Design and validation of a methodology for the characterization of RF power amplifiers

Texto completo

(1)UNIVERSIDAD POLITÉCNICA DE MADRID ESCUELA TÉCNICA SUPERIOR DE INGENIEROS DE TELECOMUNICACIÓN. TRABAJO FIN DE GRADO. TITULO: DESIGN AND VALIDATION OF A METHODOLOGY FOR THE CHARACTERIZATION OF RF POWER AMPLIFIERS. AUTOR: Alberto de la Escalera Dı́az. AÑO: 2019 DEPARTAMENTO DE SISTEMAS, SEÑALES Y RADIOCOMUNICACIONES. Escuela Técnica Superior de Ingenieros de Telecomunicación Universidad Politécnica de Madrid.

(2) 2. DESIGN AND VALIDATION OF A METHODOLOGY FOR THE CHARACTERIZATION OF RF POWER AMPLIFIERS. TRABAJO FIN DE GRADO. Autor:. Alberto de la Escalera Dı́az.

(3) TÍTULO:. Design and validation of a methodology for the characterization of RF power amplifiers. AUTOR:. Alberto de la Escalera Dı́az. TUTOR:. Jesús Grajal de la Fuente. DEPARTAMENTO: Señales, Sistemas y Radiocomunicaciones MIEMBROS DEL TRIBUNAL CALIFICADOR PRESIDENTE: VOCAL: SECRETARIO: SUPLENTE: FECHA DE LECTURA: CALIFICACIÓN:. RESUMEN: El diseño de circuitos y sistemas de microondas a partir de componentes y dispositivos comerciales exige conocer con exactitud su comportamiento en las bandas de frecuencia y temperatura en las que estos van a funcionar. El objetivo de este Trabajo Fin de Grado es la caracterización electro-térmica de dispositivos activos en la banda de microondas. La caracterización será tanto lineal a través de parámetros S como no lineal: conversión AM-AM y AM-PM, caracterización con dos tonos y caracterización multi-tono. Esta caracterización exhaustiva es básica para el diseño óptimo de subsistemas que utilicen estos componentes. En una primera fase del trabajo, se hará una breve introducción al dispositivo que se van a caracterizar, en nuestro caso un amplificador de potencia (HPA); ası́ como la metodologı́a empleada para su caracterización. En una fase posterior, se realizará la caracterización del dispositivo elegido con los instrumentos de medida correspondientes..

(4) 4. SUMMARY: Microwave component manufacturers provide in their datasheets the average performance of their components. To design new devices from these components, it is necessary to know exactly their behavior in the frequency bands and the temperature range at which they will work. The objective of this Bachelor Thesis is to elaborate a series of measurements in order to achieve the electro-thermal characterization of active devices in the microwave band. The characterization will be both linear, through S parameters, and non-linear: AM-AM and AM-PM conversion, characterization with two tones and multi-tone characterization. This exhaustive characterization is essential for the optimal design of the systems that use these components. In a first phase of the work, the brief introduction will be made to the device to be characterized, in our case a power amplifier (HPA); as well as the different measures that can be made to this device. Then, in a later phase, those measures previously presented, will be performed to an actual device. The final phase will present a proper methodology for characterizing RF power amplifiers designed regarding the information given by the previous measurements.. PALABRAS CLAVE: Characterization, Amplifier, Thermal characterization, Electrical characterization, S parameters, Noise Figure, linear characterization, non-linear characterization, M-IMR, ACPR, NPR, Multi tone..

(5) AGRADECIMIENTOS: A mis niños y a mis niñas, que han hecho que estos ”cuatro años” se me hayan pasado como si fueran 10. Quisiera agradecer a Jesús su paciencia infinita y al resto de la Escuela, tanto profesores como personal, el haber hecho que haya estado como en casa durante tanto tiempo..

(6) Contents 1 Introduction. 1. 2 Methodology for the characterization. 3. 3 EMM5074VU amplifier testing 3.1. 11. Electrical characterization . . . . . . . . . . . . . . . . . . . . . . . . . .. 12. 3.1.1. Linear Characterization . . . . . . . . . . . . . . . . . . . . . . .. 12. S Parameters . . . . . . . . . . . . . . . . . . . . . . . . . . . . .. 12. Noise Figure . . . . . . . . . . . . . . . . . . . . . . . . . . . . . .. 14. Non-Linear Characterization . . . . . . . . . . . . . . . . . . . . .. 17. One Tone . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . .. 17. Two Tones . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . .. 21. Multitone . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . .. 23. 3.2. Thermal Characterization . . . . . . . . . . . . . . . . . . . . . . . . . .. 27. 3.3. Summary . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . .. 33. 3.1.2. 3.1.3. 4 Conclusions. 34. A Synthesis of multitone signals. 35. B Project impact. 38. B.1 Ethical impact. . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . .. 38. B.2 Economical impact . . . . . . . . . . . . . . . . . . . . . . . . . . . . . .. 38. B.3 Social impact . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . .. 39. B.4 Environmental impact . . . . . . . . . . . . . . . . . . . . . . . . . . . .. 39.

(7) CONTENTS. C Estimated project budget. ii. 40. C.1 Necessary laboratory devices . . . . . . . . . . . . . . . . . . . . . . . . .. 40. C.2 Outsourcing characterization . . . . . . . . . . . . . . . . . . . . . . . . .. 41. D EMM5074VU Amplifier Datasheet. 42.

(8) List of Figures 2.1. Energy balance in an electronic amplifier . . . . . . . . . . . . . . . . . .. 4. 2.2. Illustration of a AM-AM test . . . . . . . . . . . . . . . . . . . . . . . .. 6. 2.3. Illustration of a AM-PM test . . . . . . . . . . . . . . . . . . . . . . . .. 7. 2.4. Illustration of the IP3 test with the Fundamental power curve and the IMD power curve . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . .. 8. 2.5. Illustration of multitone intermodulation ratio definition . . . . . . . . .. 8. 2.6. Illustration of adjacent channel power ratio definition . . . . . . . . . . .. 9. 2.7. Illustration of a noise power ratio test. . . . . . . . . . . . . . . . . . . .. 10. 3.1. EMM5074VU amplifier . . . . . . . . . . . . . . . . . . . . . . . . . . . .. 11. 3.2. Assembly for the S parameters measurement . . . . . . . . . . . . . . . .. 12. 3.3. Comparison between measured S parameters and manufacture S parameters 13. 3.4. laboratory assembly scheme for Noise Figure measurement . . . . . . . .. 14. 3.5. Laboratory assembly for the Noise Figure measurement and close-up to the Noise Source . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . .. 14. 3.6. Noise Figure Measurements and calculations . . . . . . . . . . . . . . . .. 15. 3.7. Noise Figure from 5,8 to 8,5 GHz . . . . . . . . . . . . . . . . . . . . . .. 16. 3.8. Attenuator’s S parameters from 5,8 to 8,5 GHz . . . . . . . . . . . . . .. 17. 3.9. Gain Compression curve at different frequencies and comparison with data provided by the manufacturer . . . . . . . . . . . . . . . . . . . . .. 18. Output power Vs Input power curve at different frequencies and comparison with data provided by the manufacturer . . . . . . . . . . . . . . . .. 19. Mean 1dB Compression curve from previous frequencies and comparison with data provided by the manufacturer . . . . . . . . . . . . . . . . . .. 19. AM-PM curve at different frequencies . . . . . . . . . . . . . . . . . . . .. 20. 3.10 3.11 3.12.

(9) LIST OF FIGURES. iv. 3.13. Variation of the AM-PM conversion curve at different frequencies . . . .. 20. 3.14. Laboratory assembly for two tones and multi-tones measurements and close-up to the Noise Source . . . . . . . . . . . . . . . . . . . . . . . . .. 21. 3.15. Measures to acquire the IP3 points at different frequencies . . . . . . . .. 21. 3.16. IP3 points and interpolated curve from 5,8 to 8,5 GHz . . . . . . . . . .. 22. 3.17. Intermodulation ratio depending on the output power and comparison with data provided by the manufacturer . . . . . . . . . . . . . . . . . .. 22. 3.18. Distortion depending on the number of tones. . . . . . . . . . . . . . . .. 23. 3.19. Multitone Input Signals . . . . . . . . . . . . . . . . . . . . . . . . . . .. 23. 3.20. Multitone output Signals . . . . . . . . . . . . . . . . . . . . . . . . . . .. 24. 3.21. M-IMR with 29 tones from -44 dBm per tone to -29 dBm per tone. . . .. 24. 3.22. ACPR with 29 tones from -44 dBm per tone to -29 dB per tone . . . . .. 25. 3.23. ACPR close-up at 7,5 GHz from -44 dBm per tone to -29 dB per tone . .. 25. 3.24. NPR with 30 tones from -44 dBm per tone to -29 dBm per tone at different frequencies . . . . . . . . . . . . . . . . . . . . . . . . . . . . .. 26. 3.25. MIMR, ACPR and NPR measurements comparison at 7,5 GHz . . . . .. 26. 3.26. Thermal Chamber . . . . . . . . . . . . . . . . . . . . . . . . . . . . . .. 27. 3.27. Input power versus Output power at different temperatures. . . . . . . .. 28. 3.28. Gain Compression at different temperatures from -15 dBm to 5 dBm input power . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . .. 28. 3.29. Gain variation from -30◦ C to 60◦ C . . . . . . . . . . . . . . . . . . . . .. 29. 3.30. M-IMR at different temperatures from 29 tones from -37,5 dBm to -23,8 dBm per tone . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . .. 30. 3.31. M-IMR variation from 29 tones through temperatures from -30◦ C to 60◦ C 30. 3.32. ACPR at different temperatures from 29 tones from -37,5 dBm to -23,8 dBm per tone . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . .. 31. 3.33. ACPR variation from 29 tones through temperatures from -30◦ C to 60◦ Cs 31. 3.34. NPR at different temperatures from 30 tones from -37,5 dBm to -23,8 dBm per tone . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . .. 32. 3.35. NPR variation from 30 tones through temperatures from -30◦ C to 60◦ C .. 32. A.1. Multitone signal generation process . . . . . . . . . . . . . . . . . . . . .. 35. A.2. Example of phase discontinuity . . . . . . . . . . . . . . . . . . . . . . .. 36.

