| Abstract | U poglavlju 1 opisan je razvoj radiodifuzije od 1887. pa sve do danas. Prvi poznati prijenos informacije pomoću elektromagnetskog vala ostvario je Heinrich Hertz u nastojanju potvrđivanja Maxwellove teorije elektromagnetskih valova. U svom eksperimentu načinio je prvi odašiljački sustav koji se sastojao od odašiljača i prijemnika s iskrištem te uspješno potvrdio Maxwellovu teoriju. Odašiljač s iskrištem i njegove izvedenice nastavio se koristiti u komercijalne svrhe, za što je najviše bio zaslužan Marconi, sve do 1907. godine i otkrića triode, vakuumske elektronske cijevi. Otkrićem ovog elektroničkog elementa dolazi do ubrzanog razvoja novih grana elektronike i elektroničke industrije, a među njima i radiodifuzije na području dugog, srednjeg i kratkog vala. Pojedini sklopovi odašiljača koji su bili izvedeni isključivo u tehnologiji elektronskih cijevi postaju otkrićem tranzistora 1947. godine postupno zamijenjeni sa poluvodičkim komponentama. Tako je do 1980. godine, kroz svoju uporabu, tranzistor istisnuo sve elektronske cijevi u pojedinim sklopovima kratkovalnih odašiljača velike snage osim u izlaznom stupnju. Tu se do danas elektronska cijev zadržala kao element koji je nenadmašiv u primjeni za konstrukciju izlaznih krugova kratkovalnih odašiljača velike snage. Razvojem novih medija za prijenos i distribuciju informacija kao što su FM radio, televizijski prijenos, satelitski prijenos, Internet, radiodifuzija na području dugog, srednjeg i kratkog vala postaje sve manje zastupljen i korišten medij. Pokušaj da se on ponovno aktualizira dolazi kroz korištenje digitalnih modulacijskih postupaka na navedenim frekvencijskim područjima i to u vidu DRM standarda. U poglavlju 2 dan je kratki pregled klasa pojačala snage te amplitudne modulacije i OFDM digitalnog modulacijskog postupka. Klasična komercijalna radiodifuzija na području dugog, srednjeg i kratkog vala dugo je godina poznavala samo postupak amplitudne modulacije kao osnovu za prijenos informacije. Razvojem DRM standarda, u ta frekvencijska područja uvodi se OFDM modulacijski postupak koji je ovdje pobliže opisan kroz njegovu primjenu upravo kod DRM standarda. Glavna mana ODFM postupka jest visok odnos vršne i srednje snage signala što zahtijeva korištenje linearnih i energetski neučinkovitih pojačala snage. Prikladnim tehnikama linearizacije i namjernim ograničavanjem visokih vršnih vrijednosti signala uz odgovarajuće filtriranje omogućeno je korištenje vioskoučinkovitih nelinearnih pojačala snage za pojačanje OFDM signala. Na kraju je ovdje dan pregled osnovnih pasivnih mreža za transformaciju impedancije koje će poslužiti u izgradnji izlaznih krugova odašiljača. Kroz poglavlje 3 opisana je arhitektura radiodifuzijskog poluvodičkog odašiljača velike snage na kratkovalnom području frekvencija. Opisan je postupak nastajanja moduliranog signala kroz outphasing princip na pojačalu snage koje se sastoji od četiri MOSFET tranzistora u punom mosnom spoju. Nadalje, na temelju prethodnih razmatranja, dana je arhitektura poluvodičkog odašiljača velike snage za radiodifuziju na kratkovalnom području frekvencija. Na početku poglavlja 4 razmatrana je klasična mreža u obliku Wilkinsonovog sklopa za zbrajanje više izvora, odnosno modula pojačala snage. Zbrajanje snage ovdje je ostvareno na paralelni način. Analiza pokazuje da je zbog prevelikih dimenzija Wilkinsonovog sklopa, na kratkovalnom području frekvencija tehnološki prikladnije koristiti mreže za zbrajanje modula pojačala snage koje su temeljene na koncentriranim elementima, tj. induktivitetima i kapacitetima. Pokazano je da se prijenosne linije koje su sastavni dijelovi Wilkinsonovog sklopa mogu zamijeniti sa simetričnim četveropolom Pi ili T-tipa koji između svojih ulaznih i izlaznih priključnica može ostvariti fazni pomak od 90 stupnjeva kao i potrebnu transformaciju impedancije. Daljnjim pojednostavljenjem mreža, ovi četveropoli svedeni su na L-četveropol. Analizirano je kako se ponaša mreža za zbrajanje temeljena na L-četveropolu u slučaju ispada jednog ili više od N modula pojačala snage. Dan je prikaz promjene impedancije na izlazu svakog od preostalih modula pojačala snage za slučaj otvorenog kraja ili kratkog spoja na mjestu ispada modula pojačala snage. Također su određeni i snaga koju daje svaki od preostalih modula na svome izlazu, koeficijent refleksije na