Design And Analysis Of 1MHz Class-E Power Amplifier - WSEAS
WSEAS TRANSACTIONS on CIRCUITS and SYSTEMSY. Yusmarnita, Shakir Saat, A. H. Hamidon, Huzaimah HusinNorezmi Jamal, Kamarudin Kh., Imran Hindustan, Sing Kiong NguangDesign and Analysis of 1MHz Class-E Power AmplifierY.YUSMARNITA, SHAKIR SAAT, A.H. HAMIDON, HUZAIMAH HUSIN, NOREZMI JAMAL,KAMARUDIN.KH, IMRAN HINDUSTANFaculty of Electronics & Computer EngineeringUniversiti Teknikal Malaysia MelakaMELAKA, MALAYSIAyusmarnita@utem.edu.mySING KIONG NGUANGDepartment of Electrical & Computer EngineeringThe University of AucklandNEW ZEALANDsk.nguang@auckland.ac.nzAbstract—This paper presents the simulation and experimental of Class-E power amplifier which consists of a loadnetwork and a single transistor that is operated as a switch at the carrier frequency of the output signal. Class-Epower amplifier is often used in designing a high frequency ac power source because its ability to perform theconversion efficiently even when working at high frequencies with significant reduction in switching losses. In thispaper, a 10W Class-E power amplifier is designed, constructed, and tested in the laboratory with operating frequencyof 1 MHz. To be specific, SK40C microcontroller board with PIC16F877A is used to generate a pulse widthmodulation (PWM) switching signal to drive the IRF510 MOSFET. This paper focuses on studying the effect ofswitching and performance analysis of the Class-E power amplifier behavior at 1MHz frequency. The performanceparameter relationship of Class-E power amplifier were observed and analyzed. The theoretical calculations,simulation and experimental results at optimum operation using selected component values are compared andpresented. The final result shows an output power is 9.45W with a drain efficiency of 98.44%. It is also shown thatboth simulation and experimental agree well with theoretical predictions.Key-Words:- Class-E power amplifier, zero voltage switching, high frequency power amplifierinclude a lowpass or bandpass filter to suppressharmonics of the switching frequency at the load, andmay transform the load impedance and accommodateload reactance.1 IntroductionClass-E switch-mode tuned power amplifier offersextremely high dc to ac conversion efficiency at highfrequencies because of a significant reduction inswitching losses [1-15]. The main idea behind switchmode power amplifier technology is to operate thetransistor in saturation, so that either voltage or current,depending on amplifier class, is switched on and off.Figure 1 is a block diagram of a single-ended switchmode amplifier. The active device acts substantially as aswitch when appropriately driven by the driver. Theactive device output is represented as a non-ideal singlepole single-throw switch. As the switch is periodicallyoperated at the desired ac output frequency, dc energyfrom the power supply is converted to ac energy at theswitching frequency. To obtain maximum fundamentalfrequency output, the switch duty ratio is madeapproximately 50 percent. To be specific, the switch is“on” for approximately half of the ac period and “off”for the remainder of the period. The load network mayE-ISSN: NetworkLoadFig. 1. Single-ended Switch-mode Power AmplifierBlock Diagram372Volume 14, 2015
WSEAS TRANSACTIONS on CIRCUITS and SYSTEMSY. Yusmarnita, Shakir Saat, A. H. Hamidon, Huzaimah HusinNorezmi Jamal, Kamarudin Kh., Imran Hindustan, Sing Kiong Nguang(a)(b)Fig. 4. Operating points for Classes A, B, AB, and C [8]fundamental voltage is generated over the resonator. Aflywheel effect is created generating sinusoidal voltageand current in the load. Theoretically, the two necessaryconditions for generating a single tone with 100%efficiency in the load are: 1. zero overlap betweenvoltage over the transistor channel and current throughthe channel; 2. blocking of harmonic currents to theload.This paper presents the design and theoretical