Research Paper DESIGN SIMULATION AND ANALYSIS OF CLASS E POWER AMPLIFIER

Transcription

1 Research Paper DESIGN SIMULATION AND ANALYSIS OF CLASS E POWER AMPLIFIER 1 Shankar R, 2 M.C.John Wiselin, 3 Divya Selvathurai Address for Correspondence 1 Research Scholar, PRIST University,Thanjavur,Tamilnadu,India. 2 Head and Professor/EEE, Vidya Academy of Science and Technology, Thrissur,Kerala,India 3 Rajalakshmi College of Engineering, Chennai, India ABSTRACT A new design methodology of Class E power amplifier is proposed in this paper which operates to broadband range of 6.78MHz-2.45GHz. The Broadband model is accomplished by high power and low power transistor. The Efficiency of the power amplifier is increased by increasing the order of the input and output matching network. A GaN-HEMT transistor is used for the high power, which is carefully modelled and characterized to prescribe the optimal output impedance for the broadband Class-E operation. GaAs HBT is used for low power design. The circuits are simulated using Advanced Design Systems (ADS-2011). 1 INTRODUCTION A power amplifier is an amplifier, which is capable of providing a large amount of power to the load such as loudspeaker, or motor. It is essential in almost all electronic systems where a large amount of power is supplied to the load. The power amplifier is used as a last stage in as electronic systems. The PA is more commonly known as audio amplifier. It will be interesting to know that a power amplifier does not actually amplify the power. But indeed it takes power from the d.c power supply connected to the output circuit and converts it into useful ac signal power. The power is fed to the load. The type of ac power developed.at the output of a power amplifier, is controlled by the input signal. Thus it is said that power amplifier is a dc to ac power converter whose action is controlled by the input signal. The power amplifiers are also known as large signal amplifiers. This term is used because these amplifiers use a large part of the ac load line for operation. Design of the art of Microwave Power Amplifier (PA) is very challenging to researchers, as it requires for providing high power with high efficiency. PA is generally a dc to ac converter which is driven by its input signal. During last few years, the field of interest of microwave design has shifted to monolithic circuit from the arena of hybrid components; this reduces the cost of PA. This change is due to the advancement of semiconductor technology which makes the monolithic designs to be delivered in the hardware as they are manufactured in bulk numbers. Thus the substrates employing monolithic circuits are selected by system requirements. The other cause is that all RF and Microwave Designs can easily be implemented by CAD methods, which improves the circuit-modeling techniques. These changes make the Microwave Monolithic Integrated Circuits (MMIC) to be time to market. Few applications involving large active element of phased array antenna design considers efficiency to be an important factor. Handheld communicating equipment has PA in the transmitter output stage. This application translates high efficiency by low power consumption, less heat dissipation, and smaller size. Researchers spurred interest on Class-E PA, which provides high efficiency topology. Class-E amplifier is a nonlinear switching type PA makes transistor to act as a switch. The design of MMIC generally employs Advanced Design System (ADS) simulator for simulation of circuits. With the increase in the use of hand held communication devices, a high efficiency power amplifier is needed at the transmitter output stage. Low power application circuit generally considered to have low efficiency when compared with high power circuits. Application such as telemetry of biological signals involves low power with low supply voltage into it. Low power PA circuits used in this transmission application require larger loads to match with the output, with high DC feed inductance when compared to high power amplifiers. Thus the power amplifiers designed for low power application has lower efficiency than high power circuits. In concern with the above application, the motivation of this research is to design a power amplifier with high efficiency and low power consumption. This is accomplished by Class-E PA designs which provides high efficiency irrespective high power or low power applications. The following are the reasons to choose class E PA. The class E amplifier is designed to provide high efficiency, and it is typically used at high frequencies. The class E amplifier uses an inductor to connect to the supply voltage, and a shunting capacitance across the switch to shape the standard waveforms for both voltage and current and decrease the power loss, thus, providing a better efficiency than Class D at high frequency. Compared to Class D, it is also less sensitive to the transition time of the switch. Class E power amplifiers can be designed with a small size, light weight and relatively tolerance to circuit variation. Furthermore, class E amplifier can be designed depending on the demand, such as specific frequency, narrow band operation. Similar to Class D, all class E power amplifiers are called power converters. In these circuits, the driving signal causes switching of the transistor, but there is no relationship between the amplitudes of the driving signal and the output signal. Thus these makes the choice of selecting the Class E amplifier over other Classes of operation.as the project aims in achieving the high efficiency of PA the option of Class E is the better choice for the work to be carried out. The proposed work has Class E PA for providing high efficiency to the low power application.

