Transformer detail or T3 of the ASX amplifier. The primary, A-B, shown here as a singlo turn, but actually uses two turns, two complete passes through the core. The secondary {also just shown as one turn) is 3 turns, three complete passes through the core. The windings end on opposite sides of the ferrite block, a BN-43-7051 Fig 2.103—100-W BJT Amplifier. This circu, originally 8) 2.38 Chapter 2 described in Motorola Engineering Bulletin, EBG3.” Is capable ff an output power of over 100 W from 3 to 30 MHz. G1 and @2 fare matched MRF454s mounted to a large heat sink. Lt is a piece of #18 wire loaded with 8 ferrite beads. Both trans- “4:1 turns ratlo with the winding, consisting of pipe, attached to the transistors. The ipped. The 4-turn input ane wvered wire wound through the transformers could be built with would use 10 beads. A larger bead and tubing size would be Detter for 72. The transformers used In our amplifier were Supplied with the kit {rom Communication Conk ‘ Beavercreek, Ohio. See QST advertisemonte for address. CCI has several other kits for power am)desirable, KSBT reports that he has hid good results with devices with severely mismatched thresholds. Equal custents of about 20 mA per teansisior are recom- mended. This umplifier has been used on the amateur bands from 3.5 to 2) ME, although the available output power is less at the higher end. “The oucpue transformer (3:2 turns ratio) presents a load of 22 Q between the Iwo drains. The resulting load is lower than might be desired for high efficiency. common tradeoff with Tinear amplifiers favoring lower distortion. The K3RT amplifier should bo built with a large heat sink, especially if experiments are planned swith variable bias currents. Careful low impedance termination of the HEXFET inputs providesstability. The power guin is still high enough to make the parts very useful, even with the reduced ain related tothe low sousee impedance. he stability problent is largely the result of internal feedback within the FETS. While extremely difficult with bipolar transistors, it becomes possible with FETs to neutralize the circuits. canceling the destabilizing effects of internal feedback, ‘These methoels were common place with vacuum tubes, but have largely been ignored with semiconductors. A neutral: ized push-pull 18-MHz linear power amplifier using IRF-SUs is ineluded in Chapter 11 A high poser bipolar transistor ampli fleris shown in Fig 2.103. This circuit was originally described in a Motorols engi- neering builletin, E63 (ref 22), and was offered in kit form from CCI. (www. communieation-concepts.com) ‘The am- plifier is capable of over 00 W of oatpat over the entire HF spectrum, A matched pair of MRFAS4s is used with a 13.5-V power supply. This circuit isa classic, similar to many of the eutpuc amplifiers in typical trans ceivers, Brass pipe transformers are used at both the input and the output. Some negative feedback Js used, along, with capacitive loading to improve gain flat- ness, This version ofthe amplifierhasbeen tested over the 2 10 30:MHy band and found to operate as described in the appli- cations note, although we did not measure IMD. The circuit has been used exten. sively on the 40-M band. It performed well as a SSB amplifier, being easily driven by a SAW QRP SSB transceiver, It has seen move service following a 1-W CW ‘wansmaitter The original version of this amplifier included an RF actuated eireuit to contsol a built-in TR relay. The RF actuated scheme ‘was found to be completely unsuitable For either CW or SSB use, When RF drive was initially applied, the relay was activated. Hut amplifier current started 16 grow before the output was properly terminated, ceausing the amplifier to draw excessive current, The power supply was current limited at 25 A. As che supply went into limiting, the voltage dropped to 7 V before starting io recover. The relay then dropped ‘ut and the eyele repeated. The relay chat tered for about half second before stabi- lizing, The RF actuated circuitry was eventually replaced with an electronic TR system with diode switching. 772, the output transformer, has a single turn between collectors with a 4-turn sec- ‘ondary, The 4:1 turns ratio cransforms the 50-2 loud to appear_as a 3-1-2 Toad. ccollector-to-collector. The load applied to ‘each collector is then 1.96 Q. Rearrange- ment of Eq 2.42 shows that an ourput of 58 1W should be available from each device at V,.213.5 V for a net output of 117 W. in spite of the T/R problems, the ampli- fier is a recommended circuit. The MRF454 is very robust, and has provided us with classic power amplifier experi- ence, We recommend modified bypassing to use parallel capacitors of equal value, A Look at some High Efficiency Amplifiers Al of the power amplifiers presented are conceptually simple, many using the Sime or similar schematic diagrams, eve though intended for differing applications Class-C amplifiers aze designed by pick ing a load resistance using Eq 2.42 and designing an output netwark to achieve that Toad at the operating frequency. The device is then biased forero current with ‘out drive, With the usual threshold, sppli- ceation of an input sine wave produces Class-C operation, Linear amplifier design is similar. An ‘ourpat network is designed for the peak ‘envelope output, sgain with Bg 2.42. Move 1 toward even lowerlond resistance may ‘enhance lincarity at the price of efficiency. ‘The linear amplifier is biused for class AB operation. This begins with class A bias, but usually allows device current 10 inereuse with applied RF drive, While cffigieney at the peak envelope power is oor, the normal voice has an average power well below the peak, providing a Useful compromise ‘An amplifier discussed earlier (the Fig 2.97 eircuit by W7EL) featured improved efficiency. Tis interesting to examine the otworks that produced this result. Fig 2.104 shows a schematic and a ‘Smith Chart impedance plot for the owipat matching network the Beginner's Trans mitter of Chapler 1. Frequency sweeps from 3.5 to 21 MHz for this 7-MEZ design, The impedance at 7 MHZ is nearly teal at about 25 ©, providing the needed load forClass-C operation. The impedance ve Forallather frequencies. This fe 760n +" tai 50 T \ Fig 2.104—Smith chart plat af the impedance “seen” by the collector of the 2NS321 2-W "Beginner's Transmitter” from Chapter 1. ior Design Basics 2.39Fig 2.108—Smith Chart plot for the Brickette of W7EL, shown in Fig 2.97. The impedance Is inductive until reaching the second harmonic. There Is @ slight Change in the plot when additional C is added at the collector to account for the Zener diode. Fig 2.108—50-0 Smith ch Impedance for a 400-W amplifier ‘operating at 13.6 MHz, See text Fig 2.107—Diploxer, bandpass-bandstop type, used for remit harmonic attenuation from a 7-MH2 12 reader ‘should consult the original OST article® for details. amplifier (T MHz, 2.2-W output, 12-volt upply) was stable and reproducible, but had only 50% efficiency ‘The contrasting amplifier was W7EL's “Brickette” of Fig 2.97, The output net work is also a r-network, and the resulting impedance plotis shown in Fig 2.108. The plot differs from the simple Class-C cir- cuit. The impedanve hasareal partof about 17 Q near the design frequency. but is inductive for much of the sweep. Ry, is about (ice that we would use for a 2.40 Chapter 2 Class-C design. Zbecomes capacitive only above the 2™harmonie, Thisamplifier has excellent efficiency (85 (0 90%) at 7 10 9-W output (7 MHz, 12-V supply) and has been stable. Class-E amplifiers have become of inoreasing imterest in the past few years Recent HEXFET offerings from Intesn- ‘ional Reetifier provide very high power capability at modest price, While the amplifiers are now used only for digital Applications (including CW.) recent work Fig 2.108—Top view of 100-W bipolar amplifier. The board is bolted to a large heat sink that 30 the top of the module. has paved the way for SSB with non-linear high efficiency amplifiers.2# The recent ‘work of greuiest interest to the experi enter evolves from the EE department at California Institute of Technology 2.106 shows na example of a high cfficieney Class-E amplifies.2° The partial schematic shows two modifications to the simple pi-network used inthe other two ci cuits, First, the normal inductor is replaced by a series LC. This provides the same inductive reactance at the 13.5-MHz.Fig 2.108—A 1. using a 2N3B66. -W 7-MHtz amplifier Fig 2.110—An RF power amlifior using an IRF510 HEXFET, The output network 's an LCC type Toe-network. Up to 10 W ‘was obtained from this circuit. sign frequency. but greater inductive re- sctance at higher frequencies. This pre- sents the needed load to the FET drain needed 10 allow the voltage to grow (“ving up") 1o values much larger than the supply and offer the phase control needed for af ciency. & Cliss-E amplifier is eharacter- ‘zed by high current flowing only when the Fig 2.111—A high effieloncy 7-MHz amplifier (circuit of Fig 2.97). voltage across the deviee is close to zero. ‘The other modification is atthe load end of the network. The usual parallel eapaei- tor is replaced with a parallel-conuected series tuned cireuit (88 nF and 390 pF) This circuit jy resonant at the 2! harmonic 2.11 A 30-W, 7-MHZ POWER AMPLIFIER ‘While QRP can be great fun, especially in a portable application, there are times when mare power can make a large differ cence in station effectiveness. The ampli fiershown in Fig 2.112is intended to boost the ouput of a QRP rig to the 30 t0 40-W level with an inexpensive HEXFET. A moderate heat sink is used, allowing extended testing and operation ‘The amplifier requizes about 1 W of rive for full output, Ifmore drive's avail- able, it may be dissipated in an input attenuator, A 3.3-dB pad is shown in the figure. This is followed by TI, a bifilar ‘sound ferrite transformer providing gate Urive for the FET. The low impedance drive is needed to accommodate the high input C of the TRFS30. A 10-2. 1-W resis tor provides a wide band termination The drain cirouit is supplied with a #25-V source through an RFC (L1) made with a large powdered iron toroid. The ‘exact value is not critical. The RF resis tance that should be presented to the drain fora 30-W ourput is 10 ©. This is realized ‘with T2, a bifilar wound ferrite trans- former. This part of the circuit is open to considerable experimentation for those so inclined, T2 is followed by & low pass ter for harmonic attenuation, Inductor LS is tuned for parallel resonance at 7 MH2, ‘An attached resistor then provides 2 {etmination for the amplifier trasistor at Frequencies other than 7 MHz when 9 fof the 13.5-MHz drive frequeney of this example. This amplifier provides xn out- put of 400 W with a drain afficieney of 86%. This circuit, which uses « 120-V supply, could be adapted to the 20-meter amateur band, The load impedance is 13.5419 Q atthe 13.5-MHy operating fe- queney. but is pusely capacitive by the time the 2° harmonic is reached. Eg 2.42 would predict an 18-02 load for this output ‘and Vag. This circuit is very similar to the -MHz design presented in OST for May 1997.7 Spectral purity is an issue with these arnplifiers, The resonant trap at twice the operating frequency included in the designs helps. One would normally insert additional low pass filters to attenuate har- monics, However, this normal low pass filter has an input impedance that is real and 50 GLat the operating frequency, but is almost a short eireuitat the harmonies. An improved harmonic reduction filter ferm is shown in Fig 2.107. This cireut js called a diplexer and has the characteristic that the input impedance is 0 at al frequen: cies, Otherdipleners aze used elsewhere in the hook: Fig 2.108 through Fig 2.111 show some of the design implementations deseribed im this section. ‘rans-match with a peaked high pass char acteristic is used. The combination emu: lates the diplexer deserihed earlier ‘ATIR system is included to supply & signal to the receiver input, As shown, this, system as & measured insertion loss of about 3 dB, the result of the low Q RF choke at L7 and the shunting effect of CL ‘This loss ef no consequence at 7 MHz. ‘An adjustable bias is available for this, amplifier, provided by 1 PNP switeb cir ‘cuit keyed with a signal from the driving transmitter. A grounding signal is applied at Tl to turn on the PNP switch. FET bias is adjusted at RI (SI open) for a few mil- umperes of rain current withno RF drive during key-dowa periods. The switching Amplifier Design Basics 2.41(asernative Surput data) ci-s pitilar cuene flee ture #20 en Fre 2B up mglae EF 22 on BH-43-20L balun Fig 2.112—Schematic for the 30-W, 7-MHz power ampli action removes bias during receive, pre- venting amplifier noise from overwhelm ing the receiver, The standing currant for SSB operation can be udjusted to larger values, up to 1 A, Monitor heat sink tem perature to be sure that it never becomes {oo hot during transmit periods. Throwing switch $1 1 the ow power position allows the power oulput to be dropped 10 levels from well below a watt upto.5 W, contrelled by a knobon R2. This scheme works well even withan output less 2.42 Chapter 2 ‘See toxt for details Fig 2.113 —The 30-W amplifier than the input drive, Initial curn-on begins by terminating the amplifier in a 50-0 load with atleast 30 W of dissipation capability. A current lim ited power supply is attached, RF drive well below the required level is applied while the output is monitored with an os- cilloseope or RF detector. Drive is slowly increased while examining the output ‘waveforms. Clean signals with smoothly varying levels should be seen with changes in drive, Any sudden change sug ‘gests Stability problems. We saw no such problems with this amplifier, ‘Monitoring drain voltage with an oscil- loscope (60-MHz bandwidth) revealed some disturbing characteristics. When CL is absent. the drain ‘voltage contained extensive harmonic current, cvident from the fine structure around the positive peaks. While these harmonies are blacked trom the outside world by the Tow pass file ter, they should be controlled or reduced atthe FET where they can compromise ef- ficiency, The low pass filter was tempo- rarily removed from the system, allowing the wideband output lead to appear at point “BY in the circuit, This immediately cleansed the signal at the drain, removing the high frequeney spikes. The low pass filter appears ns a large shunt capacitance at plane Bin the figure. This Toad is reflected through T2, allowing the trans former leakage inductance to appear at the FEL drain. This isthe load that wil allow the higher frequency currents to low. ‘The ideal solution for this situation is @ dliplexed low pass output filter. mentioned above, Sabin studied diplexer filters and presented hig_work in QEX for July/August, 1999.27 The amplifier used with these filters was described in the Now! Dee 1999 QEX:" hoth papers are excel Jent and are included on the book CD. Weelected not to use a diplexer filter in this amplifier. Rather. C1 is included at the drain, With C1 in place, the drain vali- age gees up lo about 60 V, well within the FET ratings. Although there is still distor- ton in the drain waveform, harmonic cur- rents are not excessive. Several transformer structures were tried at T2, The most interesting variation replaced the wideband transformer with a narrow band LCC type Tee-network, alse shown in the figure, This cirewit was adjusted for maximum output while slowly advancing drive power. Over 45 W of out- Put was available with this eigeuit. The drain waveform was very clean, reaching & ppeak of 75 V, CL was stil present at the FET drain during this experiment. The T-nctwork was designed to provide 10.20 the drain with a Q of 5. Experiments with other networks will allow you to move over the ill-defined border between class B or C ‘operation toward elass E, FETS with highee voltage ratings should be considered for these experiments. ‘This circuit has been used in several variations for years and on several bands up through 14 MHz, Higher bands should alsobe possible with experimentation. We hhave always been impressed with the robust character of the devices. The typi cal power supply used is 2 surplus opens frame linear regulated type with 4-AFig 2.114—Inside the 90-W ampli REFERENCES 1. W. Hayward, Introduction to Radio Frequency Design, Prentice-Hall, 1982, and ARRL, 1994, 2. P. Horowitz und W. Hill, The Art of Electronics, Second Edition, Cambridge University Press, 1989. 3. P, Gray and R. Meyer, Analysis and Design of Analog Integrated Cirewits Second Edition, Wiley, (984, 4. IBEE Standard Dictionary of Electrical and Electronics Terms, ANSUIBEE Std 1900/1984, Published by TEEE and Distributed by John Wiley, 1984. 5. See Reference | 6. See Reference | 7. The ARRL Handbook for Radio Amateurs, ARRL, 1995, pp 17.5-8, 17.10, 17.22.25. 8. D. Norton, “High Dynamic Range Transistor Amplifiers Using Lossless Feedback.” Microwave Jownal, May, 1976, pp 53-57. 9. U. Rohde, “Bight Ways to Better Radio Receiver Design”, Electronics, Feb 20. 1975, p 87. 10, See Reference 1. p 216. IL W. Carver, "A. High-Performance AGCUIF Subsystem”, OST, May. 1996. pp 39-44, 12, C. Trask, “Common Base Amplifier Linearization Using Augmentation,” RF Design, Oct, 1999, pp 30-34 13.C. Trask, “Distortion Improvement of Lossless Feedback Amplifiers Using Augmentation.” Proceedings of the 1999 IEEE Midwest Ssmpositun on Cireuits cand Systems, Las Cruces, NM, Aug, 1999, Vol 2, pp 951-254. 14, V. Koren, “A New Negative Feedback Amplifier.” RF Design, Feb, 1989, pp S4- 60. 15. R, Campbell, “A Novel High Froqueney Single-Sideband Transmitter Using Constant-Envelope Modulation”, 1998 IEEE MTT-S International Microwave Symposium Digest, 98.2. (4998 Vol [MWSYM)) pp 1121-1124. 16. H. Krauss, C. Bostian, and F. Raab, Sotid Stare Radio Engineering, Wiley 1980, 17. B, Lau, K, Chiu, J. Qin, J. Davis, K. Potter. and D. Rutledge, “High Efficiency Class-E Power Amplifiers” QST, May, 197, pp 39-42 and Jun, 1997, pp 39-42. 18, Technical Correspondence, OST, Nov, 1989, p 61. 19. R, Frey, “A 300-W MOSFET Linear Amplifier for SO MHz.” QEX, May, 1999, pp 50-54 and “Letters to the Editor.” current Limiting. Typical curremt is 2.5 A. ‘The use of slight forward bias helps to guarantee stability ‘Phe present imterest in QRP operation is generally applauded as both fon and worthwhile. However, maay folks miss some exciting experimental rewards by an overly strong adherence to a synthetic 5-W limit. This anplifier is a chance to examine the other side of the power switch, See Fig 2.113 and Fig 2.44 for tov views of the 30-W amplifier QEX, Jul, 1899, p 63. 20. N. Dye and H. Granberg, Radio Frequency Transistors: Principles and Practical Applications, Butterworth- ‘Heinemann, 1993, 21, 1, Wyckoff, “Hints and Kinks”, QST, Jun, 1993, p 50-5 2. T. Bishop. “I40W (PEP) Amateur Radio Linear Amplifier 2-30. MHz", Communications Engineering Bulletin, B63, Motorola Semiconductor Products, In, Phoenix, AZ, Jul, 1978. 23. See Reference 17. 24, R. Campbell, “A Novel High Frequency Single-sideband Transmitter Using Constani-Bnvelope Modulation, 1998 MTT-S. Internasional Microwave Symposium, Digest 98.2, (1998 Vol. I [MWSYMp): pp 1121-1124 25, See Referenee 17. 26. 1.F. Davis nd D.B. Rutledge, “A Low- Cost Class-E Power Amplifier with Sine Wave Drive," 1998 MIT'S Imer-national Microwave Symposium, Digast 98.2, (1.998 Vol. Hl, [MWSYM)): pp 1113-1116. 27. W. Sabin, “Diplexer Filters for an HE MOSFET Power Amplifier,” QEX, Jul’ Aug, 1999, pp 20-26. 28, W. Sabin, “A LQ0-W MOSFET HF “Amplifier”, QEX, Now/Dee, 1999, pp 31-40, ‘Amal 2.43CHAPTER Filters and Impedance Matching Circuits Filters constitute one of the major blocks in a communications system snd fare especially important to the radio ‘experimenter. The performance offered by a filter may well define the perfor- mance andfor cost of 4 project. The ‘experimenter who can design and build his or her own filters bas control over that 3.1 FILTER BASICS A filter is, in the most general sense, a cireuit block that linearly modifies the nature of the signals applied to it. When We say linear, we mean that the output is a replica of the input, changed in ampli tude and/or phase, However, no additional Irequencies appear ‘The term domain refers to our emphasis when describing and measuring a phe- nomenon, When filter is examined in the frequency domain, we characterize the fi ter by the way it behaves with different frequencies, We may then eliange focus fand examine the time domain eesponse. For example, we may investigate the time delay imposed upon a signal as it passes through a filter. The DSP filter designer hhas the ability o simultaneously examine and often control both the time and fre- quency domain responses, ‘The response of a filter is measured by examining the transfer properties of the circuit, The voltage transfer function is the ‘output voltage (usually across a termins- tion) divided by the input voltage that caused the output. This is just the familiar voltage gain that we used with amplifiers. Inthe case of a filter, iaat“ gain” is usually floss, anumber less than one, with a cor responding negative dB value ‘Simple filters ave built from mathemati performance and equipment cost ‘There are several ways af segmenting filters into groups. The usual scheme sog- ments filters according 1 frequency response, such as low pass vs high pass. Others methods segment by the kind of ‘components used, In that regard, this ehap- terdeals first with LC filters, and later with RC active and crystal filters. Filters can also be classified by the way they deal with impulses of energy, The filters presented in this chapter are genetully “infinite impulse response” filters, or HR, Finite impulse response liters (FIR) are detailed ina later chapter emphasizing digital si nal processing (DSP). ~ [Insertion Loss at Peak 2 2 | < | alt Cutom Frequency 6 = | a Bandwidth ‘i Frequency—> + | - Passband [Stopband— > Fig 3.1—Low pass filter charactor ies showing the passband and stopband, bandwidth, 3-48 cutot{, passband ripple, and insertion loss. This filter ha approximately 0.5 4B IL at the frequency of peak response while passband ripple ig also 0.5 dB. The vortical axis is the gain through the filter, output power Vs available Input power when tho filter is properly terminated. (Formally, the usual ‘gain used Is the forward seattering parameter, 821.) Horizontal axis is frequency. Filters and Impedance Matching Circuits 3.4: joes LLL Fig 3.2—The frequency respon! cally ideal inductors and capacitors. Such f@ filler, one without resisters. is called lossless. All of the power applied to a lossless filter is available ut the output Real filters containing resistive elements, desired or otherwise. will suffer from some loss. Loss in dB is a positive number, and lows a8 8 poser ratio is greater chan 1 The traditional filters we use are classi fied with regard to frequency domain re sponse, illustrated with @ low pass filtcr in Fig 3.1. This figuce isa plot of filter gain vs lrequency. We encountered several dif ferent kinds of power gain in Chapter 2. ‘The one usually used with radio frequency 8 of various filter forms. falvors is tansducer gain, A low-pass filter is one chat transfers all Input frequencies below a specified curoft frequency. The spectrum below the cutoff is called the passband while the region of higher attenuation above the cutoff is called the stopband, A filter dissipates some oF the available power applied, called insertion loss. The filter of Fig 3-1 has an insertion Loss (LL) of about half « AB at the highest frequency peak. TL is about 0.1 dB at very low frequency. The cutolt frequency is usually defined as that frequency where the response is 3 4B less than the peak passband response. Addi- tional variations in gain within the pass band occur with some filters; these varia tions are termed passband ripple A high-pass filter is similar to the low pass except that the regions are inter changed: the passband. the region contain: ing desired signals. is now above the stopband A bandpass filter is one that passes a given region, often narrow, while reject Jing most frequencies. The bandwidth of a bandpass filter is the difference between tivo points 3 dB below a peak. A band- reject Biller 8 the opposite, a filter that passes most of the spectrum while reject ting u specified region. Finally. an all-pass Filter is one that passes all frequencies applied to its input. The all-pass filter is Useful because itcan aller the phase of sig rnals passing through it without altering signal amplitude. The various ¢ypes (€x- ‘cept forthe all-pass) are summarized with regard lo frequency response in Fig 3.2 Passive filters conserve energy: power flowing into the input must go somewhere. TF input energy is ata frequency within the filter passhand, that energy emerges al the filter ouput where it can he usee. (A frac tion of the energy is lost in any real, passive filter, being dissipated in the losses af the inductors and capacitors that form the cir cuit) In contrast, energy in the filter stop- band is reflectod. That is, an impedance mismatch is ereated by the filter elements suich thar power is not efficiently delivered from the source, through the filter and to the output. Most LC filters display this property. allowing us 10 use input imped ‘ance match as snother way to examine filter performance, The primary performance in- dicutor remains the transfer funetion. 3.2 THE LOW-PASS FILTER—DESIGN AND EXTENSION A low passisa filter that passes frequen- cles below a specified cutntT frequency while attenuating those above, Iris vital ‘component of almost aay communications system, The low pass i also the basis for ther filler forms. Once we have 2 lows pass filter designed, cataloged, and under- stood, the properties and the component values can be extenuled to generate any of he other basic filter types. One extension changes the low pass into a high-pass eir~ cuit. Another modification changes the low pass 10-a bandpass. A band-reject file toris a direct resultof transforming high ppass circuit, itself derived from low-pass provolype. The practical application de tails of these metheds will be presemed, although many mathematical details will be ignored in this weatment, Analytic de 9.2 Chapter 3 tail can be found in Introduction to Radio Frequency Design’ ot numerous other texts AA simple theee-clement low-pass filler is given in Fig 3.3, This circuit consisis of a series inductor and a pair of shunt ca- pacitars. The filieris driven with a genera- lor with a source resistance Rs and is ter- minated in a load of RL. The source and loadare vital parto the cireits the trans- Jer funetion depeads upon having bot fends of the filter properly terminated. A filter that is terminated in resistive Toads at cach end, input and ourput, is ealled a ‘doubly-terminated filter. Mest of the LC filters that are interesting 10 us will be dou- bly terminated Figure 3.38 shows another three-ele- ment filter, This one uses two series in- ductors with one shunt capacitor. With proper design this filter will have exactly the same transfer function as that of Fig 33A. This is a common detail of fil- ters; they ofien have dual forms. We can tell by inspection thst both fil- tors of Fig 3.3 are low-pass circuits. The series inductor isa short eirenit at de and hhas reactive impedance that grows with Frequeney. Hence, it will inalbit the flow of energy through the circuit more as fre quency increases. The same argument can bbe made about the capacitors. They behave i an open circuit at de. However, as fro- quency increases, they show lower and ower impedance, more effectively shunt- ing the energy flowing in the circuit ‘A low-pass filter will have a number of clements equaling the order. The filters ofTST | eS | RY | | vers Fig 3.3—Throe element, or 3rd-order low-pass filters. 02 04 06 08 1 Frequency (M2) Fig 8.5~Butterworth filter transfer functions showing the passband details. ° ore -10 LPF NN -20 aS =e} 8 Tre] | fe ~3) bandpass fillers is similar to the DTC. Coupling between resonators (aumbered ‘mand n)is described by a normalized cou- pling coefficient, ky). Phe values will gen- erally differ for each pair of resonators. End loading, perhaps different for the two nds, is described by normalized end sec sion g values, q, and q, for a filter with » resonators, Denormalization establishes Joaded end Q values that are then estab- lished as with the DTC. The individual parullel-tuned citcuits are individually tuned tothe filter center frequency withall other parallel xesonators short-circuited. A caloulator or computer program writen for the design of double-tuned circuits may ofien be used, without modification, for the design of higher-order filters. ‘The bandpass fillers examined so fur used parallel tuned circuits. Series resona- tors may also be used. This variation is shown in Fig 3.23 with the design proce- dure given in the literature. ‘Wilh either form, values for normalized sand gare obtained from a table ef values such as those published in the classie book Fig 3.21—A riple-tuned circuit centered at 16.2 MHz with a bandwidth of 0.5 MHz. f| a cap A, | FREQUENCY. Mz 2,00 Fig 9.22—Response of triple and doub! with Qy=200. by Zverev. The values may also be ealet lated in computer programs, Sometimes ‘one encounters ables of previstorted &and 4 values. Predistortion is a process to re tainadesired filter shape, even with losses present 1011.12 Some filters are mintures between the forms presented, An example is preseated in Fig 3.24 where the familiar small cou- pling capacitor is replaced with a shunt ‘capacitor, usually large in value. A small value shunt inductor could also be used. Filters at VHF and Higher Bandpass filters are sometimes easier to realize at VHF and above than at lower frequency, the esult of higher wvailable resonator Quat WHE. Building un air-core coil with « Q of even 200 at 2 MHz re ‘vires « considerable volume. However, ‘one with such aQ at 200 MHz ean be very Smal. This results from skin effect chang ing with frequency. Filters and luned elreults built with 0.4 mH inductors The book CD includes a tutorial paper fon the DTC. That article outlines meth= ods for experimentally realizing simple bandpass filters at any frequency. The rechods outlined there are easily applica to VHF and microwave filter, including those using transmission-line resonators. Resonators can take on much different forms at higher frequency. One common and popular form is the quarter-wave- length long resonater. This is built by forming a ection of wansmnission Tine that is just slightly less than 0.25 wavelength. (One end is then short circuited while the other is open eireuited, The resenstor Q will depend upon frequency, geometry. and dielectric material. Air (or vacuum) dielectrics offer highest Q. The conductiv= ity of the surface metal will significantly aifect Q. Copper surfaces are excellent, With silver being even beter Fig 3.25 shows a method for evaluating ‘transmission line resonator. This is & schematic, yet practical scheme for build ing filler elements with, tor example Impedance Matching Circuits 3.14A 9-MHz bandpass filter required for a mixer experiment was built with available components. A triple: tuned circuit was fabricates fram top-coupled parallel tuned circuits. The filter was examined in greater detail after the experiment was finished. While the filler satisfied the immediate need, the performance was far from ideal. A deep notch appeared in the stopband at about 11 MHz, Then what should have boon an ideal filter became 2 disaster with a stopband attenuation of only 40 dB at 40 MHz. This behavior had been observed earlier in a 7-MHz bandpass filter, shown in Fig 3A. The circuit was buill on a scrap of circuit board material that was then botted into an aluminum box. Tho BNC connectors at each end were “grounded” to the board with short wires from solder lugs under the connector nut. The filler was excited with a signal generator while examining the other end with a spectrum analyzer. We observed thai the stopband attenua- tion improved slightly when a screw driver blade short circuited various spots on the circuit board edge to the aluminum box. This pointed toward grounding as a major problem with this fiter. ‘Anew 9-MHz bandpass filter was then built. The components used in the ofiginel, which was built like the 7-MHz fier "bad fiker,” were lited and used in the new one. But the new circuit was fabricated in a box built from circuit board material (Fig 3B). Tho walls were soldered to the box floor, creating a cleaner ground. One of the long walls was initially left off, easing the filter construction. Filter performance was improved even before the ath wall was added. The wall was added and the circuit was measured, revealing stopband null at 43 MHz. The depth was al ~110 dBc, near the limits of our measurement capability. The response al 70 MHz, the top of the spectrum analyzer range, was 83 dBc. Asinglo shiold was added to the fitor that removed the null and dropped the 70-MHz response to ~ 96 dB. The filter is shown in the photo “goad filter." ‘The behavior observed is easily modeled with the circuit of Fig 3C. The stray coupling, related to ground currents, is modoied by lifting all ground connsetions in the fitter and Stopband Attenuation of Bandpass Filters Fig A—Bad filter—This bandpass filter performed well around the 7-MHz agsband but had poor stopband attenuation. A vary deep attenuation notch appeared at about 15 MHz. \g 5—Good filter—A box built rom scraps of circuit board material produced {a response with good stopband attenuation. Fig 3C—The traditional bandpass filter is modified with a mutual inductor, raising the bandpass filter above ground. The resistance in series with the 1- LH inductors represents Qu of 250 at 9 Miz, 3.42 Chapter 3Fig 3D—The response of the ideal filter and that of the mutual coupling inductor are compared. The Ideal response was realized in measurement when one shield was added to the iter. attaching them te 2 common induc tor. An inductance of only 40 pico Henry (yes; pH and not even nH) produced coupling that matched the measured performance. The “before nd after” transfer responses are shown in Fig 3D. Clearly, ground integrity is 2 vital pert of an AF circuit, especially a bandpass filter using high Q resons- tors, Enclosures fabricated from soldered scraps of circuit board materia! or similar solid conductor ‘are ideal, often far superior to aluminum boxes, especially follow ing oxidation. Painted aluminum boxes are even worse. Clearly, measurements should always be performed, O.14-inch outside diameter semi-rigid coaxial cable Like that used in microwave systems. The center conductor is made available at oth ends. Iris shorted with as tle inductance as possible at one end Then, 2 50-0 generator and a 50-0 load. with detecior are loosely coupled (© the hot” end of the resonator. The coupling capacitors may be nothing more than small jeces of wire spaceda small distance (rom the high impedance end of the resonator. The couplings trom the generator and t he detecier should be on opposite sides of he line to reduce direct interaction. The coupling is adjusted for « high insertion joss and the frequency is swept until the ter frequency is found. The unloaded Q is measured by determining the 3-aB bandwidth. Center frequency may be ad jsted by adjusting line length, Trabandpass filler is to be built with the lines, the end section loading may be real- ized with the scheme shown in Fig 3.26. The “grounded” end af the resonator is attached toa coaxial connector ina ground plane. The center wire is attached to the connector und & short is created with a small inductor consisting of nothing more than a very short wire, The wire length is adjusted to set end section Q. The line shield should be carefully grounded very close to the coaxial connector. ‘Once properend section Qis established tnd resonators are tuned! te the proper eon Filters and Impedance Matching Circuits tor frequency, a working filter can be built by placing the 1wo close enough to each other that the “hot” ends are in close prox- Imity. This scheme works well for filters for the 432 and 1296-MHz bands, The line sections may be bent to fit available space. ‘The ransmission-line double-tuned er cuit just described used semi-rigid coanial cable, Another common transmission line filter uses so-called hairpin circuits. ‘Miero-strip transmission lines are printed ‘on circuit board material in his filter. The lines are each a halt wavelength long and are bent into a “U", or hairpin shape. An example of hairpin filter with three reso- nutors is shown in Fig 3.27. ‘The design of these filters is a straight- 343DTC Design Pick a center frequency, F, and a bandwidth, B, both in Hz. Pick an inductor; it can be of essentially arbitrary value, although a good “starting value” would be L=10/F where L's in Henry and F is stil in Hz. The vnloades inductor Qu should be approximately known, One must also pick normalized k and q values. For a Butterworth shape, k-0.707 and gz!.414. For a filter with some passband ripple, but steeper skirts, use 0.25 dB Chebyshev values of &-0.7154 and g= 1.79. The design equations are: o=2wF C4 =1%l Table 3A. Double Tuned Circuits using the sidebar circuit, All filters are doubly terminated in 50 @ at each end. The core designators use the copyrighted numbering scheme of Micrometals, Inc. FeMiz BWEMM2 = Core Tums Lu Ow = Cand 042. Cstuna 18501 68-2 35 aa 200 2509F 41 pF 775 pF 355 Ot 768.2 35 698 200 82 37 220 3s 02 68-2 38 98 200 93 " wT ago 758-2 35 598, 200 3 87 182 TA 02 150-8 v7 156250, 56 87 art 705 Ot 150-8 7 1156 250 35 4a 402 703 04 150-8 20 is 250 30 32 288 tot Ot 150-6 7 i156 280 14 15 189 jor 01 750-6 40 oa 250 20 44 se 141 02 150-5 10 04 260 at 32 295, 142 02 150-8 10 o4 250 34 63 ait tei 02 150-8 10 oa 200 10 15 482 21 a2 750-6 10 oa 200 61 40 195 2125 0S 150-6 10 oO 200 6 23 122 202 150-6 10 Ot 200 28 os7 96 2az 0d 1508 10 oa 150 56 08 3 agas 07 T50-6 10 04 150 38 14 68 5021.0 1506 10 04 150 35 ot 2 vai 02 750-6 5 or 200 987 0128 (1224 141 02 750-6 7 0198 200 7 er 67 wai 02 80-6 10 0 200 9 32 296 ai 02 750-6 15 09 200 3 Va 13? 