analysisfor optimum operation of the Class-E power amplifier.The effect of circuit performance when the frequency,load and duty cycle varied were analyzed. The structureof this paper is arranged as follows. Section 2 explainsbriefly the Class-E power amplifier circuit operation.Section 3 discusses analysis of Class-E power amplifierat optimum operation. Simulation and experimentalresults are shown in Section 4 to verify the theoreticalanalysis and final conclusions are drawn in Section 5.Fig. 2. (a) Current Through Switch (b) VoltageAcross Switch [8]LFLIdcVdcCio icVgQCpRLvo-Fig. 3. Typical Class-E Power Amplifier CircuitFigure 2(a) and (b) shows the desired current andvoltage waveforms of the switch for maximum powerefficiency. When the switch is open, only voltage ispresent over the transistor. When closed, current flowsthrough it. Since there is no overlap in time betweenvoltage and current, power is not dissipated and oneobtains 100% theoretical efficiency. This phenomenonis known as zero voltage switching (ZVS). According topaper [4-16], Class E power amplifier is the mostefficient inverter so far because of its ability to achieveZVS.Class-E amplification is easily differentiated fromother classes of power amplification. For example, asshown in Figure 4, linear power amplifier in which thetransistors is biased to an operating point and followsthe indicated load line, giving rise to powerdissipation and loss. In the switch-mode poweramplifier such as Class-E, an output resonator helpsshapethewaveform byblockingharmoniccomponents of the voltage and current. It keeps thesecomponents from reaching the load. Consequently, onlyfundamental current is passed to the load and onlyE-ISSN: 2224-266X2 Circuit DescriptionThe basic circuit of the Class-E power amplifier isshown in Figure 3. The details about the operatingprinciple of Class-E power amplifier are omitted sincethey had appeared in [4]. This inverter is normally usedfor fixed output voltage. However, the output voltagecan be varied by varying the switching frequency or dutycycle. The Class-E power amplifier consists of powerMOSFET operated as a switch at the input frequency,and a load network which includes a shunt capacitor, anda series-resonant RL-L-C output circuit. The chokeinductor Lf is usually high enough to force a dc current,Idc. To achieve ZVS condition, the operating frequencyshould be between the resonant frequencies fo1 and fo2given in (1) and (2), where total equivalent capacitanceCT CCp/(C Cp).1(1)f o1 2π LC373Volume 14, 2015
Y. Yusmarnita, Shakir Saat, A. H. Hamidon, Huzaimah HusinNorezmi Jamal, Kamarudin Kh., Imran Hindustan, Sing Kiong NguangWSEAS TRANSACTIONS on CIRCUITS and SYSTEMSfo2 12π LCTcurrent I through the resonant circuit issinusoidal.4) All circuit elements are ideal.(2)The switch turns on and off at the operating frequencysetting by a MOSFET gate drive. Circuit operation isdetermined by the switch when on and by the transientresponse of the load network when the switch is off. Theswitching pattern is defined as:The assumptions of the circuit are quite similar to thosepresented in [2, 3].3.2 Parameters for Optimum OperationThe parameters of the circuit shown in Figure 3 aredefined as follows. The derivation of all parameters arebased on the papers of [1-5] and [8]. These parametersare usually required to satisfy equations (3) and (4).OFF state for 0 ω𝑡𝑡 𝐷𝐷𝐷𝐷Switch ON state for DT ω𝑡𝑡 𝑇𝑇where D is the switch duty cycle and T is the period forone complete cycle. The MOSFET turns on and offalternately at ω𝑡𝑡 0 and DT. Therefore, the switchvoltage waveform that satisfies the Class-E nominalcondition, i.e., condition (3) or (4), at the switch turn oninstant as a function of the duty ratio, is expressed asfollows:The full load resistance isIo 𝑣𝑣𝑑𝑑𝑑𝑑 (𝐷𝐷𝐷𝐷) 0(3)Cp 0PVCCIoωπVCC(6) ωR Q L 1, π 2 π 1 ωR 4 21C where 𝑣𝑣𝑑𝑑𝑑𝑑 (ω𝑡𝑡) is the switch voltage. In this case, thevoltage 𝑣𝑣𝑑𝑑𝑑𝑑 across