2 The various problems that are analysed in the design of PA are stated below: Optimizing PA with high efficiency with low power level is critical. Selection of the higher order of matching network has not been yet used in the broadband design. Variation in supply voltage affects the linearity of power amplifier which needs to be taken into consideration. Switching speed plays a vital role in selection of switch topology which directly affects the efficiency of the PA. Inductance constraint has been a critical problem in the design. 2 REVIEW OF PREVIOUS WORKS Frederickh. H Raab (1977) present the idea of the class E tuned power amplifier consists of a load network and a single transistor that is operated as a switch at the carrier frequency of the output signal. The simplest type of load network consists of a capacitor shunting the transistor and a series-tuned output circuit, which may have a residual reactance. Circuit operation is determined by the transistor when it is on, and by the transient response of the load network when the transistor is off. The basic equations governing amplifier operation are derived using Fourier series techniques and a high-q assumption. Sokal (1975) gave us the idea of new class of high efficiency tuned single ended switching Power Amplifiers. The new class of amplifiers described here is based on a load network synthesized to the transient response which maximizes power efficiency even if the active device switching times are substantial fractions of the ac cycle. The new class of amplifiers, named Class E is defined and is illustrated by a detailed description and a set of design equations. Kenle Chen et al. (2011) have given the clear idea of the Broadband Class-E amplifier designs. In this paper the authors presents a new design methodology for designing and implementing high-efficiency broadband Class-E PA using high order low pass filter prototype. GaN transistor has been characterized and modelled to perform the operation of broadband Class-E amplifier. The circuit is accomplished by the matching network provided in it. A sixth order low pass filter matching network is designed and implemented for the output matching, which provides optimized fundamental and harmonic impedances within an octave bandwidth. PA is realized from 1.2 to 2 GHz with a measured efficiency of 80%-89% which is said to be the highest reported today for such a bandwidth. An overall PA bandwidth of GHz is measured with 10-20W output power, db gain and 63%- 89% efficiency throughout the band. Jun Tan,Chun et al (2012) has given the idea of obtaining high efficiency in PA. This Paper presents a fully integrated 2.4GHz PA for short distance communication has been implemented using 0.13μm CMOS technology. Here the author presents a new methodology in the design of class E amplifier with a π- matching output network. The inductance constraint is removed by DC Feed Inductance instead of RF feed. The measured output power varies from -3.2 to 5.7dBm while achieving maximum overall efficiency of about 55% including the auxiliary predriver stage. The input given to the driver is the CMOS inverter. The proposed circuit by the author is optimized for delivering low output level with high efficiency and allow for fully integrated circuits. The author also discussed the analytical equations of the Class-E amplifier. Zhisheng et al(2012),the authors describes about the switching behaviour of Class-E PA. In this work the losses in Class-E amplifier with DC feed inductance were analysed. By the combination of a dynamic supply voltage and a dynamic cascode bias voltage, high drain efficiency is achieved over a wide power control range, covering from 2.2 up to 20 dbm. Fast envelope switching is obtained by adding a single switch to the common-gate nodes of both the Class-E stage and the second driver stage. The design of the art of work is done at 2.45GHz differential cascode Class-E PA in 0.18-μm CMOS with on-chip dc-feed inductor. At 20 dbm, a power-added efficiency as high as 43.6% was measured. Kuei-Cheng Lin and Hwann-Kaeo