izlazu preostalih modula pojačala snage i pripadajući odnos stojnih valova. Drugi razmatrani sklop za zbrajanje modula pojačala snage na kratkovalnom području frekvencija koristi transformatore s feritnom jezgrom u serijskom spoju i kao takav omogućuje serijsko zbrajanje modula, tj. zbrajanje napona. Također je kao i u prethodnom slučaju analizirano kako se ponaša mreža za zbrajanje temeljena na serijskom spoju transformatora u slučaju ispada jednog ili više od N modula pojačala snage. U sljedećem koraku analizirana je izlazna sprežna mreža temeljena na T-četveropolu. Tu je pokazano kako na transformaciju impedancije utječe promjena dva od tri elementa u T-četveropolu te koji su elementi prikladni za primjenu kod ručnog ili automatskog ugađanja. Svi prethodni koraci u kojima su analizirane topologije mreže za zbrajanje i izlazne sprežne mreže omogućili su da se završno te dvije mreže ujedine u jednu cjelinu – izlazni stupanj. Kao konfiguraciju izlaznog stupnja, na temelju prethodno opisanih faktora koji određuju odabir elemenata i samih topologija mreža, predložena je građa izlaznog stupanja 10 kW odašiljača. Načinjena je detaljna analiza strujno-naponskih prilika u ovoj mreži. U poglavlju 5 opisani su realni elementi izlaznog stupnja. U odnosu na idealizirane elemente, realni elementi uključuju i fenomene elektromagnetske prirode koji se u karakterizaciji elemenata mogu opisati nadomjesnim shemama s koncentriranim i raspodijeljenim parametrima. Svojim utjecajem oni kvare funkciju idealiziranog izlaznog stupnja, no pažljivom konstrukcijom i odabirom realnih elemenata izlazni se stupanj može projektirati takav da se ponašanje određeno elektromagnetskim fenomenima pomakne u područje frekvencija iznad radne frekvencije. Tada se za područje frekvencija u okolišu radne frekvencije može primijeniti teorija s koncentriranim parametrima što znatno olakšava proračun kruga. Poželjno je da se ovaj pomak učini što više prema području frekvencija u kojem su amplitude harmoničkih komponenata signala malene, budući da će taj pomak utjecati i na rad samog modula pojačala snage. Niže vrijednosti modula impedancije na frekvencijama harmoničkih komponenata uzrokovat će veće struje harmoničkih komponenata za koje je sa stanovišta korisnosti pojačala snage poželjno da budu što manje. U poglavlju 6 opisan je prototip izlaznog stupnja ugrađen kao dio u elektromehaničku konstrukciju 10 kW odašiljača. Sama praktična izvedba sadrži nekoliko manjih cjelina, odnosno podsklopova. Tu spadaju ladice od kojih svaka sadrži tri modula pojačala snage i sklop za serijsko zbrajanje modula pojačala snage. Sklop za serijsko zbrajanje sadrži tri transformatora s feritnim jezgrama čiji su sekundari međusobno spojeni u seriju. Ladice se nalaze spojene na ulazima u paralelnu mrežu za zbrajanje. Izlaz paralelne mreže za zbrajanje spojen je na izlaznu sprežnu mrežu koja je načinjena kao T-četveropol. U postupku mjerenja izlaznog stupnja sa malim signalom određene su inicijalne pozicije elemenata prije rada sa radiofrekvencijskom snagom. To se postiglo namještanjem vrijednosti svih elementa izlaznog stupnja na vrijednosti određene proračunom u prethodnim poglavljima. U svrhu mjerenja vrijednosti impedancije elemenata korišten je vektorski analizator mreža. Tijekom ugađanja izlaznog stupnja sa radiofrekvencijskom snagom, izlazni je stupanj doveden u radno stanje određeno vrijednostima željene izlazne snage kao i željenim vrijednostima struja i napona na pojedinim elementima. U svrhu ugađanja izlaznog stupnja načinjena su i mjerenja valnog oblika napona i struje na izlazima modula pojačala snage pomoću osciloskopa. Zbog neidealnosti elemenata izlaznog stupnja na višim radnim frekvencijama, otkriven je povećani sadržaj neželjenih harmonika na izlazu odašiljača. Poradi toga, provedena je konstrukcija i ugradnja dodatnog niskopropusnog filtra na izlazu izlazne sprežne mreže kao i dodatno preugađanje izlazne sprežne mreže. Na kraju ispitnog dijela, izvršena su mjerenja ukupne ostvarene korisnosti izlaznog stupnja, linearna i nelinearna izobličenja u osnovnom pojasu frekvencija. Također su mjereni spektar i parametar MER u DRM načinu odašiljanja. Temperature pojedinih elemenata izlaznog stupnja mjerene su u trajnom radu odašiljača. Sve izmjerene vrijednosti zadovoljavaju uvjete za komercijalnu upotrebu izlaznog stupnja u sklopu 10 kW poluvodičkog kratkovalnog odašiljača za digitalnu radiodifuziju. |