the switch and the shunt capacitanceCp is zero when the switch turns on. Therefore, theenergy stored in the shunt capacitance Cp is zero whenthe switch turns on, yielding zero turn on switching loss.QRLωπ (π 4) 216(7)(8), (9).where Q is quality factor. In order to keep the currentripple in the choke inductor stays at below 10% of thefull-load DC input current Idc, the value of the chokeinductance must be greater than3 Analysis of Class-E Power Amplifier3.1 Assumptions π 2 R 1 .L f (min) 2 4 fThe analysis of the Class-E power amplifier shown inFigure 3 is carried out under the following assumptions:1) The MOSFET and diode form an ideal switchwhose on-resistance is zero, off-resistance isinfinity, and switching times are zero.2) The choke inductance is high enough so that itsac component is much lower than the DCcomponent of the input current.3) The loaded quality factor Q of the L,C and RLseries-resonant circuit is high enough so that theE-ISSN: 2224-266X(5))The component values for the load network are asfollows:Zero derivative switching (ZDS) condition:(4)(The current drawn from the DC power supply isZVS condition:𝑑𝑑𝑣𝑣𝑑𝑑𝑑𝑑 (ω𝑡𝑡) 𝑑𝑑(ω𝑡𝑡) ω𝑡𝑡 𝐷𝐷𝐷𝐷28VCC.2π 4 PRL (10)In practical terms, the choke inductance value is not allthat critical, as long as its impedance is at least an orderof magnitude higher than the load resistance and it is notself-resonant at the first three or four harmonics. It needsto look like an open circuit to these harmonics, ifpossible.374Volume 14, 2015
WSEAS TRANSACTIONS on CIRCUITS and SYSTEMSY. Yusmarnita, Shakir Saat, A. H. Hamidon, Huzaimah HusinNorezmi Jamal, Kamarudin Kh., Imran Hindustan, Sing Kiong Nguang4 Analysis of Class-E Power AmplifierThe design specifications used here are as follows; dcpower supply, Vcc 12V, operating frequency, f 1MHz, D 0.5, Q 10, and Output power, Po 10W.Based on the design specifications and assumptionsprovided in Section 3, all the circuit parameters arecalculated and tabulated as in Table I. Then simulationsare carried out using Proteus before the real circuit isimplemented. In order to validate the simulation results,the experimental work is carried out. IRF510 MOSFETis used as a switching device in the design. It is nchannel, enhancement mode and designed especially forhigh speed applications. Based on Table I, the peakswitch voltage and current are 42.74V and 2.39Arespectively. This confirms that the IRF510 MOSFET issuitable to be used in a class-E power amplifier circuit inpractical applications.In Figure 5, the SK40C microcontroller board withPIC16F887A is used to generate a 1MHz switchingcontrol signal at 50% duty cycle for the MOSFET gate.However, the microcontroller output voltage, typically.5V is not sufficient to turn on the IRF510 MOSFET thatrequires at least 10V to operate in safe operating area.Therefore, an IC gate drive TC4422 is used here toprovide sufficient gate voltage or charge to drive theIRF510 MOSFET.Fig. 5. Class-E Power Amplifier Experiment CircuitTABLE ICLASS-E POWER AMPLIFIER k)VIs(peak)AIs(rms)APo(ac)WPi(dc)Wη%4.1 Simulation ResultsFigure 6 shows the waveforms that are obtained from theProteus simulation. Based on the result in Figure 6(a),the maximum voltage across the MOSFET and the shuntcapacitor during turn off is; Vds(peak) 42.74V, which isalmost three times larger than VCC. Meanwhile, duringturn on, Vds(peak) 4.5V, nearly 11% of the peak voltage.In an optimum design yielding the maximum drainefficiency, the switch voltage Vds at the switch turn ontime is usually 10% to 50% of the peak switch voltage,which is a nonzero voltage switching condition. Inaddition, the switch voltage derivative at the switch turnon time is zero or slightly positive or negative.Refer to Figure 6(b), the switch current, Ids(peak) 2.1A, increases gradually from zero after the switch isturned on because the derivative of vds is zero at the timethe switch turns on. It should be noted that