Chiou(2013) presented the idea wafer level bonding of PA. The aim of the work is to produce a complementary metal oxide semiconductor (CMOS) power amplifier (PA) using a wafer-level bond wire spiral inductor with high-quality factor(q). The proposed inductor with 2.75-nH inductance achieves a Q of 18, which is three times as much as that of a conventional CMOS standard spiral inductor at 2.4 GHz. The Q of the inductor is over 15 from 2 to 14 GHz. The output power and power-added efficiency of the PA with the inductor are improved by 1.5 dbm and 7% as compared with those of the fully integrated CMOS PA. Munir M. et al (2008) gave us the idea of using the selection of matching network in low power application. This paper presents a fully integrated, 2.4 GHz class-e power amplifier (PA), with a class-f driver stage. The circuit was fabricated in a standard 0.18 μm CMOS technology. Measurement results show a maximum drain efficiency of 38 % and a maximum gain of 17 db.when operating from a 1.2 V supply, the PA delivers an output power of 9 mw with a power-added efficiency (PAE) of 33 %. The supply voltage can go down to 0.6 V with an output power of 2 mw and a PAE of 25 %. The circuit also has a second output to test the effects of using an on chip filter in low-power designs. It gives the idea of implementing the power amplifier to biomedical application where the concern of low power consumption is taken into account. Ray Pengelly et al (2008) the choice of selection of transistor for the broadband designs is given by the author. This presentation gave us the idea of GaN HEMT( Gallium Nitride High Electron Mobility Transistor) its characteristics,attributes compared to that of other transistor. The Analysis of the transistor such as large signal analysis and small signal analysis are derived and their schematic is simulated using Advanced Design Software (ADS) is given. The transistor modeled is checked for DC Analysis, Load Pull Analysis. K.L. Kotzebue (2011), A technique for the design of broadband microwave transistor power amplifiers is presented that utilises the powerful methods of network synthesis to achieve optimum large-signal performance. Only two large-signal transistor

3 measurements per frequency are required to achieve a good analytic model of the transistor's variation of added power with load impedance, and a mapping function is presented that translates this added-power characteristic into an equivalent linear-circuit reflection-coefficient characteristic. With this representation, methods of linear-network synthesis are used to obtain circuits which optimise the amplifier's added-power efficiency over a broad range of frequencies. The design technique has been experimentally verified by the characterisation, design and construction of a BJT amplifier of nearoctave bandwidth centred at 1 GHz, with the largesignal performance in good agreement with that predicted by the design theory. Acar (2007) presented the analytical design equations in this paper for Class-E power amplifier taking into account both finite drain inductance and switch on resistance.the analysis indicates the existence of infinitely many design equations; and the conclusions given by them are based on Class-E conditions (e.g. zero voltage and zero slope) can be satisfied in the presence of switch-on resistance. The drainefficiency (η) of the Class-E power amplifier is upper limited for a certain operation frequency and transistor technology. Using a finite dc-feed inductance instead of an RF-choke in a Class-E power amplifier can increase η by 30%. 