14102 750.6 20 16 200 25 os 68 14102 150-6 2 25 200 76 os 48 141 0B 750-6 30 36 200 Ba ose 2B a1 02 750-6 35 49 200 5a 02 © 208 Note: Oni 2 couple of core types are needed to cover the entire spectrum from 1.8 to 90 MHz. The last eight table entree describe ine Tete: fae 2 Ta Pie ctcut nih 2 209 khiz bandwidth The number of tune fs allowed to vay, lystrting ha {raegor availabe to tho Fare saseaner, tie buldar with'a compete program set up for daign Gan vary Inductance and bandwidt {D realize 8 desired fier with Standard fane junk-box avalisle) componem values. 3.14 Chapter 3Small Numeric Value Capacitors Top coupled LC bandpass filters often use capacitors with small numeric value. These fare becoming increasingly difficult to obtain. However, a simple substitution will provide the seme coupling, but with larger more ‘convenient values, picked with the equa- tions shown. For example, assume a fiter design calls for a capacitor with C,x=1 2 pF. ‘The substitute network can use any value of Coen that is greater than 2.4 pF. Assume we use series capacitors of 10-pF value. The parallel capacitor is then Coan=63.3 pF. A practical value would be either 56 or 68 pF. ‘The new network will have an equivalent parallel component al each end; you must reduce the capacitance that tunes the resonators accordingly forward chore with 2 modern computer, although it’s a job for professional-level microwave simulation software. ‘The total length of each section is 0.5 wavelength for proper toing. The twoend sections are usvally identical. The lengths pf the end sections are 2(X4) + XS while thac for the middle seetion is 20364) 4X3. End section loading is determined by X2. ‘essentially the spaving from the center of the end resonators, virtual ground point. Coupling between resonators is estah~ lished across the “pap” shown in Fig 3-27, analyzed by considering the overlapping sections as dircetional couplers, It is important for the computer anlysis 10 in= clude the junctions to the 50-2 lines (Tee junctions} and a proper model for the open line ends, The designer must also have ‘2004 information about the board material including loss, dielectric constant, and thickness herween the pattern layer andthe ground foil below: ‘Thehairpin teri generally alossy struc ‘ure when builtenconventional creuit board materials used by amateurs, This material ‘generally has @ Toss tangent of 02, produe- ing resonator Qo 50, As suet, nasTov fillers are not possible, Hairpin filters generally Ihave 10 10 20 % bandwidth unless built on some of the more exonie materials, Hairpin filters have responses at har ‘monics fzequencies, A halt wave resona- lor is resonant at frequencies where the line is 1, 2. 3, ete wavelengths long. “Another popular structure for higher fe- quencies is the helical resonator, These were very popular for UMF FM mobile ralios of just a few years ago. A belical resonstor is a section (usually one quarter ‘wavelength of line using a helical trans- mission line, A helical line is a solenoid coil-like structure placed inside a shielded ‘enclosure. We ean think of a wave as prope: {gating along the wire atthe speed of light. Hence, tae propagation velocity parallel to the axis is much less than that of light, This is a slow wave structure, Cutting « quarter ‘wavelength section, grounding one end ‘with the other open circuited, forms a eso nator. The usual telieal resonator is just under a quarier-wavelength Jong, The e: tra length required for resonance is com pensated by adding a stall adjustable ca: pacitor to the end, often nothing more than A grounded metal serew close (© the “hot end of the eenter conductor. Numerous review articles have appeared deseribing the helical resonator and fiers using them, Equations are often given for resonator dimensions, an impli- ‘cation that they must conform to w well- defined structure, Generally, there is much isreater freedom available io the builder. A Iiclical filter may still work well iFbuilt ia, ‘a volume that is “too smal Fig 3.24—Double-tuned circuit with huni capacitor for coupling between resonators. This illustrates one of ‘numerous bandpass filter topologies that are mixtures of the two methods presented. Fig 3.25—A quarter wavelength of Irangmission line forms a resonant tuned circuit in this example, N=4, ig 3.26—Loading (coupling to the ‘outside world”) can be controlled with small wire inductors. Filters and Impedance Matching Circults 3.15A casual glance may not reveal a seus identity. That i, ahelival resonator with a twning capacitor Looks like a shielded LC resonator. However, the difference becomes clear if widehand measurements are done with Ioosely coupled probes like the ones that have been described for Q measurement. Such measurements wil show a high Q at the fundamental fre {quency and additional responses (alsa hay= ing high Q) at3, 5. and other odé harmon- ics of the fundamental. In contrast, a pure LC resonator will mot show’ these departures. If capacitance is added 10 a helical resonator to decrease fundamental Frequeney, the higher frequencies will not move as fast. Slight capacitive loading ‘might move the first “spurious response" to 4 FO with greater departure as leading grows, Q remaios high and excellent fil ters can sill be built Helical resonators are coupled to each other with a variety of methods, although the most popular is through apertures, or holes in the walls between adjacent reso: ators, As with other filter types, the cou: pling ean be relaied 10 the frequency spread between peaks when te resonators are unloaded, End section loading is real ized in 4 variety of ways with helical reso rators, A small line from a coaxial con- neetor can be tapped onto the helix, The 3.16 Chapter 3 Line wth ur x1 | Gap @ ob une with xa Fig 8.27—Three resonator Hairpin ype bandpass filter. x2, L x8 usual tap point is very close to the grounded end, often a small fractionof one tum, Again, the loading may be adjusted to establish an end section loaded Q. We have only scratched the surface with some filler types we have built, A detailed review of the literature will reveal numer= ‘ous other filter topologies of interest. The bandpass fillers presented here are trans: formed from simple low-pass filters, the so-called all-pole low-pass circuits with nothing more than series inductors and shunt capacitors, Other low-pass filters such a the Elliptic ean be transformed 10 bandpass form to generate bandpass cit~ cults with transmission zeros next 10 the passhand Another variation injects a transmission zero in a passband with po additional in- ‘ductors. This is realized by an additional ‘coupling capacitor that couples energy between non-adjacent resonators. This method was used in a 144 Metz wansceiver discussed later in the book.'* There is. a ‘reat deal of work available to be done by the curious experimenter.3.4 CRYSTAL FILTERS No element is more intimately related 1 radio receivers than the quarl2 crystals used in fillers. The early super heterodynes of the 1930s obtained single- signal selectivity with a crystal filter using but one crystal, u practice that continued ‘through the 1970s, ‘The use of high quality fillers using a multiplicity of erystals be- came popularin the 1950sias SSB replaced classic AM as the radiotelephone method of choice, Crystal Fundamentals A modem quartz crystal is usually a round disco single crystalline quartz with thetlization on each sige. The metal films serve iocreate (and sense) an electric field within the quartz. The basic structure is Shown in Fig 3.28. "The basis for the interesting circuit properties of quartz erystal i the piez- Slectric effect. This effet is a material shuractenistic where an clectic field causes a mechanical displacement. The Iechanical motion istright angles to the electri ied inthe quartz crystal. An clec- tric field occurs when a voltage is placed between the two metelization layers atteched tothe ergstal. The opposite efest ares an electric Feld. “The action ofa quart crystal when sub- jected toan electrical impulseis analogous to striking a bell or chime witha hammer: the energy ofthe impulse causes an osei- samme way, the resonant frequency of a uarizerysial ie elated tothe orystlthick= ness, The Q of aquartz crystal ean be very high, from 10,000 to over one million. The motions of a quartz erystal arc transverse with the erystal vibrating parallel to the surface, This allows the Q and resonant frequency to be altered by surface effets ‘The reader with an interest in the physics ‘of quartz erystal is referred tothe lassie text by Virgil Bottom.'* The quartz crystal is modeled as the LC tuned ctcut shown in Fig 329.1, andC,, ace ternied “motional” parameters or they relateto the mechanical mation ofthe erys- {al The equivalent series resistance, ESR is an element representing losses: it is re" lated to the erystal Q. The final element, Gy. is the parallel, oF holder capacitance. This Cisa simple consequence ofthe erys- tal construction as a parllel—plate eapaci- tor. This valve is the sum of the patallel plate C (the dominant element) and some Stray C related tothe package housing the crystal, ‘The parallel and the motional capacitance are related in the usual AT cut erystal. (AT cut refers to the crystallographic orientation ofthe rystal Many of the crystals we deal with in radio are AT cat) The relation berween eapaci= tors is approximately ‘Table 3.4 shows some measured repre sentative values for some. junk-box crystals. A crystal placed between a S0-02 signal generator and 50-22 load shows a response like that of Fig 3.30. 1 the crystal was a simple series tuned circuit ‘without the parallel expacitor, Cy. the re= sponse would be a simple peak ‘Acrystal filter can be built with « single erystal with the scheme of Fig 3.31 L-networks at each end transform 5010 :0 present 500 @ at the crystal. Transformer TI provides an out-of-phase voltage to drive a phasing capacitor. This signal combines with the energy flowing through the exystal parallel capucitance to control the position of the noreh. The 10-pF ea- pacitor increases the effeetive parallel C ‘of the crystal. moving the notch closer 10 the peak while the 25-pF capacitor res0- nates the ferrite transformer, Fig 3.32 and 3,33 show the result of tuning the phasing capacitor Changing the terminating L-networks can alter the filter response. The band- width will decrease if the terminating impedance is dropped. A link could be used on T1 to replace the input L newwork while an output could be terminated with another wideband transformer. The modi- fied circuit would then funetion well with awide variety of erystals, Bandwidth will, of course, vary considerably as the com. lation to oceur, a ringing that dies out in time. The resonant frequency ofthe chime ° is related to mechanical dimensions. In the quae : 40 Thickness \ je —— wires —] 80 Fig 3.28—Cross section of a quartz 4886 siete SmIe crystal. Fig 2.29—Symbol a quartz orystal Fig 3.20—Crystal in a 0-0 systom with response. This crystal has a 5-MHz series resonant frequency, L, 098 H, 0=240,000, and Cy=5 pF. Table 3.4 Frog. Me byw H Om PF Cy pF ESF, 358 ois ots2 335 50,000 Ba 50 098 0194-2275 BaooD0 128 10.0 020 © 187 28 ogq0a ae Filters and Impedance Matching Circuits 3.17ponents are changed. This filter type could ‘even be used ahead of a rocciver Crystal Measurement and Characterization Earlier we swept an LC tuned circuitthat vias loosely coupled vo generator snd detector. A bandwith measurement pro duced 2 Q,, Loose coupling t parallel tuned circuit cured wth «igh impede ance source and load, The erystal is & Series tuned circuit and needs low im- pednce environment for the loose coi Dling required for measurements, We can 50-0) system measure a crystal in the shown in Fig 334, The signal generator should be well buffered and extremely stable. The input ‘of the cireuit shown begins with a 20-dB ‘ad, compensating for mismatch, The load can be a 5-82 terminated oscilloscope, a spectrum analyzer, or a sensitive power meter. (See Chapter or O57 June, 2001.) A 50-0, switched, 1-dB step attenuator is ‘useful aid in determining bandwidth, A crystal is inserted in the test set (Fig 3.34) and the generator is tuned fore peak output. Nete the peak response amplitude and the frequency FO where it Having measured peak response remove 3-4B attenuation from the system increasing the response, Tune the genera- tor upward until the response drops to the level af the previous peak and record the Frequency. This is one of the -3 dB fie- ‘quencies. Repeat this step by fading the lower -3 dB point. The frequency dilfer- tence, AF, is the 3 dB-loaded bandwidth in He forthis test setup. which will be greater ‘haa the unloaded cxystal bandwidth, Knowing AP. rewurn the generator tothe Frequency of peak response, Remove the crystal and plug the 100-2 pot into the test set. Adjust the pot For the same meter resd- ing: remove the pat from the test setup and measure its resistance with a digital volt- meter. This is approximately the ESR of the crystal oe 4.48uE, isipe] | f 10 ur oe f10 um | 5-30 pF 25.3 pF 4 ~ “| 3:80.71 is 2 bilar A turns #26 on @ FT- 50-61 ferrite toroid. This filter has a 3-48 bandwidth of 1a kHz, per ‘of 5 pF, The solid line represents t tease of exact balance when the phasing ‘capacitor equals the crystal Cy. 3.18 Chapter 3 Fig 9.98—Response of the sing! crystal iter of Fig 3.31 when the phasing capacitor Is at maximum value 030 pF. The solid ine represents the tage of exact balance when the phasing Capacitor equals the crystal Cy. Some experimenters have mounted the pot in a pauel and switched it inte the circuit as needed. This may give inaccu: rate results owing to stray inductance. The pot should be mounted 10 a suilable “dummy crystal” with short leads. A detailed analysis of the method reveals errors, These can be reduced substantially by shitting colowermeasure- ment impedance. ‘The test set of Fig 8.34 is complete, pro- viding both motional parameters and Q information, However, measurements with this apparatus become tedious. A simple erystal oscillator can provide the motional parameters. This circuit Fig 3.38. includes a series capacitor chat may he Switched into the circuit to pro- duce a frequency shift. Related equations © included with the figure. "The required Qy, for filter applications swill depend upon the filter bandwidth and center frequency as well as on the filter shape and the number of resonators. A reasonable tule of thumb for most filters (LC and erystal) is that the “normalized Q" mustexceed twice the number of reso- hators. Normalized , qy, is defined as Qu Fig 3.34—Simple test set for crystal ‘measurement. The pad is a 20-dB, 50-2 circuit. Tho output should be terminated in 50.Q. A maximum input power from the generator would be about ~10 dBm, resulting in a maximum to the crystal of “30 dBm. The 100-0 pot ‘or the crystal for ESR. See text. Approximat motional parameters are: 12-108 .F ay AF As ais AF 326.10. SS Fodivided by the filter Q, or Qu -B wae Eq 3.0 A 500 Hz bandwicth tilter at 5 MHz would have filter Q of 10,000, It erystal Qy=100,000, gi=10 and the filter would be practical with 5 crystals. Generally, the most pructical way to build crystal filters in the home lab begins with a large number of essentially identi= calerystals, These can sometimes be found at focal surplus houses. often for very low prices. Equally good sousces ate mail order catalogs selling microprocessor crystals. Measurements (by W7AAZ) con firmed chat many crystal brands offer good (Qg with a minimal frequeney spread. But this is changing, evem at this writing. The experimenter might consider ordering a small lot (pethaps 10) of a given crystal type. He o she cun then messuce them for Qand frequency distribution. If results are Suitable, another order can be placed for a larger number. Typical cost for these crys- tals is around $1 each, so a batch of 10 crystals is still much less expensive than ordering even one special crystal Ceystals should be matched 10 within $ to 10% of the filter bandwidth co build effective filters. Hence, erystals for a '500-Hz wide CW filter should be matched within 25 to 50 Hz ofa nominal frequency. ‘The recommended measurement proce- dure begins by numbering and marking all crystals in a Set with stick-on labels. The crystals are measured for oscillation fre= quency in the same oscillator. If the “G3UUR" oscillator is used, be sure you specify which switch position is used, and record it in the notes. Measure motional Fig 8.85—The G3UUR method for measuring quartz crystal motional par the motional f eters A meters of fundamental mode quartz Tm the elrcult at Y1 and oscillation is ‘confirmed. The frequency is measured. Then the switch is thrown and the frequency Is measured agein. Typical values are C,=470 pF and C,=33 pF. C,, will hhave same units a8 C.. Be sure that C, includes the stray capacitance of the switeh as well the circult part. Then: " Cs Cp then AF Cy ¥ 2:65 -— us 20g and 1 tw w? Cy where w=2nF with F now in Hz. AF is the F difference observed when the switch Is ‘activated. Example: Use capacitors jentioned above, 10 MHz cryst n7.0239H and C,=10.6 (F, (1000 {F = 1 pF) “=axi0", Filters and Impedance Matching Circuits parameters for several erystals to guaran- tee that there is small spread between erys- tals, Itisalso worthwhile to measure afew crystals for Qy. The data is then entered into # computer spreadsheet where it is sorted according (0 frequeney. making it easy to select matched crystals fora filter. How many erystals shoud be purchased to make one filter? The answer is diffi- cult, forit could vary a great deal with the crystal manufacturer. Generally, the pur- chase of 2 or 3 times as many crystals as the number of filter resonators is a good start, More is always useful, A Larger ot, perhaps 100. almost guarantees a large selection of filters using most of the ery tals. Left over crystals will be used in oscillators. It is rarely practical to build homebrew filters for already existing ‘equipment. Designing Simple Crystal Filters Having characterized a set of exystls, we can now consider a filter design, The procedure will depend on the quality of the filter to be built Some filters are e4s) while others may require extensive and very careful measurement us well 3s eom- puller simulation, Both extremes will be discussed Most of the filters we will discuss use the lower sideband Inder topology. An cexampleis presented in Fig3.36. Thecrys- tals are series elements ina ladder, Shunt capacitors couple energy between adja- cent crystals, A mesh ise loop of a f= der, oneerystal andthe wo shunt coupling capacitors on either side of it A mesh could ilso bea Toad, a mateling capacitor a crystal, and one coupling capacitor Some meshes include a series capacitor to tune the mesh othe same frequency tthe other meshes inthe filter The first method presented ignores the parallel crystal capucitance, testing the crystal as u simple series LC eireuit. This Scheme is suitable for simple CW filters, (Although we think of naurow filters 28 beeing more exotic than wide ones, it is generally easier to build narrow crystal filters.) This willbe illustrated with an cexumple, a ath-ordor fer at 5 MEL with 2.400 Ha bandwideh and a Butterworth shape. The n=4 Butterworth is 4 sym: metrical filter with q.=9,=0.7654, 20.8408, K,y=0.4512, and ky=0.8408, The crystals have @ 5-MHz center fre- queney, a motional induetance of 0.098, H, parallel C of 3 pF, and Qy, of 240.000. Normalized Q is qj=19.2, s0 this is areal dzable filter. Calculating the motional C from resonance at 5 MHz. we find C,,=0.010339 pF. We calculate the cou- 3.49pling capacitors with ey et BUR Bo deky Ber Fela Fq3.lt where B isthe bandwidth; F and B are both in Hz, Substituting. Cyyal54 pF and C, 300 Q yielding the preliminary filter as shown in Fig 337A. The filter has yet to be tuned. ‘The filter would, otherwise. be finished if wwe wanted to terminate in this resistance. ‘To illustrare the general case, we will ter- ‘minate in a larger value, 450 82 A termination Ry will “look like” a smaller value Re if itis shunted with a parallel capacitance, Cy. where [Go -Rp) “YRe 6 -Ry Eg 3.3 Using the values from above, we obtain an end capacitor of 47 pF, producing the next version of the filter as shown in Fig 3.37B. Only filter tuning remains ‘The end meshes, | and.4, re terminated. in a parallel RC cirouit, The equivalent series RC consists of the original end resistance, R,,.andacapacitanee C'where bee ‘T° ‘i Fig 8.37—Evolution of a bandpass filter showing the steps in the design. See text for details. 3.20 Chapter 3 Goa RE Eq 3.4 Cis 193 pF, Ry is 4500 and Ry is 309 0 far this example ‘The end meshes are shown, isolated from the other meshes, in Fig 3.37C while the interior meshes are showa in isolation seties Cf 76.7 pE while the interior ones have a net series C of 100.1 pF. Both will be detuned frem the nominal erystal 5 MHz, but the meshes with the smallest Capacitance will be detuned by the largest amouat. The lower meshes can be prop- erly tuned by added series C so that chey hhave the same net series C as the highest ‘Trequency one, This will occur with a tune ing Cof Eq 3.8 Using Cyjgq)=!00.1 pFand Cy, 576.7 pF. proper tuning capacitor is 328 pF. The final filter circuit is shown in Fig 3.37E. ‘The computer generated response for ‘09 00 Frequency 7200 Fig 3.38—Response for the crystal filter designed in Fig 3.37. Accounting for Parallel Crystal Capacitance ‘The quartz crystal model of 9 is generally an accurate one. C, has little ef» oct in filters that are sufficiently narrow. so was ignored in the previous, design, The 5-MHz CW filter just pre- sented was designed for @ 400-Hy hand with with a Butterworth shape. The shape is very close to an ideal Butterworth Problems inerease as the filter bund- widths grow. This is illustrated with Fig 3.39 which shows the response of two different 3-kH, bandwidth filters using 3.58-MHy TV color burst cxystals. TheFrequency 2) Fig 8.39—The response of two crystal fliters Built trom 3.58-MHz color Burst crystals. One uses ideal crystals with zero CO to produce a symmetrical solid curve is the response we would like, designed with ideal erystals with zero parallel capacitance. Cy=4 pF produces the ‘her response, The filter bandwidth is to0 narrow and the attenuation is markedly in- creased, Itis for this reason that this circuit is named the lower sidebund ladder Filter. Response distortion results hecause the parallel Cy makes the series resonators behave as if they had a larger motional L than is measured. This effect js plowed in ig 3.40 forthe 5-MHz crystals used in the earlier CW filter design. The lower curve shows the etfvet of a 2-pF parallel capaci tance while the upper curve is tor Cy 5 pP. Here, X is the atio of Lyy to Ly. The horizontal axis in the curve is BF, the off set from the series resonant frequency. ‘These effects were discussed in greater Shope. other (with Gasned ine) detail in QE for June, 1995, where dses Glad pr crystal. x¢ke) xa(ee [ po | My 10 ia isi a Fig 3.40—X, defined a5 Lilly is plotted for frequency offset, f, above crystal series resonance in Hz. These 5-MHz crystals had parallel C of 2 and 5 pF. [wt 2a Fig 8.41—Experimental crystal iter. 1.2.3.4 = 3.56-MHz surplus color burst crystals. (Ly=0.117H, Co=4 pF) = 151 yh, 48 turns #30 on FT-50-61 Ferrite toroid.(Amidon) Grim = 3-12 pF coramic trimmer. S00 the referenced QEX paper for adjustment procedure. Filters and Impedance Matching Circuits detailed design equations are given. The corrections related to the effective induc: tance are inchided in the program XLAD.eve. Both the program and the 1993 (QEX paper are included on the book CD. The effective inductance is larger than the normal motional L hy a faclor of 2 or ‘more. This reduces the effective motional cupacitanee by the same factor, Avcord- ingly. the coupling capacitors must be reduced by the same factor. The change also alters the calculation of end resis lance. The new terminations and reduced ‘coupling capacitors will then alter the fil- ter toning. One can build symmetric filters if the effect of parallel eapactiance iseliminated. ‘One way to do this paallels each cxystal with a large inductances. The value required is one that resonates with Co. ‘orming a parallel tap that is then bridged by the series resonant portion of the ery tal, An experimental filter was built 10 examine this idea, The inductance used ‘was smaller than requieed for resonance, s0 small trimmer capacitors were added. ‘Tho filter, built with 3.58-MHz.color burst, crystals fora 3,5-kHiz bandwidth, isshown in Fig 3.41. The measured response is pre- sented in Fig 3.42. Crystal filters built with paralleted in- ductors suffer from degraded stopbandt response. Although the performance around the filter center is as designed, it degrades « few hundred kHz away from center, necessitating the crystal filler be supplemented with an LC bandpass. The Min-Loss Filter of Cohn and other Simplified Forms A simplified non-mathematieal scheme for building erst filters uses the Mine Loss circuit. This circuit isthe resul of fimdamental ork by S-B. Cohn where he described a fomily of coupled resonator filtersthatachicved very low insertion loss ‘hile mainaining good stophand attenuas tion. A really interesting property of these filters was the Tact thal ey used identical resonators that were coupled 0 each other with equal valves of coupling. This means that al shunt coupling capact- tors ina Mineo erystal ier are equal Ifthe filters are designed without shut and loading capacitors tning is erestly simplified. A Min-Loss type crystal filter is properly tuned if + all erystals have the same frequency, + all coupling capacitors are of the same value, C ‘series capacitors having the same capaci tance asthe coupling Care placed in series 3.24Butterworth Crystal Filter, 3.58 MHz Relative Response, dB 00 20 40 60 80 Relative Frequency, KHz 190 120 140 22 Chapter 3 Fig 3.42—Measured response for the filter shown in Fig 8.41, A three element crystal filter at 10 MHz. The metat can crystals have ‘small wires soldered to them ‘that are then grounded to the ‘oll. Throe experimental crystal filters. The top circuit uses 10 crystals in a circuit ‘with equal coupling between resonators (Cohn). The bottom filter is that from Fig 3.41. Fig 8.49—Min-Loss type crystal filter uwith equal coupling and simplified tuning. With both end erystals + both terminations are equal and properly related to coupling, A crystal filter of this type, with five resonators, is shown in Fig 3.43.17 ‘This filter topology often appears with the name "Cohn Filter,” titled for the ‘original cizcvit theorist who contributed so extensively to our design methods, Other filters have also appeared with the Cohn name. Here we have divorced the ‘name from this Simple crystal filter. for it is but one example from Cohn’s body of ‘work, a collection that is much richer and more extensive than bas been presented in the amateur literature. ‘While most of the Min-Loss erystal fil- ters we build are fabricated without design Ge., without any mathematical analysis). they may certainly be studied and designed fn the computer. The normalized coupling coefficients and end section Q tor this fil ter type are approximately given by 1 ‘where mis the number of resonators, These values are tabulated for » from 2 to 10 in ‘Table 3, (The first few points appeared in the original Cohn paper, while & and g for N>S are extrapolations via our above equations.) Shown in Fig 344A are wansfer func tion plots for to different fillers of this ype. The wider, lowerloss one has 3 reso~ nators while the other has 8 crystals. Both circuits were designed for S MHz with a S)0-Hz bandwidth using bigh Q erystals ‘with Lyj=0.098 H. Part A of the figure shows close-in details while Fig 3.44B shows the response to the -80 dB level Part C of the figure shows the group delay for the filter with 8 resonators. (More will bo said about group delay shortly.) All three plots are computer generated re- Table 3.5 yok a 2 707 1414 3 oss 1.587 4 0595 1.683 5 O54 L741 é set 1.782 7 0552 181t a 0545 11834 9 O64 1.852 10 0836 1.866sponses, although taey are in good & ment with measurements on similar filers We have built Min-Loss crystal filters up to 10th order, ‘The data of Fig 3.4 illustrate the salient properties of the Cobn ker. The passband shape is smooth with miaimal ripple for the low order filters (N=3), but becomes distoried as the number af res0- nator grows beyond five. The ripples on the passband edges near the skirts become extreme with wider bandwidth filters. The data of Fig 3.448 illustrate the excel- lent shape afforded by the Min-Loss filter However, the time domain performance as depicted in the group delay plot suggests Frequency) “ 20 a0| Len - say a | 0 | Frequency ® 000 Frequency © 0007800 Fig 3.44—Min-Loes crystal filter responses. A and B compare 3rd and 8th order filters in responses to -20 and “80 dB. C shows the group delay for the 8th order filter, that chis iter may have severe einging if built for narrow (CW) bandwidths Although the ewo filters (N=3 und N=8) described in Fig 344 have different responses, they ate remarkably similar in component values. The N=3 filter used 146.pF capsicitorsnd 181-0 terminations while the N=8 Gller used 168 pF and 155 Q. A filter designed with uso or three crystals can be extended with the same capacitor values and terminations This becomes extremely useful for the experimenter. ‘The Min-Loss crystal filter has virwes of low insertion loss und good skirts. but at the price of poor passband shape with higher orders. Some other filters offer similar non-mathematieal simplicity and better passband performance, witha group of erystals all at the same frequency. Fig 3445 