| Abstract (english) | Chapter 1 describes the development of commercial AM radio broadcasting since 1887. to present times. The first known radio broadcasting was done by a German scientist Heinrich Hertz in 1887. In his experiment, Hertz was trying to confirm J.C. Maxwell's electromagnetic theory by transmitting and receiving an electromagnetic wave in a primitive radio broadcasting system that included a spark gap transmitter and a spark gap receiver. In his experiment, Hertz successfully verified Maxwell's electromagnetic theory. In the following years, the spark gap transmitter was further technically improved and has been used for wireless telegraphy by G. Marconi until the discovery of vacuum electron tube in 1907. The introduction of this new electronic element led to a rapid development in the field of electronics. Electrical circuits for radio broadcasting systems in the AM band, that covers long, medium and shortwave frequencies, were massively developed and produced. As time passed, shortwave transmitter output power level rose to 500 kW of average RF power. Due to wave propagation characteristics in the shortwave band, with this amount of output power and an appropriately designed antenna system, it is possible to cover any part of the Earth with radio signal. Therefore, for the first time in history it was possible to achieve global information coverage by using an electromagnetic wave as a carrier. After the discovery of transistor in 1947, vacuum electron tubes that served as active elements in transmitter circuits were gradually replaced with solid-state devices. Until 1980, in the majority of transmitter circuits, vacuum electron tubes were replaced with transistors. The last remaining circuit where the electron tube has not yet been replaced with a transistor is the transmitter output stage. This output stage design is still in active use today. Today’s high power shortwave transmitters contain only one electron tube in the output stage. The reason that this design is still active lies in the superiority of the electron tube in terms of achievable output power when compared to a single transistor. Currently, there is not a single transistor device on the market that can provide the same amount of RF power as it can be achieved with the use of a single electron tube. With the development of new media broadcasting platforms such as FM, TV and Internet, commercial AM radio broadcasting started to decline globally, or it has even been completely abandoned in some of the more technologically developed countries. In 1998, an attempt to make the existing AM radio bands interesting to modern audience was made by DRM (Digital Radio Mondiale) consortium. DRM standard uses modern digital modulation schemes in the AM broadcasting band that co-exist with the original analog modulation schemes. By using OFDM modulation in DRM, it is possible to enhance the existing audio quality and make it more FM like. DRM also introduced novel features such as broadcasting of text, image or short video to a wide audience. In that manner, DRM broadcasting provides a good foundation that AM broadcast bands can stay active in the future and this valuable part of electromagnetic spectrum can be used for covering a wide global audience. In chapter 2 a short review of power amplifier operational classes is given. In classes A, B and C the active element operates in the linear mode, and can be represented as a current dependent source. Classes D, E and F use the active element in the switching mode and the active element can be ideally represented as a simple electronic switch. During the historical development of transmitter broadcasting circuits all of the classes have been used in the circuit design. Classical AM broadcasting used only analog AM until the introduction of OFDM modulation in DRM broadcasting. A review of OFDM modulation scheme in the DRM standard is given here. This modulation scheme has proven itself worthy in AM broadcasting bands. Wave propagation in these bands is subjected to wideband impulse noise and multipath propagation effects and OFDM modulation is a very good choice to mitigate these effects. Looking from a power amplifier perspective, the main disadvantage of OFDM modulation scheme lies in the fact