(10) LIST OF FIGURES. v. A.3. Signal generated by Matlab . . . . . . . . . . . . . . . . . . . . . . . . .. 37. A.4. Signal in the frequency domain with number of samples and sampling frequency not being multiples of each other . . . . . . . . . . . . . . . .. 37.

(11) List of Tables 3.1. Test’s Conditions Table . . . . . . . . . . . . . . . . . . . . . . . . . . . .. 12. 3.2. Test Results Table . . . . . . . . . . . . . . . . . . . . . . . . . . . . . .. 33.

(12) Chapter 1 Introduction In telecommunications, electromagnetic waves are the most used due to their accessible nature to be worked with and the perk of them not being harmful to humans. Many applications of the electromagnetic waves in the bandwidth between 300MHz and 30GHz exist due to the properties of these waves. Some of the reasons are: [1] The gain of the antennas is proportional to their electrical size. Since high frequency waves have smaller wavelength, smaller antennas are needed to have the same gain. Greater bandwidth can be achieved at higher frequencies. For example, a 1% bandwidth at 600 MHz is 6 MHz which can provide a data rate of about 6Mbps, while at 60 GHz a 1% bandwidth is 600 MHz, allowing a 600 Mbps data rate. Microwave signals are not bent by the ionosphere allowing satellite and terrestrial communications with very high capacities. The radar cross section of a radar target is usually proportional to the target’s electrical size. This fact, coupled with the frequency characteristics of the antenna gain, generally makes microwave frequencies preferred for radar systems.. Transmitters and receivers in the microwave band are generally built of a chain of monolithic microwave circuits (MMIC). Due to its many advantages, mentioned in [2] and [3], the use of this technology has spread in the research and industrial communities. Among the great variety of circuits that can be integrated by MMIC technology (mixers, diodes, oscillators. . . ), the present Bachelor Thesis focuses on the development of a methodology for the characterization of microwave amplifiers..

(13) 1. Introduction. 2. This characterization of microwave amplifier consists of an electrical characterization and a thermal characterization. The electrical one emphasizes the importance of predicting the distortion generated by nonlinear devices, since these disturbances generated by the nonlinearities deeply affect multi-carrier transmissions and broadband communication systems, which are new transmission methods currently used. Thermal characterization answers the need to know how these devices are going to behave in a particular temperature range..

(14) Chapter 2 Methodology for the characterization The need for power amplifiers in communications systems is very clear, both to increase signal levels from the near-rim values in receiver systems and to achieve high signal levels in the transmitters. In general, an amplifier is needed whenever an increase in the signal level is sought. Some characteristics that define the power amplifiers are the working frequency, the relative bandwidth, the maximum power levels, the noise figure and the transfer function. All of these characteristics should be given by the manufacturers in their data sheets. Power amplifiers have traditionally been characterized using power swept continues wave (CW) signals to find the AM/AM and AM/PM distortion and the 1 dB compression point, or simple two-tone measurements to find the third-order intercept point (IP3 [4]. The AM/AM distortion is the amplitude distortion as a function of the amplitude of an input CW signal and AM/PM is the phase distortion as a function of the same amplitude. This characterization is usually carried out with one frequency at a time and there is a strict one-to-one relationship between the input power and the AM/AM and AM/PM. These measurements are poor as a characterization tool of amplifiers, because non-linear systems subject to a sinusoid can only produce output spectral components that are harmonically related to the input frequency. To complement these previous tests, a series of multi-tone measurements will be presented. As is mention in [5] the following measures will be required to make a good characterization of an microwave amplifier:.

(15) 2. Methodology for the characterization. 4. Figure 2.1: Energy balance in an electronic amplifier Electrical characterization. Linear characterization: S parameters One of the most common techniques for characterizing linear behaviour of microwave components is the use of S parameters. While the parameters Z, Y and H are highly recommended at low frequencies, this type of approach is often problematic at high frequencies. This is due to the fact that at very low frequencies, the wavelength of the signal is much greater than the circuit elements. However, as the frequency increases, this wavelength will become smaller, so smaller that the Kirchhoff Laws cease their validity, because the phenomenon of propagation occurs along the circuit. At microwave frequencies, measuring current or voltage is very difficult unless there is a pair of clearly defined ports. This terminal may be present in TEM lines (coaxial, microstrip ...), but it will never exist in lines that do not support TEM modes (Rectangular wave guides, circular wave guides ...).[1] The S parameters are obtained by measuring the incident and reflected power waves at the input and output ports of the device. Equation 2.1 represents the matrix S of an n-port device, where the incident power waves correspond to a1 , a2 ... an , and the power waves reflected, with b1 , b2 , ..., bn . Thus it can be seen that the parameters Sii represent the output matching of the device when the rest of the ports are match with Z0 . On the other hand, parameter Sij represents the transmission from the j port to the i port, while the parameter Sji represents the transmission in the opposite direction. As mentioned above, to characterize the device through the parameters S, it is necessary to measure the relationship between the incident and reflected power waves. For its measurement, all ports, except for which power is injected, should be loaded with the characteristic impedance Z0 ..

(16) 2. Methodology for the characterization.    b1 S11 S12  b2   S21 S22    ·  · ·  = ·  · ·    ·  · · bn Sn1 Sn2. 5. · · · · · ·. · · · · · ·.    · S1n a1   · S2n   a2     · ·  · ·    · ·   · · ·  · · Snn an. (2.1). To perform this measurement, a vector network analyser is used. In order to be able to measure the S parameters with a networks analyser, calibration is required. This calibration eliminates system errors and compensates the losses and mismatches of the transmission lines needed to perform the measurement. The technique used in this Bachelor Thesis is known by the acronym SOLT (Sort Open Load Through) and consists of loading the ports with a short circuit, an open circuit, a standardized load and finally, to end the calibration, both ports are connected [6]. Linear characterization: Noise Figure Noise limits the sensitivity and dynamic range of the receivers. In microwave frequencies, the characteristic of the noise figure requires the measurement of its power. To characterize the noise figure there are different techniques, but the one selected is the factor Y. This method has been selected because it is the most common in microwave equipment. The noise figure is determined by two noise power measurements, N1 and N2 , corresponding to two source impedance temperatures (T0 , usually 300 K, and Th , greater than 1000K). Manufacturers specify temperature Th with the excess noise source EN R, defined as [7]: EN R =. Th − T0 T0. (2.2). Both noise measurements can be presented as: Tube Off → N1 = kG(ρsc )B(Tc + Te (ρsc )). (2.3). Tube On → N2 = kG(ρsh )B(Th + Te (ρsh )). (2.4). Where B, G, and Te stand for bandwidth, gain of the power amplifier and equivalent temperature respectively. ρsc and ρsh are the reflection coefficients of the tube in both states (on and off). Knowing that ρsc = ρsh , the noise figure can be calculated by: Th + Te N2 = (2.5) Y = T0 + Te N1 Th − Y Tc Te = (2.6) Y −1 EN R F = (2.7) Y −1.

(17) 2. Methodology for the characterization. 6. Non-Linear characterization: AM-AM This characterization describes the relationship between the output power and the input power for a fixed frequency. Therefore, this characterization has a particular importance in communication systems based on amplitude modulations [5]. Moreover, the AM-AM conversion also characterizes the compression or gain expansion of a non-linear device. Thus, it serves to evaluate a figure of merit of great importance: the compression point at 1 dB (Figure 2.2). This can be defined as the input or output power at which the signal obtained also at the output, has been compressed 1 dB, compared with the signal that would be obtained by extrapolating the linear behaviour in small signal of the device.. Figure 2.2: Illustration of a AM-AM test Non-Linear characterization: AM-PM One property of non-linear systems is that vector addition of the output fundamental with distortion components also determines a phase variation od the resultant output when the input level varies (Figure 2.3). These variations of the output signal phase might be interesting for phase modulation systems. AM-PM characterization consists of studying the variation of the output signal phase, with input signal amplitude changes for a constant frequency [5]..

(18) 2. Methodology for the characterization. 7. Figure 2.3: Illustration of a AM-PM test Non-Linear characterization: IP3 When a microwave signal is composed of signals at different frequencies, the output signal of a non-linear circuit will be composed of spurious [5]. If the non-linear circuit is excited by a two-tone signal like [8]: x(t) = Acos(w1 t) + Acos(w2 t). (2.8). the output would be given by y(t) =. ∞ X. AOr cos(ωr t + φOr ) where ωr = mω1 + nω2 ∀m, n ∈ Z. (2.9). r=1. which shows that the output would be composed of a very large number of mixing terms involving all possible combinations of ±ω1 and ±ω2 . The ones that are not m + n = 1 are called intermodulation products. The most important intermodulation products are those of the third order, 2ω1 −ω2 and 2ω2 − ω1 , given that they have the highest level of power and that are very close to the signals that generate them, being very difficult to filter them out. The third-order intercept point IP3 is a fictitious point that is obtained by extrapolating the slope of the fundamental frequency (ω1 or ω2 ) that increases 1 dB for each decibel that the input signal is increased, and extrapolating the slope of the intermodulation products, that increases 3 dB for each decibel. The point of intersection of both lines is the third order intercept point (Figure 2.4). The IP3 can be calculated theoretically by the expression: IP 3dB = P0dB +. P (ω1 ) 1 · 10 log( ) 2 P (2ω1 − ω2 ). (2.10).