shows such a filter. This design is a Butterworth design at 10 MHz with nor- malized parameters of g=0.765, ky: koj20.841, and kyy=0.941. This filter is designed with a pure resistive termination atthe ends (no shunt end capacitors.) The equations predict the end resistance and the shunt capacitors. The series tuning ea- pacitors are yet to be established. How- ever, the values are elear from inspection. Ifthe end capacitors are set to the value of the center capacitor (85 pF.) each mesh bas the same capacitors inthe elated loop. Design with the equations docs not take he parallel erystal capacitance effects into account. This is done with curves Tike those of Fig 340 that establish an increased effective induetance value that can then be applied with the equations. Approximate designs without the curves will still result in practical fillers at the higher frequencies (8 MHz and up} the bandwidth will be a bit nar- rower than the design values. Ringing, Group Delay and Filter Passband Shape All serions receiver experimenters have their Favorite effons, receivers with speci- fications differing litte from oibers, but with a “erisp sound” chat sets them apart from the ordinary. There are numerous phenomenon that tend to degraded peckor- ‘mance and remove “crispness.” One that ccan ruin an otherwise excellent receiver is an IF filter with excessive group delay. Al filters have time delay. a sth that canaot be avoided. The filters that “sound” the best are those that have small delay for a given bandwidth and, of greater import, behiave like a transmission line with litle variationin group delay over the passband. The group delay of an eighth order Min- Loss filter was presented in Fig 3.44C. The delay was high, exceeding [0 rniliseconds Jn part of the passband. ‘The group delay variation over the passhand was also severe. This filter, although very selective, ‘would probably not sound goad, espe- cially with noise pulses. ‘Two S-MHz filters were designed for a bandwidth of 500 Hz. each with five crystals, One filter used a 0.1-dB ripple ‘Chebyshev response while the other used a linear phase response. The Chebyshev r2- sults are shown in Fig 3.46 while the linear phase response is given in Fig 3.47, Both plots overlay group delay and gain, The “ears” of the Chebyshev group delay plot line up sith the 3-dB edges of the pass- band, so all delay variations are heard. In contrast, the region of low group delay in the linear phase filter extends well beyond the filter bandwidth edges. Both of these filters have been built and wied in an experimental CW receiver. The linear phase filter was more difficult to build, but sounded much better. The skists were steep. in the Chebyshev. so it presented adequate selectivity. We found the linear phase filter in need of more skit selectivity. Altha not shown in the figures, the Chebyshev filter group delay was 2.5 times as large as the linear phase Titer delay ‘We have also had good results with an intermediate filter shape, the Gaussian: 10-6 dB response. This is a filter with a rounded peak shape for the top 6 dB, but With steep Chebyshev-like skirts Transi tional filters (Ganssian-to-6 dB, Gaussian: 10-12 dB, linear phase, und maximum flat delay) are slightly more difficult © build than the Min-Loss, Butterworth, or Chebyshev filters, for they luck che Sym. eee “Taye 00 Fig 3.45—10-MHz SSB bandwidth filter using crystals with identical frequen has a Butterworth shape; the simplified tuning and “easy” tuning. This f method often works well with Ne4 Chebyshev filters. Fillers and Impedance Matching Circuits 3.23Fig 3.48—Group delay and gain for a Chebyshev crystal filter. The gain is plotted over a 20-48 range. metry ofthe traditional types. Ifthe transi- tional filters were commercially available, they would probably be very expensive. ‘On the other hand. they offer a challenge that is well worth the effort for the ad vanced experimenter. The reader should 3.5. ACTIVE FILTERS While most reecivers ate super-hetero- dyne designs with an TF, some simple superhels as well as virtually all direct, conversion receivers obtain much of their selectivity from audio filtering, Audio fre queney inductors have become available jn recent years, making traditional LC designs viable at low frequencies, Even prior to the arrival of those parts, some builders had built audio filters with sur- plus elephone toroids, Still, the most com: ‘mon method for audio filtering uses RC active cizeuits. An RC active filter com: bines gain with resistors and capacitors to synthesize induetor behavior The Low Pass Filter Figure 3.48 shows an active low pass file ter form known as the voltage controlled voltage source (VCVS). It uses an opers- tional amplifier configured as & non- inverting amplifier. usually with « gain of one, Two resistors and two capacitors com- plete the circuit, Fig 3.48 shows pact values for the two resistors, here assumed equal, and one capacitor. The other capucitar is & ‘multiple of the frst. A representative set of responses is shown in Fig 3.49 where A has 3.24 Chapter 3 Fig 3.47—Group delay and gain for a linoar phase crystal filter. The gain Is plotted over a 20-d8 range. review the work of Carver Intuition would suggest that a FIR (finite impulse response) filter, usvally realized with DSP, would have signifi- canily reduced ringing. Some do, but some others still show significant ringing avalueof 1,2, 5, and 10, A peak appears in the response. A exceeds 2, The circuit po vides a voltage gain of 1.T when A=10. ‘The filter has a two-pole Butterworth response when A=2. For A <2 and for equal R, the 3 4B cutoff frequency is given by -4at Qe RCA yaq24q2ea Bq 3.18 where A is the capacitor ratio, C2/C1, For example, with R=10 0, C1=.01 {uF (01 WE = 10 nF), and A=1 (equal capacitors), the cutoff is 1024 Hz. Eq 3.18 can be solved for R for an arbitrary cutoff frequency. If Acxeceds 2the fltersakes on apeaked response. It is then more convenient to ‘work withthe peak frequency as a function ofR,C, and A, the capacitor rat. IF A>2, the peak frequency is given by Freak Eq 3.19 2a ACR Extreme selectivity always seems to bring some ringing. Generally, it is the less selective schemes with smooth peak shapes that always sound the best, without regard to the method used to achieve it, traditional nardware or digital signal processing. Fig $.48—RC active low-pass filter. Tho ‘op-amp Is assumed to be powered from ‘round ground. Other ie ie configured of 1. C2, the. ‘noninverting feedback capacitor, is Ax C1 where A is a value greator than 1. Table 3.6 A Voltage Gain = A__ Voltage Gain 22 1.008 ba tat 24 1ota 0 t67 33 1.088 a 24 38 ine 3 28 3a 14 478.48 47 12B20 04 02 08 040508 08 10 Frequency (#2) vet say vas) = sa) Fig 3.49—Response of the filter shown in Fig 3-48 with A=1, 2, 5, and 10. These curves, and several others in Compact Software, The solid line corresponds to A=t whi 0. Some values of low pass voltage gain at this type is shows in Fig 3.80. Three two- the response peak are tabulated vs A, the pole sections with A: Table 3.6. orm a 6-pole filter suitable for SSB recep: ‘There are numerous ways 1 design tion, The response fortis filteris shown in practieul low-pass filters with the equa- Fig 3.51. The dip at low frequeney results tions. A cascade of sections like those in from the 1-F input coupling eapactior. Fig 3.48 would form Butterworth or Cascades of peaked low-pass filters Chebyshev filters of high order. Each (A>2)can be very useful. The gain can be capacitor corresponds to one pole in the considerable when several stages are cas- response, one L or Cin the traditional fil- caded. These filters take on a bundpass like ter. Generally, each (worpole low-pass shape, offering an attractive response for section will differ from the othersinhigher direct conversion receivers intended for ‘capacitor ratio, order Butterworth or Chebyshev filters. CW use. For details, see the text by Johnson, etal.” ‘The filter shown in Fig 3.50 is biased Alternatively, several identigal low-pass for single power supply operation, This sections can be cascaded (0 form « useful Scheme is especially attractive with the cirouit. These filters are easy c analyze lov-pass filler, foran entire ehain of Filter and design, and offer excellent perter- sections may be biased with only one mance, especially with simple direct con- divider. IT LM-358 or LM-324 op-amps version receivers. An example ofa fillerof are used, a pull down resistor should be Is section, were generated with Super Spice from tho highest peak is for are cascaded 10 connected from the amplifier output to ground. The resistor should pass a stand- ing current of about I mA, Severe cross- over distortion will result without chis louding. High-Pass Filters Figure 3.52 shows a VCVS type high+ pass filler, This circuit is the dual of the Tow pass just discussed. Iris designed with ‘equal valued capacitors. The resistors now differ by a factor “A”. The usual filcers have the grounded resistor asthe one with, larger value, Fig 3.53 shows the response 0 20 329 8 So CC ae vinay — Frequerey tz) Fig 9.51—Response for the cascade of identioal low-pas: in Fig 250. This although we have but eeveral designs Fig 2.52—Valtage controlled valtage ‘source high-pass filter. The operational amplifier is again set for a closed loop. gain of +1 “2 20nF +12 Fig 3.50—Practical low-pass filter that can be bullt with common op-amps, such as the 741, 1458, 358, 324, 5532. Filters and Impedance Matching Circuits 3.25quency domain response is where s is now the complex (LaPlace) frequene: the shunt capacitor while AxC is the feedback capacitor. The conesponding Ire ‘The VCVS low-pass filter with equal resistors has a crunsfer function of 7 x Bq 3.20 yin the frequency domain. Cis Eq 32 vay vines wa mee T | 03 040506 0810 20, Frequent) 30 405060 80700 Viz} = = vse Fig 2.52—Transler funetions for four versions of the high pass section of Fig 3.52. The resistor rat corresponds to for four different filters, all with 10-nF ‘capacitors and a 20-49 ungrounded resis tor. The grounded resistor varies to set ‘gain and peaking, The values used are 20 KO. 10 KO, 4 kO, and 2k, ‘The characteristics ofthe high-pass see- tion are much like those of the low pass, ‘The eircuit begins 10 take on a peaked re- sponse when A exceeds 2. A peaked high ‘ass will huve a peak frequeney given by L 2m CRA Eq 322 There is np peak if A<2. The pure high pass then hes a 3B cutettfrequeney given by 3.26 Chapter 3 varies, taking on values of \=1 while the highest peak is for 2,5, and 10. The solld line Ey 3.23 The VCVS high-pass seetions donot have a de path through them that allows the easy biasing afforded by the low pas. A high-pass section may be biased with the methods shown in Fig 3.54 when dual power supplies are not available, ‘The high pass and low-pass forms may ‘be combined in acaseude to form bandpass filters with excellent stopband attenuation, An example response is shown in Fig 3.55 ‘where four peaked low-pass sections are 5532. The tpoloy Fig 3.54—Biasing method for high-pass filler sections. A voltage divi ‘a synthetic ground at half of the single supply, DF 10 “00 Freausncy hz) vig2) — Fig 3.55—The 4x4 filter, a cascade of {our peaked low-pass sections (6.8 KO, 10 nF, and 50 nF) followed by four eaked high-pass sections (20 nF, 27 ko, and 8.6 KO) cascaded with four peaked high-pass sec tions Active Bandpass Filters A bandpass-filter section is shown in Fig 3.56 using an operational amplifier in ‘an infinite gain multiple feedback circuit. ‘The IGMFR cireuit is practical with com. mon op-amps such as the 741, 1458, and is represented with two equal valued capacitors and three resistors, One of the resistors allows the user to specify cizcuit gnin as well us cen: ter frequency and Q or bandwidth. The de- sign begins by picking these values for voltage gain K (a dimensionless ratio}. Q, (qin Ha, and C in Farads. The required resistors are then kg 3.24bn pet me lt ‘Output Fig 3.56—Infinite gain, maltip feedback (IGMFB) bandpass filter. This topology is capable of moderately high and gain with practical components. Frequency Fig 3.57—Calculated gain in dB for the IGMFB bandpass filter shown in ig 2-56, This version used the resistor and capacitor values calculated in the text for Q=5 at 600 He with a gain at resonance of 2. The solid curve represents the nominal response while the dashed curve shows the result of tuning R2 to a lower value. Changing R2 to 1 k= variable in serios with @ 560-0 fixed istor would produce a tunable bandpass characteristic with ‘essentially constant gain and idwidth, This tuning scheme works ‘woll only when R1>R2. This sweep was generated with Super-Star Professional trom Eagle Software. 2-9 ae R where ay=2xmxf, We see from Equation 3.25 that the gain should be less than 2Q2. For example. a filter using 22-nF capai tors with a venter Frequency of 800 Hz, 0 of 5, and a gain at resonance of 2 is built with R1=22,600 Q, R2=942 Q, and 390.4 KO, The transfer fumetion for this, filter is shown in Fig 3.87 The IGMFB bandpass filter must be biased with the method shown ewalicr fora high pass filter if a single power supply is to be used, This filter form is ideal if Fig 3.58State-variable audio filter for CW receiver applications, All op-amps are 741 or 1458. The op-amp pin numbers are not shown. The b F must also connect the power supply line to the Vee point on the op-amps. This circuit was inserted between the audio gain controf and the output amplifier in a high Pertormance CW receiver. several sections are 10 be cascaded. It is sometimes useful to provide a rotary ch allowing the user the ability to select one of several ourpunrs ina cascade, Each section of a IGMEB filter can have a Qas high as 10 oF 20. Other bandpass circuit forms are alse suitable, An especially interesting one is the so called state-variable filter, which uses three operational amplifiers. The one circuit will simultaneously provide low pass, high pass, and bandpass outputs Adding one more op-amp will even allow 4 notch filter function. An example is shown in Fig 3.88, This circuit is tunable over the normal range used for CW notes and has variable Q. The notch is not included in the version that was built, but could be added with the circuitry shown, The reader interested in more informa- tion on the stale-variable filter should examine the article by Howard Berlin? ‘The state-variable filler is an especially interesting cixeuit for chose with a math: ‘ematical inclination, forthe circuitry is an ccxact replication of the equations The All-Pass Filter An especially interesting, bul very simple RC active filter cireuit is the all-pass of Fig 3.59, This citeuit uses an ‘ep-amp, a single section RC low pass file ter, and a pair of resistors. Although we analyze the cigcuit with mathematics, much of the behavior is clear from inspec= tion. At very low frequency, the capacitor is an open circuit. The op-amp input impedance is very high, so the input vel: age is also that appearing at the point marked “E," ‘The negative feedback fetion forees the inverting op-amp input toalso be E. The only way for this to hap pen is for the output ioalso equal E. Atlow frequency the output is in phase with the input and has the same magnitude for unity gain, In contrast, at very high frequency, the capocitorisa short circuit. The ep-amp 1 Fig 3.59—Basie, single section all-pass filler. This circuit has unit for phase shift networks such as those used for the phasing method ot single sideband. Filters and Impedance Matching Circuits 3.27Freqney i) ase vis —™ Fig 3.60—Phase response tor an allepass filter then behaves as the familiar inverting ‘amplifier (180deyrees of phase shift) with unity gai, ‘The transfer funetion for this cizeuit is oq 3.27 nxt with Fin Hz. This circuit a all where has an umplitude response «of unit all frequencies and « phase shift given by Fy 3.28 where @ = f/f, with f being the fre- {gueney where the network has a 90 degree phase, fj is given by Eq 3.29 ‘The phase response of the network is presented in Fig 3.60 for the case of R= 10 KO and C=10 a ‘A common application for the all-pass network is to generate the audio phase shift needed in a phasing type SSB receiver or wansmitter. Examples are found in Chapters $ and 9. A FIR Bandpass Filter ‘The all-pass filterserves asa frequency dependent delay element for a variety of applications. An unusual one is in a spe- vial bondpass filter, one with a finite impulse response, The asic, repeated clement in this filter is a delay element, shown in Fig 3.61. The delay arises from a eascude of qwo all-pass networks, The RC in the al-pass is picked for 90 degrees ‘of phase shifl at 800 Hz, Hence, the ex ccade of two has 180° shift at $00 Hz, The Shi is less at lower frequeney. but more at higher lrequency. The circuit of Fig 3-61 behaves Tike # transmission linc with length of one half-wave at 800 Hz The haliwave lines are repeated and eas- cexded to forma line thatis, in this example, 4.5 wavelengths long at 800 Fl, shown ia Fig 3.62. The line is tapped at each half ‘wave point, Because the line is built from several operational amplifiers, the tap points are low impedance und can be loaded without interaction or other adverse consequence, difficult with a real Uansmission Line. A sinusoidal audio signal at 800 Hz is applied to the input. The signal looks the sane atall points along the line except for changes in phase. Ifwe extract (wo signals from two taps on the line that are separated by one full wavelength, the (wo signals Will be in phase, If the owe signals are added, they will produce a signal that is tweice the original, If, however, the two tapsare one (orthree, orfive,..)halfwave- lengths apart. the result is complete can- callation, for the two components are then ‘equal in magnitude, but out of phase. The cancellation can be turned into positive reinforcement if we add 180 degrees of phase shift co one before addition: this results from an inverter. Fig 3.62 shows a complete filter. All taps with even numbers are summed to- gether in a summing amplifier U1. U2 serves a similar role for signals from odd numbered taps, U3 inverts one resultant signal with the final output extracted trom Us as the sum of the two. An output re- sponse is presented in Fig 3.63. This filter has a characteristic tha di fers from the typical audio filter, the finite naruze of the impulse response. The usu bandpass audio filter, such as described Fig 3.61—Half wave transmission line emulator. 3.28 Chapter 3 earlier, will ring virtually forever when Subjected 1 a noise impulse, The long ringing is evident from the mathematics: it is ulso evident from listening to such a fil- ter. In contrast the FIR filter has a impulse ‘response that is limited to the total delay of the all pass scrueture. A filter Tike this one ‘will still “color” noise, but that noise will rnot bring ubout the sometimes tereible ringing that would occur with «cascade of high Q resonators, Note the rounded peak shape: it’s similar to that founel with filters with the better time domain responses, ‘The filter circuit shown in Fig. 3.62 is rot completely impractical. although it is rot recommended as a construction pro- ject. One of the authors built several FTR audio bandpass filters in the late 1970s. In some, the signals from the taps had uuncqual weighting, accomplished by changing the summing resistors from each tap. The number of tps grew to impractical extremes, (Don't ask!) ‘Taps ean be added as the delay length pros, The results were mixed with the eventual conclusion that a Filter ofthis type was aot practical in simple analog form, The experiments were, none theless, among the most enlightening that ‘we have ever experienced! ‘A lage number of taps is possible and completely practical today in FIR filters based upon digital signal processing. Its, informative te continue the analogy, + A DSP audio filter begins by sampling the incoming signal. The incoming sig- nal is merely a voltage that changes with ‘ime. Sampling means that the signal is captured atone instant in time. This must ‘aceur quickly and often, at least twice forevery cycle forthe highest frequency that our audio system will process + Bach sample is applied co an analog- to-digital converter, The A to D provides 1 stream of data that ean he processed. It ‘can be done in a high speed gen pose computer orin special circ signed specifically for this task. The digitized data is stored in computer + Computer memory also contains data that ‘was stored from eszliermoments. Remem-= ber that we are sampling the signal atleast Istice per eyele for the incoming data we wish to process. The memory has the data Just sampled, that from one sample peril back, from two periods buck, and so forth, ‘extending into the past by number of taps” commensurate with our ability to store and process. + Aceach interval in the process, we will multiply each of the stored numbers by ‘a constant, weighting the samples in the same way that they sre weighted by the somuming resistors in the analog filter, ‘They are then added together to obtain aPN ee al Fig 3.62—A Finite Impulss Response, or FIR bandp. 35 filter built from a cascade of all-pass filters. This filter has 9 taps. Op-amps Ut through US serve to add final result + The digital output “word” is applied toa DAC, a digital-io-analog converter that provides a signal chat ean be injected {nto an audio amplifier and, eventually. headphones. + Data is climinated from memory at each step in the process, We only go as far back in time as our computing power will allow. Among the significant lessons that emerge from a study of FIR filters is the 3.6 IMPEDANCE MATCHING NETWORKS ‘Most filters built from inductors and capacitors were designed to achieve 9 desired frequency domain result: They accepted an input consisting of many fre> ‘quencies, but allowed only a few to emerge at the output. Other LC cireuits are designed for impedance transformation. An impedance transforming or matching net ‘work is one that accepts power from a gen- erator with one characteristic impedance. the souree, une! delivers virtually al ofthat power fo a different impedance, the loud. Both source and load may be complex with both real and imaginary (reactive) pacts Simple designs ate peciormed at only one frequency. More eefined methods can encompass a wide band of frequencies. Impedance transforming networks en erally have filtering properties, even if they are not designed for that charaeteris- tic. We found earlier, for example, that a modified low-pass filter could be termi: rated in an impedance that differed from the original design value. serving a wideband matching role ectional Impedances Consider point A in the circuit of Fig 3.64. A frequent question we hears, “What isthe impedance at point A?” This question does not have a good answer, for wwe did not ask the right question. Imped ances are dircetional, A better question \would have been, “What isthe impedance looking into the amplifier from the plane marked by A. The cireuit in the figure is a simple amplifier operating a, for example, SO MHz, The input impedance looking into the base is 20 -/10.0. This value would be reasonable for an RF transistor biased to a few mA and operating at Fy/10. Wishing [RC Antes Temavoreal kor 20 Tape Fig 3.83—Transfer function of a 10-tap FIR filter. realization that filtering is a comparative process; « signal is compared with a rep- Tien from an earlier point in time. The nature of the comparisons direct and clear in the FIR filter. Its present in the simpler filters, be it single LC resonator or erys- tal, or an active version with an identical function, The signal components from ear Fier times vanish from the resonator as they dissipate in the tuned circuit losses to transfer as much power inte this umpli fier from the source as possible, we will strive for a eonjugare input match by de: Signing a suitable input network. One of many possible networks that will realize such a transformation is the L-network shown, transforming frem 50 down to 20 G2. If we then add an inductance with 10-Q2reactance in series with the inductor of the L-network, we will have trans: formed the 50-82 source 10 look like the desired 20 +10 needed by the amplifier. ee a] e Fig 3.64—An amplifier with matching networks at input and output illustrating inees. See irectional impet Filters and Impedance Matching Circuits 3.29We were cateful to match the input, but will not seek a conjugate match at the out- put. This often oceurs with, for example, power amplifiers where we present a spe- eifle load, Zyqap» to the collector in order to realize 4 well defined output power. But this load will generally be different than a conjugate mateh to the amplifier output impedance, Zgyr. Al- though a conjugate output maich may well provide the highest gain and the maximum ‘output power for small signal conditions, ‘that output Toad could produce limiting ‘that constrains large signal output power. [nput matching resulted froma low-pass type Lenetwork, Aninput blocking capaci (or is an integral part of the amplifier Output matching is performed with high: pass type L-network, whieh serves double uty by providing a route for Vt reach the transistor. There is no “perfect” march anywhere through the output. Recall also that changing the load presented (o the amplifier will probably alter the input im pedanee, We often build transforming networks that will present impedances for reasons ‘other than matching. Output loading for power wus mentioned, We sometimes present impedances at the input of low noise amplifiers that will optimize noise igure, usually different than those that provide best guin, We must be clear in etining our goals when designing match- ing circuits, and exercise similar elarity ‘when talking about such eircuits. The L, x and Tee- Networks Pethaps the mos! cammon LC imped tance transforming network is the L, so named hecause it uses two elements, one as aseries element with the other as a par allel one, resembling the capital L on it’s side. Both L-network forms ate shown in Fig 3.65, The lower value resistor, R,. is transformed by adding a series reactance ‘The higher value, reactive impedanee, is resonated at one frequency with a parallel reactance, yielding a load that looks like 3 real impedance of value Ry ‘The sume equations apply if we wish to transform a higher resistance, R, 10 "look like” a lower one, R, This bilateral nature is general characteristic of all lossless net= ‘works. The derivation of these equations is outlined in Chapter 4 of Freroduction to RF Design! We can define a network Q as the ratio of the parallel resistanee, Ry in this example, to the reactance of the parallel clement. That is, we treat the network as if i was a parallel tuned cireust. Newwork Q is related to the voltage transformation of the network, but is not always a good indicator of network bandwidth £q 3.30 oy? Ri =X, Xp X : Eq 3.31 Eq 3.2 Consider an example: We wish to trans- form a 10-0 resistance 1 look like 50).0.4¢ 7 MHz. The series reactance, from the equations. is 20 £2 and the parallel one is 25 0. The low-pass form, the L-network with a series inductor, would use L= (0.455 4H and 909 pF. The high-pass form would use 0.568 HH and 1137 pF. Both networks offer essentially identical per formance ut the design frequeney, hut dit fee in theie filtering properties. The Q of this L-network is 2. Qisa characteristic of the L-network that is established by the transformed impedances, Another popular network is the pi named because its three elements resemble the Greek x. This network is shown in low pass form in Fig 3.66. Aguin, Ry is resteicted is now a network parameter that the designer must pick. Trean take on a wide variety of values, although they are bounded. The lowest Q allowed is defined by Eq 3.32. presented above for the L-network, Tf you used this value. the Fig 8.65—L-Network with design equations when R, < Re, 3.90 Chapter 3 Fig 3.56—Schematic and corr {design equations for the popular network pi-network equations collapse to these for the L, Low Q values are generally preferced with the Tow impedances usually found With solid-state circuits, offering more practical component values and lower net- work foss. Higher Q tends to restrict hand- ‘width, just as it would in a simple tuned circuit. It also exacerbates the effects of loss in the network L and € pauts As an example, we examine the same 10-Q lowe that must be transformed to 50 Q; we pick a network Q of 5. The results are Xe=10 Q, Xey=4.88 2, and X,= 13.5602, AU? MHz, the respective compo- rent values are 2274 pF, 4660 pF. and 0.308 HL A high-pass variant ofthe pi network is tulso possible. The pi-network component values may not be as practical as thase in some other circuits, especially when Q is high, R,2R, Eq 3.93 Although less common, a very practics! ‘and usefil network is the Tee using to capacitors and one inductor. Component values are practical amd loss is low, espe- cially for the Tow impedances found with solid state circuits. The design begins by picking a network Q. ‘The T-network has the sume minimum Qas the pi network, which is the Q of the L-neiwork given by Eq 3.32. The Tee cireuit is shown in Fig 3.67. Intermediate variables, A and B. are used in these calcu lations. We pick the sume example used beforea op Fig 3.67—LCC type Tee-network and design equations. With R, #10, Ry=50, and Q=5. ‘The tesulting reactance values become Xo =88.12, Xo =102.5, and %,=50, all in &. At 7 MHz, these values correspond 10 258 pF, 222 pF. and 1.137 fH, respec- tively. ‘These components are especially practical for both input and output net- ‘works of RE poweramplifiers if micacom- pression variable capacitors are used, R,>R, 8-R,-(a? +4) £99.06 £q 3.37 £q 3.38 £43.09 Xe, = A-Re Eq 9.40 Inereasing the inductor, then adding a series capacitor that cancels the added inductive reactance, may modiiy all the networks described. The modified net- works are more easily adjusted and can provide narrower bandwidth, We often view m or T-neiworks as back to back L-networks, transforming from a nominal impedance to another, and then back, This has the effect of increasing overall circuit Q of selectivity. Cascaded, L-networks can have the opposite effect of decreasing selectivity, an extremely powerful tool when building circuits 10 function over wide bandwidth The Transmission as a Transformer Transmission tines have well known impedance transforming properties. A ter- mination of value R, is transformed to a new value, Ry by atcansmission line that is a quarter of a wavel 8 characteristic impedance Z,, given by Fg 3 1, forexample, we wished to transform 4 10-0 toad to appear as 50 Q at 7 MHz, ‘we would use a line with a characteristic impedance of 22.4 2. The length would be AUS ut 7 MHz, about 25 ft in cable with a velocity factor of about 0.7. This charac- teristic impedance isimpractical, butcould ‘be approximated with parallel sections of higher impedance lines. (Line with Zy=25 Q ean be purchased.) Transmis- sian line transformers ure sometimes prac- Seal at this low irequency. especially in antenna systems where the lines are needed anyway. Coaxial transmission lines can be coiled with virtually no impact on their behavior so faras the fields within the Tine. The quarter wavelengtt lines are often called “Q-Sections.” A transmission line need not have a iv 10 serve as a transformer. A Smith Chart is often used forthe design of these elements ‘Transmission lines become more prac sical circuit elements at higher frequen- cies. One printed line form is microstrip, shown in Fig 3.68, The lower conductor is, 2 ground plane on the back of a circuit board while the upper conductor is a printed run. Electric field lines between the conductors are found in the dielectric as well as in air. Hence, these transmission Tines have a velocity factor part way between that of air and that of the higher dielectric constant insulator. Microstrip is versatile, for it can be designed for about any characteristic impedance in the 10 to 100-2 vegion, or more. The wider lines have lower Z, Robert Wilson, KL7ISA and Hal Silver ‘man, W3HWC, in “Wire Line—A New and Easy Method of Microwave Circuit Construction,” described a wonderful Fig 3.68—Microstrip transmission line shown in cross section. The dielectric material Is the ingulated portion of printed circuit board. The lower Conductor is usually a solid ground plane. The drawing is not to scale, variation that the experimenter can build without etching in the July 1981 QST.28 Another practical transmission line form is a simple twisted pair of insulated wires. Wire insulated with plastic often produces lines witha characterisicimped- ance around 100 ©, Bnameled #24 wire will produce line with an impedance neat 50 £2 when tightly twisted. A variation on the quarter-wave line ‘matching uses synthetic transmission Tnes. Here, a transmission line is replaced by & picnetwork using inductors and capacitors [A sidebar earlier in this chapter discussed the half-wave filter, a variation of this c= cuit. Fig 3.69 shows « synthetic quarter- ‘wave example, the same case considered earlier at? MHz. Transforming from 10 0 502 occurs with a 22.4.0 line Powdered Iron Toroid Inductors and Transformers TInduetors ate realized with many struc tures, ranging from straight wire picces 1 solenoid sind toroid coils. The solenoid is easy to wind and can exhibit high Q, expe~ cially at VHF, However, the magnetic field of a solenoid extends well outside the ceil ~{ et we “or i Fig 3.69—A synthetic quarter wavelength line is formed at 7 MHz with thre reactance values of Z, of a Q section. Filters and qual Impedance Matching Circuits 3.31dimensions. leaving it free to couple to other circuit elements in close proximity. including conductive walls that can alter Q.Incontrast, the toroid inductor has most (but not quite all) ofits magnetic field con- fined to the core interior, allowing « toroid to be mounted directly against a ground ane with minimal change — in