both theswitch voltage and the switch current are positive foroptimum operation. Therefore, there is no need to addany antiparallel diode to the switch [8].According to the theoretical predictions, the outputvoltage is sinusoidal at high Q. Based on Figure 6(c),peak output voltage, VRL(peak) 12V at optimum load.E-ISSN: 740.831.55114.03128.892.391.2810.0110.00100.12f 1MHzSimulatedDifference 1.951.261.638.6213.919.841.5687.5612.55The dc power input is Pi(dc) VCCx Idc 9.84W. Thesimulated ac output power, Po(ac) (IRL(rms))2*RL 8.62W. In term of efficiency, it can be seen that thecircuit produces 86.16% efficiency, slightly lower thanthe calculated value. However, the simulation results areconsistent with the theoretical ones. In fact, all results inthis section show that the Class-E power amplifiercircuit satisfies the ZVS conditions since there is nooverlap in time between voltage and current.4.2 Experimental Result375Volume 14, 2015
WSEAS TRANSACTIONS on CIRCUITS and SYSTEMSY. Yusmarnita, Shakir Saat, A. H. Hamidon, Huzaimah HusinNorezmi Jamal, Kamarudin Kh., Imran Hindustan, Sing Kiong Nguang(a)Figure 7: 1MHz Class-E Power AmplifierExperimental SetupTABLE IICLASS-E POWER AMPLIFIER )Po(ac)Pi(dc)η(b)(c)Fig. 6. PROTEUS Simulation Waveforms: (a) Vgsversus Vds at 1MHz (b) Vgs versus Ids at 1MHz (c)Output Voltage Across RL,VRLnFnFuHuHVVAAVVAAWW%f 83.961.6812.5V which is 3.02% lower than the theoretical value.The experimental value for the maximum voltage acrossthe MOSFET during turn off is Vds(peak) 39V which is8.76% lower than the theoretical ones. The dc powerinput is Pi(dc) VCCx Idc 9.6W. The simulated ac outputpower, Po(ac) (IRL(rms))2*RL 9.45W. In term ofefficiency, it can be seen that the circuit produces98.44% efficiency, which is 1.68% lower than thetheoretical value. There are slightly differences inFigure 7 shows the experimental setup for 1MHz ClassE power amplifier circuit. Figure 8 shows the waveformsobtained from laboratory circuit experiment. As shownin Table II, the measured value of the resonancefrequency is 1MHz. The peak output voltage, VRL(peak) E-ISSN: 76Volume 14, 2015
WSEAS TRANSACTIONS on CIRCUITS and SYSTEMSY. Yusmarnita, Shakir Saat, A. H. Hamidon, Huzaimah HusinNorezmi Jamal, Kamarudin Kh., Imran Hindustan, Sing Kiong Nguangmeasured values due to power amplifier losses occurredin the parasitic resistance of each components, switchinglosses and dissimilarity in component selections.The main reason is the difference value of the loadresistor. Since the measured load resistance is 1.08%greater than the theoretical value (RL Ropt), the optimumoperation cannot be achieved. The amplitude of thecurrent through the L-C R series resonant circuit is lowerthan that optimum operation, the voltage drop across theshunt capacitor Cp decreases, and the switch voltage Vdsis not equal to zero at turn on. Therefore, the energystored in Cp is dissipated in the MOSFET as heat afterthe switch is turned on, resulting in a turn on switchingloss.In order to transfer a specified amount of power Po at aspecified dc voltage Vdc, the load resistance must be ofthe value determined by equation (5). However, inpractical applications, the load resistance is available andsometimes the practical load resistance is different fromthat given in (5). Therefore, there is a need for matchingcircuits that could provide impedance transformation inorder to increase the measured efficiency near to thetheoretical ones. This impedance transformation can beaccomplished by tapping the resonant capacitance,parallel with the load resistor. Furthermore, in realcircuit implementation, a transistor is not an idealswitch, overlap does happened and limit the efficiency.5 ConclusionAn analysis of the Class-E power amplifier operationhas been presented in this paper. The switch controlsignal for IRF510 MOSFET using microcontrollerPIC16877A has been proposed and the results indicatethat ZVS