(19) 2. Methodology for the characterization. 8. Figure 2.4: Illustration of the IP3 test with the Fundamental power curve and the IMD power curve. Non-linear characterization: M-IMR With a multitone input, the Multitone Intermodulation Ratio (Figure 2.5) defines the relationship between the power of one of the tones that form the desired signal, and the power of a component present in the adjacent higher or lower bands [5]. M − IM R(r) =. P0/T PL/U (ωr ). (2.11). being PL the power ot the tone at ωr1 and PU the power of the tone at ωr2 .. Figure 2.5: Illustration of multitone intermodulation ratio definition.

(20) 2. Methodology for the characterization. 9. Non-linear characterization: ACPR The Adjacent Channel Power Ratio measures the amount of power that is generated by the adjacent bands due to the non-linearity of the modulated signal [5]. It compares the total power in the desired band, P0 to the integrated power in an adjacent band, upper band (PU A ) or lower band (PLA ). These upper and lower bands have the same bandwidth as the desired band (Figure 2.6). R ωU 1 S0 (ω)dω P0 ωL2 Rω = R ωL2 ACP RT = (2.12) PLA + PU A S0 (ω)dω + ωUU12 S0 (ω)dω ωL1. Figure 2.6: Illustration of adjacent channel power ratio definition The ACPR can also be defined separately for the upper adjacent channel and the lower adjacent channel. In this case, only the integral of the corresponding adjacent channel appears in the denominator. R ωU 1 S0 (ω)dω P0 ω (2.13) = R ωL2 ACP RT = L2 PLA S0 (ω)dω ωL1 R ωU 1 S0 (ω)dω P0 ω ACP RT = = R ωL2 (2.14) U2 PU A S0 (ω)dω ωU 1 Non-linear characterization: NPR Noise power ratio (NPR) was proposed as an indirect means of characterizing cochannel distortion (Figure 2.7). The noise power ratio test eliminates some of the fundamental components from the zone where the test is to be made creating a ”notch”. Any frequency component, or power spectral density function, observed at the output within the notch position, constitutes spectral regrowth, and is thus the desired cochannel distortion [5]. S0 (ωT ) NP R = (2.15) Swd (ωT ).

(21) 2. Methodology for the characterization. 10. The numerator is the arithmetic mean power of the two tones closest to the eliminated bandwidth , while the denominator is the output power of the eliminated bandwidth.. Figure 2.7: Illustration of a noise power ratio test Thermal characterization Due to the different temperatures that the devices can face, it is important to know how they will perform in a wide range of temperatures [9]. Different important measures, like gain compression or the multitone measures, will be performed inside a thermal camera in order to see the variation of its performance with the temperature..

(22) Chapter 3 EMM5074VU amplifier testing. Figure 3.1: EMM5074VU amplifier The EMM5074VU amplifier (figure 3.1)has been selected for a complete characterization taking into account the previously defined linear and non-linear tests The EMM5074VU amplifier is, according to the datasheet provided by Sumitomo, a wide band three-stage power amplifier MMIC internally matched, for standard communications band in 5,8 to 8,5 GHz frequency range..

(23) 3. EMM5074VU amplifier testing. 12. All tests were made with the conditions of the following table 3.1 VDD. 6V. IDD. 1.216 A. VGG. starting from -1.2 V to 0.26 V to turn on the amplifier. Power Frequency Range. 5.8 - 8.5 GHz. Table 3.1: Test’s Conditions Table. 3.1 3.1.1. Electrical characterization Linear Characterization. S Parameters For this measurement, the set-up is described in the figure 3.2a.. (a) Assembly for the S parameters measurement characterization (b) Laboratory assembly for the S parameters measurement. Figure 3.2: Assembly for the S parameters measurement.

(24) 3. EMM5074VU amplifier testing. 13. S11. S21. -10. 31 Manufacturer Data Lab Data. -12 30.5. -14. 30. -18. S21 (dB). S11 (dB). -16. -20. 29.5. -22 29. -24. Manufacturer Data Lab Data. -26. 28.5. -28 -30. 28 6. 6.5. 7. 7.5. 8. 8.5. 6. 6.5. 7. 109. Freq (Hz). (a) Param S11 measured and manufacturer data from 5,8 to 8,5 GHz. 7.5. 8. 8.5 109. Freq (Hz). (b) Param S21 measured and manufacturer data from 5,8 to 8,5 GHz. S12. S22. -50. 0 Manufacturer Data Lab Data. -5. -55. -10 -60 -15. S22 (dB). S12 (dB). -65 -70. -20 -25 -30. -75. -35. Manufacturer Data Lab Data. -80. -40. -85. -45. -90. -50 6. 6.5. 7. 7.5. Freq (Hz). 8. 8.5 10. 6. 6.5. 9. (c) Param S12 measured and manufacturer data from 5,8 to 8,5 GHz. 7. 7.5. 8. Freq (Hz). 8.5 109. (d) Param S22 measured and manufacturer data from 5,8 to 8,5 GHz. Figure 3.3: Comparison between measured S parameters and manufacture S parameters Before performing this characterization, the network analyzer must be calibrated to include the effects of the attenuator. It must be there in order to take care of the network analyzer and not overpower it. From figure 3.3 it is worth highlighting some aspects: The amplifier meets main specifications as can be seen in appendix B. The S21 differs approximately 1 dB at frequencies from 7 GHz to 8 GHz to the value specified by the manufacturer. The isolation measure by the S12 is even better than the manufacturer specified, differing around 5 dB. Return Losses are greater than expected..

(25) 3. EMM5074VU amplifier testing. 14. Noise Figure For the noise figure measurement, a noise generator (346CK01 Keysight), and a spectrum analyzer are required.. Figure 3.4: laboratory assembly scheme for Noise Figure measurement. (a) Laboratory assembly for the Noise Figure measurement. (b) 346CK01 Source. Noise. Figure 3.5: Laboratory assembly for the Noise Figure measurement and close-up to the Noise Source In figure 3.6a the noise with the noise generator on and off is measure. These measures are needed to calculate the Y Parameter as shown in equation 2.5. The ENR(Excess Noise Ratio) is a parameter of the noise source. The manufacturer gives some data about it on certain frequencies (Figure 3.5b). For the inter frequencies needed, linear variation of the ENR is supposed, therefore these ENR values are.

(26) 3. EMM5074VU amplifier testing. 15. ENR and Y Parameter Noise Power Measures -95. 18. -100 16. 14. Power (dB). Power (dB). -105 -110 -115. 12. -120. 10. -125 8 -130. Floor Noise of the Spectrum Analizer Noise with power OFF Noise Power ON. -135 6. 6.5. 7. 7.5. 8. ENR Parameter Y Parameter. 6 6. 6.5. 7. 8.5. Frequency (Hz). 10. 7.5. Frequency (Hz). 8. 8.5 109. 9. (b) ENR Parameter and Y Parameter from 5,8 to 8,5 GHz. (a) Measured Noise from 5,8 to 8,5 GHz. Noise figure. 12. Chain Noise Figure DUT Noise Figure. 10 8 6. NF (dB). 4 2 0 -2 -4 -6 -8 6. 6.5. 7. 7.5. 8. 8.5 109. Frequency (Hz). (c) Calculated Noise Figure from 5,8 to 8,5 GHz. Figure 3.6: Noise Figure Measurements and calculations approximated by the following equations:  −0.42f + 20.71    −0.5f + 21.19 EN R = −0.33f + 20    −0.39f + 20.48. if if if if. 5<f 6<f 7<f 8<f. <6 <7 <8 <9. GHz GHz GHz GHz. (dB). (3.1). Once the ENR is set and the Y Parameter is calculated, as shown in figure 3.6b, the noise figure can be calculated following equation 2.7. Results are presented in figure 3.6c. Finally, an interpolated curve is made with the previous results to show the Noise Factor Figure (Figure 3.7) According to Friis Formula (equation 3.2), it is observed that the noise measured by the spectrum analyzer is not that of the amplifier (DUT). It measures the total noise of the chain. Knowing the noise figure and the gain of the components of the assembly provided in each manufacturer’s datasheet, a correction factor is calculated to obtain F2.

(27) 3. EMM5074VU amplifier testing. 16. (Figure 3.4), which is the parameter corresponding to the noise figure of the amplifier under study. Connector 1 → F1 = 0.2 dB. G1= -0.2 dB. Amplifier → F2 = Noise Figure Connector 2 → F3 = 0.2 dB. G= S21 Parameter G3= -0.2 dB. Spectrum Analyzer Pre Amplifier → F4 = 9dB. FT otal = F1 +. F2 − 1 F3 − 1 F4 − 1 + + G1 G1 G2 G1 G2 G3. (3.2). Interpolated Noise figure. 12. Chain Noise Figure Points Interpolated Chain Noise Figure Interpolated DUT Noise Figure. 10. NF (dB). 8. 6. 4. 2. 0. -2 6. 6.5. 7. 7.5. Frequency (Hz). 8. 8.5 109. Figure 3.7: Noise Figure from 5,8 to 8,5 GHz.

(28) 3. EMM5074VU amplifier testing. 3.1.2. 17. Non-Linear Characterization. One Tone These measurements have the same assembly (Figure 3.2a) as the S parameters measurements. To protect the network analyzer the measures were made taking into account the attenuator, therefore, all measurements should be corrected by adding the attenuator’s S21 (Figure 3.8). This sum can be made because both devices, attenuator and amplifier are adapted to the same impedance (50 Ω). Attenuator S Parameters -15 S11 S21 S12 S22. -20. Power (dBm). -25 -30 -35 -40 -45 -50 -55 5.5. 6. 6.5. 7. Frequency(Hz). 7.5. 8. 8.5 109. Figure 3.8: Attenuator’s S parameters from 5,8 to 8,5 GHz.