uluetance or Q. The Q available for low volume coils is generally much higher for toroids up through 30 MBz, ‘oroids are more difficult than sole- noids to wind, creating apprehension among beginning experimenters, Tt is, however, straight forward, even if time consuming Toroid inductance is almostexactly pro- portional to the square of the number of L=Ken Rg 3.42 A commen core is the T30-6 from Micrometals with inévetance coastant, K. (of 3.6 uH/t2(mano-benry perturn squared.) ‘Various manufacturers use other units tnat can berelated directly to the K we find con- venient for RF parts. A coil with 15 turns evenly wound around most of this core has 1 predicted inductance of 810 nH, 0° 0.81 UH. Generally, the highest Q will result ‘when the cores use the largest wire that ‘will fit in one layer. [tis important for Q. and especially for temperature stabili that the wize be tightly wound against the core, A more temperature-stable coil can often be built with a Wire size smaller than that producing the highest Q. Micrometals, Ine copyrights the usual toroid numbering seheie. illustrated here with T30-6. The ~6 indicates « specific core material or “mix,” while the 30 indi- cates un outside diameter of 0.30 inch. A manufucturer or vendor catalog might list the inductance constant for the T30-6 as 36 LH per 100 turns. The user ean convert these constants to whatever Form he or she prefers. toroid is wound by counting the mum ber of passes through the center hole, ‘While solenoids can havea fractional num ber of tums, this does nor happen with tor= colds. A single tum on a toroid consists of the wire passing through the hole just one ‘We built the inductor montioned by winding 15 1urns of #28 wire over about 90% of a T3046 core. Using an Almost Al Digital Electronics L/C Meter TB, the inductance was measured as 872 nH, 85% above the prediction. Partof the difference was probably the result of slight bunehing ‘of some of the turms. The permeability ol erance normally associated with these 32 Chapter 3 3 volts Fig 8.70—Circult llustrating the transfer characteristics of an ideal transformer. uy L Fig 8.71—Methed for connecting fings that allows cou ing coefficient to be calculated. This method is general and can be applied with powdered iron of ferrite ‘core transformers. The results bocome less accurate when coupling is strong, and itis not unusual to calculate fot. This is usually an cores is +/-5%, The accuracy is usually betier as induetance and care size grow: The windings were then compressed to cover only 60% of the core, increasing inductance to 1.039 lH. This 15 10 20% inercasc is typical and offers a convenient means for adjustment ‘This inductor can be used directly in impedance matching networks, oF as part of a L/C filter. The reader should consult the extensive data available from Amidon Ine. Thisis found atan excellent Weh site, ‘www.amidon-inductive.com/. A common impedance matching net ‘work uses a powered iron indueror with a second winding. forming a transformer. The inductor we just described was modi- ied by adding a $ turn link of #26 wire on the eemaiing bare portion of the core, The measured inductance was 206nH. This is much higher than the 90 nH the formula would predict, bt the coil is severely com- pressed, (Even with the Sturns spread over the complete core, L=121 nH.) The 15 tum winding L was unchanged at 1039 nH, ‘Weexpect RF voltage oinerease in pro ication of capacitance. portion (othe 1ums ratio and impedance to transform with the square ofthe turas ratio in:an ideal transformer, Hence, a50-C2gen- erator attached to the S-turn link should provide three times the voltage across the 15-tum winding with the combination looking like «450-2 souree to the follow- ing circuitry, as shown in Fig 3.70. If it was terminated ina 450-Q load. the im- pedance match looking into the Tink should be perfect. This uansformer mightbe used to match between a $0-0 amplifier and a 450-0, 10-MHz erystal filter But, these ideals are not realized. First, the impedances are highly reactive. This is remedied by taning the secondary with a parallel eapacitor, 244 pF at 10 MHz, This brings the voltage gain nearly up to the predicted 3 when the output is terminated, but impedance match is still poor. This is 11 result of less than ideal coupling ‘The coupling coetficient is easily mea- sured with the same instruments used 10 measure inductance. This is shown in Fig 3.71. L, and Ly are the S and 15 turn ‘windings and are measured with the otherwinding open circuited. The two windings are then connected as shown in Fig 3.71 and the composite inductance values are ‘measured as L, and I. The coupling coef ficient is then given: bt) Bq 3.43 4eqby Ly This method was presented by Bill Carver, WIAAZ. in the January, 1998 issue of the ORP Quarterly.2* When the method! was applied to the test transformer, we measured 1.51533 nH and 2 nH, leading to a coupling coeffi- cient of k=0.357. The input VSWR exceeds 2:1 for this ronsformer, even when tuned ‘and properly terminated. Ideally, all inductors should be mea sured after they are wound. While the traditional tuned transformer is still a practical component, it may require more design effort than an impedance transform= ing network built from discrete elements, The Ferrite Transformer The powered iron core transformer dis- cussed above had to be resonated to func tion as desired. Even after tuning, ic suffered for a lack of coupling, Both prob= Jems re overcome with higher inductance, which occurs with the much higher perme ability found in ferice cores. The toroid is the most common form, but balun cores, with their binocular shape, are also popu= lar. Most of the powered iron cores We use have initial permeability under 10 while typical ferrites show yi values between 40 and 5000. Recall he classic inductor, component that “tries”to maintain whatever current is flowing at any instant. Ii the dual of the capacita, which does not allow voltage 10 change instantly. Consider a switch that connects a battery to an inductor. The in- auctor current is zero before the switch closed, so it must be zero immediately afterward. There is no restriction on the voltage, The voliage impressed on L changes quickly, soon reaching the battery value. The current conserving charactetis- Tic ofthe inductor isa result af the magnetic field, When the switch is closed, current begins to flow. But as soonasthe feld starts to build up, the changing magnetic tield generates an electric field (hence, « vol tage) that opposes the electric effect that caused the current inthe first place. This is a non-rigorous statement of Faraday’s Las, one of Maxwell's equations. The induetor is shown with curves ilustating the behavior in Fig 3.72 Inductor current inereases without ‘pound in the ideal, lossless case. Losses, resistance within the wire and the battery ‘would limitihe current toa finite. but large level in a practical circuit. Consider now a modified structure, The single winding inductor is replaced with 3 pairof windings, shown in Fig3.73, thacare very close together, The wites, although iso- lated from each other, occupy virally the same space and sce essentially the same magnetic field. we let the second winding (BB) open circuited, voltage from A to A” ‘builds upin the same way thatitdid with the simple inductor, Measurement across eter winding will show the same voltage profile But, no current flows in BB" when itis open vireuited ‘The behavior changes when we repeat the experiment with # Toad at BB" As the voltage builds, load current will begin to flow. Transformer action begins. The cur: rent in the second winding will generate a ‘magnetic field, just as that in the primary winding did. Bue the field from the sec: ondary is in a direction opposite to that from the first winding, Because the net ‘magnetic field has been reduced (nearly) to zero, current flow is determined by R, the extemal load. ‘The transformer deseribed (Fig 3.72) with the two wires in close proximity, is said to be bifilar. Bifilar windings are often twisted. One manufacturer supplies Moltifilar® wire with strands of differing colors, simplifying transformer eonstrue~ tion. (Multifflar® parallel banded magnet wire from MWS Wire Industries.) The dots on the ansformer schematic tre wseful. An increasing voltage at one dot produces an increasing voltage at the other. Current entering the A dot equals that leaving the B dot. This bebavior arises because the magnetic Hild vanishes within the core. Ifthe primary (AA") had Ny turns while the secondary (BB") had Ng turns the currents would obey the more general boundary condition that Nps Bq 3.4 Bifilar winding and the use of a bigh permeability magnetic material produce tight coupling. approaching £=1. Coupling is measured for a ferrite transfesmer with the same method outlined for a powdered iron design, Fig 3.71, Strong coupling ‘means that all of the magnetic field lines created by the primary also couple into the secondary. In a practical transformer, some of the primary field loops out from the core, only to return without communi cating with the secondary ‘The transformer is often modeled as an ideal one with udded components, shown in Fig 3.74, The ideal transtormer has & voltage “ratio propertional to the turns ratio and a cusrent ratio defined by Eq 3.44. Ly isthe primary inductance, the value we would measure ifthe primary was examined without 2 secondary termina T nattery, vit) ro) Va Wty if R=0 Fig 3.72—Principles of an ideal inductor, with waveforms. The current would grow linearly forever In an Ideal ‘component. Resistance astablishes an Uhimate value. i(t) with finite R| fs ‘ume Filters and Impedance Matching Circuits 3.33atery Class Billa Tarstormer Fig 8.73—Current flow in a bifilar wound transformer. RF transformers can be built by placing forrito beads over brass tubing that forms a single turn winding. Circult board material connects tho tubing ‘ends with a short at one enc. A multiple ‘wire winding is then threaded through the middle of the tubing, guarantesing tight coupling, tion, The L-leakage is the inductance accounting for the magnetic flux that does not pass through both windings. R} and R2 aecount far losses. The transformer is aa bandpass circuit with Lp presentin shortat de and very low frequencysL-leak= age, a series element, presents a high impedance at high frequency. A practical transformer will have a pri mary inductance witha reactance at least $ times the terminating resistance at the low frequency limit and a leakage inductance reactance less than 1/5 che resistance a the 3.34 Chapter 3 a oy Eel Ne @ Yse@ pik WY 4 j be ideal Fig 8.74—A transformer model highest frequency, and loss resistances small with respect tothe source and load Tnductance of windings on fersite cores is proportional to he square of the turns, although the higher permeability offerte produces dramatically higher “k” eon- Sants for use with Eq 3.42, For example. the popolar FT37-43 ferste toroid has k of bout 360/nFft?, Core loss can be modeled 8 a parallel resistance, whieh is also pro- portional tothe square of n although this formulation isnot in general use. Examples of practical transformers are found throughout the text, A wonderful treatment of the modeling of this “simple component is presented by Clarke and Hess Amore complete review of tans former modcling is presented by Chris Trask.2° We generally use powdered iron toroid cores for high-Q inductors with good temperature characteristics while ferrites are relegated to low-Q wideband transformer, application. However, this uistinetion 1s mot required. Some pow dered iron cores are suitable for wideband transformers while some ferrites have ex- cellent Q at HE. A good example of the later is -63 material irom Faie-Rite Prod ucts Corp ww fai-rite.com), often r0= dlucing Q values of several hundred at HE. Ferrite Transmission Line Transformers ‘The example presented above 10 illus trate basic transformer action used a biti lar winding. sith one wire as primary and the other a8 a secondary. A pair of wires also forms a transmission Tine. Ax such, i cam operate as transmission Line trans- former such as « Q-section aecording to Eq 3.41, Even if it is not a proper A/a length, it will still wapsform the ianpedanee seen at one end from that pre= sented at the other. The transmission ling properties persist if the line is wound in the shape ofa coil, including a toroid, But the structure then assumes a different extended behavior. summarized in a clas sic paper by Ruthroff. “The sienplest ferrite transmission line transformer is that shown in Fig 3.75. This Steuctore, formed with a bifilar ninding on toroid was atone time called a balua. A bolum is a structure that gencrates a bal- anced voltage from one that is single ended, This connection dos not force such balance ard is, henee, not strictly & helun, even thouh it docs perferm some Of the isolation chores that we might ask fof ahalun, Pochaps a better name is isola. sion transformer, ‘Transformer action, described above. does force equal currents in thetwo windings. so his ircuitissomo- times also called a curren: hatur. ‘The isolation transformer is labeled AB at one end of the winding while the other endis A'S’. Wires A and are not attached to each other, a useful detail to keep in mind when winding such transformers without wires of differing color. Viewing this structure as a transmission line, cur- Fig 3.75—Part A: Basic ‘toroid. This structure he load connected to a single-ended drive while © shows polarity inversion.rent at point A'is delayed from that at A However, toe ferrite core and traditional transformer behavior would force equal current through a winding, and indeed, in the other winding. R= ” Fig 8.76—A 4:1 step-up balun transtormer. stop down transformer. The isolation transformer of Fig 3.75 hasa single ended input. The single ended drive will appear ay a balanced output on ‘balanced load such as that in part B. In this sense, it isa balun structure. However. if the load becomes unbalanced, as in Fis 3,75C, the input may still b applied tothe ‘ermination, It is instructive to mentally connect the two wires at one end (A and B) together. doing the same thing atthe other (A' and BY end. The result is an inductor. Several turns on a high permeability ferrite would produce considerable inductance. This is termed a common mode inductance, Sepa rating the wires, a load placed across fone end, A'B', is then seen differentially (between A and B) at the other end, This structure is often ealled a common mode choke for common mode signals atone end are isolated from the other by the large inductance, while differential signals are not impeded ‘The isolation properties ofthis structure allow us to drive one end while seating the other end as if it were a separate gen: erator, An isolation transformer (Fi 3,75C) can produce a polarity reversal Tis useful to connect the ourput of an isolation transformer in series or parallel with the input. An interesting example is shown in Fig 3.76 where 2 load is con- nected between the input ancl the inverted output. The composite input will carry Fig 3.76—A 1:1 impedance ratio true balun transtormer. (wo Coes) Re a Fig 8.79—Illustration of a 9:1 Unbalanced transformer, n (wor) 4 alana talons Tanner Fig 3.80—A 4:1 balancod-to-balanced transformer. tovice the current that one transformer ‘winding carries, resulting in a true balun, for it forees equal, but out of phase volt ages to appear between the ends. This is & 4:1 impedance transforming balun, The same structure is reapplied in Fig 377. The transformer forces twice the ceurrent to flow in the output asat the input ‘The isolation properties of the transmis: sion line transformer are used to parallel «an output with a “direct connection” to the input, This cigcuit now serves an unbal- aneed-to-unnbalanced role. This cireuit is used for transforming from 50 Q clown to the 12.5-0 input on a RF power amplifier. We also saw il used extensively to cause & 512 load to Took ike 200 Sat the eollee- tor of a feedback amplifier. ‘These wideband transformers may be viewed as either ransmission Line circuits for as conventional transformers, Their ‘operation is consistent with either set of boundary conditions. The transformers are designed with about M8 10 A/4 of trans mission line atthe upper frequeney of the citeuit, The charactcristic impedance of the line is consistent with line behavior for the (erminations considered. If, fo example, we built 4:1 step down from 50 to 12 Qusing Fig 3.76, Z, should be 25 2. ‘This could be realized bY paralleling two 50-0. windings on the core, A 50-2 wind- ing consists of a tightly pwisted pair of #24 ‘enamel wires, The transformer of Big 3.78 isa trac I:1 balun, The termination impedance is that seen at the input, but the circuit creates ‘ovo voltages that are equal in magnitude, bout out of phase, ‘A useful step down cireuit for high power single ended amplifiers is the 9:1 Circuit of Fig 3.79, This transformer uses ‘wo cores to drop from 50.2. down toabout 6 Q. Series connections at the input side drive parallel ones atthe output. A similar series/parallal circuit is presented in Fig 3.80 where two cores form a balanced to balanced 1-4 impedance ratio step up transformer Numerous other kinds of transmission line transformer can he built, some almost iabolic in their cleverness. The reader ig refened to Motorola Applications Note AN-593% for further interesting ‘examples. Some Multiple Port Networks Allorthe networks presented in this see- tion have used but two ports, an input and fan output. These ate, however, several multiport networks thar are of special interest to the radie amateur. The first is the so called “Spliter/Combinet™ shown Filters and Impedance Matching Circuits 3.350 1 250mm | ne $ Input | ia a Fig 3.81—An in-phase splilter/combiner network, Use 10 bifilar turns on a FT-37- 43 fertile toroid for the HF spectrum. Fig 3.92—First-order low-pass/high- pass diplexer. in Fig 3.81. This circuit, using nothing more than « bifilar winding on a ferrite toroid, accepts energy front a single gen- erator with a 25-0 characteristic imped- ance and supplies that energy co (wa ‘outputs, each with @ $0-Q impedance. A. 50-2 input can be transformed down 10 25 Q with any of the matching schemes presented above. Variations of this net work use transmission Lines or L-Net- works. The 100-2 resistor absorbs excess ‘power that becomes available when one of the two output ports is miss-terminated. common application splits the output of & local oseillator chain to drive two mixers, ‘The cireuit isolates the Uso ousputs. This circuits called a3-dB hybrid transformer, for the power in each output, neglecting losses, is 3 UB below the inpul, while Hybrid refers to wansformer-lke circuits that provide isolation berween two of three ports. Hybrids were used in early tele- pphoncs to isolate the microphone from the eaephione, 3.82 shows atheee porteireult where 3.36 Chapter 3 ou, Pon’? Inout se Pot O degrees + out Y Fig 9.83—Phase shift notwork for RF phasing in simple $SB equipment. tovo outputs receive drive from a single input. This cireuit, a diplexer. is similar 10 a crossover network used in audio sys tems. Frequencies below 2 cutoff pass through the inductor and are dissipated in the elated termination, Signals above cute ff pass through the capacitor to the related resistor. The L and C are picked with regard (o the source impedance such that there is always a perfect impedance match presented to the generator, Ifthe cutoff frequency is f, then the related angular frequency is @.=2nf, Then, the L and C fora perfect match are Eq 345 ee oR ‘The diplexer is applied where mixers (e.g, diode rings) must be terminated in a wideband $0 £2 to minimize distortion. The iplexer shown is an especially simple ‘one where each arm isa one pole low pass corhigh pass filter. Nic Hamilton, G4TXG, thas described high order low pass high pass diplexers 29 A third-order example of this design is shown in the diplexer sidebar. Diplexers can also be built with combinations of band-pass and bandstop networks, also summarized in the sidebar, An interesting. yet simple phase shift network is shown in Fig 3.83. A gencrator ives two one pole filters that are termi hated at their output in open ciscuits. The Iwo capacitors, egual in value, are picked to have a reactance at one frequency equal 0 R, the resistor value used in each arm, The phase difference for this network is, 90 degrees at all frequencies, However the ‘wo output amplitudes are equal only at the design frequency ‘An especially interesting four-port cir- ccuit form is the directional coupler. The coupler fas an input and output, usually ‘with Tow loss between them. A third is called the “forward” coupled port, for the energy available is proportional to the nergy flowing from the “input” (© the ‘output.” A fourth is the “reflected” coupled port with energy proportional to that flowing from the “output” te the “input.” Fig 3.85 shows a schematic rep- resentation of adirectional coupler, which is alse 1 practical topology in mictostrip form. Par B of Fig 3.85 shows a wideband variation using fecrite wansfermers.°° A practical version of the wideband coupler Using three transformers was designed by Roy Lewallen’! and is included on the book CD. The directional coupler is extremely useful fore variety of applications, When used with a power meter or speettum ana- Iyzer, reflected energy is a measure of the Impedance at the output port, leading 1 popular in-Tine power meters such as the WTEL design. But the coupler can also be used to inject signals on a line. The cou- pling valuc is the power ratio hetween the ‘output and the coupled ports and is JN? for the ferrite version. Most directional couplers have coupled enerry that is in phase with the output. The microwave lit- ‘erature abounds with interesting couplers. ‘A coupler is also characterized by slirectivity. Assume thatthe vary path ister ‘inated in an open (or short) cirevit and a power PI is measured in the seflected port I the main path is now loaded with a per fect matet, the reflected power will drap to 2. The ratio of PI te P2is called the direc- tivity. We consider direetivity with a num- ber of bridge circuits in Chapter 7 Directional couplers can be built with lumped components, even al VHF. A Jumped element example with -28 dB cou- pling with 20-48 directivity at M44 ME is included in a design discussed later in the ‘book and included on the book CD. That design is a quadrature coupler, dis- cussed below.*? There ure numerous refer ences inthe literature to directional couplers. See, for example, Andre Boulouard. The owvisted-wire quadrature hybrid directional couplet isavery useful variation This circuit was described by Reed Fisher, W2CQH.M Fisher's OST anicle is 1 cluded on the book CD-ROM. Also see.*°7 For information on distributed couplers, see This is a 3-dB coupler for the ‘coupled output is below the inpur by 3 dB, producing two outputsof oqual strength. The Circuits called aquadrature couplerbecause there is » 9O-degiee phase difference between the two output pons. A HE varise tion. buile for the 7-MHz band. is shown ia Fig 3.84, The design equations forthe coupler are identical to those presented for the Uiplexer. Bq 3.45. However, in this cuse, the capacitance is the total C inthe circuit.45 degree Input out, @ Patt 20 pF Port 2 O decrees SS degrees: # Y | ue out, Port 4 zaipe a “T wnws | A mid : : be oa Re ay This must be halved to build he circuit. As Fisher points out. the capacitance af the tightly wound bifilar pair (12 pF im his ‘exumple) is measured and removed from he calculated C before construction. The inductance is that of the two windings in parallel, essentially the same as that of a single winding on the core of interest Fisher used low-permeability ferrite core. while we have generally used pow- ered iron cores, owing primarily (0 avail- ability, Small povider irom cores such as he T25 in the -6, -12, or -17 materials are suitable through 150 MY tthe design frequoney, the circuit is a 3-4B coupler, providing equal power at port 2 and 4. However, the coupling is Giffereut at ether Gequencies. The very rnveresting properties of the quadrature aybrid are summarize: 1. There is power transfer fzom port Lc 2. Fig 3.85—Part A shows a general schematic for a directional coupler while B presents a wideband version using ferrite core transformers. The coupling on 8 1s 20 dB owing to the 10:1 turns ratio used. This Is a practical circuit if wound with FT37-43 or FT37-75 cores. A single binocular core can be used for both transformers. Fig 3.86—Some applications for quadrature hybrids. Identical ‘amplifiers (A) oF fillers (B) are combined to form tormination insensitive linear circuits, The extra terminations required ‘are shown In the circuits. jpper freq with 1 dB amplitude balance, fine, Linu, Cin pF Fig 3.87Extonded bandwidth quadrature hybrid network Filters and Impedance Matching Circuits 9.37at the cutof frequency. [roger unposnn gy - t a eal, 4 i, Land C values shawn are reactance @ (from 1 to 10) 2 @a2eE 50-0. Bendpass-Bandstop Diplexer 7 se : 50 ? 50 . i 1. Pick cutoff frequency F and REFERENCES LW. Hayward, Jnsraduetion to Radio Frequency Design, Prentice-Hall, 1982; ARRL, 1984. Also see The ARRE Handbook, 1995 or later editions. 2. GPLA accompanies Introduction to Radio Frequency Design (see Rel. 1) 38 a DOS program. GPLA 2002 isu Windows version included on the book CD. ARRL Radio Designer was formerly available from ARRL 3. W. Hayward, Ham Radio Mage Jun, 1984, p. 96 4, D. Johnson and J. Johnson, “Low Pass 3.88 Chapter 3 Filters Using Ultrasperical Polynomials,” IEEE Transactions on Circuit Theory, Vol CT-13, No, 4, Dec, 1966, pp 364-369, 5. Tortorella, RF Design, MarfApr, 1983. 6. Zverev, Handbook of Filter Synthesis, Wiley, 1967 7. M.— Dishal, “Alignment __and Adjustment of Synehronously Tuned Multiple-Resonant-Cireuit Filters,” Elect. Commun. un, 1952. pp 154-164 8. $B. Cohn, “Dissipation Loss. in Multiple-Coupled-Resonant Filters.” Proc. IRE, Aug. 1959, pp 1342-1348, 2. Power is transferred from port T 0 4. 3. There is no power transfer from port | o 3 when all ports are properly termi- nated. 4. There is no reflected power back out ‘of port 1, again with proper termina- tions. 5. The phase difference hetween ports 2 and 4 js 90 degrees, ‘The characteristic of greatest interest will depend upon the application. The phase difference is important in the con- struction of pbasing-metiod SSB equip: ment. However, it is the isolation from reflection problems, item 4, that leads 10 some of the move subtle applications. Two ‘examples, each using a puir of couplers, are shown in Fig 3.86. In part A, vo amplifiers are combined, while in B, we filters are combined. In both eases, thecwo elements must be identical. However. the networks to be combined necd not be impedance matched for a good match 19 cexist at the input, For example. the two amplifiers could be FET eizcuits that have ‘an L network at the input. Such a citeuit produces a very poor input impedance match, but an excellent noise figure Alternatively, Wo conditionally-stable amplifiers can become an unconditionally stable circuit when imbedded in quadra- ture hybrids, This balanced scheme is attributed to. Engelbrecht and Korokawa.*041 4 termination invensi- tive crystal filter ig desevibed! in Chapter 6 where quadrature couplers are applied ‘The circuit of Fig 3.86 is narrow band- width with identical output amplitudes at only one frequency. However, the band- width can be extended to an oetave by cas caging two identical quadrature hybrids with a pair of pi-networks between. This topology, with telated design equations, is shown in Fig 3.87. 95. G.Matthaci, L, Young, E.M.T. Jones, Microwave Filters, impedance Marching Neiworks and Coupling Structures, McGraw-Hill, 1964, 10, See Reference 6, ILA. B. Williams, Elecironie Filter Design Handbook, McGraw-Hill, 1981 12, W, Hayward, Introduction 10 Radio Frequency Design, ARRL, 1994, Ch 3. 13.W. Hayward, “The Double-Tuned Cireuit: An Experimenters Tutorial.” QST, Dec, 1991, pp 29-34 14, R.Larkin, "The DSP-10: Au All-ModsMeter Transceiver Using a DSP IF and PC-Controlled Front Panel, Part 1," OST Sep. 1999. pp 33-41. 15. V. Bowom, Introduction to Quartz Crystal Unit “Design, Van Nostrand Reinhold, 1982, 16.8. B. Cohn, “Dissipation Loss in Multiple Coupled Resonators”, Proc IRE, Aug, 1959. 17.W. Hayward, “Designing and Building Simple Crystal Filters", QS. Jul, 1987, pp 24-29. 18, Carver, K6OLG. “High-Performance Crystal Filter Design.” Commminications Quarierty, Winter, 1993. 19, D. B. Johnson, J. R. Johnson, and H. P. Moore, A Handbook af Active Filters Prentice-Hall, 1980. 20.H. Berlin, “The State-Variable Filter.” QS7. Apr. 1978, pp 14-16, 21. W. Hayward, Introduction 10 Radio Frequency Design. ARRL, 1993, Ch 4 22. G.L, Matthaei, “Tables of Chebyshev Impedance-Transforming Networks of Low-pass Filter Form.” Proc IEEE, Aug. 1964, pp 939-961 23.R. Wilson and H. Silverman, “Wire Line = A New and Easy Method of Microwave Ciscuit Construction.” QST, Jul, 1981, pp 21-23. 24, W. Carver, “Measuring Capacitors and Inductors.” ORP Quarterly, Jan, 1998, p 37. 25, Clarke and Hess, Communications Cireuits: Analysis and Design, Addison- Wesley. 1971 26. C. Trask, “Wideband Transformers: ‘An Intuitive Approach to Models, Characterization and Design,” Applied Microwave and Wireless, Nov, 2001 27.Ruthroff, “Some Broad-Band “Transformers”, Proc. IRE, Aug, 1959. 28.N. Dye and H. Granberg, Radio Frequency Transistors: Principles and Practical Applications, Butterworth Heinemann, 1993, Ch 10. 29. Hamilton, “Improved Direct Conversion Receiver Design”, Radio Communications, Apt, 1991, Appendix, 30. W. Hayward, Invroduction to Radio Frequency Design, ARRL, 1984, Ch 4, 31. R. Lewallen, “A Simple and Accurate ‘QKP Directional Wattmeter.” OST, Feb, 1990, pp 19-23, 36 R. Larkin, “An §-Watt, 2-Meter “Brickette’.” OST, Jun, 2000, pp 43-47 33. A. Bouloward, “Lumped-Element Quadrature Couplers.” RF Design. Jul. 1989. 34. R. Fisher, ‘Broudband Twisted-Wire Filters and Impedance Matching Circuits (Quadrature Hybrids." /EBE Transactions on Microwave Theory and Techniques No. 5, May, 1973, pp 355 35, R. Fisher, “Twisted-Wire Quadronure Hybrid Directional Couplers,” O57. Jan, 1978. pp 2 36.1. D, Cappucci_ and H. Seidel, US Patent 3,452,300, Four Port Divecitve Coupler Having Electrical Ssmmetry with respect to Both Axes, issued Jun 24, 1969. 37.1. D. Cappucel and H. Seidel, US Patent 3.452.301, Lumped Parameter Directional Coupler, issued Jun 24, 1969. 38.B, M, Oliver, “Directive Eleetro- Magnetic Couplers,” Proc. IRE, Oct, 1954, 439. $. B, Colin, “Shielded Coupled Strip ‘Transmission Line,” MIT, Oct, 1955. 40.K. Kurokawa, “Design Theory of Balanced Transistor Amphfiers,” Rell System Techaical Journal, Vol. 44, No. 10, Oct, 1965, pp 1675-1698. 41. RS. Engelbrecht and K. Kurokawa, “A Wideband, Low Noise. L-band Balanced Transistor Amplifier.” Proc. IEEE, Vol $3, Mar, 1963, pp 231-246. 42.R. S. Engelbrecht, US Patent 3,371,284, High Frequency Balanced Amplifier, Feb 27, 1968, 3.39CHAPTER Oscillators and Frequency Almost all of he Amatcur Radio equip ‘ment we build will contain at least one oscillator. It may be a simple erysial con- trolled circuit, & tuned LC variable fre- quency oscillator, or even a direct-digital synthesizer, a circuit that provides an out- put similar to what we might expect from a simpler cirenit. A basic oscillator might be a simple one tuned by a mechanical variable capucitor. Alternatively, it might be voltage contealled. Combinations ofall of these are possible and are commer modern communications equipment. The local oscillator (LO) is a critiesl part of any communications system. Mod- ern transceiver performance is often com: Synthesis promised by LO systems that suffer from excess phase noise, effectively limiting the receiver dynamic range. While quter os- cillators, those with low phase noise, can be built using traditional methods, these circuits often lack the thermal stability of a synthesizer. Beyond their practical importance, o5- cillators are extremely interesting circuits An effective oscillator ean be built with single transistor, Yet, this simple, prim tiveciteuit will include both positive feed back, causing oscillation to stact at the desited frequency. and negative feedback that maintains operating amplitude con stant with time, 4.1 LC-OSCILLATOR BASICS Oscillators may be classified in a num- ber of ways. One categorizes the eitcuit by the devices used for the active element and the resonator, such asthe bipolar transis tor, crystal cowtrolted oseillainr and the IFET LC oscillator. One can also classify oscillators according to a historie ciceuit form, such as the Colpitts or Hartley. An oscillator can be classified by the active device configuration, such as common emitier. Finally, it can be classified ac: cording to the method used during design, such as w negarive resistance oscillator. ‘The first question we ask (or should ask) is if anoscillator will indeed oseillate when power is applied Fig 4.1 shows a block diagram of an oscillator, The circuit is segmented into two elements: a resonator or tuned circuit, and an amplifier. The tuned eireuit output is applied to the amplifier input, But, the amplifies output is routed back tothe input of the tuned cireuit Assume thatthe circuit hus a power sup- ply attached, but through soiie means or another the resonator is short-circuited A frequency synthesizer offers outstand- ing thermal stability and frequency aceu- racy. A synthesizer using a liandful of inte~ grated eiteuits, each containing hundreds Df transistors, is less expensive to manu- facture than a high quulity mechanically tuned LO system, Itis more eliable, owing to u reduced number of moving parts, Fre- queney synthesis is not, however, the an ‘wer toall ofthe LO problems presented 1 the experimenter. Some PLL synthesizers are burdened by excessive phase noise ‘Those using DDS, while quieter emit spu- sious outputs, often in profusion. Both use aan excess of digital eizcustey that ean often corrupt a receiver environment, with a switch or otherwise altered so that the citCuit is not oscillating. The switch is then opened, restoring resonator function: ality, The amplifier is operational with nnorinal operating bias applied: hence, it ‘generates noise, ‘The noise present at the Fig 4.1—Block diagram of an oscillator. Part A shows the basie osolllator while part B illustrat Oscillators and Frequency Synthesizers the method used for analysi tlther LC or crystal oscillators, or even circuits using RC or. Amplifier input and output is labeled with “i” and “0.” This analysis can be applied to 44input is amplified to appear at the output with preateramplitude, This noise is spread moze or less evenly over a wide bandwidth. ‘The amplifier output is applied to the tuned cireuit where iti filtered and phase shifted. ‘The resulting signal emerges where it is ‘again appliod to the amplifier input. For each frequency, the signal that has tra- versed the amplifier-resonator loop emerges with a new amplitude and ew phase. Ifthe amplifier has a net gain at the Fesonator center frequency, the signal st that frequoncy is lurger alter having tra versed around the circuit. I will continue to grow with each round tip. There will be one unique frequency where there is no net phase shitt as energy at that frequency traverses the loop. TBs ‘eventually establishes the oscillator oper- ating frequency. Energy at frequencies ove und below the center cariier ire ‘queney will be shifted further in phase with each tip around the Toop, eventually merging 90 degrees away where it no Tonger contributes to the power. We have just deserihed oscillator start ‘urg, Oscillation will begin if Ue signal grows in amplitude with each pass around the loop and ifthe phase is the same as it was in the beginning. These are the so-called Bark hhausea criterion, They are measured or ana lyzed withthe systeminthe Figure. The loop thas been broken at °X" in part “a” of the Figure. A signal source und load re inserted that allow the gain to be measured, shown in part “b."