condition can be achieved successfully. In thelaboratory experiment, the Class-E with optimum loadand D 0.5 has achieved 98.44% efficiency at 10Woutput power for 1MHz operating frequency. Moreover,the agreement between experiment performance andtheoretical performance can still be considered excellent.For future development, this circuit will be applied at thetransmitter side of a capacitive power transfer system.Acknowledgement:Sincerely to express appreciation to Universiti TeknikalMalaysia Melaka (UTeM) for funding this research workunder PJP/2014/FKEKK (2A) /S01299 grant.References:[1] Sokal, N. O., & Sokal, A. D. (1975). Class-E-A newclass of high-efficiency tuned single-endedswitching power amplifiers. IEEE Journal of C.1975.1050582[2] Raab, F. (1977). Idealized operation of the Class-Etuned power amplifier. IEEE Transactions TCS.1977.1084296[3] Raab, F. H. (1978). Effects of circuit variations onthe Class-E tuned power amplifier. IEEE Journal JSSC.1978.1051026[4] Kazimierczuk, M., & Puczko, K. (1987). Exactanalysis of Class-E tuned power amplifier at any Qand switch duty cycle. IEEE Transactions CS.1987.1086114[5] Kazimierczuk, M. K., Member, S., & Wang, S.(1992). Frequency-Domain Analysis of SeriesResonant Converter for Continuous ConductionMode, 7(2).(a)(b)Fig. 8: Experimental Waveforms: (a) Switch Voltage andGate Voltage (b) Output Voltage Across RL,VRLE-ISSN: 2224-266X377Volume 14, 2015
WSEAS TRANSACTIONS on CIRCUITS and SYSTEMSY. Yusmarnita, Shakir Saat, A. H. Hamidon, Huzaimah HusinNorezmi Jamal, Kamarudin Kh., Imran Hindustan, Sing Kiong Nguang[6] Zulinski, R. E. (1986). Frequency Multipliers VdCb.[7] Zulinski, R. E. (1987). Class E Power Amplifiersand Frequency Multipliers with Finite DC-FeedInductance, (8715397).[8] Kazimierczuk, M.K., and Czarkowski, D.: Resonantpower converters (Wiley, New York, 2011)[9] Mohsen Hayati, Ali Lotfi, Marian K. Kazimierczuk,Hiroo Sekiy. Analysis and Design of Class-E PowerAmplifier with MOSFET Parasitic Linear andNonlinear Capacitances at Any Duty Ratio. IEEETransactions On Power Electronics, Vol. 28, No. 11,November 2013[10] T. Suetsugu and M.Kazimierczuk. Design procedurefor lossless voltage clamped class E amplifier with atransformer and a diode. IEEE Trans. PowerElectron, vol. 20, no. 1, pp. 56–64, Jan. 2005.[11] T. Suetsugu and M. Kazimierczuk. Off-nominaloperation of class-E amplifier at any duty ratio.IEEE Trans. Circuits Syst. I, Reg. Papers, vol. 54,no. 6, pp. 1389–1397, Jun. 2007.[12] T. Suetsugu and M. Kazimierczuk. Designprocedure of class-E amplifier for off-nominaloperation at 50% duty ratio. IEEE Trans. CircuitsSyst. I, Reg. Papers, vol. 53, no. 7, pp. 1468–1476,Jul. 2006.[13] Norezmi Jamal, Shakir Saat & Y. Yusmarnita(2014). A Development of Class E Converter Circuitfor Loosely Coupled Inductive Power TransferSystem. WSEAS Transactions on Circuits andSystems. ISSN:11092734[14] Norezmi Jamal, Shakir Saat, Noorazma Azman &Thoriq Zaid (2014). The Experimental Analysis ofClass E Converter Circuit for Inductive um on Technology Management andEmerging Technology (ISTMET)[15] T. Suetsugu and M. K. Kazimierczuk. Maximumoperating frequency of class-E amplifier at any dutyratio. IEEE Trans. Circuits Syst. II, Exp. Briefs, vol.55, no. 8, pp. 768–770, Aug. 2008.[16] Kamarudin.Kh, Shakir Saat, Y.Yusmarnita,Norezmi Jamal. Analysis and Design of WirelessPower Transfer: A Capacitive Based Method forLow Power Applications. WSEAS Transactions onCircuits and Systems. ISSN / E-ISSN: 1109-2734 /2224-266X, Volume 14, 2015, Art. #26, pp. 221-229E-ISSN: 2224-266X378Volume 14, 2015
Class-E switch-mode tuned power amplifier offers extremely high dc to ac conversion efficiency at high frequencies because of a significant reduction in switching losses [1-15]. The main idea behind switch-mode power amplifier technology is to operate the transistor in saturation, so that either voltage or current, depending on amplifier class .