that OFDM signal has a high peak-to-average power ratio that requires linear and low efficiency amplifiers for amplification of such signal. Using linearization techniques and PAPR reduction techniques it is possible to use highly efficient nonlinear amplifiers for the amplification of OFDM signals. At the end of this chapter, a theoretical review of impedance matching networks that are usually used in the design of transmitter output stage circuits is given. Chapter 3 gives an architecture of a novel completely solid-state shortwave broadcast transmitter. An outphasing RF power amplifier that contains a MOSFET full-bridge and is suitable for the use in the shortwave broadcast band is described. A novel high power shortwave broadcast transmitter architecture that uses this power amplifier module is given here. This architecture contains a number of power amplifiers that are connected to a power combining network. The total number of RF power amplifiers depends on the transmitter output power. This network is further connected to an output matching network. Usual means of lossless RF power combing employ Wilkinson transmission line combiners. Due to its large mechanical dimensions, this approach is not suitable for the use in the shortwave frequency band. A better approach here is the use of power combining networks that contain lumped circuit elements (inductance and capacitance). These elements can be physically realized in the form of inductors and capacitors that are usually used in the shortwave frequency band. In the next chapter a detailed study of the power combining and output matching networks that use lumped components in the shortwave frequency band will be given. The idea of this approach is to integrate the power combining network and the output matching network to a single stage, the RF output stage. Using this approach, it is possible to minimize the effects of parasitic capacitance and inductance in the design process and doing so make the design more flexible. At the beginning of chapter 4, a classical N-port Wilkinson transmission line combiner is analyzed. Analysis shows that due to its large mechanical dimensions, this parallel combining technique is not technologically suitable for the use in the shortwave frequency band. A better approach is to use a parallel power combining network that contains only lossless lumped components (inductance and capacitance). It is shown here that transmission lines in the classical Wilkinson power combiner network can be replaced with symmetrical 90 degrees Pi or T networks. Power combing is formed by summing currents from each of the symmetrical networks at one node. By further simplification, it is possible to convert these networks to a simple L matching network that contains only two lumped components. A detailed analysis of a parallel power combiner network that includes L matching networks is given here. In the case of one or more power amplifiers dropout, in either open circuit or short circuit mode, original operating conditions for the remaining power amplifiers will not be fulfilled any more. This will cause a change in the power amplifier output VSWR and the total output power. These expressions are given here together with their graphical representations. An alternative power combining network is based on a series combining circuit that employs voltage summing. This network uses ferrite core transformers suitable for operation in the shortwave frequency band. This circuit is formed in such way that their secondary windings are connected is series, while each of the primary windings is connected to a singe power amplifier output. A detailed analysis of a series power combiner network that includes N series connected transformers is given here in the same way as for the previously mentioned parallel combiner network that uses L matching networks. In the next step, an output matching network that is based on a symmetrical T matching network has been analyzed. For this network to be used for manual or automatic tuning it has been determined which elements affect impedance module and phase. All of the previously conducted analysis have made it possible to unite these two networks to a single network – transmitter output stage. Based on the previous findings, architecture of a 10 kW shortwave solid-state DRM transmitter