! ‘The amplitude cannot continue to grow without bound. Something must occur ‘within the circuit that will reduce the over= all gain tothe level just needed to maintain 4 stable amplitude. This usually oecurs through current or voltage limiting, with current limiting generally preferted. (Au- tomatic gain control can also be used.) Biasing details usually establish limiting and set oscllstor operating level. A high operating level is generally desired We rarely analyze starting in an HE os cillator we wish 1 built for a project. Rather, we merely build and examine the pseillator to see if there is an output. The Colpitts and Hartley Circuits While thete ure numecous named LC cscilltors, they con generally be eatezo tized as Colptts or a Harley variations ‘vith both eifeuits named for thir inven- tors, early radio pionsets from the Bell abst the 19206 and 1930sera, The basic forms are showin Fig 42, A and B. The only difference between the wo is in the tneans for feedback, The Harley (B) ses {Tapped inductor while the Colptts (A) 4.2 Chapter 4 de tT a = EG =f O fg 42—Colpits (A) and Hatley (8) ona of the thr iminated from the last Iwo eirulte. Alfmough illustrated with FETs, bipolor 's. The versions at (C) and (D) have FET terminals to be grounded. The Fig 4.3—The Golptts (A) evolves Into the Clapp (B) and then the Seller (©). The variation on the Colpitis where the base is Vackar oscillator at (D) is yet anoth ae © a 1 Sai © (). The driven from a lower impedance, achieved with a capacitor tap across one of the usual “Colpitts Capacitors or bipolar transistors, uses capacitors The Hartley and the Colpits oscillators of Fig 4.2 A and B use « source follower amplifier. This distinction is an arbitrary fone, as is ilusteated with the wo varia tions of Fig 4.2 C and D, which are drawn without a ground. The ground and biasing can then be inserted as needed by the designer. The operation of the Hartley is often explained with transformer action. The source follower of Fig 4.28 has «high in- put and relatively lovs output impedance, and a voltage gain close to 1. The ammpli- fier output signal is applied to the cap on ‘he tuned circuit. Transformer action then Jnereases the voltage that appears at the gate. Breaking the loop at either the FET gate of source will show the required These oscillators can be designed with either FETS greater-than-unity, zero pase shift start= ‘The Colpitis cireuit (Fig 4.2) may not be as iawitive, Detailed circuit analysis will show that driving the capacitive tap ‘witha low impedunee source will produce the required voltage step up in the com= posite tuned circuit, Indeed, a similar analysis shows thatthe same action oceurs in the Hartley oscillator even if there is ne ‘magnetic coupling between the two indue- tor sections, Transformer action is not re~ quired! A Hartley is easily built with wo separate coils, an occasionally. useful ‘The Hartley oscillator with positive feedback resulting from inductors can bavean advantage over the Colpits:Ifitis tuned with a variable eupacitor with minioe SI wii ‘This Hartley Oscillator is mounted in stamped box. A vernir drive Is attached to the capacitor shaft and is fixed to the box with a single bolt that prevents rotation. Spade lugs allow a lid to b mal fixed capacitance, it will produce # wider tuning range than is easily realized with a Colpitts, There is no other fund- mental advantage of one over the other. ached to the box. ‘The Colpitts oscillator! lar variations shown in circuit (A)is the basie Colpitts, now shown, With a bipolar transistor, Part B shows the Clapp oscillator, alse called a series ined Colpitis. The Clapp starts with a Colpits the VCO output. This filtering charac- istic is not available to one building the nore conventional heterodyne system, Schematies are presented fora practical implementation of the system of Fig 442, design we used fora 10-year period, Two utpul frequency bands were available: 7 97.1 and 14 to 14.2 MHz. The 14-MHz utput Was alse frequency doubled to pro- ea 28-MHz signal, Tite basic cicuit is, 2 |4-MHz PLL. bur the output is digitally sivided to produce the 7-MEz componeut, ‘The 14-MBz VCO isshown in Fig 4.83, § 2N3006 PNP (QI) oseillator is tuned sith a MV209 abrupt junetion varactor siode, The grounded collector facilitates, diode biasing. The emitter current in the PNP guarantees an operating level that never forward biases the tuning diode. A ‘buffer increases the ousput to +2 dBm. There are oe large bypass capacitors ‘within the shielded VCO, for the +12-¥ supply is keyed. ‘The VCO output drives a passive power splitter where the (wo applications are isolated, shown in Fig did. One path routes L4-MHz energy te Q3 where it 3s amplified to a 2.5-V pk-pk level to serve as the LO for Q4, a dual gate MOSFET mixer. The 12.5-MHz signal is generated with QS. The level reaching the mixer is adjusted to prevent overdriving the mixer. ‘The mixer output is filtered in a 1.7-MH2 Jow pass filter ‘The other splitter outputs applied a Q6. 44 stage providing 14-MHz output, Some energy is “Stolen” athe emitier‘odrive Q7 and Ul. a D-flip-tlop operating as a divider, The resulting squate wave is fur ther butfered in Q8 and is Low pass filtered to produce a clean 7-MHz signal. The low pass is a peaked (ultra-spherieal) desig, offering greater than aormal harmonic attenuation. A band-switch selects the appropriate oulpul. Even though the 7= MHZ circuitry continues to operate when the T4-MH? band is in use, the 4D-meter output is stil 70 dB below the desired out put. The O-dBm output was used to drive a ‘ova stage, 1-W power amplifier. This was ow pass liltered and used on the air fur QRP activity, or applied to a FET power amplifier for more aggressive effors ‘The 1.5-MHz output from Fig 444 is applied lo the phase Irequeney detector shown in Fig 448. This then drives a loop filter using & LM3D1 op-2mp. The loop was designed for a 10-kHz loop bund- \idth. The reference VFO (not shown) for the phase detector was a JEET Hartley butfered with a MOSFET. ‘Keying and timing details, although not shown, are eritical in this system, The VCO was keyed with a “+12T" voltage that started as soon as the Key was pressed. 12,13: Su, 32t#26, 150-2 73,74: 12 bifilar 37-43 #23, 12: 20teze ¥T37-43, 4t Link. 74Ls74 0 dBm ourene T 14 Mie Fig 4.44—Mixer section for the tracking PLL. Oscillators and Frequency Synthesizers 4.23va paste Fig 4,45—Phasa-trequeney detector and loop filter for the tracking PLL. Programing Fy or ope a xy vide by | —* , Phase wool” Frequency a Detector ry) Px Ne Free wide = ‘Loop Filter Lape) Fig 4.46—A single loop Divide-by-N PLL. Careful listening and examination with an oscilloscope showed that pase lock was fast and always occurred before a signal ‘was applied to the antennal, A “hold-ofT" circuit was included that prevented the keying voltage from reaching the power amplifier mati the key had been down for 5 milliseconds. This was applied only on snitial key closures, VOX-like circuitry then maintained the system in teansiit mode (VCO on and TIR relay closed) for halfa second or so. This system would be ‘compromised if the VCO locking bad not been quick ‘A number of changes would be imple mented if this sysiem was rebuilt today The dual-gate mixer would be replaced ‘with a balanced virenit. The op-amp would hecome an up-to-date aleernative, such as the OPA-27. High speed CMOS would replice the LS-TTL used. Finally, the VCO would run continnously without eying, bur would operate at a different frequency. This could be 28 or 56 MHz 4.24 Chapter 4 where direct division would produce the desired outputs. Divide-by-N Phase Locked Synthesis ‘The most common scheme for frequency synthesis isch divide-by-N PLL Fig 446, A ceystal oscillator at Py. is divided by & (usually) fixed integer R, producing « reference signal atthe phase frequency de tector at Fy/R. The VCO is divided by a programmable integer, N. The divided VCO mustalso appear Fx/R, so Fy=NFx! R. Cousider an example: We wish to build a symlesizer for the 9 to 9.-MEz range, We divide a 2-MHz crystal oscillator by 00 to generate a 10-LH2 reference. N must be se 10 900 to produce a 9-MELzsig- ral. Tnereasing N eauses Fy t9 increase in 10-kHz steps, reaching 8.5 MHz with 30, This system would work well asa rans- ceiver local oscillator (LO) inn envizon Phase-trequency doteotor using LS-TTL logic. Thi cireutt is shown in Fig 4.45. ‘ment where signals were spaced at 10-kHz intervals. [t would not, as shown, be very ‘useful us a general purpose LO. Modifications could improve resolu tion, For example, increasing R to 20400 produces [-KHz reference. N ranging ‘rom 8000 to 9500 would then cover the desired range in I-KFTz steps. (A means of pulling the 2-MHz crystal oscillator by 2 ‘mere 222 Hz would then generate all LO frequencies within the desired range.) Generally, 100-Hz resolution produces understandable SSB while 10-Hz steps yield natural sounding voices. But div {15 of 90,000 oF 900,000 are impracties even though they are easily achieves witb digital Logie. Consider the L-kHz step system with N=9000 to 9500, The detector reference frequency is L-kHz, the step value. The loop filtering (plus balance effects) must produce considerable attenuation at Tktiz, Generally. a system with a 1-kHz step would use a loop bandwidth of 100 Hz of less, The de ftom the loop filter includes @ small 1-kHz component that frequency modulates the VCO carrier at 1 kHz, The spectrum is a carrier with I-kHz sidebands, These would be transmitted if the LO was part of a transmitter If part of a receiver, the contaminated LO would cause a strong signal to be reccived in a couple of extra Frequencies, albeit at re= duced strength Timing problems oveur when N is inereniented to tune such a synthesizer. While the N change is instantaneous, the esult is not, A filter with 100-Hz band. width is capable of change in a time com mensurate with /B where B is the loop ‘bandwidth, here 10 millisceonds. The ef- fect can he a “chirpy” sound with tuning, ‘There is yet another problem, a degra dation in phase noise. The PLL with 2 division-by-N is a frequency mutdiplicr Assume, for example, that 1-Hz changes the reference at the detector. With N=9000, thar 1-H shift becomes a 9-kH2,Sift in the VCO. If we think of the I-Hz erence shift as a noise, the result after Sequency multiplication by N is 8 noise ease by 20LOg(N) dB, 79 dB for this esse. Clearly, PLL synthesizers with large N should be avoided. PLL synthesizers are still practical. With large frequency steps, perhaps 10 Hz or more, tuning seems instantaneous while keeping reference sidebands well suppressed. Gaps between steps can be Siled in with schemes using additional PLLs, VXO tuning of the reference, or die digital symthesis—a method that we will discuss later Numerous schemes are available for cogrammable frequency division, limited y the exparionce of the designers. One is vv in Fig 4.47, The incoming signal is sand applied tothe down counting input of an Up/Down counter. a SHC193, The state of the counter decre- is by 1 with each clock pulse. When it ches 0, the “borrow” line goes low. This fed tothe data input of a D-FF. When the Q of that part goes low, the “load” eom- Sand on the “193 is executed, causing the Sata on the “jam” inputs, J4 (0 Jp, to be loaded in the couater, beginning the eycle anew, This averall circuit will divide by the number loaded at 1, 10.15 (010 15) plus 2, Several T4HC193s ean be cascaded 10 realize large divisors. The 74HC74 forces the output to be synchronous with the in pt clock. Many PLL frequency synthesizers use a presealer, a divider chat divides by a fixed amount before reaching programmable circuitry. This reduces the complexity of the programmable parts, but has the disad- vantage of multiplying the synthesizer step size by the pre-scale value, This diffieulty is eliminated with a vati- able modulus prescaler, a chip that divides by one of owo different values, depending ‘on the status of a control pin. For example, the Motorola MC12015 is a divide by 32/33: it divides the incoming frequency by cither 32 or 33. Extra circuitry is re quired in the programmable part of the synthesizer 1 uecommodate prescaler programming, but the programmable cir- cuitry is relatively slow, easing design and redueing power. Numerous commercially manufactured LSI (large-scale integration) chips are aif it P yanei9s Bor iene fawn, rosa 2— sa) 3} = Fig 4.47—A simple programmable divider. See text. BF output, Le-Lew Pasa pues a4 piviae ny wf “Lk Programing nected Fig 4.48—A simple PLL synthesizer featuring frequency steps much smi the reference frequency, ler than Oscillators and Frequeney Synthesizers available for phase locked loops. Ono ex- ample is the Motorola MC145170, which includes programmable N and R dividers, phase-frequency detector, erystal oscilla (or, and digital control and memory circuits.!® This IC functions up 10 160 MEE, receiving instructions as a 16-bit serial word. While the use of a this chip simplifies a synthesizer it often means that f microprocessor ar computer must be present in equipment using such a synthe: Sizer. The MC145170 and the National LMXISOLA are used in & synthesizer on the book CD, the DSP-10 transceiver, ‘The frequency multiplication and the resulting phase noise degradation between the reference and the VCO is a fundamen- tal property of a divide-by-N synthesizer that cannot be avoided with “improved” design. For this reason, it is becoming common for manufacturers of PLL inte- grated circuits to specify the phase noise Df their ICs at the phase detector. Spectral noise density in the -160 dBe/Hy region is ‘common. The final system desium is then degraded by 20Lox(N). It will be even worse if other noise soureos come into play, such as a poor VCO. A VXO Extending Synthesizer ‘A simple PLL. synthesizer with a single loop can be used in conjunction with a ¥XO for numerous special applications This could be & divide-by-N design like that of Fig 4-46, ora modified design chat ineludes a mixer, shown in Fig 4.48, The crystal oscillator (VXO) now serves asthe LO fora mixer and as u divided program able clock for the phase detector, The step size is no longer uniform, x conse uence of the variable reference divider However, the scheme is eapable of pro ducing very small steps with a relatively high reference Frequency. Cotsider an eximple: A 6.892-ME oscillator is placed in the circuit of Fig 4148 with N ranging from 32 0 64, Some Abut not al output Frequencies, sepsizes, and reference Frequcneics are given in Table 4.1 The reference frequency vaties acvord- Jing to the crystal frequency dividea by N ‘wile he step size vaies with Fx/N2, Con- Table 4.1 N VCO Output Step Size Rot Frog. 32 TIO T HH: ~BSkH2 2154 ke 3 7012 8.3 208.8 63 700171741094 64 70000 9 1681077 4.2509 06 “sn 4 “08 To -08 oe ‘ 19 Fig 4.49—A\ stepped, or verting the crystal oscillator toa VXO fills the gaps. When this is done, it may not be necessary «© use all possible N numbers. Synthesizers of this kind are useful as a means of extending the range of a VXO 10 cover a larger band. However, they are best used with an independent Frequeney ‘counter that provides readout, A practical project using this scheme is given else where inthis chapter. A practical, general purpose counter is also presented." Direct Digital Synthesis DDS, or direct digital synthesis is very powerful and is easily implemented with special. large-scale integrated ciscuits ‘The concept is deceptively simple: Digi- tal approximations to values for a synthe- sized sine wave are calculated or looked- up from memory. These values are loaded imo a digital-to-analog converter (DAC) witha new value being periodically gener- ated alter a fixed sample time Atypical DDS IC might be clocked with a 40-MHe crystal oscillator. This signal serves as a clock for updating the eulput ‘with a new sample that will persist for 25 nanoseconds (1/40 MHz) until the next update arrives. To illustrate the operation, assume we Want (0 generate a 3-METz sine wave witha | V amplitude. This is given as Vesin(Qemx xt) f=3MHe Eq 44 4.26 Chapter 4 wave is generated in DDS with a stepped approximation. Both the fampled” wavetarm and the desired eine wave result are shown, _Attimezere, the desired, outputsine wave will have zero amplitude, But 25 n§ later, it ‘will have un samplieude ealculeted by insert- ing 25 a8 into the equation, 0.454 V, At 50. nS, the signal will be 0.809 V. and so tort. ‘One could plot these values against nto ‘obiain the usual sine wave, However, this isnot what you would see when examining the DAC with a high-speed oscilloscope. Rather, you would see a line that is flat and level for 25 nS. Ir would then jump almost instantaneously to 454 millivolts and re- main there for another 25 nS. Av $0 a8 3t would jump to 0.809 V, and so on. This behavior is shown in Fig 449 where a sine ‘waveis sampled about 10.7 times per eyele It we had used an even 10 samples for teach cycle of the sine wave being gener- ated. the lowest frequency in the overall signal would be that of the output. The only distortion would be harmonies. Con- sider a slightly different case, one where wwe use 10.333 samples for euch cycle af the final oscillation. Three cycles of the ‘output waveform would then be generated with 31 samples. There is a longer peri- ‘odie character tothe overall waveform that ‘would create spurious outputs at one-third the output requeney. All harinonics of the low frequency axe also available. The spurs become more numerous as the peri couds become longer Fig 4.30 shows the measured output of san Analog Devices AD-9831 residing on a deme-board from Analog Devices. The part used a 25-MHz clock. An output of Fig 4.50—Measured output of a direct digital synthesizer using the Analog Devices 9831. Measurements were performed with a Tektronix 4948 Spectrum analyzer sel for a center frequoney of 7.0 MHz. The signal Is at TA MHz, This DDS device uses 2 10-bit Dito-A converter end the manufacturer reporis similar spurious respon 7.1 MHz was synthesized fortis example, producing spurious outputs over a wide spectrum. Other examples preduced spurs confined to limited regions, There areeven some “sweet spots,” output frequencies ‘that ace virtually free of spurs! Limited DAC accuracy is a common reason given to explain spurs in a DDS synthesizer. While this is usually domi- ‘ant, iis not the only source of spuss. The analysis presented above assumed a per- feet DAC and sill generated spurs. The very star-step waveform of Fig 4.49 is an approximation 9 a more ideal sampling waveform reconstruction.!5 ‘The wideband phase noise in the output of a DDS synthesizer is alten very low. comparable with the best Divide-by-N PLL. systems. However, this is of [ttle consequence if the noise is merely replaced by a family of coherent spurious responses, ‘Most current commercial transceivers use a combination of PLL and DDS tech- nology. Unfortunately. itis very ditticult togain even abasic understanding of these systems from the sketchy manuals, Rohde described an excellent example of a dual technology synthesizer.) That design used DDS to generate a 10,7-MHz signal that was tunable in small steps. The result ‘was bandpass filtered with « 10-kH2 wide crystal filler and chen frequeney divided 10 10 kHz wineee it served as the reference for 9 PLL controlling a 73 to 105-MHz vo,4.8 THE UGLY WEEKENDER, MK. The “Ugly Weekender” is a viable project for both the beginner and the sea- soned builder. The major feature, snd the source of the name, is the construction ‘method outlined in Chapter 1. This section Gescribes a version of that transmitter that uses frequency doubling to achieve improved oscillator isolation ‘The transmitter (Fig 4.$1) begins witha 3.5-MHz variable frequency oscillator. The familiar Hartley topology is used, al- though others would work as well. The oscillator, QL, runs continuously to avoid repeated warm-up drift, oscillating a few KHz above the normal Trequeney, but is shifted to the desired frequency during Uwansmit intervals. The VEO is tempera- ture compensated with a combination of NPO and polystyrene capacitors in the 3.5-MHz tank eireuit, The combination owas picked and confirmed with repeated temperature runs in a home-built environ- mental chamber ‘The VFO is buffered with a keyed dual- gate MOSFET amplifier. Q2. A JFET source follower driving a feedback ampli- fier would also provide the needed 10-milliwatt output needed 10 drive the frequency doubler. The 2X-frequeney multiplication occurs with a pair of diodes, as discussed in greater detail in Chapter 5. The doubler ‘output is selected with a single tuned circuit. A 10% banawidth double tuned circuit would be a better choice in this, position. The power lost in the passive fre- ‘quency multiplication is regained with a butter amplifier using Q6 and Q7. ‘The 7-MHz output from Q7 isupplied to & 500-0 drive control with output to a keyed feedback amplifier. Q8, shown in Fig 4.52. The keying voltage is derived from Q4, an integrating waveform shap- ing circuit ‘A feed-through capacitor in the two box version ofthis circuit routes the Q4 collee- 1, A 7-MHZ VFO TRANSMITTER tor voltage between modules. This eom- ponent wag eliminated in the single compartment version. The output power amplifier, Q9, un exer-reliable 2N3866 with 2 small heat sink, isshown in Fig 4.52, Numerous other Outside view of “Ugly Weekender” transmitters for 7 (left) and 3.5 MHz. U1, 36t #22, 768-6 tap até 2.5 me 224/0Eq, Li 100/Poly ** insis2 3n2iL a2 Keyed to Driver 1 72 3 16 1S with § bifilar turn output, Fr97-: ast #26, 90-2, 5 turn Links(z). 15t #28 on FT37-43, 4 t link. 15t FT37-43 Fig 4.51—VFO and frequency multiplier for the Mk Il Ugly Weekender, Oscillators and Frequency Synthesizers 4.27The power amplifier for the 7-RHz version. parts will function in this position with circuit details discussed in Chapter 2 Curput power is just over two watts with the drive control at maximum. A TIR sys: tem is included for QRP applications QS isa transistor switch thal generates & grounded line when the key is pressed. This signal is timed to hold for a short period afterthe key is opened to control an electronic transmit-receive switch with a 100-W power amplifier sometimes used with this exciter A Digital Dial ‘The frequency counters we see in the amateur literature are either general-pur pose lest instruments or special designs, intended as a readout for a receiver oF tansceiver. This unit falls into the later category, but itcould be expanded to serve ‘general applications ‘We waitled this design to use standard parts, Excellent special purpose counter chips areavailable, but they are often expen- sive and difficult to fing, Miero-processors, such as the popular PIC and Basic Stamp Series, can be configured as counters while serving all related display chores. But 4 4.28 Chapter 4 Inside frequency doubler. The VFO portion of the transmitter, Including diode 1 of a single board version of the 7-MMe trans- mitter. A receiving converter is atthe rear (left) of the box. ah sae #22, 68-6 Fig 4.52—Driver and power amplifier portion of the Mk ll Ugly Weekender. simpler solution was sought, one that was usable without special programming skills. This circuit uses a small number of readily available, inexpensive integrated cireults, including the four-LED displays. ‘The design was intended to be cheap cuough for repetitive use in a variety of projects, The approximate $10 parts cost included the time base erystal but did not include a PC boasd.2? ‘This counter avoids multiplex methods, which are prone to RF noise generation Frequency resolution is 100 Hz. Figure 4.53 shows 2 functional block diagram for the frequency counter, Sig nals to be counted aze applied to a single tuansistor conditioner that drives a gate controlled by the counter time base. For 100 Fz resolution, the gate rust he “open” for LO milliseconds. However. this design has an extra divide-by-10) to suppress last digit flicker, so» 100-mS count window is used. After the counting is finished and the gate is closed, a “strobe” signal is ap- plied to ICs that remember the counted result and decode itto a format suitable to drive the 7-segment light emitting diode displays. This is followed by a pulse that resets the counters to zero, ready for the next cycle ‘The time base. shown in Fig 4.54, be- ‘ins with a crystal controlled bipolar tran.cate strobe Bier mowaer 1] oi wowaer je] connected, sense eT | Fig 4.54—Time base portion of frequency counter. sistor oscillator operating at 3.2768 MHr, ‘The crystal isa readily available, off-the shelve pari. The oscillator is divided by 215 in Ul and U2, a pair of 74HC4060 timer” ICs, resulting in the desired 100: millisecond gate window. Further division in U2 provides a chain of additional 100 mS windows. These are decoded in U3 1 generate strobe and reset pulses, ‘The rest of the counter is shown in Fig 4.55. The signal to be counted is condi tioned with QI with the resulting log applied to U4A, part of aguad NAND gate with other sections serving as inverters ‘The ourpur is then counted by Ul 1a, US, and U6, 7AHC390 dual decade counters These drive the decoder drivers, U7 through U10, using 45118 decoder-driver ICs. This configuration will display kH2 tothe lett of a decimal place and tenths of a KFD to the right of the decimal place. We have used ICs from two families in Oscillators and Frequency Synthesizers Fig 4.53—Block dlagram for counter. ‘A.clean way to fabricate an LC oscillator uses a Hammond 15908 box, offering excellent shielding. DC enters through a foodthrough capacitor and AF leaves on coaxisl cable. This oscillator used a differential ca citor, but only one this design. Most ofthe system uses "HC" high-speed CMOS parts. This allows the circuit t9 function to 50 MHz or beyond. However, there is no need for high speed in the display function, so the decoder drivers use the slower standard CMOS ‘parts, Using slower parts here should help lo minimize RF noise and current con- sumption. We used common cathode, seven segment LEDs, type MANA7S0, Early versions of this counter used only two digits of display, showing only 01099 kHz. While this worked well as a digital substitute fora mechanical dial. i bee frustrating in some uppliculions. found ourselves wanting more resolution, ineluding « digit to the right of the KH? ecimal place, A more complete display with digits to the left allows complete elimination of mechanical dials in many systents. The lower current two-digit for mat is available by eliminating the related 4511 drivers and LEDS in the design pre sented, oval current depends upon the digits being displayed. With S-MPI2 input sig- nals, current was about 80 mA when the display read “S88 8", dropping to 30 mA with “111.1 The sensitivity was excel- lent with « 5:MH input, counting reliably ‘with an input of less than ~40 €Bm froma 50-02 generator. The counter continues 10 Iunetien to over $0 MHz, but requiring higher RF drive power. 4.29oan He rags) am Das ion 4.30 Chapter 4 Fig 4.55—input circuit, counter detail, and display portion of frequency counter. Frequency counter installed in a recelver. Ut was added "dead bug style" to oliminate flicker.4.9 REGARDING COUNTER ACCURACY Simple counter deseribed above is sble of good accuracy so long as the al and the oscillator components ace ‘The capacitor in series with the | shuld be adjusted to produce the seoper count when a known fgequoney is 4 to the counter iaput, outer as shown is suitable for use The wth simple direct conversion transceivers © supethet systems where the intermedi an even multiple of 100 chen functions accurately 2 the LO alone is counted, except for lft most digit. [Fa “less triendlly” LF is cd, other schemes must be applied, The sual transceiver might have several inter- Sediate frequencies, all of them with seven values, The corresponding osc sors, including BFOs or carrier oscilla- could all be counted. A mixture of up {down counting might be required with e Various oscillators, depending wpon the say the final frequency is calculated or measured. Clearly, this would be « good application for a microprocessor. ‘A simple counter that would stil be accu- sate over 2 wide frequency range could be builewith cireuiny much like that in Fig 455, exen ifthe IFis “unfriendly.” The simple up counters would be replaced with presetable ‘up-down counters, Instead of resetting the counters to 2er0 atthe end of each cyele, the counter would be loaded with an appropri- ate digital word that causes the LO coumting to produce the right readout. Ir is possible in some applications te ‘obtain reasonable results over a narrow tuning range merely by changing the crys- tal frequeney. This counter uses a clock ‘oscillator of 3276.8 EHz. That value is di- vided by a fixed valve t produce a time ‘windows that drives the counting gate. The final count is the number of cycles that ppass through the gate during the time in terval. ‘The display is a number that is 2 ‘constant multiplied by the ratio ofthe two Frequencies. IV the erystal frequency chunged, the “dial” can still be exactly right for one frequency. It might not be too Tae off at others that are close. Consider an example, a 7-MHz trans ceiver using a crystal filter at 1.98 MHz ‘The VFO will then be tuned to 5.02 MHz when the transceiver is at 7.000 MHz. Using the counter with the standard 3.2768-MHz crystal would produce a count of 20.0" instead of the desired “00.0.” Ifthe clock crystal waschanged to 3.2899 MH2. a 13.1-kH2 difference. the count would be proper at? MHz. The exror at 7.1 MHz would be 0.4 kHz, This may be tolerable for some applications. ‘There are several options available t0 the builder wanting to use a microproces- sor controlled counter. Simple units are available in kit form, ready for installation in QRP rigs and the bike, with references found on the web, Some examples are also included on the book CD.2! 