(29) 3. EMM5074VU amplifier testing. 18. Gain Compression Curve (Figure 3.9). Gain Compression 32. 31. S21 (dB). 30. 29. 28. 27. 26 -20. 5,8 GHz 6 GHz 6,5 GHz 7 GHz 7,5 GHz 8 GHz 8,5GHz Sumitomo @ 7,1 GHz. -15. -10. -5. 0. 5. Power(dBm). Figure 3.9: Gain Compression curve at different frequencies and comparison with data provided by the manufacturer The gain varies as a function of frequency, as can be seen in figure 3.9 . It is higher than the 28,5 dB reported by the manufacturer by 0,5 dB to 2,5 dB From the data of the gain compression curve (Figure 3.9), the output power versus input power curve (Figure 3.10) and the compression point at 1 dB (Figure 3.11) have been calculated. Given that the measured gain is higher than it was first expected, it is logical to assume that these two last figures (3.10 and 3.11) were also higher in the laboratory than in the data provided..

(30) 3. EMM5074VU amplifier testing. 19. AM-AM Conversion 40. Output Power (dBm). 35. 30. 25. 20. 5,8 GHz 6 GHz 6,5 GHz 7 GHz 7,5 GHz 8 GHz 8,5GHz Sumitomo @ 7,1 GHz. 15. 10. 5 -20. -15. -10. -5. 0. 5. Input Power(dBm). Figure 3.10: Output power Vs Input power curve at different frequencies and comparison with data provided by the manufacturer. 1 dB compression Point. 35. Lab Data Manufacturer Data. 34.5. OutputPower (dBm). 34 33.5 33 32.5 32 31.5 31 6. 6.5. 7. 7.5. 8. 8.5. Frequency (GHz). Figure 3.11: Mean 1dB Compression curve from previous frequencies and comparison with data provided by the manufacturer.

(31) 3. EMM5074VU amplifier testing. 20. AM-PM curve (Figure 3.12) is one of the fundamental contributors to BER, and therefore it is important to quantify this parameter and its variation (Figure 3.13) in communication systems. This measurement has been corrected with the attenuator’s S21 values. Corrected AM-PM 400. 300. 5,8 GHz 6 GHz 6,5 GHz 7 GHz 7,5 GHz 8 GHz 8,5GHz. S21 (degrees). 200. 100. 0. -100. -200 -20. -15. -10. -5. 0. 5. Power(dBm). Figure 3.12: AM-PM curve at different frequencies. Variation of the AM-PM conversion 18 16. Phase Offset(degrees). 14 12. 5,8 GHz 6 GHz 6,5 GHz 7 GHz 7,5 GHz 8 GHz 8,5GHz. 10 8 6 4 2 0 -2 -20. -15. -10. -5. 0. 5. Power(dBm). Figure 3.13: Variation of the AM-PM conversion curve at different frequencies.

(32) 3. EMM5074VU amplifier testing. 21. Two Tones With this assembly (figure 3.14b) we get the measurements of the tones and their inter modulation products, which are necessary to be able to create the curve of thirdorder intercept points.. (a) Laboratory assembly for two tones and multi tone measurements. (b) Laboratory assembly for two tones and multitone measurements scheme. Figure 3.14: Laboratory assembly for two tones and multi-tones measurements and close-up to the Noise Source Following the procedure for calculating the IP3 point for different values of the frequency band as presented in figure 2.4, figure 3.15 was obtained. Third Order Intermodulation Products 30. 20. (dB). 10. 0 Pout @ 5,8 GHz Pout @ 6 GHz Pout @ 6,5 GHz Pout @ 7 GHz Pout @ 7,5 GHz Pout @ 8 GHz Pout @ 8,5 GHz I3out @ 5,8 GHz I3out @ 6 GHz I3out @ 6,5 GHz I3out @ 7 GHz I3out @ 7,5 GHz I3out @ 8 GHz I3out @ 8,5 GHz. -10. -20. -30 -16. -14. -12. -10. -8. -6. -4. Input Power(dBm). Figure 3.15: Measures to acquire the IP3 points at different frequencies.

(33) 3. EMM5074VU amplifier testing. 22. Third Order intercept points These points were obtained following the equation 2.10 and then a curve was interpolated with them. (Figure 3.16) Third Order Intermodulation Products 39. Power (dBm). 38.5. 38. 37.5. IP3 (interpolated) IP3 (measure). 37 6. 6.5. 7. 7.5. 8. 8.5. Frequency (GHz). Figure 3.16: IP3 points and interpolated curve from 5,8 to 8,5 GHz The ratio between the power of the desired tones and the intermodulation products as a function of the output power and the frequency compared to that indicated by the manufacturer are shown in Figure 3.17 IM3 -15. -20. (dBc). -25. -30. -35 5,8 GHz 6 GHz 6,5 GHz 7 GHz 7,5 GHz 8 GHz 8,5GHz Sumitomo @ 7,1 GHz. -40. -45. -50 20. 21. 22. 23. 24. 25. 26. 27. 28. 29. 30. Output Power (dBm/per tone). Figure 3.17: Intermodulation ratio depending on the output power and comparison with data provided by the manufacturer.

(34) 3. EMM5074VU amplifier testing. 3.1.3. 23. Multitone. With the same assembly as in the two tones measurements (Figure 3.14b ), a signal with 29 (Figure 3.19a) or 30 (Figure 3.19b) tones spaced 1 MHz is amplified by the EMM5074VU amplifier. According to [5] the results do not vary significantly when the number of input tones is higher than 10.. Figure 3.18: Distortion depending on the number of tones. As can be seen in figure 3.18, Input Signal. -30. -30. -40. -40. -50. -50. -60. -60. -70. -70. -80. -80. -90. -90. -100. -100. -110 7.475. 7.48. 7.485. 7.49. 7.495. 7.5. 7.505. Frequency(GHz). Input Signal. -20. (dB). (dB). -20. 7.51. 7.515. 7.52. 7.525. -110 7.475. 7.48. 7.485. 7.49. 7.495. 109. (a) 29 tones signal with 50 MHz span. 7.5. 7.505. Frequency(GHz). 7.51. 7.515. 7.52. 7.525 109. (b) 30 tones signal with 50 MHz span. Figure 3.19: Multitone Input Signals.

(35) 3. EMM5074VU amplifier testing. 24. These measurements give the necessary information to get the following merit figures: Output Signal. 0 -10. -20. -30. -30. -40. -40. -50 -60. -50 -60. -70. -70. -80. -80. -90. -90. -100. -100 7.46. 7.47. 7.48. 7.49. 7.5. 7.51. 7.52. 7.53. Frequency(GHz). 7.54. 0 dBm -5 dBm -10 dBm -15 dBm. -10. Power (dB). Power (dB). -20. -110 7.45. Output Signal. 0 5 dBm 0 dBm -5 dBm -10 dBm. -110 7.475. 7.55 10. 7.48. 7.485. 7.49. 9. 7.495. 7.5. 7.505. Frequency(GHz). 7.51. 7.515. 7.52. 7.525 109. (a) 29 output tones signal with 100 MHz span (b) 30 tones output signal with 50 MHz span. Figure 3.20: Multitone output Signals Figure 3.20a has a 100 MHz Span because it is needed to calculate the Adjacent Channel Power Ratio. M-IMR This measure is the relationship between the power of one of the tones, in this particular case, the tone used was the 27th ; and power of the upper band, specifically the power at f0 + 0.018GHz, f0 ∈ [5.8 6 6.5 7 7.5 8 8.5] GHz. M-IMR 50 5.8 GHz 6 GHz 6.5 GHz 7 GHz 7.5 GHz 8 GHz 8.5 GHz. 45. (dB). 40. 35. 30. 25. 20 -44. -42. -40. -38. -36. -34. -32. -30. Input Power (dBm/per tone). Figure 3.21: M-IMR with 29 tones from -44 dBm per tone to -29 dBm per tone..

(36) 3. EMM5074VU amplifier testing. 25. ACPR The total ACPR is defined as the ratio between the integrated output power in the desired frequency band and the integrated output power in the lower and upper adjacent channels (Equation 2.12). It has been considered that the size of the adjacent channel of each side is the same as the size of the desired channel. ACPR 60 5.8 GHz 6 GHz 6.5 GHz 7 GHz 7.5 GHz 8 GHz 8.5 GHz. 55 50. (dB). 45 40 35 30 25 20 -44. -42. -40. -38. -36. -34. -32. -30. Input Power (dBm/per tone). Figure 3.22: ACPR with 29 tones from -44 dBm per tone to -29 dB per tone. ACPR 7,5 GHz 60 55 50. (dB). 45 40 35 30 25 20 -44. -42. -40. -38. -36. -34. -32. -30. Input Power (dBm/per tone). Figure 3.23: ACPR close-up at 7,5 GHz from -44 dBm per tone to -29 dB per tone.

(37) 3. EMM5074VU amplifier testing. 26. NPR This figure of merit is used to characterize the co-channel distortion. It is defined as the ratio between the spectral output density at the frequency f0 where the gap was made and the output power of the closest fundamental components that were not eliminated (Equation 2.15) NPR 45 5.8 GHz 6 GHz 6.5 GHz 7 GHz 7.5 GHz 8 GHz. 40. (dB). 35. 30. 25. 20. 15 -44. -42. -40. -38. -36. -34. -32. -30. Input Power (dBm/per tone). Figure 3.24: NPR with 30 tones from -44 dBm per tone to -29 dBm per tone at different frequencies Comparing the results of the previous measurements (Figure 3.25) with those of figure 3.18, it can be seen that those carried out in the laboratory comply with what is expected theoretically MIMR ACPR and NPR comparison. 60. MIMR ACPR NPR. 55 50. (dB). 45 40 35 30 25 20 15 -44. -42. -40. -38. -36. -34. -32. -30. Input power per Tone (dBm). Figure 3.25: MIMR, ACPR and NPR measurements comparison at 7,5 GHz.