3 PROPOSED BROADBAND PA DESIGN The Class E PAs are classified into different types depending upon the Input and output Matching Network they are designed depending upon the requirements. The Designs are as follows Narrowband amplifier Design Broadband amplifier Design High gain amplifier Design Maximum gain amplifier Design Low-noise amplifier Design Minimum noise amplifier Design Multistage amplifier Design The Proposed Design chooses the matching network of Broadband amplifier Design and Narrowband amplifier Design. Broadband amplifier is essentially a flat response (generally no resonant element) over a wide range of frequencies one to several octaves or decades of bandwidth. In the Narrowband design as the series LC resonator has a limited frequency response. The Fourier series expansions of Class E indicates that the optimum load impedance is frequency varying. The Topology of Broadband Amplifier is given in Figure 3.The Broadband Class-E PA when it is implemented two problems is to be addressed. Matching network is needed to transform Z 0 to Z E for the switch capacitor tank over a significant bandwidth, e.g.,>50% A Broadband filter, whose pass band covers the desired bandwidth, is needed to provide infinite Impedance as Harmonic frequencies replacing the LC resonator. The switching behavior, characteristics of Class E amplifier is not achievable at the gigahertz frequency. This is due to the non-ideal effects of transistor and non-square wave driven signal at the input. But in practical case the transistor is biased at the pinch-off point. Fig 1. Topology of Broadband Class E PA So the input signal should be sufficiently large to make the transistor switched on during the positive half duty cycle and make it off during the negative half duty cycle. The Design of the Broadband Amplifier is very critical to the researchers as it needs to work over for an octave bandwidth. It is notorious that the bandwidth of the conventional topology of broadband amplifier s matching network design is limited to only one octave, as the second harmonic of the low-end fundamental will occur in the pass band if the bandwidth is above one octave. To achieve a multi octave bandwidth, one can design multiple matching networks for different bands. The Proposed design which produces higher efficiency by increasing the order of matching network. The design methodology is carried out as the steps described in power amplifier design. The First Step in the design of a Power Amplifier is to select a proper device for the operation to be carried out. The suitable transistor for Broadband Amplifier is GaN-HEMT. To date GaN HEMTs have been applied in numerous high-efficiency broadband PA designs. In order to choose a transistor for designing a Power Amplifier it involves a various design procedure into it. DC and Load line analysis Bias and Stability Load Pull Analysis Impedance Matching Gallium nitride power transistors have very high RF power densities which range from 4 to 12 watts per mm of gate periphery depending on operating drain voltage. Even though GaN/AlGaN on SiC substrates have high thermal conductivity it is necessary to be aware of channel temperature rise incurred by both DC and RF stimuli when designing power amplifiers. Thermal management is even more important for broadband amplifiers (a very popular application) where drain efficiencies can vary considerably as a function of frequency. In general, GaN transistors exhibit high breakdown voltage, low and voltageindependent output capacitance, and low turn-on resistance these characteristics are very important for realizing high-efficiency switch-mode PAs. However, current commercially available GaN HEMTs that are appropriate for discrete designs are typically packaged. Their packages often introduce non negligible parasitic. The GaN-HEMT transistors are manufactured by two foundries such as Cree and RFMD. 4 SIMULATION ANALYSIS OF THE PROPOSED PA The Proposed design of broadband amplifier the efficiency of the circuit is increased by increasing the order of the matching network. The input matching