f ANALOG MODULATION A built-in DDS generator provides Sine, Square, Ramp, -Ramp and triangle waves, these can be applied in the forms of AM, FM or PM from the internal modulation source; at frequencies ranging from 1mHz to 1MHz. External analog modulation signals can be applied to the carri
Dielectric Strength volts / mil ASTM D149 4,700 6,000 5,500 4,600 7,000 Dielectric Constant, 1 Khz ASTM D150 2.1 2.6 — 3.3 @ 1MHz Dissipation Factor, 1 Khz ASTM D150 0.0005 0.0003 0.0008 0.005 — 0.004 @ 1MHz Surface Resistivity ohm/sq ASTM D257 1x1015 15 1x1015 14 — 1x1016 OPTICAL PROPERTIES Refractive Index ASTM D542 1.35 1.4 — 1.7
The receiver is a zero intermediate frequency (IF) type which makes channel filtering possible with low-power, integrated low-pass filters. The receiver consists of a low . 1MHz ,19.2kbps, Main filter cut off frequency 133kHz 56 dB Offset 1MHz 61 dB Offset 2MHz 58 dB Offset 5MHz 46 dB Offset 10MHz 62 dB Blocking Desired signal: 19.2 .
ment. This manual is meant to give a brief introduc-tion to the Theratouch 7.7 and its operation. The Theratouch 7.7 offers six stimulation wave-forms and two different sized ultrasound transduc-ers with the ability to output both 1MHz and 3MHz. The Theratouch 7.7 touch-screen and director
Silicon conventional planar devices range from voltages under 100V to greater than 1000V, . (Switch Mode Power Supplies (SMPS)) and in . controlled devices such as MOSFETs that have the ability operate at high frequencies (up to and over 1MHz). Another disadvantage of conductivity modulated device such as IGBT is the 0.7V voltage knee .
P L(1dB) output power at 1 dB gain compression 2110 MHz d f d 2170 MHz 27 30 - dBm IP3 O output third-order intercept point 2110 MHz d f d 2170 MHz; P L 19 dBm per tone; tonespacing 1MHz 40.5 44 - dBm Table 1. Quick reference data continued 4.75 V d V SUP d 5.25 V; 40 qC d T case d 85 qC; P i 20 dBm; R3 523 : (tolerance 1 %); input and .
The buck converters support 100% duty cycle operation to extend battery life. If the PWM pin is held low, the buck converters automatically transition from Burst Mode operation to PWM mode at high loads. With the PWM pin held high, the buck converters remain in low noise, 1.1MHz PWM mode. The buc
used when setting windows hooks . Relies on the user api hook . User Api Hook Special hooking mechanism introduced to support Windows themes RegisterUserApiHook Can only be registered by privileged processes Requires the TCB privilege Caller must be running as SYSTEM Allows Windows to load a theme client module into every GUI application . Smashing the Atom . Theme Subsystem Introduced in .