4.10 A GENERAL PURPOSE VXO-EXTENDING FREQUENCY SYNTHESIZER 4.86 shows the block diagram for a unique frequency synthesizer. Although his example was built for 14 MHz using an off-the-shelf TV calor-burst crystal in he VXO at 14.318 MHz, the system can be adapted for many other applications. VHF examples are given later. This example used the VXO design presented tn Fig 4.30, The VCO used sith the syn- hesizer is thal of Fig 4.34, which can be scaled to other frequencies, We only discuss he synthesizer circuits in detail here, The VCO provides the niceded output, It wil usually be split in a hybrid swith one component used in an in tended output while the other drives the synthesis cireuitry. A level of -6 dBm is, needed by the synthesizer at both the VCO and the VXO inputs ‘The programmable Frequency divider is a version of the circuit shown in Fig 4.47 using two 74HCLU3 chips, allowing dlivi- sion by up to 258, The detailed circuit is shown in Figs 4.57 and 4.574. The divi- sion ratio is derived from two more 74HC193 chips, now operated as un up- down counter. Pulses to the “up” or the down” inputs increment or decrement the frequency by ene step. The user must es lablish the division range, controlled by Four hazd wired pains below U2, marked A.B, C, and D in Fig 4.57. The four inputs are connected to logie 0 (ground), logie 1 (45 V), or to the outputs from Us, Some possible variations are shown in Table . The frequency determining up-down counter, U3 and U4, may also be loaded ‘with an often-used setting, such as a rec ognized ealling frequency. Eaeb line must bbe hard-wired by the user to establish this frequency. The LjvDown commands are buttered with U6. Grounding an input Fine (P9 oe PLO) will cause an up or down pulse to appear at U3. A ground command on J8 also causes the “calling frequency” wo be loaded. The user may Wish to add mere 1 > 50 kakz to {f Low Pass mee \ @ a |_ Jenene 14.32 GIy +] Programmatie Drager iii ae P| es vents won NS LITT TP omecuns Rapir "Down’ va,04 vil toad ! internal pata “ening Prequency” Fig 4.56—Block diagram for the VXO extending synthesizer. Oscillators and Frequency Synthesizers 4.31Table 4.2 avaiable A B CD Siates Bio Us Us Usd 2t0130 Ud Ud Us 0 6610130 Ud Ud 1 0 saio6s Ud 1 0 0 interface circuitry to the Up/Down line: standard CW keyer circuits work well, as does a Keyer paddle or an computer mouse san input device. ‘The VXO and VCO are both applied to mixer U7, an NE-612, The low frequency jutput is iow pass filtered and impedance transformed with a pi-network using LT In the example, 2 200-KHz signal is trans- formed from 1.3 k02 to 300 0 wich the pi network formed by L1,C18, and C19. The 600-[H inductor consists of 22¢ #26 on a FB43.6301 ferrite bead. The lox pass ti ter components will change with other applications. The low pass filter ootpat is amplified ad conditioned for digital lev- Fig 4.57A—Continuation of the schematic in Fig 4.87. (70 wiser) asa)" vancasa BOF FJ] 2 5 op Toad SP sa © a i eacra apy 74H01 193 ba soos | any bat al Sa 23904 10K Ly oa 40K +5 Fig 4.57—Schemat 4.32 Chapter 4 for the experimental synthesizer. See text for details.els with Q2 and Q3. ‘Two programmable jumpers are provided at FPDI and J-PD2. While pin 3 of UR is normally driven from US in applications With thecrysial helow the VCO frequency. it may change to drive from Q3 in other sys- tems. The frequency scheme shown has the crystal above the VCO. A VCO tuning po- larity may also require a change. “The loop filter uses premium op-amp, the OPA-27. This fast, low noise partis iden) fortis application, The four input resistors are all 47 kQ while the feedback elements are 10 KQ amd 1.0 UF for the 1d-MF example. All of these components are sub- ject to change with ether applications and are marked TBD in the schematic for “to be determined.” They ure picked with the PLL computer program that accompanies Inaroduction to Radio Frequency Design. Phase lock loops must he designed with some eareandeamponent values ate nor well sited to casual selection, ‘The 14-MHz version of this design is summarized in the equation sheet of Fig 458, The programming sets N for values rom 34 1 66 with some frequencies listed in the table. The design equations use a ml eon tes) Ye wept al TRIS a Fig 4.58—Summary of available frequencies and isles of the 14-MHz “VXO extender.” This generated with MathCad 7. 7 nin stron =i 50-MH2 transceiver based upon a 10. Generated with MathCad 7.0. Oscillators and Frequency Synthesizers 4.33‘minus sign for this ease, Por the exystal is shove the VCO. ‘The synthesizer board is housed in a milled slaminum box (Hammond 1580BB) with elther coaxial cables oF feedthrough capacitors forall interfaces, The VX and he VCO are cach housed in individual nilled boxes (Hammond 15908.) While Lis possible to include both digital and RF/analog citeuiney on a single board, the solated and shielded approach is less orone fo spurious responses and is recom- ‘mended, ‘Once the boards are functioning and checked out. the system is turned on with relative ease. An oscilloscope senses the de on the control line while the VCO g is adjusted. 4.59 shows a design for the f6-metor band. It is intended to be used in a mono-band super-heterodyne trans- REFERENCES 1.W. Hayward, Introduction to Radio Frequency Design, Chapter 7, Prentice Hall, 1982; R. Rhea, Oscillator Design nd Computer Simulation, Second Ea tion, Noble Publishing, 1995, 2, Fora discussion of the squeeging prob- Jem, see Clarke, JEEE Transactions on Circuit Theory, Vol CT-13, No. 1, Mar, 1966. 3, W. Hayward, “Measuring and Compen- sating Oseillator Frequeney Drift,” QST, Dec, 1993, pp 37-41 4.K. Spaargarem, “Crystal Stabilized VFO" RatCom. Jul, 1973, pp 472-473, 5.1. Makhinson, “A Drift-Fres VFO,” OST. Doc, 1966, 9p 32-36: K. Spasrgaren, requency Stabilization of LC Oscilla- 088," QEX. Feb, 1996, pp 19-23, 6, U, Rohde, Digital PLL Frequency Sym shesizers Theory and Design, Prentice Hall, 1982, 7. °The RF Oscillator”. Radio Communi ations Handbook, Sixth Edition, RSGB, 1994, p 6.36, 8. U- Rohde, Digital PLL Frequency Syn- shesizers Theory and Design, Prentice- Hall, 1982; U. Rohde, “Designing Low: Phase-Noice Oscillators.” QEX. Oct. 1094, Fig 15, p 10; H. Johnson, personal correspondence with author. "Demphano =A Device for Measuring Phase Noise.” J 4a 4 Chapter 4 ceiver with a 10.0-MHz IF. ‘The synthe sizer operates in the 20-MHz range with a 19.847-MHz VXO. Ie is then frequency doubled and filtered to provide a 300 kHz range at 4D MHz The circuit could also be adapted for 25-MHz operation: frequency doubling would then allow use with a f-meter phasing transceiver. AA similar version could be built for the Lwo-meter band where an injection fre= queney of 144-10=134 MHz is needed. An especially useful scheme would use synthesizer operating at a tenth of this frequency, 13.4 MHz. IDN varies from 66 to 130, the required VXO would operate at 13.298 MHz, The synthesizer ourput ‘would be multiplied by 5 with a 74HCO$ and bandpass filtering, followed by 4 X2 ‘diode muhiplige and 134-MHz filter. The 10X scheme leads to simple frequency ‘counting, The system can alse be adapted Makhinson, Communications Quarterty, Spring, 1999, pp 9-17 9.D. B. Leeson. “A Simple Model of Feedback Oscillator Noise Spectra.” Proc; IEEE, Vol 54, Feb, 1966. pp 329-330. 10.U. Rohde, personal correspondence with author. 11, W, Hayward, “Variations in a Single- Loop Frequency Synthesizer,” O87, Sep, 1981, pp 24-26 12. bttp://www.gslnet/7n3 vm/ supervxo.htmt 13,W.S, Mortley. “Frequency-Modu- lated Quartz Oscillators for Broadcasting Equipment," EEE Proceedings, Part B. May, 1957. pp 239-249: W.S. Mortley, Circuit Giving Lincar Frequeney Mod lation oF Quartz-Crystal Oscillator,” Wire ess World, Oct, 1951, pp 399-403: V. Manassewitseh, Frequency Senihesi Theory and Design, Thied Edition, John Wiley & Sons. 1987, pp 401-405, 14, See, eg, U. Rohde, Digital PLL Fre- quency Synthesizers: Theory and Design, Prentice-Hall, 1983; U, Rohe and D. P. Newkirk, RF/Microwave Circuit Design for Wireless Applications, Chapter 5,1ohn ‘Wiley & Sons, Inc., 2000, 15, FM, Gardner, Phaselock Techniques. Second Edition, Wiley, Apr, 1979; V. Manassewitsch, Frequency Synthesizers for direct phasing at 144 MH. Nearly one full MHz of range is available ot the 2-meter hand ‘The “WXO Batender” isan experimen- tal synthesizer something of a departure from the normal schemes in use. The method is one that provides relatively small step sizes with much higher refer ence frequency, hu atthe price of uneven stop size Single loop synthesizers can be config- teed in & more tational format with mod est sep size while stil being used for seneral-purpose applications. For example, the Eleratt K2 CWISSB transceiver uses 2 single loop synthesizer with WKH seps The “clock” is « voltage controlled crystal xcillator thats then dsiven by a DAC, al: Tossing all gaps 10 be filled with small steps. Clever firmware on the part of the designsrs remove taning ambiguities. Theory and Design, John Wiley & Sos, 1976, 16, CMOS Application-Specifie Standard ICs, Motorola Ine, Publication DL130/D,, 1991, pp 5-101. Data sheet has a good set of references. Soe also design equation page. 17, W. Hayward, “Vaciations in a Single- Loop Synthesizer.” OST, Sep. 1981. pp 24-26; Talbot, “N-over-M_ Frequency Synthesis," RF Design, Sep, 1997 TREO. Brigham, The Fass Fourier Transform, Section 5.4. “Sampling Theorm." Prentice-Hall, 1974, pp SO+ 510. 19, U. Rohde, “A High-Performance Hy- brid Frequency Synthesizer,” OST. Mar, 1995, pp 30-38, 20. This cicuitis similar to one deseribed by G. Adcock, GABUK. "A Simple Fre- quency Counter for DC Receivers.” Sprat 73, Winter, 1992/93. p 10. 21. For the ultimate, high performance circuit, see W. Carver, “The Modular Dial." Communications — Quarterly Spring. 1998, pp 35-44. See also N Heckt, “A PIC-Based Digital Frequency Display.” QST. May, 1997. pp 36-38: and D. Benson, “Freg-Mite—A program. able Morse Code Frequency Readout,” OST, Dec. 1998, pp 34-36,CHAPTER Mixers and Frequency 5.1 MIXER BASICS Nearly all of the equipment we build ses at Least one mixer. Even the simplest ect conversion roeeiver uses a product ‘or, whieh is one form af mixer. Fig I shows the block-diagram symbol for a AA mixer isa three-part circuit with ‘0 input signals and one output occur ata frequency that is the sum andfor sisterence of the two input frequencies. De input, the Jacat oscillator (ox conver on osediator) is. usually much stronger an the other, the RF fap. The output in pical receiver applications is called an mediate frequeney. or IF, for it is fien part way beUveen u higher input fre- guency and baseband. While this historic lationship dacs not alway’ apply to mod systems, the JF term remains We begin our examination of mixers with an experiment designed 10 nnalyze a imple mixer with the goal of extracted understanding, What are the device char cieristies that allow mixing (difference and sum frequencies) and what are the ‘sulting signal levels? Are there undes- ed output signals? Our experimental mixer is the single JFET citcuit of Fig 5.2, Bath local oscilla- tor and RF are applied at the gate, While this may not be the most common scheme, it lends itself to analysis, Examination begins with the bias circuit of Fig3.3. Our goal isto model the FET and to then bias it half way between pinchoft and full drain current. The Fig 5.3 circuit is built without a “test” resistor, peaducing a source voltage uf 3.74 V. (These ate actual measured results with a J310 FET.) The FET current is very Jow owing to the high value source resistor, so the FET pincholt voltage will be close to 3.74 V. Test resis: tors from 10 KQ down to 15. were then Multipliers RE (Anput) I< TE (output) oC 3) tr = ratte cm Fig 6.1—Block diagram ig 5.2—Basic JFET mixer with LO and RF applied at the gate. The drain will then have all available outputs. It can be tuned to emphasize one mixer product. wet] Tr “Fceeet “il Fig 5.9—Blasing setup for FET modeling. Mixers and Frequency Multip! placed in the circuit, measuring source voltage for each, This allowed us to form a curve of drain current vs gate-source vol i, Fig $4. The data scatter (che bumps) resulted from thermal effects at higher cur rent levels. The smooth curve is calculated for an ideal JFET with a ~ 4.2 V pinchoff and Ipss=45. mA. These parameters pro- duced good ft to the measured data aver most of the range This exercise provides « mathematical ‘model, something to use to study the mix- ing process. A 150-8 resistor provides the desired bias that sets the source voltage al 2-V, about half way between full current ‘and pinchott ig §.5 is « modification of the smoath, modeled data. The zero voltage point has ‘been shifted to the middle of the graph, the bias point chosen with the 150-0. souree resistor. The vollage i the actual value appearing atthe gate in Fig 5.2. The total current has been split into three seg- iments. The frst sa1constant, the bias eur rent with no signals present. The second is the linear term, a straight ine. The third is ‘ parabola, The three components add to form the previous eurve. We now consider each of the three curve segunents by themselves as signals are ap- plied to the mixer input. The bias isa fixed ‘value: the tixed eutrent does no depend on any applied signal. This is evident in the bias curve in Fig 5.5, whieh is Ma ‘The linear term becomes more useful. I 1we apply a sine wave to ie gate that causes the voltage to oscillate henween Sand 0.5 V.a LV peak-to-peakswing te curtent wil vary by about HT mA peak-peak. A high im: pedance in the drain allows the signal eur- Fentto develop an output voltage. This isthe characteristic weseek when weuse the JFET rs Bto Calculate 7 / Weasured Drain Current, mA, Drain Curent ma 35 Gate-Source Voltage [y Pavabolie cate Votege Fig 6.4—Curve {it of data for FET modeling, The bumps are the result of thermal effects in data, whil Is calculated. as an amplifier, Consider the linear curve when two sig- nals are applied o the inpor: Two sine wave volrages atthe gate produce two sine wave currents, but nothing more: no mixing ‘occurs asa result of the linear term. There is also no distortion, This isthe behavior we intend when we speak of linearity Tis the parabola that becomes interest- ing. taking us beyond amplifier behavior. ‘A low amplitude gate signal causes no ceurtent. for the parabola is ze70 every~ where near 0 V, But current flows as the signal grows, Moreover, it is distorted This is evident; for a positive excursion ‘will produce the same positive current that is generated by 4 negative excursion. A large amplitude sine wave will produce tsv0 output current pulses per eyele as the signal swings both positive and negative about the bias point. We have built a tre quency doubler. We now apply the sum of «wo signal voltages to the gate. Again, the bias curve produces nothing. The Tinear curve will ‘enerate two response currents, cach a ‘replica ofthe input, but nothing more. No mixing occurs from the linear response, Butthe parabolic curve generates interest- ing results. Not only do we see cach input Frequency doubled, but we now see sum and difference products. This is not evi- dent directly from the curves, but follows dicectly from the related mathematics This is available on the book CD as a Marhead ile, mixer jietLmed. & file is also available (miemath pdf) that can be viewed even if the reader does not own Mathcad. ‘The two-component inputusesone part, the “local oscilla." at higher level than the other, the “RE” When this term is §.2 Chapter 5 ‘he smooth curve ed bias, applied to the parabolic curve, the result is, 1 product of two sine waves. Multiplica tion is the reason our mixer symbol. Fig 5.lusesalarge multiply sign, Highschool trigonometry identities convert the prod- uct of two sine waves into sie waves at the sum and difference Irequencies, the mixing result that we seek. The sum is cojien called the upper sideband, while the difference isthe lower sideband, erminol- ogy lelt over from modulation theory Most of the circuits that we call modula- tors are actually mixers, The power ampli- fier ina classic amplitude modulated (AM) transmitter operates asa power mixer. The cizcuit traditionally called a “modulator” Js really just an audio power amplifier. Fig 5.6 shows a practical version of the circuit we have designed. We use a 1-V local oseillator signal at 10 MHz with RF amplitude of 0.2 V at 14 MHz. The drain is terminated in 50 Q by way of a wideband transformer with a 5:1 turns ratio, result ing in a drain Toad of 1.25 kQ. The caleu- Jted output powers for all frequencies appear in Table 8.1. These are very close to those measured when we built de cit= cuit with the FET we had characterized. ‘The caleulations are in the Marhead file mentioned earlier. ‘The twa converted, or mixed outputs at 4 and 24 MHz have equal amplitudes, Which are much less than the amplified RF ‘output, The amplified LO isa large signal lose othe maximum possible from aJ310 FET with a 12-V supply with the drain impedance used. This mixer topology is ‘normally buile with tuned output, Tuning would eliminate the large drain voltage at the 10-MHz LO frequency. This would then allow a larger LO power to be used, which would increase conversion g: Fig 5.5—The FET current is split into three components: @ linear term and a parabola, Fig 5.6—JFET mixer with a wideband ‘output termination using a §:1 turns ratio transformer. LO power is applied to the ‘source, but this still results in LO between the source and drain, making this clreult the equivalent of Fig 5.2. Table 5.1 Freq. Power Description (ats) Ba aBi Lower Sideband mixod {own convertce) output 10 +189 Arpliiee LO {feedthrough} 8 Ampliieg Br {feeathrough) 20-01 Frequency doubled LO 24 “8° Upper Sideband mixed {up-converted) output 26-28 _Frequenoy doubled RF Generally, FET mixers (including those using MOSFETS) will have an optim conversion gain that is below the amplifier gain by 12 dB when the same terminating impedances are used. ‘Tho JEET example presented is but one of many devices that will produce mixing action, Mixing usually arises ftom nonineur device behavior. Mixing can also be produced ina system with time-dependent parameters, But, an ideal linear amplifier will never produce mixing. Even-order ‘curvature in a device characteristic is the nonlinearity needed for mixing ‘The simple single ended JBET mixer of Fig 5 6 becomes a practical cizeuit when the drain ic ned. But, it suifers from the wide spread in PET ehatuctersties, making it dif- fleult to use in a “plug-and-play” mode, A builder really needs to examine the FET to determine pinchoff and Ips, t0 establish bigs, and to pick the right LO level. The fol- lowing procedure may be used: (1) Build the mixer with u 100-KQ source resistor, Measure the source voltage to approximately establish the pinchott (2) Place a small resistor or even a short citeult across the souree resistor to in- fet Ipgs. (optional) (3) Find (mathematically or experimen- tally) a source resistor that sets the de Source voltage at half the magnitude of the pinchott (4) Apply LO power from a low Z source ‘and increase LO amplitude until the peak voltage approaches the de bias value. In the J310 example, the opti mum LO signal would be nearly 2-V peak, or 4-V peak-to-peak. A high- speed oscilloscope is required ‘The low impedance LO drive allows the FET to “look like” the souree is grounded for RF input signals. Similarly, the RE tuned circuit should be one where the gate Fooks back into low impedance atthe LO Froqueney. ‘The single JFET mixer, when earefully done, is capable of excellent performance. We have measured + to 6-4B NF with in put intereepis (third order) from 0 to +10 4Bm with a 2N4416, The J3L0 is more Gifficult to drive owing to the increased Ings, bul is vapable of higher IPS. ‘A bipolar transistor ean be operated as i single-ended active mixer, shown in Fig 7. Lowest distortion will result from higher standing current, bu this produces very low Fig 5.7—Simple bipolar mixer. input impedances presented to the local oscillator, making drive difficult. Emitter degeneration reduces drive power, but cun compromise noise figute. We have not per- formed carefial measurements on this mixer. Fig 5.8 shows a mixer using a single diode. Such mixers were once very com- men, especially for microwave appli tions. They have largely disappeared in modem times, ‘The usual diode mixer has no bias applied, but the LO signal is large enough that it causes the diode 10 conduet, When the diode conducts, it looks like & small resistance, allowing current. flow as the result of the applied RF, We envision the diode as a switch that is controlled by the LO. The switch is “on” for half of the LO eyele, and off forthe rest. When on, virtu- ally all of the RF power available can be selivered 1o-a load atthe IF port. But when the switeh is off, none of the power ean reach the load, With the RF reaching the TF load only half of the time, the voltage developed across the load from the RF ‘generator is only half as high as it would be if present all of the time. Accordingly. the mixer has a 6-dB loss. Fig .9 shows wavelorms for a single diode switehing ‘mode mixer Switching mode mixers are extremely common, with most af the mixers we use in communications operating in this way. ‘These mixers are typically passiveand use ‘50 power supply’ they offer no ain. The diode mixer of Fig 5.8 uses. series switch, “6 input Fig 5.8—A simple diode mixer. RF and LO inputs ‘generate an IF output, but the output is rich in ‘signal teodthrough. a0 INIA 0 T 7 myo 5 ‘ % Fig 5.9—Time domain waveforms for a single diode switching mode mixer. The IF output at any instant is the RF Input if the LO voltage Is postive, but 0 when i LO is 0 or negative. Mixers and Frequency Multipliers 5.3Fig 8.10—Switching mode mixer using a single FET. Although a JFET is shown, the mixer can also be Implemented with f bipolar transistor, a MOSFET. or 3 Gods FET. This clrcult typically has a ‘conversion loss of 6 dB, Input intercept {third order) can be from 0 to 420 dBm, sdoponding on the FET type. LO eneray at the RF port is typically reduced by 10 to 15 dB. Operating frequency wil dictate the components in the dipl filler, C1 and Lt. See text. hur shunt switehes also work well. FETS and bipolar transistors can be used in switching mode mixers Fig 5.10 shows a single FET as a shunt switch mixer. Steve Maas presented this circuit in detail ina 1987 paper.) We have used this mixer extensively in integrate formin GaAs imegrated circuits? The FET often las bias applied to the gate, a nega tive voltage equaling the FET pincholl. The 10 is typically a sine wave with a peak value equal to or just over the pinelioff, All three ports are terminated in 50 2. but the LO presents a severe mismatch. The con figuration shawn is 4 down-converter with am IF below the RF and LO, Up-converters exchange the RF and IF ports. ‘Tae diplexer filter. Cl and Lt in Fig 5.10, isolates the IF from the RF port ‘The capacitor isu single element hizh pass filter while the inductor is a low pass cir- cet, A coramon application might use an TF much lower than the RF. One can then calculate a “cvossover” Mrequeney that is the geometric average of the IF and RF.L1 land C1 are then picked t have a reactance 41 the crossover equal to the terminations. Higheroriler diplexer filters willbe needed if the IF and RE are closer. A bandpass? bandstop diplexer can alse be used. Mixer Specification and Measurement ‘We now examine mixers in more detail. secking the propertics needled to specify and understand mixers for use in a com- Chapter 2 included some vital, yet less common specifications for amplifiers ‘neluding noise figure and IMD. These phenomenon. which also occur in mixer 5.4 Chapter 5 Fig 5:11—Partial i= block diagram of a LP | athe recoiver. The IF le 3.6 Re, | proaesim' ‘o.4-Miz locel oscillator. cizcuits, ar illustrated by the system of Fig 5.11, a CW receiver for 14 MEiz with 10-4-MMe LO and 3.6-MULz IP, IMAGES, SIDEBANDS, SUMS AND DIFFERENCES ‘The example receiver mixer is preceded by a M-MHz bandpass filter hat ideal passes only frequencies clase to the ‘Wh-meter band. The 10.4-MHlz LO drives the mixer to produce an TF omtput at the 3.f-MHz difference between the RF and LO Irequeney. 14 -10.4 Temporarily rensove the input bandpass Filter and atach a wide range signal gen- trator atthe receive mixer RF input. There is now also a response at 6.8 MHz, for 14 — 68 = 26. The response to u 6.8 (MM? inpue is called the image response We ovaluate the receiver. now with the bandpass filter reconnected, by attaciing & signal generator to the input. Tune the gea- «erator to 14 MH~, deactivate reesiver AGC. land measure the receiver output signal This measurement works best with a mod- es input signal, perhaps -100 dBm, Note ‘he audio output, then tune the generator to 6.8 MHz, Increase the generator level until ‘he receiver output is identical tothe origi- nal. The ratio of generator power levels Is the receiver iinage suppression. It's straightforward fo build a bandpass filter at 14 MHz that will suppress 6.8-MHz signals hy 100 dB or more. Early receivers, the old insiruments now sought by collectors, used intermediate frequen: cies near 500 kHz, allowing 14 MHz 10 be received with a 13.5-MHz LO, The image response would then be at 13.0 MHz. Tt ‘wasdlfficult to obtain signifieant (by mod com standards) suppression of 13 MHz in a LMI filter ‘The receive mixer example hus Wo inputs: W4 and [4 MHz. We use the 3.6-Miz difference output response, But the mixer output will also contain su response. 10.4 + L4 = 24.4 MHz, The 5.6-MHz response is terminated in the usu ally ccasonable impedance match of the R6-MHz bandpass filter. But all 244-MED energy is generally reflected by the TF filter. That energy can got back int the mixer “output” where it might be reconverted back to 14 MHz, but in a dit ferent phase chan the criginal signal where it ean alter conversion gain and distortion performance. These problems are espe- cially insidiows with the popular diode ring mixers. I is for chiy reason that we often see exira resistive pads used with such mix- cers. They are often used in all three pots. Active mixers such as the FET discussed earlige age much fess prone co this problem. Assume that the incoming L4-MHz signal is modulated, containing a single uuppot sideband at 14.002 MH. We ana- lyze the bentiviar of the sidcbarl by cone sidering it 1a he an independent signal. It will be mixed down to IF without uny dis- turbance from the original carrier. The sideband ends upat 3,602 MH, sillabove the 3,600-MHz. cageier appearing atthe TF: iis sull « USB signal Our seeciving mixer would funetion just as well if we used @ 17.6-MHz LO, 3.6 MBz above the input. Aw upper-side= band st 14,002 MHz applied to such a receiver would produce an IF response at 3,598 MHz, now below the 3.6-MHz car Tier. Sideband inversion hax vecuered. This possibilily should be investigated in any SSB system, The analysis is equally valid when a carrier is supprcssed. Sideband inversion is often u practical advantage 10 the boilder/designer. For example, a popu lar erystal filter form is the lower sidchand ladder with greater stophand atenuation on tone side than the other. ISOLATION We ate always concoraed about the out put at one port of a mixer as signals are applicd 10 the others. For example, we might ask how much LO signal appears at a mixer's RF port. This would be impor lantin a receiver: we don't want a large LOsignal tobe radiated for the mixer RF port may be attached w the antenna with mini mal filtering. Even without radiation con- siderations, isolation can be important, If excessive LO wus present. it could be reflected by a filter to re-appear at the inixer RF port where it would be converted to produce # de output component. This could, in some mixers, alter the bias to change the mixer propertics. Isolation is easily messured for a mixer thats no already imbedded within &picce of equipment. If you are concerned with, for example, LO to RF port isolation, apply LO ata known level while examin. ing the output at the RF port by attaching ito a spectrum analyzer or measurement receiver. The LO power at the RF port will be lower (we hope!) than that available from the LO source. The difference is the suppression. This wil] depend on mixer luning in circuits such as the JFET described earlier. Often we hear Folks tal ing abeut “mixer balance” in éB. Usually, they are concerned with port-to-port iso= lution, which can be enbanced with bal: anced circuits, a method discussed later, SPURIOUS RESPONSES Consider the transmitter application shown in Fig 8.12, In this example, we ‘want to build a 7-1-MHe transmitter that ‘works with an existing receiver using a S-MHz IF, This will be accomplished by mixing the signal from 2.1-MHzLO with that from a 5-MH2z crystal oscillate. The output is Filtered with a bandpass filter: produce the desired output ‘Theidea! outpuicesponse feomthis mixer, assuming that the output filters removed. is that shown in Fig $.13. The desired sum product at 7.1 MHZ is accompanied by 1 difference response at 2.9 MHz. “The ideal is rarely realized. Fig $.14 shows, what we might actually see. This 2 result of ‘hamonic responses. Specifically, the output ‘ofa mixer excited by an LO at L MH and RE aR MHZ will be at FMF, 7 te PAater H>- he. / Eq ‘where n and m are integers. This spurious response, of spur generation relates 1 harmonies created within the mixer, even ‘when the inputs are free of harmonies. The upper patt of Fig 5.14 presents what we would see ifn and m were allowed to take con values from (to 7 with the bandpass Filter missing. The lower display is even more extiome, allowing values of 9 and sm up through 15. (These data were gener- ated with Spurtune.exe. a program distrib tuted with Iniroduerion 70 Redo Frequency Design.) ‘These uncalibrated displays are dis- couraging. Undesired outputs in such abundance would discourage anyone from ever using @ mixer in atransmiter! Fortue nately, not all spurious responses aze of ‘equal magnitude. The spurs tend w get Weaker asthe total order (n4m}inereases, Further suppression can occur with some spurs as @ consequence of balance that ‘might be used in toe micr. Spurs are also less with some system architectures over others. For example, if the transmitter considered here used a 12.1-MF@ LO instead of 21, the outputs of Fig 5.15 result A spur related 19 order “na” for the RF Fig §.13—Idealized mixer output for the clrcult of Fig 5.12 Frequency 15.0000 0.0000, | Mlhout ns output | al Ucfol inh laf de 0.0000, frequency 15,0000 ; : 0 ra je % f {dal hla dd hiatal ail com 0.0000 sro--> frequency -- 15,0000, freouency 45,0000 Fig 5.14—Mixer outputs with a variety of orders allowed, n ‘and m to 7 in the upper curve and 15 in the lower. Mixers and Frequency Multipli Fig 5.15—Spur spectrum for the same transmitter, but with a 12:1-llH2 LO. Spur orders through 7 are shown. 