4 network is taken to be the order of six and the output matching network is chosen to be the order of three. 4.1 MODEL OF GAN-HEMT As the model of this transistor is not available in the simulator the modeling of transistor is done with the large signal analysis of FET-Model as shown in Figure 5.5 with the values specified in the datasheet. But in the simulator the modeling of transistor is not possible with the design. It can be done if the specifications are provided by the manufacturer the particular voltage is found using this analysis. The characteristics of transistor are verified using this analysis. Fig 4.DC Analysis of RF3931-Schematic Fig 2. Large Signal Modelling of GaN-FET The design kit of RF3931 provided by the RFMD manufacturer is imported into the Simulator for further simulation designs is shown in Fig 5.6. The transistor resembles the same as it is commercially available in the market. Fig 3.RF3931 Transistor in ADS. 4.2 DC AND LOAD LINE ANALYSIS This analysis is performed to verify the minimum and maximum voltage that the transistor requires to perform the operation. The Id-Vgs curve that is obtained for that particular transistor is plotted. The load line plot helps to find the power consumption at Fig 5.DC Analysis of RF3931-Results 4.3 BIAS AND STABILITY Proper Bias network design is essential for any nonlinear circuit design as it is essential to ensure that right amount of bias reaches the device and also it doesn t load/leak the desired RF energy. Choice of bias network topology is pretty much dependent on the frequency of operation. For lower frequencies designers can use Inductors/Choke in the DC bias path and for higher frequencies high impedance quarter wavelength line is the preferred choice. Necessary and sufficient conditions for device to be stable are: Stability Factor (K) > 1 Stability Measure > 0 Fig 6.Biasing of RF3931-Schematic

5 over certain section in smith chart to characterize Output Power, PAE, IMD (with 2-tone LoadPull) etc., to find out our optimum impedance to be presented to device and then accordingly perform impedance matching network design. Fig 7.Stability Factor-Results Fig 8.Stability Measurement-Results 4.4 LOAD PULL ANALYSIS Load Pull is a very commonly used and preferred analysis for PA design applications. Load Pull is the technique during which we keep source impedance and source power is kept constant at certain level and then sweep impedance / reflection coefficient of load Fig 9.Loadpull Measurement RF3931- Schematic 4.5 IMPEDANCE MATCHING AND S PARAMETER ANALYSIS This is usually done with the smith chart in order to find the input and output matching network for the amplifier design. The circuit is designed with the help of impedance calculated with the particular frequency.this Analysis is performed to find the input reflection coefficient (S11), reverse voltage gain (S12), forward voltage gain (S21), output reflection coefficient. The gain at the particular frequency is found using with this analysis. Fig 10.Results of RF3931-Loadpull Measurement required values. The output obtained for the different frequencies are given in the table 1 and 2. Table 1. Forward voltage gain of RF3931 Table 2. Input Reflection Coefficient of RF3931 Fig 11.S Parameter analysis of RF3931- Schematic The S2P files are imported from the datasheet of vendor, to check whether the data coincides with the

6 4.6 Power Amplifier Designed With Rf3931 With the various analysis performed as mentioned above the designing of Power amplifier with RF 3931 is made. The circuit is tested with the test bench of power amplifier in the simulator is shown in Fig.The circuit works to the broadband range of GHz. Fig 12.3GPP FDD RF power amplifier power added efficiency test bench 5 RESULTS AND DISCUSSION Fig 13.Results of RF Power added vs Source power The power amplifier designed is worked to the range of GHz the input and output reflection coefficient and the forward and reverse transmission coefficient measured is shown in Fig 14 (a) and (b). (a) (b) Fig 14 (a)input Reflection coefficient of RF3931 Power Amplifier (b)reverse Transmission of RF3931Power Amplifier 6 LAYOUT GENERATED FOR RF3931 POWER AMPLIFIER DESIGN The schematic mentioned above involves input and output matching network to balance the transistor. The layout of the above schematic is shown in Fig 5.19 Fig 16.Layout of Lumped Components Fig 15.Layout of Transistor via Grounds Fig 17.Layout of the complete design with measurement