(38) 3. EMM5074VU amplifier testing. 3.2. 27. Thermal Characterization. To perform these measurements, the assembly carried out is analogous to that required for the measures of the electrical characterization, but this time, the amplifier is within a thermal chamber to fix the temperature. There are numerous studies that reflect the temporal dependence of the response of MMIC amplifiers to variations in working conditions. In these studies, it is observed that since a change in the amplifier operation point is applied until the temperature in the MMIC reaches a stable value, a certain time elapses related to the thermal capacity of the materials that make up the entire structure. The main objective of the thermal chamber is to keep the temperature of these materials constant, so that the measurement is accurate and reliable.. Figure 3.26: Thermal Chamber. The losses of the cables connected to the amplifier and other devices were measured previously, in order to correct the measures given by the spectrum analyzer. As can be seen in figure 3.26, the chamber that is been used to perform these measures is the Votsch VTL 4010. For all these measures, the temperature range has been set from -30 to 60 ◦ C. The selected measurements to be performed are: AM-AM conversion Gain Compression M-IMR ACPR NPR. With these measurements, an approach to the thermal behaviour in terms of linear and non-linear performance and with one tone input signal or multi tone input signal will be obtain..

(39) 3. EMM5074VU amplifier testing. 28. One Tone Characterization. – AM-AM Conversion For the AM-AM Conversion, instead of using a network analyzer as in the electrical characterization, a simpler assembly with a signal generator and a spectrum analyzer will be utilized. 34. Input-Output at different temperatures. 32 30. Output Power(dB). 28 -30 º -20 º -10 º 0º 10 º 20 º 30 º 40 º 50 º 60 º. 26 24 22 20 18 16 14 -15. -10. -5. 0. 5. input Power (dBm). Figure 3.27: Input power versus Output power at different temperatures – Gain Compression at different temperatures 31.5. Gain Compression at different temperatures -30 º -20 º -10 º 0º 10 º 20 º 30 º 40 º 50 º 60 º. 31 30.5. Gain(dB). 30 29.5 29 28.5 28 27.5 27 -15. -10. -5. 0. 5. input Power (dBm). Figure 3.28: Gain Compression at different temperatures from -15 dBm to 5 dBm input power.

(40) 3. EMM5074VU amplifier testing. 29. – Gain Compression variation through different temperatures Gain Compression 31.5. interpolated -5 dBm interpolated 5 dBm input manufacturer data (-6 dBm) -5 dBm input 5 dBm input. 31 30.5. Gain(dB). 30 29.5 29 28.5 28 27.5 27 -30. -20. -10. 0. 10. 20. 30. 40. 50. 60. Temperature (ºC). Figure 3.29: Gain variation from -30◦ C to 60◦ C As can be seen in figure 3.29 the variation of the amplifier’s gain with temperature is wider when input power is lower. With -5 dBm of input power, the slope of the interpolated gain is -0.0115 dB/◦ C. Meanwhile, with 5 dBm of input power, the slope of the interpolated gain is -0.0066 dB/◦ C Both are lower than the slope given by the manufacturer, that is -0.024 dB/◦ C..

(41) 3. EMM5074VU amplifier testing. 30. Multitone Characterization These three measurements are performed as the ones executed in the electrical characterization and with the same laboratory assembly (figure 3.14a) but with the amplifier inside the thermal chamber.. – M-IMR MIMR at different temperatures. 70. -30 º -20 º -10 º 0º 10 º 20 º 30 º 40 º 50 º 60 º. 65 60 55. (dB). 50 45 40 35 30 25 20 -36. -34. -32. -30. -28. -26. -24. Input power per Tone (dBm). Figure 3.30: M-IMR at different temperatures from 29 tones from -37,5 dBm to -23,8 dBm per tone. Variation of the MIMR with the temperature. 70 65 60 55. (dB). 50 45 40 35. -37,5 dBm per Tone -32,6 dBm per Tone -27,9 dBm per Tone -23,8 dBm per Tone. 30 25 20 -30. -20. -10. 0. 10. 20. 30. 40. 50. 60. Temperature (ºC). Figure 3.31: M-IMR variation from 29 tones through temperatures from -30◦ C to 60◦ C.

(42) 3. EMM5074VU amplifier testing. 31. – ACPR ACPR at different temperatures. 60. -30 º -20 º -10 º 0º 10 º 20 º 30 º 40 º 50 º 60 º. 55 50. (dB). 45 40 35 30 25 20 -36. -34. -32. -30. -28. -26. -24. Input power per Tone (dBm). Figure 3.32: ACPR at different temperatures from 29 tones from -37,5 dBm to -23,8 dBm per tone. Variation of the ACPR with the temperature. 60 55 50. (dB). 45 40 35 -37,5 dBm per Tone -32,6 dBm per Tone -27,9 dBm per Tone -23,8 dBm per Tone. 30 25 20 -30. -20. -10. 0. 10. 20. 30. 40. 50. 60. Temperature (ºC). Figure 3.33: ACPR variation from 29 tones through temperatures from -30◦ C to 60◦ Cs.

(43) 3. EMM5074VU amplifier testing. 32. – NPR NPR at different temperatures. 45. -30 º -20 º -10 º 0º 10 º 20 º 30 º 40 º 50 º 60 º. 40. (dB). 35. 30. 25. 20. 15 -36. -34. -32. -30. -28. -26. Input power per Tone (dBm). Figure 3.34: NPR at different temperatures from 30 tones from -37,5 dBm to -23,8 dBm per tone. Variation of the NPR with the temperature. 45. 40. (dB). 35 -37,65 dBm per Tone -32,8 dBm per Tone -28,3 dBm per Tone -24,4 dBm per Tone. 30. 25. 20. 15 -30. -20. -10. 0. 10. 20. 30. 40. 50. 60. Temperature (ºC). Figure 3.35: NPR variation from 30 tones through temperatures from -30◦ C to 60◦ C As can be seen in figures 3.31 3.33 and 3.35, the higher the input power, the less variation of the electrical performance with temperature. The NPR is the least temperature dependent of the three figures of merit..

(44) 3. EMM5074VU amplifier testing. 3.3. 33. Summary. Table 3.2 is a summary of the measurements presented previously. All measures has been taken at 7,5 GHz Small signal Gain Noise Figure 1 dB Compression Point IP3 M-IMR with 29 input tones (@ -37 dBm per tone input power) ACPR with 29 input tones (@ -37 dBm per tone input power) NPR with 30 input tones (@ -37 dBm per tone input power) Gain Variation with Temperature (@ -5 dBm input power) M-IMR peak to peak Ripple with Temperature (@ -32,6 dBm per tone input power) ACPR peak to peak Ripple with Temperature (@ -32,6 dBm per tone input power) NPR peak to peak Ripple with Temperature (@ -32,6 dBm per tone input power ) Table 3.2: Test Results Table. 29 dB 4,7 dB 34 dBm 39 dBm 35 dB 44 dB 32,5 dB -0,0115 dB/ºC 5,6 dB 5,6 dB 0,4 dB.

(45) Chapter 4 Conclusions In this Bachelor thesis, the characteristics that define the RF power amplifiers have been analyzed, a measurement plan has been designed in accordance to those characteristics, and finally the validity of these measurements plan has been verified. Nowadays telecommunication systems often use multi-frequency waveforms. This fact has been taken into account by analyzing the co-channel and the adjacent channel distortion they generate. A measurement plan has been designed based on the traditional characterization with one tone and two tones for the linear and non-linear aspects of the amplifiers, adding the multitone characterization. In addition to the traditional characterization, the thermal behaviour of the power amplifiers has been considered too in these measurements..

(46) Appendix A Synthesis of multitone signals To carry out the multitone tests in the electrical characterization of the EMM5074 amplifier, a multi-carrier signal generated in Matlab was used. The process followed is described below.. Figure A.1: Multitone signal generation process First, the N tones that compose the multi-carrier signal are synthesized. This signal is passed as a file of I-Q samples to the Agilent N5182A MXG generator. However, phase continuity of the input signal must be guaranteed when this is introduced into the signal generator, as the signal generator repeats the samples of the signal generated in Matlab periodically. If this phase continuity is not guaranteed, a spectral widening occurs and distortion is introduced. Figure A.2 shows a sampled sinusoid generated through a Matlab code. It is clear that the signal generated is a sinusoid, but the problems arise when the signal generator repeats it periodically, since when generating it, it has not been guaranteed that the sinusoid segment that is introduced in the generator occupies a full period of the signal. Therefore, it is not only necessary to generate the sampled signal, but also to guarantee that the waveform segment that is being generated is a complete period thereof..

(47) A. Synthesis of multitone signals. 36. Figure A.2: Example of phase discontinuity Figure A.3 shows the generation of the 14 tones in frequency domain (Figure A.3a) and in time domain (Figure A.3b) in Matlab. The lower frequency of this signal is 1 MHz (observing the positive frequencies). This signal is generated as a set of complex exponentials summed in time, being able to choose their starting frequency, their separation and the number of tones that compose it, both for positive and negative frequencies. The generated signal fulfills its continuity in phase, since the lower frequency, the separation between the tones, the number of samples, and the sampling frequency are multiples of each other. On the other hand, said signal must be sampled with a frequency equal to that used, subsequently, in the MXG generator. This frequency is chosen so that it is greater than twice the maximum frequency of the signal and is also make that the signal generated has a full period. Given that along the characterization of the multitone distortion signals with different bandwidths are generated with a sampling frequency of 64 MHz. The resulting spectrum, it is obtained through an FFT of 8192 samples which is the maximum number of samples allowed by the Agilent N5182A MXG generator..

(48) A. Synthesis of multitone signals. 37. FTT of the Signal. 100. Time domain tones modulated by a 200 MHz carrier. 8. 50. 6. 0. 4. -50 2 -100 0 -150 -2 -200 -4. -250 -300. -6 -15. -10. -5. 0. 5. 10. 15. 7.8. 7.9. 8. 8.1. 8.2. 8.3. 8.4. 8.5 10-5. Frecuencia (MHz). (a) Signal in the frequency domain. (b) Signal in the time domain. Figure A.3: Signal generated by Matlab If the number of samples and the sampling frequency are not multiples of each other, the samples will not be taken in the right places of the waveform and the FTT will not show ”pure tones” (Figure A.4) . This is just for the visualization of the signal in Matlab. If this signal is inserted in the MXG generator with the right sampling frequency, the signal will be the desired one. FTT of the Signal. 70. 60. 50. 40. 30. 20. -15. -10. -5. 0. 5. 10. 15. 20. Frecuencia (MHz). Figure A.4: Signal in the frequency domain with number of samples and sampling frequency not being multiples of each other.