55‘will generally have a strength proportional to the “m th” power of the input atthe R mixer port. Hence, decreasing the RF ‘input by | 4B will drop a m-order spur by m dB, Mixer overdrive should be judi- ciously avoided, The worst possible cases are those where the IF is related to the ‘output by a small integer, IF = k x RF, or IF = REAR LO DRIVE LEVEL Most commercial mixers are specified with regard to LO drive level. For example, the typical diode ring mixer is Specified for #7 dBm. This is nor the power that is actualy delivered to the mixer port. Rather, itis the power avsilable to a 50-2 termination from the source that wll even: tually drive the mixer. Oscilloscope ccxamination ofthe LO drive to diode ing shows a severely distorted signal with less amplitude than the original sine wave dsiv- ing a pure 50-01 load. Many of the mea- surements we do with RF spplications are substitutions rather than the familiar in-situ ‘measurements of analog electronics, Various mixers behave differently as Lo power is varied. A small change in LO power makes almost no detectable ditfer= cence with the typical diode ting, In con- trast, the JEET studied earlier will show ‘output decreasing almost linearly ws LO drive drops, CONVERSION GAIN (OR LOSS) Miners are all characterized by # com ‘version gain, meaning thatwe examine the converted output power vs that available to the RF port, The method of specifying the gain will vary slightly. A diode ring imixer, a passive circuit, might be speci- fied with a loss, with 6 dB being a typical vvalue, Active miners suel s the JFET con sidered earlier will be specified by power ‘gain in a well-defined circuit, or peshups by a conversion transconductance, “Terminal impedance is specified for a mixer. Most passive mixers show an RF ‘input impedance that equals the IF termi- nation while the IFET mixer at the begin- ning of this chapter shows a nearly open circuit asthe input impedance at the gat, ‘ora low impedance al the source like that ‘of a.common gave amplifier. Output (TF) 5.6 Chapter 5 port impedances are usually high with ae- tive mixers, but related to other port termi- nations with switching mixers. That i, the impedance seen at the IF port equals the value presented to the RF port NOISE FIGURE Mixers all exhibit noise chat can be ear= acterized by noise ligure. Te measurement js similar o that of an amplifier. A wideband resistive termination at 290 K is frst pre= sented ta mixer input and the noise output is noted. Then, a sironger but knows noise source is applied to the input, again while observing output noise, The “noise gain” is compared sith normal available power gain te infer a noise figure ‘The procedues, both for definition and for measurement, are nearly identical to those used with an amplifier. Two differ cent mixer noise figures are available during uny given measurement, as shown in Fig 5.16. with the difference being the image-steipping filter. (An image-strip- ping filter is one that provents an image ‘rom reaching the input ofamixer.) Single sideband noise figures the desired p ler, for most systems use filters 10 eliminate the imge, Care is required 10 ‘puurantee that $SB NF is measured, for noise figure is defined only for a single signal c Passive mixers usually have a noise fig ture equaling the numeric value of the loss. Hence, the usual diode ring with a 6-dB conversion Joss will have a naise figure at 6 UB, or usta bit more INTERMODULATION DISTORTION AND GAIN COMPRESSION While noise figure limits the weakest signal mixer can process, inter= modulation distortion and gain compres- sion usually define strong signal behavier. IMD measurement is the same as is used with an amplifier, except that the output signals are observed at the converted fre- quency. T0 RF signals ot tones are com= bined in a suitable hybrid circuit with the ‘result applied to the mixerbeing ested. The butt tomes are then observed atthe mixer ‘output frequency. along with che distortion products, An intermodulation ratio is es tablished by the measurement, allowing a Fig .16—Scheme for measuring mixer noise figure. The upper cireuit determines the usual single sideband INF, The lower applies noise at two frequencies and establishes what is ‘often called double sideband noise figure, The bandpass filter eliminates any Image response from the mixer Input, DSB noise figure is typically 3.0B lower than the desired SSB noise figure. input or output intercept to be calculated. Gain is a constant for small signals, but eventually decreases as the RF level in creases. A useful parameter is the avail- able RF input power where the gain is below the small signal value by 1 dB. Mast mixer manufacturers specify their tigers by an input intercept value. This is indireet contrast tothe amplifier folks who focus on the output, Both forms are fine, so long as the reader understands what is being specified. Implicit in a mixer inpucintercept speci fication isan impedance. The usual spect ication uses 50-0 terminations atl ports. ‘and those terminations are wideband ones. This usually implies that the mixer was driving the input of a specttum analyzer uring the measurement, an instrument ‘with a good 50-2 input impedance at all frequencies. This oceurs when the ani- Iyzer is set for at Jeast 10 6B of input at- tenvation, This becomes very important with switching mode mixers where a poor ourpor termination can destroy otherwise excellent IMD performance.5.2 BALANCED MIXER CONCEPTS Some intrinsic mixer problems ean be reduced or eliminated when circuits are modified by adding balance. Consider Fig 5.17, part A, where we start with the familiar JPET active mixer. Local oscilla- tor energy is applicd at the source. FET je-source capacitance couples the Source voltage to the gate, degrading LO to RF isolation. Connecting a spectrum analyzer to the RF port reveals consider able LO energy at the RF port ‘The term balance implies symmetry. a circuit with two sides or parts. A circuit becomes a balanced mixer through dupli cation, shown in Fig 5.17. The duplication presented in part B did not improve 1.0 to RF suppression, bur that in C does. The sources in C are in parallel, but the two 108 are differentially driven. LO energy transferred to the gate of the first FET is exactly duplicated by that at the second FET, resulting in gate voltages that are in phase, But the transformer gate connee~ ‘ion results in no net custeat, and no LO frequency signal at the tansformer pri- mary. The LO 1o RF port isolation is now excellent, Practically, one might expect <8 improvement with balance, ‘The reverse, RF to LO isolation is also improved. A signal applied at the RF port results in gate voltages that are out of phase. But the sourees ure poralleled, resulting in seduced output at the LO port. RF to IF isolation is similasly improved. forthe drains are paralleled. However, LO ‘o IF isolation isnot altered. LO is applied a8 an unbalanced or single-ended signal, with IF extracted form a similar single- ended connection. There are no balanced currents that can produce any cancellation. Generally, balance improves isolation be- ‘ween ports that have differing termination forms, differential vs single ended ‘The mixer of Fig 5.17, part C. isasingly balanced circuit because balanced cir ceuitry is used in but one place. ‘The IFET balanced mixer could use other connections to obtsin similar results. For example, a transformer caus ing differential LO energy to be applied 10 the sources, while keeping single ended RF at the gates improves LO to RF isola- tion, I would also aid LO to IF isolation, bur would not improve RF to IF isolation. ‘A variation of the previous mixer might use a drain uansformer at the IF port shown in Fig §.18, A basic mixer, Q1.is duplicated in Q2, with a differential output conection through the wansformes. The LO is still single ended, but is now a cur- rent from the drain of Q3 applied to the sourees of QI and Q2. Althougit RE is applied only to the QI gate, this isu ditfer- ential excitation, for QL and Q2 are a dif ferential pair. As such, RF at the QI gate causes RF signal currenls in QI and Q2 that are equal, but out of phase. Bslance in this mixer improves LO to IF suppression (single ended to differential ports), but does not help RF to IF isolation. ‘The active balanced mixers presented, ‘ze all assumed to be built from identical transistors, Aluhough best when the cir= cuits are fabricated in integrated form, they can still be practical with discrete devices. Fig 5.19 shows balanced diode mixers. Past A presents a simple, yet very useful ‘Wo-diode mixer circuit. LO is applied toa transformer and causes the diodes, now behaving as switches, ¢© turn on during the positive half of the LO cycle, The diodes are off for the other half cycle. This, mixer is configured as a down-converter: ‘higher frequency RF signal is applied to the diode junction through C, while lower Frequency IF energy moves from the june- tion to the TF port It is instructive to examine the trans- ormer action in greater detail. LO power ceauses, at one instant, a positive voltuge at 44 dot on the transformer. But positive ‘voltage on one dot causes a positive signal ‘on the other. The windings are wired to ‘eenerate the polarities shown, one positive Fig 5.18—A JFET balanced mixer with single ended LO and differential IF ports, This mixer is similar to a bipolar Classic, the RCA CA3O20A. The RF at LO poris ean be imerchanged with little performance difference. vaa Fig 6.17—Evolution of balanced JFET miner. Mixers and Frequency Multipliers 5.7ind the other aegative at one instant in ime. ‘The diodes arc identical, with vatched on-resistance. Voltage divider astion then causes the junction (© be at round, or zero LO voliage. Even when the LO polarity reverses, the identical diode reverse capacitance values generate 2010 LLO voltage at the junetion. LO to RF and LO co [F suppression are both enhanced ‘The Land C values form a diplexer til- ter (see Chapter 3) in Fig 5.19. The usual sossover frequency used is the geometric moan of the RF and JP, the square root of (ge fg) Then, if the RF and IF imped ices are 5022, Land Care picked to have 51) Q of reactance at the crossover fre- jqueney. More complicated diplexer filters ty he needed if the IF is not small with gard 10 the RF. Diode LO eurrent is established by the ode characteristics and the souree mpedance provided by the LO system. The open circuit voltage must he hi enough 10 cause the diodes to frm oa. Greater available LO power produces ‘erdiodc current, which means that the inde on resistance is lower and conver sion loss is lower, Hot carrier diodes are normally used in mixers of this sort, for oy usually turn on with less voltage than 4 siligon junction type, The absence of a junction climinates charge storage effects allowing quicker diode turn-off, improv- ing UHF performance. This mixer is sill ry practical at HF with silicon switching. diodes such as the INA, The diodes in a mixer should all be matched for volta {rop when forward biased to a few mA The local oscillator essentially causes the diodes to switch on ane! off. This eambines Wwith the transformer behavior to generate Jow impedance between the transformer center tap und the diode junction when the diodes are conducting. The impedance is high when the diodes are ff. This behavior is extended to form a widehand mixer with the circuit of Fig 5.19B. “The mixers in parts & and Bot Fig 5.19 present a poor load to the LO generator, For LO current only flows on half of each cyele, The acldition of two more diodes, Hig 5 19C, provides a load on both halves of the LO waveform, With this connec- tion, the LO action can be thought of as & square wave. “These three misers (Fig 5.19. parts A. B, and C) are singly balanced with differ- etal connections only atthe LO port. But they evolve into a doubly balanced mixer in Fig 5.19D, whieh is labeled with LO polarity, During the polarity shown, diodes A and d2 conduct while diodes d3 and dt tre open circuit. The diode roles inter change when the LO polarity changes. The switching action is further illus 5.8 Chapter § 5.20 showing the two LO polarities, Diodes dl and 2 conduct with 43 and d¢ off in part A. Transformer ction generates a low impedance eonnec- tion between the diode junction and the T1 ‘conver tap. Bold lines in Fig 5.20 empha size the current that now flows 28 a result of upplied RF, Part B of the figure is the same, excopt for un opposite LO polarity ‘The diode ring mixer essentially creates & direct connection between the RF input, through the RF transformer T2, to the TF load, However, the polarity of the cannec- tion changes in synchronism with the applied LO. This process is called eor- ‘mutation: the diode ring is the classic example of a commutation mixer Fig 5.20 reveuls another interesting property of this circuit: The RF trans former. T2, communicates the LF termina- tion through co she RF port wirhour impedance transformation. ‘The tra former used at T2 is often thought of a having a 4:1 impedance ratio, and it ean certainly Tunetion this way in some appli- cealions. Bur this is not consistent with the figure. Rather, one half of the center- tapped seconilary carries eurrent for ec polarity of the LO. The inactive side has vvollage across it from transformer action, bbut no current other than that needed to charge stray capucitanee. (Care must be exercised whenever transformers with more than two windingsare used with non- linear deviees.) Time demaia waveforms for 8 coramut tion mixer are shown in Fig 5.21. The LO does no mare than to commute polarity of as Uy ting, doubly balanced mixer in part D. Fig 5.19—Evolution of diode mixers. Parts A and B show narrow and wideband versions of a to-diode mixer. The mixer is expanded to 4 diodes in part C, a circuit offering a better termination for the LO generator. These evolve it 10 a diode pe cred Fig §.20—Diode ring commutating balanced mixers. Seo text fornnd |] =m { t Fig 5.21—We ‘The AF and LO signals are those examined into resistive loads. Th s negative. ee it Fig 5.23—Evolution of the Maas mixer ‘whore balance Improves LO to RF Isolation, ‘when the sources are ‘signal is moroly the RF waveform, except that the polarity is reversed when the LO. ial appearing ut the IF port, Field effet transistors ean also be used in switching mode commutation mixers as shown in Fig 5.22. Patt A isa doubly bal- anced FET ring deseribed by Ed Oxner of Siliconix? Oxner’s mixer originally uscd fan integrated array of | MOSFETS. the Siliconix SD8901, Many quad analog switches are also suitable in this applica tion, although one should use those Featur- ing low on- resistance MOSFETs. Discrete MOSFETs will also function in this cir- cuit, A detailed analysis shows that exaerly the Same commutation action oecursin this mixer as we saw with the diode ring. Osner’s mixer is an exellent per- fomer, offering third order input inter cepts in excess oF +30 dBm. This low IMD occurred witha conversion loss of about & to 9 GB, The mixer functions well at HF, but degrades significantly at VHF. The FET ring mixer can be extended to higher frequencies with other technologies. In Mixers and Frequency Multipliers Fig §.22FET ring mixers using MOSFETs. Thi that originally desoribe by Oxner while that at B is a minimum transformer topology. ircuit at A is some measurements we saw conversion loss under 6 dB with large area monolithic GaASFETs, but IMD was not as low as observed with the MOSFETs* ‘The variation in Fig 5.22 part B uses only one transformer, Performance is similar to the other zing. although the incervepts are usually nor quite as high, ‘The passive FET mixer using shunt FETs, Fig 5.23. can also be extended ‘with balance. Duplicating the circuit with differential LO and TF, but a single ended RF resulls in a singly balanced mixer, Fig 5.238. Typical LO to RF isolation is 40 4B, ever at low microwave frequencies Balance is an extremely powestul and general design tool that ean often. be applied to enhance port-to-port isolation AE any mixer is lacking in, for example, LO-W0-RE isolation. placing «wo of them in a balanced pair will often enhance iso- lution by another 30.dB, with « bonus of a 3 dB increase in IIP3.>5.3 SOME PRACTICAL MIXERS The Gilbert Cel £By larthe most popularintegrated mixer circuit available isthe Gilbert Cell, named for Barrie Gilbert of Analog Deviees. Gil- bert developed a “four quadrant” multi- plier cireuit as an extension of a circuit presented earlier by Jones in US Patent 3,421,078 issued in 1966, The revised cit- cuit is described in more decal in the toxt by Gray and Meyer* The Gilbert Cell is based upon the sim. pler mixer eiseuit shown in Fi drives me base of QI to preduce the com- bined de and RF current that is then applied to the commen emitters of 2 dif fereniial amplifier, Q2 and Q3. LOenersy applied differeatially tothe dif-amp bases jeauses the RF to be toggled from one col- Tectorto the other. The IF termination is a hhalanced load, usually crested with @ teansformer. This topology improves RF (0 IF aud LO 10 RF isolation, for the RF input is single ended while the IF ouput and LO input are ditterential, This circuit ‘was available from RCA in IC form asthe CA3028A. This mixer suffers from poor LO to IF isolation, for ditferential drive atthe bases of Q2 and Q3 proxiuce directly amplified responses at the differential collectors. ‘The Gilher! Cell in rudimentary form, shown in Big 8.28, contains 2 pair ofthese Gifferential amplifier mixers. RE is applied o the lower differential amplifier, QI and Q4, producing two currents con taining de bias and the RF signal. These dtive the emitters of identical differential pairs that are switched by the same LO signal The Q3 and Q5 collector currents are in phase with each other with regard to LO drive: Q2 and Q6 share the other phase. However. one of the two output collector connections is “twisted” before attachment, producing a connection that cancels LO appearing atthe IF. Portto port isolation is now excellent for all combinations. Most Gilbere Cell mixers are imegrated. ‘The popular MC1496 and similar devices have heen replaced with ICs that include incersal biasing resistors. The most popular ‘of these isthe NE-602 shown in Fig $26 This version includes load resistors as wel 1 input biasing. One can actually measure the collectorresistors withan Ohmmeter.the RF input resistors do not really appear tbe there, although network analyzer measure- ments show the resisiors to representa goo! model. The test circuit of Fig $.27 was lab- reated t9 evaluate the NE6O2. ‘The conversion gain for this mixer was 20 B with LO drive of 0 dBm (632 mV 5.10 Chapter 5 pk-pk at pin 6) with the test circuit of Fig 5.27. Barly Sigmetics data recom. mend & minimum LO of 200 mV peak peak, -10 dBm in our test circuit. Conver Experimental discrete transistor version of a Gilbert Cell Mixer. sion gain dropped to 14 dB at this level in ‘our measurements, Both the RF and JF ports were floating inthe test circuit, allowing balanced drive Fig 5.24—The basic bipolar diferential amplitior mixer that is the basis for ine Gilbert Cell, This mixer can be built with 8 CA3028A, oF fabricated from discrete transistors. The 2N3904 would be Sultable for HF applications. Bi resistors (not shown) set the 2 ang Os ‘bases al approximately mid supply. Fig 5.25—Fundamental Gilbert Cell mixer. ‘The collector load is sometimes realized with resistors, although this will degrade Intercepts, for internal load resistors bsord power that would otherwise be avaliable to an extemal load,© balanced loads. This balance could be altered experimentally by bypassing one end of the transformer. Bypassing pin 2 educed gain by 2 UB and degraded the put impedance match. A similar exer 'se wt the output (pin 5) degraded gain by 4 4B. OF greater import, unbalanced ter- mination at either port degraded portto- port isolation, Balunecd RF drive will also alter product detector performance Our best IMD performance resulted with a single ended RF drive, IP3in was hen 17.5 dBm with conversion gain of S dB and 0 dBm LO drive Single sideband noise figure was mea sured at 7 dB for this test vireuit, This sveasurement was realized with a L5-MHz ow pass RF filter and a 19-MHz 1.0. We usually think of the Gilbert Cell as sn integrated eireuit, However, there is othing fundamental to preclude building these mixers in diserete form, A discrete Dilbert Cell mixer built From 2N3904 tran wistors is shown in Fly 5.28. No special ransistor matching was used, although all ransistors came from the same bag with entical manufacturer and date codes The chance is reasomahle that they came from the same silicon wafer, The eirevit presented some VHF oscil ation difficulty when power was initially applied. Although the problems occurred VEE, LO harmonics mixed with the ‘VHF signal 10 produce a low frequency utput that moved in frequency as our and was moved close to the circuit, The Sequency could also be tuned with chang supply voltage, The oscillations were suppressed with the 1- and 36-0 resis ors included in Fig 5.28. nt circuit of the Phillips NEGO2/NE612.° Fig 5.27—Test circuit used to evaluate the performance of the NE602. Most measurements used a 14-MHz RF, 19-MH2 O.dBm LO, and an IF of § Miz. The output 1 dB bandwidth extended from 0.5 to 10 Miz with the transformer shown. The AF port impedance match was a return loss of 19 dB while thet at the IF was 15 48. The internal oscillator was not used in these experiments. ‘The mixer was biased to either 5 or 15 mA with most experiments pertormed al the higher level. Single-ended drive is appear in the wideband IF output with both ‘about 14 dB below the respective input levels. Numerous other spurious outputs used for both RF and LO inputs, slightly romising porlo-port isolation ig 5.29 shows the IF port output spectra Conversion transducer gain for this circuit was 18 dB (1S mA, P-LO = 0.dBm, F-LO 10.4 MHz and RF= 14.3 MHz.) Inereas- ing LO drive by 10 dB made no difference in gain, but a drop to-10 dBm produced a -dB gain decrease. RF and LO signal sare present, all expected mixer spurious responses, Mast would he lawer in magni tude if the circuit was setwally integrated. This circuit bad a third-order input inter- copt of +11 dBm with 1S-mA bias and 0-Bm LO power. Decreasing the standing currentto 5 mA, produced a 1P3in=-2 dBm, with 16-B gain, still dramatically fetter than the (ete 2) Fig 5.28—Gilbert Cell mixer built with discrete transistors. A resistor (300 oF 62 2) Mixers and Frequency Multipliers bbotlom sets the bias current for the overall circuit 5.44— 2 _ 2 : ba o ® . ™ prequency, Hz" ™ Fig §.20—Output spectrum observed with the mixer of Fig 5.28. Seo text for details. E602. A diode noise source was used t0 measured DSB noise figure of 10.8 dB. ‘This extrapolates to a SSB NF of 13.8 dB Degeneration (22-0 resistors in the emitters of QS and Q6) was needed in the [RE input stage © reduce IMD. However, this degraded the noise figure, Although the main tool used to improve IMD performance in a Gilbert Cell is to increase current, feedback can also be applied. The experimenter should exam ine the work of Trask.7 ‘Some ofthe integrated Gilbert Cell mixers that were once popular (e.g., MCI496. E602) are becoming dificul to find. The ropalogy remains popular and is often found as partofalarger, multiple function TC. Some Gilbert Cells are available internationally, although design data is sometimes difficult to obtain. One example is the SNI6913P, from Texas Instruments Japan. This device is slated for discontinuation at thi writing It Fig 5.30—Part A shows 2 mixer using a dual gate MOSFET. Best gain occurs with ‘around 5 V pkeok at gate 2 for LO ‘2 cascode connection. This mixer I FET type. See text. 5.12 Chapter 5 ction. The mixer at B asily fabricated with nearly any available a pair of JFETs in appears similar to snother discontinued TL par, the TLA42. The Toshiba TA7ASSP is siill'in-produetion and could be 8 viable replacement in-new designs. (Thanks to JGIEAD and JA3FR for information on Japa nese parts.) There is ample challenge avail- able to the experimenter Dual Gate MOSFET Mixers JPET mixers were discussed earlier. A related device is the metal oxide silicon Field effect ransistor. or MOSFET. While the usual JFET isa depletion mode device, the ical MOSFET is an enhancement mode part, See the References chapter of any recent issue of The ARRL Handbook for definitions and further information MOSFETs were, at one time, often built with (wo gates with that closest t0 the source termed “gate 1." When one of the gates is forward (positive) biased with rospect to the source, the device behaves much ike a JFET with the remaining gate as the controlling element. These devices are often modeled as a cascode connection of single gate FETs. Mixers can, of course, bbe built with MOSFETs, for they exhibit the same quadratic transfer characteristic seen with the JFET. Fig 5.304 shows 2 mixer type that was very popular from the mid 1960s until about 1990. This circuit uses a dual gate MOSFET. an insulated gate topology with ‘ovo parallel gates. A rule-of dumb is that dual gate FET will display a nartow band ‘conversion transconductance of '/sthe gm expected for an amplifier biased at a simi Jar current with similar terminating imped- ances. (This guideline is consistent with ‘more refined analysis.) Traditional dual gate MOSFETs required an LO dive of about 5 V pk-pk at gate 2 to realize opti ‘mum gain Dual gate MOSFETS, although still available, are not as ubundantas they once ‘wore, The altemative mixer of Fig 5.308 inses a caseode-connected puir of FETS in a similar circuit. This connection was evaluated for noise figure, gain, and inter cept, The 2NS454 FETs from our junk box are similar 10 the popular 2N4416, TIS- 88, MPF-102, 2NS485, 25486, and many ‘other components; any of these parts should perform sell inthis topology. Our initial attempt with this cireuit presented a stability problem with an oscillation ocetsring ul the resonant frequency of the input cireuit. This was observed with a power meter attached to the TF output. The pseillation was eliminated when RI was inserted across the uansformer primary. A broadband IF ousput transformer is wound 1 relatively low loss type 61 ferrite core with a turns rato to present a good outputsrae7 watenas € ¥737~ tos Fcc ee sm Fig 5.31—Schematic for a low noise 10.1-MHz converter. match 10 50. An alternative winding would allow matching to a erystal filter. ‘The mixer shown, biased for 34 mA at 12, V. has measured conversion gain of & dB, with a noise figure of 10 dB and TIP3 of +5 Bm, There is nobalance in this circuit, so LO and RF energy is available at the IF port. This mixer is used in a simple Superhet receiver appearing later in the book. Many dual gate MOSFETs show very low amplifier noise figure with values of 1.4B being common, They can also func tion well in mixer applications. Fig $.31 shows a receiving converter with a meax sured NF of 6.6 dB and a conversion gain of 22 dB. This circuit needed an LO of 14.1 Miz 1o convert 10.1 MHz to.4 MHz. An available 7.05-MHz junk box crystal was used with a frequency doubler. The oscillator provides 10 mW to drive the passive diode doubler. The single tuned Circuit then increases the voltage to the required level. This mixer has a ow noise alow imped ss other than that of figure because gate 2 "sees tance at all frequen! the LO injection. Hence, noise energy within the LO system at the 4-MHiz IF and at the [0..-MHz RF does not reach the mixer output. The same miner with a wideband LO drive circuit will usually have a noise figure closer to 10 to 12 dB. ‘We did not measure IMD with this circuit ‘The traditional dual gate MOSFET mixer biased for 5 mA at about 10 V will have OIP3 of around +20 dBm. The input intercept will be this value ceduced by the conversion gain. The best dynamic range for mixers this sort will occur when the impedance presented (0 gate 1 (RF input) produces lowergain. Lowerimpednces will {Also alter noise figure. The advanced experi imenter (the one willing to measure amd optimize results) can expoctoutstanding per formance from either mixer in Fig 5.30. Diode Ring Mixers and Related Circuits The diode rin wcome the work: horse for the communications industry Although the mixer has Toss, noise figure 4 agie 24 ate 6m mK TRS at al Fig 5.92—A 14-Miz reoeiver front enc illustrating the /u0 oie Bandpass Filter problems of terminating a diode ring mixer Mixers and Frequency Multipliers is low and intercepts are generally high, making #1 the best choice when dynamic range is critical, The lack of gain is not, it itself, a problem. Is is important to use the tng with care if best perfermance is to he realized, Probably the most eritical characteristic ofa diode ring, and most other switching ‘mode mixers, is the need (0 curetully ter ‘minate the IF port. A proper termination (usually $0 ©) means that output energy available from the mixer is absorbed. IT power is reflected from the TF, it them impinges back upon the mixer IF port where i can be reconverted back to the RF, orto image frequencies. Reconverted components ean then exit the mixer RF post where they are yet again available for absorption or another reflection. With ‘each reflection can come phase shift and distortion, Fig, $.32 illustrates the termination problem, A diode ring is used in a 14-MHz receiver where a 10-MHz LO converts the desited signal toa4- MHz IF, But the mixer ‘output also contains a 24-MHz signal. The mixer is terminated in-an IF amplifier with the tirst selectivity appearing ajter the amplifier. Typical amplifiers have an put impedance that varies with fre- ‘quency. Evenif the amplifier inputiselose to 50.Q at 4 MH, it probably will not be 51) 24124 MHz as well. The 24-MHzcom- ponent will then be scattered from the ‘amplifier input back to the mixer output where itcan participate in further conver- sions, all undesired. ‘The mixer needs to be properly termni- nated for any and all signals that emanate from il, Assume the receiver is tuned © 14.00 MHz, buta strong signal appears at 14.01 MHz, That signal, once translated 1w the TF. is probably out of the erystal fil- ter passband. It will then be reflected by the filter and returned to the amplifier ‘output, possibly creating excess distortion there. IF the amplifier uses negative feed hack, the poor output termination for the |4.01-MHz signal will be reflected back 1 the amplifier input, creating an improper termination for the mixer. ‘The ebvious question that arises when a good impedance match is specified is “How g00d?" Generally, we look for an LF cerimination that is better than a 2:1 VSWR, ora 10-dB return loss. This match is easily measured in the home lub with a return loss bridge. signal generator, and sensitive detector. The detector could be a special receiver, a spectrum analyzer, poner meter, or even an oscilloscope (sce Chapter 7), The match should be exam- ‘ned over a wide frequency range, and with «signal level low enough to guarantee that the terminating circuitry is notoverdriven. 5.13Fram Mixer Fig 5.33—A post mixer amplifier using a Junction FET. A high Igy FET Is roquired such as the J810. See the text for transformer discussion. Tn many situations the IF port termina- tion requirements may be relaxed if the ‘match is improved at the RF port, Gener ally, distortion and gain measurements Will reveal the problems. The aggressive ‘experimenter can build the instrumenta- tion needed for these measurements Idoally, the best amplifier for terminat ing a switching mode mixer is onc with excellent reverse isolation anda frequen invariant (*flat") input impedance. Th ‘amplifier must have good distortion prop erties for it is often subjected to an entire band full of signals. The noise figure should be low, for it will add dicectly to the mixer loss 10 set the noise figure look ing into the mixer. Finally, the gain should be high enough vo compensate for mixer Joss and Joss inthe filter that will fallow, but nota lot more, Excess gain means that the signals become too large, stressing the following filter (erystal filters can be dam aged hy excessive signals, and can gener- ate their own IMD) and stressing the dis: lortion properties of the amplifier. A erounsded gate 1310 JFET amplifier suitable for post mixer applications is shown in Fig 5.33, This cirewit has good reverse isolation, so acrystal filter may be driven directly, The output transformer determines gain. A drain impedance of ‘shout 1200 2 yields 2 gain of about 10 dB. We measured third-order output intercept of +28 dBm for this ampliticr ‘when biased for [y= 14 mA. A noise figure cf less than 3 dB is possible with a slight 5.14 Chapter 5 Fig $.34—Post ‘mixer amplit Using a medium power, high F-t bipolar transistor, See text. turns, FTS7-33 input mismatch, The amplifier will or impedance is very close to 50 @ and is ‘mally yield an input match (return Toss) — fairly flat through the HF spestram. Typi- better than 10 JB. Good input match and cal OIP3 is +41 aBm ifthe attenuator is imodestinterceptsare ound only withhigh not partof the measured cireuit. The 6-13 current, whies happens only with fairly attenuator decreases the overall output high Ipss FETs. inercept o +35 dBm. The guin is 21 UB, ‘A favorite amplifier of ours (Pig 8.34) dropping to 15 dB with the 6-4B pad. for terminating & switching miner is a This particular amplifier uses the feed bipolar transistor feedback amplifier fole back resistor for transistor biasing, so losied by a -B pad, Negative feedback is changing circuit elements will alter bias- used to set the gain and to stabilize the ing as well as feadback. Altering feedback input and outpur impedances, This izeuit witheoostant bias current will maintain the ‘wus discussed in devil in che amplifier output intercept while chunging the gain chapter. The output termination on & feed- Input intercept will change accordingly. back amp will sttongly influence the input Noise figure for the amplifier of Fig impedance. As such, one should avoid 5.34 will ary with ransistor type and bias, diving actystal filter directly with sichan _butalues of SdB aretypical, Careful mea. amplifier, The filter impedance changes surements on one version of this eireuit rapidly with frequency, especially in the showed lower NF with reduced curren, region atthe passband edges, What may be offering some DR optimization, ‘fine termination in the passband becomes An attenuator a the input ofa feedback an open or short circuit in the skies and amplier will generate stable por imped Stop and. The resulting mixer termination ances as well as good output intercept may cause severe [MD problems However, the input pad degrades noise fig These problems are largely avoided by ure plicing a 6 dB pad inthe amplifier ourpat. Some recsiver designs (with high level ‘his then guarantees an amplifier with a mixers) demand amplifiers ith higher stable, Frequency independent input imped- intercepts, This is possible with higher tance (o terminate the miner, labo guuran- current. However, the output pad compro tees agood source impedance forthe crystal mises efficieney. A better solution uses filter, another vital consideration. {wo feeback amplifier stages with atenu ‘The amplifier of Fig 5.34 uses a ansis- ation between. The impedances ae stable tor usualy specified for RF powerorCom- and noise figure and intercepts are main munity TV service. They are bipolar tained. devices with a TW or better output capa- ‘There are some situations where no bilityand withanF-tharisatleast IDtimes amplifieris required. Its stillimportant to the highest feequency IF where they will maintain the proper mixer terminations. beused. The 2N3866.and 2NS109 are both An example might be the front end of a available at this writing and work well in spectrum analyzer, shown in Fig5.35. The this service, Many other parts ure suitable. first mixer is preselected with a low pass Paralleled 2N390ds of similar plastic filter and produces a Grst IF of 1.5 GHz. cased devives are also suitable and are ‘The pad in the mixer output stabilizes im- shown later. The amplifier in ihe figure peduncein both directions, ensuring mixer uses a bias emitter current of 50 mA anda and filter performance. The second mixer collector termination of 200 ©, provided produces 2 50-MEtz IF where an amplifier with « bifilar transformer. The input witha pad is now used. This topology hasa much higher noise figure than the usual reveiver, but is capable of excellent IMD performance, the parameter of greater interest for measurements Fig $.36 shows a different approach 10 the problem. Here, a mixer is followed by a diplexer filter that then drives a post mixer amplifier using a dual gate MOSFET. (40673, or 3N211 used.) The 2.2-KR2 gate resistor is transformed to look like 30 0 to the mixer through an L-net- ‘work, LT and C1. This only provides a ter= mination at the IF. 1.9 MHz in this example, Sum products are terminated ‘with a high pass filter paralleling the L-network. The preselecior filter was @ triple tuned cireuit in this example with about 3-dB loss while the MOSFET ampli- fier has « noise figure of about 3 dB, for a net NF of 12 4B. Overall gain is 9B, Mea- sured input intercept for the system was +15 dBm, This two-decade-old scheme is not as strong as others, but cam be an effi- one for battery operation, The broad: band impedance match is marginal? Perhaps the ultimate TF termination for the switching mixer is a special crystal fil- ter that presents a proper impedance at all amplifiers result from anow classic method described by Kurokawa. etal! Sucha f- teris discussed inthe next chapter aut lke the MiniCireuits SBL-1, TUF- 1, and ADE-1, 8 SMT part, epresent the Standard diode ings. There ure, of course ‘any more listed in then catalogs. These tnixers are specified fora LO drive power of 47 dBm. (Recall chat this avatlable power from the LO source.) The mixer is usually wall saturated at this 47 dBm and LO drive changes do not ater gain, The “sT-dBin” mixers will continue 19 fune- tion with LO drives es low as O10 43 dBm. withrediced gain and deyradedinterceps Some Mini-Cireuits pars are available for LO power as low as 0 dBm, Mini-Cirewits +7 dBm mixers ate speci fied for an input | aB compression power of +1 dBm, Arule of thumb states thatthe input intercept of a diode mixer is 10 to 15 dB above Pyp, placing IIPS at +11 10 4#16dBi, These ilves ar inline with oa Ieasucements for the TUF-1 and SBL-1 Most mixer manufacturers also build mixers. speciticd for LO power of 417 dBm, These miners usually use to series connected diodes in each leg of an Hequeneicn: Ths hter "and ‘simi thetwne coventona og, Ose ex an Ke) | KA_N_+ PASS ARS VF amplifiers. Fig 5.35—Front end of a spectrum analyzer showing ring mixers without ample, the TUF-1H, has +14 dBm value for Pp. placing IP3in at +24 dBm or higher. Even higher power mixers are available, including some “level 27-