7 7 CONCLUSION The Idea of designing a Power Amplifier involves various analysis procedures which help to design a perfect impedance matching network is discussed in detail. As per the objective of this project, the narrowband design of 2.4GHz and the Broadband design of GHz were designed. Both these designs were simulated using Advanced Design Software (ADS ) and analysed the Parameter such as Power Consumption, gain, output power, S- Parameter, Impedance, Efficiency etc. The various analyses such as DC Analysis, Stability Analysis, Load Pull Analysis, Impedance matching analysis, Smith chart analysis were performed to measure these parameters. The Narrowband Power Amplifier is designed using BSIM model transistor, and the Broadband Power Amplifier for high power Application is designed using RF3931 GaN-HEMT and for low Power Application SXA3489BZ is used. From the DC Analysis of Narrowband and Broadband Power Amplifier, the Power consumption is measured. It is found to be 0.010W with 1V power supply for BSIM model Transistor, with 48V supply for RF3931 GaN HEMT power consumption is measured. An efficiency of 55% is calculated and plotted for Narrowband Design. for broadband power Amplifier design 65% drain efficiency is calculated for RF3931and 43%for SXA389BZ.The maximum gain is 18.5 db at 850MHz for SXA389BZ,and the maximum gain 15dB at 2GHz for RF3931. The layout is generated for both RF3931 and SXA389BZ power amplifier designs. As the broadband is fulfilling the narrowband range only the broadband power amplifier design is taken for fabrication. REFERENCES 1. Acar.M, A,Annema.J, and Nauta.B,(2007) Analytical design equations for class-e power amplifiers IEEE Trans. Circuits Syst. I, Reg. Papers,vol. 54, no. 12, pp Jun Tan,Chun-Huat Heng,Yong Lian(2012) Design of Efficient Class-E Power Amplifiers for Short Distance Communications IEEE Trans.Circuits Syst..Ivol.59 no.4 pp Kazimierczuk.M.K and Puczko.K(1987) Analysis of Class E Tuned Power Amplifier at any Q and Switch Duty Cycle, IEEE Transactions on Circuits and Systems, Vol CAS-34, No. 2, pp Kenle Chen,Dimiritios Peroulis(2011) Design of Highly Efficient Broadband Class-E Power Amplifier using Synthesized Low-Pass Matching Network IEEE Trans on Microwave Theory vol 59 no 12 pp no Kuieu-Cheng Lin,Hwann-Kaeo Chiou,Po-Chang Wu,Wei-Hsien Chen(2013) 2.4-GHz Complementary Metal Oxide Semiconductor Power Amplifier using High Quality Factor Wafer-Level Bondwire Spiral Inductor IEEE Trans.Components Packaging and manufacturing tech..vol.3 no.8 pp Lee.O, An.K.H, and Kim et al.h.,(2010) Analysis and design of fully integrated high-power parallel-circuit class-e CMOS power amplifiers, IEEE Trans.Circuits Syst. I, Reg. Papers, vol. 57, no. 3, pp Raab. F.H(1977), Idealized operation of the class E tuned power amplifier, (1977)IEEE Trans. Circuits Syst., vol. CAS-24, no. 12, pp Sokal.N.O and Sokal A.D(1975), Class E A Class of High-Efficiency Tuned Single-Ended Switching Power Amplifiers, IEEE Journal of Solid-State Circuits, Vol SC-10, No. 3, pp Zhisheng Li, Guy Torfs, Johan Bau Jan Vandewege welinck, Christophe Van Praet, Peter Spiessens, Huub Tubbax, and Frederic Stubbe (2012) A 2.45-GHz +20- dbm Fast Switching Class-EPower Amplifier With 43% PAE and a 18-dBWidePower Range in 0.18-μm CMOS IEEE Trans.Circuits Syst..II vol.59 no.4 pp Saad.P, Nemati. H. M., Thorsell. M, Andersson.K, and Fager.C(2009), An inverse class-f GaN HEMT power amplifier with 78% PAE at 3.5 GHz, in 39rd Eur. Microw. Conf, vol. 1, pp Pengelly.R, B. Millon, D. Farrell, B. Pribble, and S. Wood(2008), Application of non-linear models in a range of challenging GaN HEMT power amplifier designs, presented at the IEEE MTTS Int. Microw. Symp.Workshop. 11. P.Sambath,K.Gunavathi(2012) Class-E Power Amplifier and its Linearization using Analog Predistortion Indian Journal on Engineering Material Science,vol.19,pp George D.Vendelin,Anthony M.Pavio Ulrich(2011) Microwave Circuit Design Using Linear and Non Linear Techniques 2nd edition Wiley. 13. Munir M. El-Desouki,M.Jamal Deen,Yaser(2008) A Low Power CMOS Class- E Power Amplifier for Biotemetry Applications 14. D. M. Pozar, Microwave Engineering, 3rd ed. Boston, MA: Wiley, application