(49) Appendix B Project impact In this appendix the impact of the Bachelor Thesis will be analyzed in the ethical, economic, social and environmental fields. To put in context, the objective of this project is to design and validate a methodology for the characterization of RF Power amplifiers.. B.1. Ethical impact. The ethical impact of this work depends on the use of the system in which the RF amplifier will take place, since in itself, the amplifiers will lack any impact on people’s lives.. B.2. Economical impact. This project does have an economic impact since once you know the operation of the amplifier, you can optimize the resources to create the system in which it will be placed so that it maximizes the benefits of the amplifier and minimizes the manufacturing errors it may have. If the optimized systems work better by knowing better the operation of the amplifiers that compose it, more buyers may be interested in that particular product, encouraging the market of microwave devices.

(50) B. Project impact. B.3. 39. Social impact. The social impact of this project is again the optimization of systems that use RF amplifiers. The more optimized they are, the better they will work and greater will be their impact in society.. B.4. Environmental impact. Since optimization of the systems also brings a better efficiency in terms of energy this bachelor thesis does have an environmental impact. Better energy efficiency means that less resources are consumed when the systems, where the RF amplifiers are placed, are powered on..

(51) Appendix C Estimated project budget This annex provides an economic budget of the approximate costs involved in carrying out this project. The human costs as well as the material and service resources that have had to be used will be taken into account. Two assumptions will be shown, one in which it is necessary to obtain all the necessary devices to carry out the characterization and another in which the characterization is outsourced. It is considered that the person who carries out the project is a graduate of engineering in telecommunication technologies and Services Engineering. A graduate in this area is estimated to have a gross annual salary of about 24,000A C, working 20 days a month for 8 hours. Given that when hiring a graduate, taxes must be paid, gross annual salary has been approximated by being 1/3 of the human cost. With these data, the human cost is 37,5A C/hour. Taking into account that the project dedication is 330 hours, the total human cost is 12375A C. Regarding material resources, a laptop has been used to represent the measure results with Matlabr . A laptop with a 500Gb SSD, Intel i7 CPU, and 16Gb RAM to run this software has a cost of 700A C; plus Matlab’s license, which has a price of 800A C/year. These previous costs are common to both assumptions and add up 13875A C.. C.1. Necessary laboratory devices. As it has been shown in chapters 2 and 3, lab equipment is required to perform the electrical and thermal characterization of RF amplifiers. For the measurements plan design in this project, the require devices and its costs are: Network analyzer → 27795A C Spectrum analyzer → 32500A C Noise generator → 5438A C.

(52) C. Estimated project budget. 41. Signal Generator from 0 to 3.8 GHz → 35347A C Signal Generator from 0 to 18 GHz → 37217A C Thermal Chamber → 16995A C Lab material (wrenchs, cables, etc) → 150A C. All these devices total an amount of 152442A C. Adding the common costs previously shown, the estimated project budget is 166317A C.. C.2. Outsourcing characterization. Assuming all measures can be performed in 3 weeks, it is usual to outsource characterization to a laboratory saving the costs of purchasing all lab equipment. The cost of renting the lab space with this equipment per week is approximately 700A C depending on the cuality of the given equipment. Given that for this project, the measures can be performed in 3 weeks, the total amount for renting the lab is 2100A C. Adding the mutual costs previously shown, the estimated project budget is 15975A C..

(53) Appendix D EMM5074VU Amplifier Datasheet.

(54) EMM5074VU C-Band Power Amplifier MMIC FEATURES ・High Output Power: Pout=33dBm (typ.) ・High Linear Gain: GL=27dB (typ.) ・Broad Band: 5.8 - 8.5GHz ・Impedance Matched Zin/Zout=50Ω ・Small Hermetic Metal-Ceramic SMT Package(VU) DESCRIPTION The EMM5074VU is a wide band power amplifier MMIC that contains a three stage amplifier, internally matched, for standard communications band in 5.8 to 8.5GHz frequency range. Eudyna’s stringent Quality Assurance Program assures the highest reliability and consistent performance. ABSOLUTE MAXIMUM RATING Item Drain-Source Voltage. Symbol VDD. Rating 10. Unit V. Gate-Source Voltage. VGG. -3. V. Input Power Storage Temperature. Pin Tstg. +26 -55 to +125. dBm o C. RECOMMENDED OPERATING CONDITIONS Item Drain-Source Voltage Input Power Operating Case Temperature. Symbol VDD. Condition <=6. Unit V. Pin TC. <=10 -40 to +85. dBm ℃. ELECTRICAL CHARACTERISTICS (Case Temperature Tc=25oC) Item Frequency Range Output Power at 1dB G.C.P.. Symbol f P1dB. Test Conditions. Limits Typ. Max. 8.5. VDD=6V. Min. 5.8. IDD(DC)=1200mA typ.. 30*1. 32*1. -. Zs=Zl=50ohm. 31*2. 33*2. -. Unit GHz dBm. Power Gain at 1dB G.C.P.. G1dB. *1:f=5.8~7.1GHz. 23. 26. -. dB. Power-added Efficiency at 1dB G.C.P.. ηadd. *2:f=7.1~8.5GHz. -. 18*1. -. %. *2. -. Third Order Intermodulation* Drain Current at 1dB G.C.P. Input Return Loss (at Pin=-20dBm) Output Return Loss (at Pin=-20dBm). IM3 IDD. ∗:∆f=10MHz ,. -37. 22 -42. *1. 1800. dBc *1. 2-Tone Test ,. -. 1400. Pout=20dBm S.C.L.. -. 1450*2 1800*2 15 15 -. RLin RLout. mA dB dB. G.C.P.:Gain Compression Point, S.C.L.:Single Carrier Level ESD. Class 0. ~ 249V. Note : Based on JEDEC JESD22-A114C CASE STYLE. VU. RoHs Compliance. Yes. Edition 1.2 April, 2007. 1. http://www.eudyna.com/.

(55) EMM5074VU C-band Power Amplifier MMIC. Output Power vs. Frequency. Output Power, Drain Current vs. Input Power @VDD=6V, IDD(DC)=1200mA. @VDD=6V, IDD(DC)=1200mA. 36. 2500. 34. 2300. 32. 2100. 30. 1900. 28. 1700. 26. 1500. 24. 1300. 22. 1100. 20. 900. 36 34. 30 28 26 24 22 20 5. 5.5. 6. 6.5. 7. 7.5. 8. 8.5. 9. 700. 18. 9.5. -12. Frequency [GHz]. -10. -8. -6. -4. -2. 0. 2. 4. 6. 8. 10. 12. Input Power [dBm] Pin=-4dBm. 0dBm. +4dBm. +10dBm. P1dB. 5.8GHz. Power Added Efficiency vs. Frequency @VDD=6V, IDD(DC)=1200mA. 35. Power Added Efficiency [%]. 30. 25. 20. 15. 10. 5. 0 5. 5.5. 6. 6.5. 7. 7.5. 8. 8.5. 9. 9.5. Frequency [GHz] Pin=-4dBm. 0dBm. +4dBm. +10dBm. P1dB. 2. 7.1GHz. 8.5GHz. 14. Drain Current [mA]. Output Power [dBm]. Output Power [dBm]. 32.

(56) EMM5074VU C-Band Power Amplifier MMIC IM3 vs. Frequency. IMD vs. Output Power @VDD=6V, IDD(DC)=1200mA. -20. -20. -25. -25. -30. -30. -35. -35. -40. -40. IMD [dBc]. -45. Solid Line : IM3 Dash Line : IM5. -45. -50. -50. -55. -55. -60. -60. -65. -65 -70. -70 5. 5.5. 6. 6.5. 7. 7.5. 8. 8.5. 9. 16. 9.5. 17. 18. 19. 20. 21. 22. 23. 5.8GHz. P1dB, G1dB vs. Frequency by Drain Voltage. 25. 26. 27. 28. 7.1GHz. 29. 30. 31. 32. 8.5GHz. P1dB, G1dB vs. Frequency by Drain Current. @VDD=6V. @IDD(DC)=1200mA 42. 36. 42. 34. 40. 34. 40. 32. 38. 32. 38. 30. 36. 30. 36. 28. 34. 34. 28. Solid Line : P1dB Dash Line : G1dB. 26. 32. P1dB [dBm]. 36. G1dB [dB]. P1dB [dBm]. 24. 2-Tone Total Pout [dBm]. Frequency [GHz]. 26. 32. Solid Line : P1dB Dash Line : G1dB. 24. 30. 24. 30. 22. 28. 22. 28. 20. 26. 20. 26. 18. 24. 18. 24. 22. 16. 16 5. 5.5. 6. 6.5. 7. 7.5. 8. 8.5. 9. 5V. 5.5. 6. 6.5. 7. 7.5. 8. 8.5. 9. 9.5. Frequency [GHz]. Frequency [GHz] VDD=4V. 22 5. 9.5. 6V. IDD(DC)=800mA. 3. 1200mA. 1600mA. G1dB [dB]. IM3 [dBc]. @VDD=6V, IDD(DC)=1200mA, @Po=20dBm S.C.L..