You might also like

• ARRL Solid State Design For The Radio Amateur

  

  No ratings yet

  ARRL Solid State Design For The Radio Amateur

  258 pages
• Experimental Methods in RF Design

  

  100% (6)

  Experimental Methods in RF Design

  503 pages
• Exp - Meth - in - RF Des

  

  No ratings yet

  Exp - Meth - in - RF Des

  501 pages
• A Simple Dual Band SSB Transceiver

  

  100% (1)

  A Simple Dual Band SSB Transceiver

  10 pages
• Learn Navi Pocket Guide

  

  No ratings yet

  Learn Navi Pocket Guide

  34 pages
• Solid State Shortwave Receivers For Beginners R A Penfold

  

  80% (5)

  Solid State Shortwave Receivers For Beginners R A Penfold

  93 pages
• Nice

  

  100% (3)

  Nice

  503 pages
• Experimental Methods in RF Design

  

  No ratings yet

  Experimental Methods in RF Design

  503 pages
• Microwave Projects

  

  88% (8)

  Microwave Projects

  195 pages
• ARRL Propagation

  

  100% (2)

  ARRL Propagation

  191 pages
• Ikfast Paper Solutions PDF

  

  100% (1)

  Ikfast Paper Solutions PDF

  4 pages
• Printed Circuit Board Designer's Reference PDF

  

  No ratings yet

  Printed Circuit Board Designer's Reference PDF

  6 pages
• Radio Enthusiasts' Project Guide

  

  100% (1)

  Radio Enthusiasts' Project Guide

  13 pages
• Robot Kinematics

  

  No ratings yet

  Robot Kinematics

  33 pages
• ARRL Handbook For Radio Communications 1980 57th

  

  80% (5)

  ARRL Handbook For Radio Communications 1980 57th

  580 pages
• Solving Kinematics Problems of A 6-DOF Robot Manipulator PDF

  

  No ratings yet

  Solving Kinematics Problems of A 6-DOF Robot Manipulator PDF

  6 pages
• ARRL The Care and Feeding of Trasmission Lines PDF

  

  100% (2)

  ARRL The Care and Feeding of Trasmission Lines PDF

  111 pages
• Ikfast Paper Solutions PDF

  

  No ratings yet

  Ikfast Paper Solutions PDF

  4 pages
• 4 5900066337668990814

  

  100% (3)

  4 5900066337668990814

  390 pages
• W4RNL Antennas Made of Wires Vol 3

  

  100% (2)

  W4RNL Antennas Made of Wires Vol 3

  789 pages
• Radio Handbook 18 1970

  

  100% (6)

  Radio Handbook 18 1970

  897 pages
• Experimental Methods in RF Design

  

  100% (5)

  Experimental Methods in RF Design

  503 pages
• Kinematic Design and Description of Industrial Robotic Chains PDF

  

  No ratings yet

  Kinematic Design and Description of Industrial Robotic Chains PDF

  36 pages
• Computers in Amateur Radio

  

  100% (1)

  Computers in Amateur Radio

  257 pages
• Matrix Energetics - DR Richard Bartlett

  

  100% (6)

  Matrix Energetics - DR Richard Bartlett

  105 pages
• Misc RF Topics 2011

  

  100% (2)

  Misc RF Topics 2011

  41 pages
• Solid State Radio Engineering

  

  100% (3)

  Solid State Radio Engineering

  544 pages
• Upthecoax 3

  

  100% (2)

  Upthecoax 3

  6 pages
• Understanding LF and HF Propagation PDF

  

  100% (2)

  Understanding LF and HF Propagation PDF

  16 pages
• NanoVNA User Guide

  

  100% (1)

  NanoVNA User Guide

  13 pages
• Robot Kinematics Powerpoint PDF

  

  No ratings yet

  Robot Kinematics Powerpoint PDF

  45 pages
• ARRL Antenna Compendium Vol. 3

  

  100% (3)

  ARRL Antenna Compendium Vol. 3

  238 pages
• T2FD - The Forgotten Antenna: by Guy Atkins

  

  No ratings yet

  T2FD - The Forgotten Antenna: by Guy Atkins

  6 pages
• Amateur Radio Transceiver Performance Testing

  

  No ratings yet

  Amateur Radio Transceiver Performance Testing

  101 pages
• ARRL Radio Amateur Satellite Handbook

  

  100% (4)

  ARRL Radio Amateur Satellite Handbook

  360 pages
• Fundamentals of EMC Design

  

  No ratings yet

  Fundamentals of EMC Design

  12 pages
• Antennas

  

  100% (1)

  Antennas

  42 pages
• Facts About SWR and Loss PDF

  

  100% (1)

  Facts About SWR and Loss PDF

  62 pages
• Antenna Impedance Models

  

  100% (1)

  Antenna Impedance Models

  71 pages
• The ARRL UHF Microwave Projects Manual Vol 1

  

  100% (3)

  The ARRL UHF Microwave Projects Manual Vol 1

  349 pages
• Balun All Part

  

  100% (2)

  Balun All Part

  28 pages
• Kurt Speaks Out - Digital Version - Palomar Engineers

  

  100% (2)

  Kurt Speaks Out - Digital Version - Palomar Engineers

  140 pages
• W4RNL Antennas Made of Wires Vol 1

  

  86% (7)

  W4RNL Antennas Made of Wires Vol 1

  535 pages
• CQ Amateur Radio October 2020

  

  No ratings yet

  CQ Amateur Radio October 2020

  116 pages
• Handbook Amateur Satellite Service

  

  50% (2)

  Handbook Amateur Satellite Service

  84 pages
• APRS Manual

  

  No ratings yet

  APRS Manual

  182 pages
• HF Radio Guide EN

  

  No ratings yet

  HF Radio Guide EN

  26 pages
• Helix Antenna Performance Analysis

  

  No ratings yet

  Helix Antenna Performance Analysis

  48 pages
• Rockless VFO Transceiver Guide

  

  100% (3)

  Rockless VFO Transceiver Guide

  8 pages
• How To Build Advanced ShortWave Receivers

  

  100% (7)

  How To Build Advanced ShortWave Receivers

  119 pages
• (Magazine) QEX. 2004. Number 4 (2004)

  

  100% (1)

  (Magazine) QEX. 2004. Number 4 (2004)

  68 pages
• A High-Performance Homebrew Transceiver

  

  No ratings yet

  A High-Performance Homebrew Transceiver

  35 pages
• FPGA Radio for HF Applications

  

  100% (1)

  FPGA Radio for HF Applications

  76 pages
• ARRL Low Power Communication - QRP Radio Guide (Excerpted)

  

  33% (21)

  ARRL Low Power Communication - QRP Radio Guide (Excerpted)

  282 pages
• From Russia With Love: The Latest News

  

  No ratings yet

  From Russia With Love: The Latest News

  74 pages
• Yaesu FT-857D Operating Manual

  

  No ratings yet

  Yaesu FT-857D Operating Manual

  132 pages
• DIY Directional Coupler for Ham Radio

  

  100% (2)

  DIY Directional Coupler for Ham Radio

  3 pages
• Feature Dongle PDF

  

  No ratings yet

  Feature Dongle PDF

  6 pages
• Baluns and Symmetric Baluns

  

  100% (1)

  Baluns and Symmetric Baluns

  12 pages
• 1296 MHz Yagi Design for DXers

  

  No ratings yet

  1296 MHz Yagi Design for DXers

  5 pages
• The G5RV Antenna System - An Analysis: Rick Hiller - W5RH

  

  100% (1)

  The G5RV Antenna System - An Analysis: Rick Hiller - W5RH

  49 pages
• W1FB - Design Notebook

  

  100% (1)

  W1FB - Design Notebook

  200 pages
• The ARRL Antenna Compendium Vol. 3 PDF

  

  No ratings yet

  The ARRL Antenna Compendium Vol. 3 PDF

  238 pages
• Tactical HF Field Expedient Antenna Performance Volume 1

  

  No ratings yet

  Tactical HF Field Expedient Antenna Performance Volume 1

  78 pages
• Eznec 4nec2: Antenna Design For Free Using Mmana-Gal Software

  

  No ratings yet

  Eznec 4nec2: Antenna Design For Free Using Mmana-Gal Software

  8 pages
• EZNEC Modeling Tutorial by W8WWV

  

  100% (1)

  EZNEC Modeling Tutorial by W8WWV

  106 pages
• CQ Amateur Radio 08-2019

  

  No ratings yet

  CQ Amateur Radio 08-2019

  116 pages
• Urban Antenna Design

  

  80% (10)

  Urban Antenna Design

  214 pages
• K7QO - QRP - Laboratory PDF

  

  100% (2)

  K7QO - QRP - Laboratory PDF

  136 pages
• HF Propagation

  

  No ratings yet

  HF Propagation

  7 pages