has been proposed. A detailed analysis of currents and voltages in the transmitter output stage has been made. This provides a foundation for selecting inductors and capacitors that will be used in the design process for the transmitter output stage. Taking into account electromagnetic effects, real word inductors and capacitors will inherently exhibit lumped circuit behavior only at the beginning of the operating frequency range. Chapter 5 gives an insight on the wide frequency behavior of real-world inductors and capacitors that will be used in the design of the novel solid-state shortwave transmitter RF output stage. When compared to idealized lumped inductors and capacitors, real-world elements can be described using more complex lumped or distributed circuit models. Lumped circuit models will contain two or more lumped elements, while distributed circuit model will contain transmission lines. In the design process, first an idealized lumped circuit is formed without taking into account real-world elements. This circuit is then expanded by including real-world inductor and capacitor models. We have analyzed the broadband frequency behavior of this more complex model. The analysis shows that lumped circuit theory can be applied at carrier frequency. At harmonic frequencies, it is necessary to apply transmission line theory. To make the RF output stage design process more simple and effective, real-world capacitors and inductors are chosen and designed so that in regards to frequency bandwidth, their lumped-circuit behavior frequency range is maximized as much as possible. On the other hand, the design goal is to translate the transmission line behavior to higher harmonic components of the carrier signal and make the power amplifier loading impedance module high at harmonic frequencies. This will affect the class D half-bridge power amplifier operation. From the transmitter efficiency perspective, the goal is to minimize harmonic components in the current that flows from the output of the RF power amplifier module towards the power combining circuit. After choosing the optimal architecture, a 10 kW prototype of a commercial shortwave solid-state DRM broadcast transmitter was designed and constructed, measured, tuned and tested. The prototype is described in chapter 6. The design process should satisfy the demands for optimal frequency behavior and the necessary power handling of the complete RF output stage. Twenty-one PA modules are arranged in a 3×7 configuration, where 3 PA modules are combined in series via ferrite transformer combiner and stacked horizontally. Seven of these are then combined in a parallel manner through the horizontal inductors that are stacked vertically and then connected to a common capacitor battery on their ends. The output matching network is a T section that consists of two separate L sections. The initial tuning of elements in the RF output stage, without RF power applied from the output of the PA modules, was done using a vector network analyzer and observing the matching elements impedance. The transmitter output was then connected to a dummy load and tuned. After the initial tuning, PA modules were turned on and the nominal transmitter output RF power was reached gradually. Voltage and current waveforms at PA module output were measured. Higher operating frequency current waveforms show the existence of not only the carrier frequency but also harmonic components that are caused by the non-ideal behavior of the RF output stage components. Although the new solid-state transmitter prototype circuits and elements have been designed and optimized with the goal to satisfy lumped theory behavior, the effects that arise from the electromagnetic nature of real-world inductors and capacitors cannot be avoided, what is true especially when operating at higher carrier frequencies. Overall transmitter efficiency, ratio of average RF output power and input AC power that includes all transmitter losses and auxiliary systems consumption, was found to be in the range from 60 to 70 % in AM mode, 50 to 60 % in DRM mode. Other parameters such as linear and nonlinear distortion, output RF spectrum and MER were also measured. It was found that all of these parameters satisfy the conditions for transmitter application in commercial shortwave radio broadcasting. |