(57) EMM5074VU C-band Power Amplifier MMIC Output Power, Drain Current vs. Input Power by Drain Voltage. Output Power, Drain Current vs. Input Power by Drain Voltage. 2500. 36. 2500. 34. 2300. 34. 2300. 32. 2100. 32. 2100. 30. 1900. 30. 1900. 28. 1700. 28. 1700. 26. 1500. 26. 1500. 24. 1300. 24. 1300. 22. 1100. 22. 1100. 20. 900. 20. 900. 700. 18. 18 -12. -10. -8. -6. -4. -2. 0. 2. 4. 6. 8. 10. 12. Output Power [dBm]. 36. 14. 700 -12. -10. -8. -6. -4. Input Power [dBm] VDD=4V. 6V. VDD=4V. 36. 2500. 34. 2300. 32. 2100. 30. 1900. 28. 1700. 26. 1500. 24. 1300. 22. 1100. 20. 900. Drain Current [mA]. Output Power [dBm]. @f=8.5GHz, IDD(DC)=1200mA. 700. 18 -10. -8. -6. -4. -2. 0. 2. 4. 6. 8. 10. 12. 14. Input Power [dBm] VDD=4V. 5V. 0. 2. 4. 6. 8. 10. Input Power [dBm]. 5V. Output Power, Drain Current vs. Input Power by Drain Voltage. -12. -2. 6V. 4. 5V. 6V. 12. 14. Drain Current [mA]. @f=7.1GHz, IDD(DC)=1200mA. Drain Current [mA]. Output Power [dBm]. @f=5.8GHz, IDD(DC)=1200mA.

(58) EMM5074VU C-Band Power Amplifier MMIC Output Power, Drain Current vs. Input Power by Drain Current. Output Power, Drain Current vs. Input Power by Drain Current. 36. 2500. 34. 2300. 34. 2300. 32. 2100. 32. 2100. 30. 1900. 30. 1900. 28. 1700. 28. 1700. 26. 1500. 26. 1500. 24. 1300. 24. 1300. 22. 1100. 22. 1100. 20. 900. 20. 900. 700. 18. 18 -12. -10. -8. -6. -4. -2. 0. 2. 4. 6. 8. 10. 12. Output Power [dBm]. 2500. Drain Current [mA]. Output Power [dBm]. 36. 1200mA 1200 A. 1600mA 1600 A. 36. 2500. 34. 2300. 32. 2100. 30. 1900. 28. 1700. 26. 1500. 24. 1300. 22. 1100. 20. 900. Drain Current [mA]. Output Power [dBm]. @f=8.5GHz, VDD=6V. 700. 18 -10. -8. -6. -4. -2. 0. 2. 4. 6. 8. 10. 12. 14. Input Power [dBm] IDD(DC)=800mA 800 A. 1200mA 1200 A. -8. -6. -4. -2. IDD(DC)=800mA 800 A. Output Power, Drain Current vs. Input Power by Drain Current. -12. -10. 0. 2. 4. 6. 8. 10. 12. Input Power [dBm]. Input Power [dBm] IDD(DC)=800mA 800 A. 700 -12. 14. 1600mA 1600 A. 5. 1200mA 1200 A. 1600mA 1600 A. 14. Drain Current [mA]. @f=7.1GHz, VDD=6V. @f=5.8GHz, VDD=6V.

(59) EMM5074VU C-band Power Amplifier MMIC IMD vs. Output Power by Drain Voltage. IMD vs. Output Power by Drain Voltage. @f=5.8GHz, IDD(DC)=1200mA. @f=7.1GHz, IDD(DC)=1200mA. -15. -15. Solid Line : IM3 Dash Line : IM5. -20. -20. Solid Line : IM3 Dash Line : IM5. -25. -30. -30. -35. -35 IMD [dBc]. IMD [dBc]. -25. -40 -45. -40 -45. -50. -50. -55. -55. -60. -60. -65. -65. -70. -70 16. 18. 20. 22. 24. 26. 28. 30. 32. 16. 18. 20. 2-Tone Total Pout [dBm] VDD=4V. 5V. 6V. VDD=4V. IMD vs. Output Power by Drain Voltage @f=8.5GHz, IDD(DC)=1200mA -15. Solid Line : IM3 Dash Line : IM5. -20 -25 -30. IMD [dBc]. -35 -40 -45 -50 -55 -60 -65 -70 16. 18. 20. 22. 24. 26. 28. 30. 32. 2-Tone Total Pout [dBm] VDD=4V. 5V. 22. 24. 26. 28. 30. 2-Tone Total Pout [dBm]. 6V. 6. 5V. 6V. 32.

(60) EMM5074VU C-Band Power Amplifier MMIC IMD vs. Output Power by Drain Current. IMD vs. Output Power by Drain Current. @f=5.8GHz, VDD=6V. @f=7.1GHz, VDD=6V. -15. -15. Solid Line : IM3 Dash Line : IM5. -20. -25. -30. -30. -35. -35 IMD [dBc]. IMD [dBc]. -25. -40 -45. -40 -45. -50. -50. -55. -55. -60. -60. -65. -65 -70. -70 16. 18. 20. 22. 24. 26. 28. 30. 16. 32. IDD(DC)=800mA. 1200mA. @f=8.5GHz, VDD=6V -15. Solid Line : IM3 Dash Line : IM5. -25 -30 -35 -40 -45 -50 -55 -60 -65 -70 16. 18. 20. 22. 24. 26. 28. 30. 32. 2-Tone Total Pout [dBm] IDD(DC)=800mA. 1200mA. 20. 22. IDD(DC)=800mA. 1600mA. IMD vs. Output Power by Drain Current. -20. 18. 24. 26. 28. 30. 2-Tone Total Pout [dBm]. 2-Tone Total Pout [dBm]. IMD [dBc]. Solid Line : IM3 Dash Line : IM5. -20. 1600mA. 7. 1200mA. 1600mA. 32.

(61) EMM5074VU C-band Power Amplifier MMIC OUTPUT POWER, DRAIN CURRENT vs. INPUT POWER by Temperature. OUTPUT POWER, DRAIN CURRENT vs. INPUT POWER by Temperature. 36. 2300. 36. 2300. 34. 2100. 34. 2100. 32. 1900. 32. 1900. 30. 1700. 30. 1700. 28. 1500. 28. 1500. 26. 1300. 26. 1300. 24. 1100. 24. 1100. 22. 900. 22. 900. 700. 20. 20 -6. -4. -2. 0. 2. 4. 6. 8. 10. Output Power [dBm]. 12. 700 -8. -6. -4. -2. Input Power [dBm] -40°C. +85°C. -40°C. 2100. 32. 1900. 30. 1700. 28. 1500. 26. 1300. 24. 1100. 22. 900. Psat [dBm] @Pin=10dBm. 34. Drain Current [mA]. Output Power [dBm]. 2300. 700. 20 0. 2. 4. 8. 10. 12. 6. 8. 10. +25°C. +85°C. 36. 36. 35. 35. 34. 34. 33. 33. 32. 32. 31. 31. 30. 30. 29. 29. 28. 28. 27. 27. 26 25. -2. 6. @VDD=6V, IDD(DC)=1200m A(Tc=25°C). 36. -4. 4. OUTPUT POWER, GAIN vs. TEMPERATURE. @VDD=6V, IDD(DC)=1200mA(Tc=25°C), Freq=8.5GHz. -6. 2. Input Power [dBm]. +25°C. OUTPUT POWER, DRAIN CURRENT vs. INPUT POWER by Temperature. -8. 0. 24 -50. 12. 26. Solid Line : Psat Dash Line : Gain. -25. 0. 25 25. 50. 75. 24 100. Input Power [dBm] -40°C. +25°C. Temperature [°C]. +85°C. 5.8GHz. 8. 7.1GHz. 8.5GHz. Gain [dB] @Pin=-6dBm. -8. Drain Current [mA]. @VDD=6V, IDD(DC)=1200mA(Tc=25°C), Freq=7.1GHz. Drain Current [mA]. Output Power [dBm]. @VDD=6V, IDD(DC)=1200mA(Tc=25°C), Freq=5.8GHz.

(62) EMM5074VU C-Band Power Amplifier MMIC ■ S-Parameter VDD=6V, IDD(DC)=1200mA. Input/Output Return Loss [ dB ]. Input/Output Return Loss vs. Frequency. 0 -10 -20 -30. - Input Return Loss - Output Return Loss. -40 5. 5.5. 6. 6.5. 7. 7.5. 8. 8.5. 9. 9.5. Frequency [GHz]. Small Signal Gain vs. Frequency. Small Signal Gain [dB]. 34 32 30 28 26 24 22 20 5. 5.5. 6. 6.5. 7. 7.5. 8. Frequency [GHz]. 9. 8.5. 9. 9.5.

(63) EMM5074VU C-band Power Amplifier MMIC ■ S-Parameter VDD=6V, IDD(DC)=1200mA Frequency [GHz] 5.0 5.1 5.2 5.3 5.4 5.5 5.6 5.7 5.8 5.9 6.0 6.1 6.2 6.3 6.4 6.5 6.6 6.7 6.8 6.9 7.0 7.1 7.2 7.3 7.4 7.5 7.6 7.7 7.8 7.9 8.0 8.1 8.2 8.3 8.4 8.5 8.6 8.7 8.8 8.9 9.0 9.1 9.2 9.3 9.4 9.5. S11 MAG ANG 0.262 30.2 11.5 0.157 0.072 -12.5 0.020 -99.5 0.057 179.4 0.095 160.6 0.122 150.2 0.137 141.4 0.147 133.8 0.155 127.1 0.163 121.5 117.8 0.170 113.0 0.174 0.175 108.1 0.177 102.5 95.8 0.180 0.180 90.1 84.3 0.180 78.5 0.176 0.172 71.2 0.169 64.1 56.7 0.167 50.3 0.164 42.7 0.159 34.8 0.154 26.4 0.152 0.152 18.4 11.0 0.154 4.1 0.154 -3.7 0.156 0.163 -12.2 0.171 -19.8 -27.4 0.182 -35.4 0.192 0.201 -45.0 -56.0 0.213 0.224 -67.9 0.231 -80.6 0.233 -95.2 0.231 -111.2 0.230 -130.1 0.229 -150.3 0.229 -172.4 0.238 163.4 0.260 139.1 0.295 116.2. S21 MAG ANG -56.0 17.543 -80.8 20.174 22.695 -105.3 25.097 -129.3 27.472 -153.0 29.615 -176.8 31.529 159.6 33.070 136.2 33.937 112.8 34.309 90.0 34.227 67.7 33.559 46.3 25.8 32.721 6.0 31.764 30.689 -13.1 29.747 -31.3 28.885 -49.1 28.009 -66.3 27.365 -82.9 26.923 -99.6 26.479 -115.9 26.371 -132.1 26.189 -148.3 26.180 -164.6 26.344 178.9 26.509 162.4 26.756 145.6 27.090 128.5 27.338 111.2 27.642 93.5 75.5 27.971 28.114 57.1 28.302 38.5 28.314 19.2 28.082 -0.2 27.775 -20.1 -40.2 27.174 26.284 -60.4 -80.9 25.281 23.980 -101.2 22.667 -121.4 21.239 -141.7 19.653 -161.8 18.126 178.3 16.569 158.1 14.970 138.4. 10. S12 MAG ANG 0.000 -108.8 0.000 -3.4 0.000 -141.0 0.000 -45.2 0.000 -8.7 0.000 -155.6 0.000 -168.0 0.000 171.9 0.000 -51.8 0.000 -49.8 0.000 -111.6 0.000 -138.4 0.000 -163.6 0.000 161.8 0.000 -122.8 0.000 -156.8 0.000 -162.7 0.000 -140.0 0.000 171.1 0.000 -175.8 0.000 -114.8 0.000 -148.5 0.000 -148.1 0.000 -179.7 0.000 -76.2 0.000 -101.1 0.000 -122.4 0.000 -124.8 0.000 -126.7 0.001 -127.1 0.001 -127.1 0.001 -138.4 0.001 -146.1 0.001 -170.0 0.001 -168.1 0.001 -157.6 0.001 -175.4 0.001 176.9 0.001 159.3 0.000 161.7 0.000 156.4 0.000 -169.6 0.000 -146.4 0.000 -143.6 0.000 -131.6 0.001 -119.5. S22 MAG ANG 0.370 -74.5 0.342 -89.3 0.313 -105.0 0.278 -122.2 0.242 -141.3 0.206 -163.2 0.173 171.9 0.148 142.6 110.6 0.134 80.0 0.131 0.136 53.3 0.141 31.7 0.145 14.5 0.6 0.144 -11.1 0.141 -20.3 0.135 -28.1 0.129 -33.9 0.124 0.117 -39.4 0.113 -44.5 -48.9 0.108 -53.5 0.105 -59.2 0.102 -65.0 0.101 0.099 -72.3 0.098 -80.2 0.094 -89.0 0.090 -98.2 0.086 -109.0 0.079 -121.1 0.071 -136.5 0.056 -151.4 0.042 -173.5 0.032 145.6 0.036 89.2 0.054 48.3 0.083 25.5 0.117 7.0 -7.9 0.152 0.195 -20.6 0.228 -32.6 0.266 -42.6 0.300 -52.1 0.330 -61.1 0.358 -69.0 0.383 -76.4.

(64) EMM5074VU C-Band Power Amplifier MMIC ∆Tch vs. DRAIN VOLTAGE (Reference Data) IDD(DC)=1200mA 60. o ∆ Tch[ C]. 50 40 30 20 10 0 3. 4. 5. 6. 7. VDD[V]. Note : ∆Tch : Temperature Rise from Backside of Package to Channel MTTF vs. Tch 1.0E+12 1.0E+11 Ea=1.56eV. 1.0E+10. MTTF (hrs.). 1.0E+09 1.0E+08 1.0E+07 1.0E+06 1.0E+05 1.0E+04 1.0E+03 1.0E+02 1.0E+01 50. 100. 150 Tch (deg-C). 11. 200. 250.

(65) EMM5074VU C-band Power Amplifier MMIC. ■ Block diagram. PIN ASSIGNMENT 1 : VGG 2 : RF in 3 : VGG 4 : VDD 5 : RF out 6 : VDD. ■ Recommended Bias Circuit. D N G. 1000pF 1uF. VGG. 4. 2. 5. 1. 6. 1000pF. VDD. VDD. D N G. 1uF. 3. t u o F R. n i F R. VGG. 1uF 1000pF. 1000pF. 1uF. Note 1: The capacitors are recommended on the bias supply line, close to the package, in order to prevent video oscillations which could damage the module. Note 2: Two pins named VGG are internally connected. Note 3: Two pins named VDD are internally connected.. 12.

(66) EMM5074VU C-Band Power Amplifier MMIC ■ Package Outline. 13.

(67) EMM5074VU C-band Power Amplifier MMIC ■ PCB Pads and Solder-resist Pattern. Notes : 1.LAMINATE : Rogers Corporation RO4003, Thickness t=0.2mm, Cu Foil 18μm Finish to copper foil ; Ni 0.1μm min./Au 0.1±0.08μm (Both side) 2. : Resist. 14.

(68) EMM5074VU C-Band Power Amplifier MMIC ■ Mounting Instructions for VU Package for Lead-free solder Mounting Condition. 1. For soldering, Lead-free solder (Sn-3.0Ag-0.5Cu)*1 or equivalent shall be used. (*1:The figure displays with weight %. A predominantly tin-rich alloy with 3.0% silver and 0.5% copper.) 2. A rosin type flux with a chlorine content of 0.2% or less shall be used. The rosin flux with low halogen content is recommended. 3. When soldering, use one of the following time/ temperature methods for acceptable solder joints. Make sure the devices have been properly prepared with flux prior soldering. * Reflow soldering method (Infrared reflow / Heat circulation reflow / Hot plate reflow): Limit solder to 3 reflow cycles because resin is used in the modules manufacturing process. Excessive reflow will effect the resin resulting in a potential failure or latent defect. The recommended reflow temperature profile is shown below. The temperature of the reflow profile must be measured at the device lead.. Temperature (deg-C). Reflow temperature profile and condition:. 260 250 220 200 140 (2) RT. (1). (4) (3) Time. (1) Temperature rise: 5deg-C/sec. (2) Preheating: 140 - 220deg-C (3) Main heating: 220deg-C over. (4) Main heating: 250deg-C over. * Measurement point: Device lead.. 60 - 120sec. 10 - 40sec. 10 sec. (260deg-C max.). 4. The above-recommended conditions were confirmed using the manufacture’s equipment and materials. However, when soldering these products, the soldering condition should be verified by customer using their equipment and materials.. 15.

(69) EMM5074VU C-band Power Amplifier MMIC. Eudyna Devices USA Inc. 2355 Zanker Rd. San Jose, CA 95131-1138, U.S.A. TEL: +1 408 232-9500 FAX: +1 408 428-9111 Eudyna Devices Europe Ltd. Network House Norreys Drive Maidenhead, Berkshire SL6 4FJ United Kingdom TEL: +44 (0) 1628 504800 FAX: +44 (0) 1628 504888. CAUTION. Eudyna Devices Inc. products contain gallium arsenide (GaAs) which can be hazardous to the human body and the environment. For safety, observe the following procedures: ・Do not put these products into the mouth. ・Do not alter the form of this product into a gas, powder, or liquid through burning, crushing, or chemical processing as these byproducts are dangerous to the human body if inhaled, ingested, or swallowed. ・Observe government laws and company regulations when discarding this product. This product must be discarded in accordance with methods specified by applicable hazardous waste procedures.. Eudyna Devices International Srl Via Teglio 8/2 - 20158 Milano, Italy TEL: +39-02-3705 2921 FAX: +39-02-3705 2920. Eudyna Devices Inc. reserves the right to change products and specifications without notice.The information does not convey any license under rights of Eudyna Devices Inc. or others.. © 2007 Eudyna Devices Inc.. Eudyna Devices Asia Pte. Ltd. Hong Kong Branch Suite 1906B, Tower 6, China Hong Kong City 33 Canton Road, Tsimshatsui, Kowloon Hong Kong TEL: +852-2377-0227 FAX: +852-2377-3921 Eudyna Devices Inc. 1000 Kamisukiahara, showa-cho Nakakomagun, Yamanashi 409-3883, Japan (Kokubo Industrial Park) TEL +81-55-275-4411 FAX +81-55-275-9461 Sales Division 1, Kanai-cho, Sakae-ku Yokohama, 244-0845, Japan TEL +81-45-853-8156 FAX +81-45-853-8170. 16.

(70) Bibliography [1] Davic M. Pozar. Microwave Engineering. fourth edition, 2013. [2] E.C Niehenke, R.A Pucel, and I.J Bahl. Microwave and milimeter-wave integrated circuits. [3] I. Bah and P. Bhartia. Microwave Solid State Circuit Design. [4] David Wisell. Measurement Techniques for Characterization of Power Amplifiers. [5] Jose Carlos Pedro and Nuno Borges Carvalho. Intermodulation Distortion in Microwave and Wireless circuits. 2003. [6] Agilent Technologies. Advanced Calibration Techniques for Vector Network Analyzers. [7] Agilent Technologies. Fundamentals of RF and Microwave Noise Figure Measurements. [8] Agilent Technologies. Two-tone and Multitone Personalities for the E8267C PSG Vector Signal Generator. [9] A.Y. Tang, E. Schlecht, R. Lin, G. Chattopadhyay, C. Lee, I. Mehdi, and J. Stake. Electro-thermal model for multi-anode schottky diode multipliers. [10] Mario Ramı́rez Torres, Marta Ferreras Mayo, Gorka Rubio Cidre, and Jesús Grajal de la Fuente. Caracterización de componentes para el programa elsa + artes 5.2..

(71) Firmado: Alberto de la Escalera Dı́az. Graduado en Ingenierı́a de Tecnologı́as y Servicios de Telecomunicación.. Madrid,. Enero de 2020.

(72)

(73)