DESIGN METHODOLOGY FOR HIGH EFFICIENCY ACTIVE RADIATORS.
High efficiency power amplifiers have been developed over the last few years, largely due to their use and application in mobile communications. In a similar way this fact has caused the development of active antennas and, more specifically, active radiators. Some authors have tried to unify the concept of high efficiency in power amplifiers associated with antennas.  However, it is only recently that a true high efficiency active patch (HEAP) antenna has been reported. [2,3] A schematic of this structure is shown in Figure 1.
To understand the basic principles and basis of these antennas, it is important to note that the fact that the amplifiers are an integral part of the antenna must be taken into account, and the operation of the amplifiers should be correctly described. For this purpose, the shape of the current and voltage variations in these devices must be known. To obtain these variations, it is necessary to synthesize the optimum output impedance for each of the harmonics to ensure that the amplifier is working in the appropriate mode.
Table 1 shows the optimum impedances that should be synthesized at the output of the device to ensure proper operation. However, other necessary parameters such as the excitation or the feed of the device to obtain the desired mode of operation are not shown.
After the impedances that should be synthesized for each one of the modes are obtained and the right excitation and feeding parameters are known, the radiator type to be used should be selected.
Carrying out a detailed study, it is easily concluded that the only appropriate radiators for use as high efficiency active radiators are the patch and linear wire (dipoles) antennas. By modifying the length and the width (in the case of the dipoles) or the shape, feeding point and excitation type (for the patches), the desired impedances can be obtained. The impedance conditions make the use of other types of radiators, such as horns, for example, inappropriate.
Moreover, although linear wire antennas, due to their inductive character, seem very adequate for systems working in C-E and E-class, this type of antenna cannot be used for systems working in other classes because it would be necessary to present other types of impedance. Therefore, their use as active antennas is not possible.
Because patches offer higher integration possibilities than linear wire antennas, the realization of HEAP antennas with patches will be discussed.
HEAP DESIGN METHODOLOGY
A possible procedure for the design methodology is described below.
In this stage, the characteristics imposed as main parameters for the design include operating frequency, radiation pattern, polarization, configuration, minimum required efficiency, EIRP and linearity.
The previous data is a guide to select the adequate solution to the problem with the aid of the displayed information shown in Table 2. In this table the relationship between the shape of the antenna and the viability for working in the desired class for high efficiency is summarized. Although the use of more complicated shapes has been proposed, it is believed that classical geometries are preferred because of their simplicity and versatility.
If Tables 1 and 2 are analyzed in more detail, it can be concluded that the most suitable modes of operation for these types of antennas are E, C-E and BAR.
Active Device Characterization
The active element is selected according to the amplification class, the effective isotropic radiated power (EIRP) and the frequency of operation. After the active device has been chosen as a function of the first stage objectives, it is necessary to characterize its behavior with different output loads not only at the operating frequency but also at the harmonics, due to the inherent lack of this knowledge in the commercial devices that are available in the market. To carry out this process, techniques based on load pull methods are usually employed.
Passive Device Characterization
The object of this stage is to characterize the impedance of the antenna in order to find the adequate calculated load impedances and to check if the appropriate shapes of voltage and current that define the amplification mode are obtained. In addition to these simulations, a measurement of the load impedance should be made.
Construction and Simulation
After the calculations and measurements of the different elements have been carried out with the help of a computer, and linear and nonlinear simulation programs, the HEAP antenna should be adjusted for good operation. With these last simulations, the value of the components is adjusted, the good operation of the design is verified and the design of the matching input network is completed. All these simulation processes end with the construction of the initial prototype on which measurements and adjustments are made.
Measurements, Adjustment and Redesign
The measurement of the main characteristic parameters of the HEAP antenna, especially the EIRP, efficiency and obtained polarization, should be performed.
After the antennas have been measured and the pertinent adjustments have been made, a simulation should be made to try to model correctly the final design for the construction of the definitive prototype.
EXAMPLE OF A LOW BIAS HEAP ANTENNA
The construction of the antenna follows the steps proposed in the previous section and in Radisic, et al.  The first step is the determination of the specifications, as listed in Table 3.
The second step is the characterization of the radiating element impedance. The selected antenna is a short-circuited ring, fed by a coaxial line due to its good performance characteristics. The input impedance of this antenna has been obtained based on a cavity model resulting from the circuit of Figure 2. It is composed of an input inductance and different resonant equivalent circuits corresponding to the radiating modes. It must be noted that [L.sub.coaxial] results from the impedance of the feeding system (remember that the drain is directly joined to the patch), which is modeled as a coaxial line with its inductive impedance. It can be seen that at high frequencies the equivalent circuit at the first resonant mode presents a high inductive impedance. Then, in accordance with Table 1, it can be concluded that the suitable impedance for even and odd harmonics can only be reached through E or BAR modes. The BAR mode has been chosen to construct the active antenna because of the low bias voltage required, a lthough the efficiency is not very high. 
The third step is the selection and characterization of the active device. The selected transistor used in the HEAP antenna is the SHF-0589 from Stanford Microdevices, which is an AlGaAs/GaAs heterojunction FET, housed in a low cost, surface-mount plastic package. HFET is an ideal choice for high dynamic range requirements because of its low output capacitance resulting in low transient time.
The design of the amplifier starts with an input matching network at the center frequency. This network is based on a high pass filter consisting of a series capacitor and shunt inductor to reduce noise effects. The selection of the device must be done according to the efficiency and output power contours as previously suggested.  As a result, it can be concluded that there is a wide set of loads that satisfy the requirements.
After the active device has been characterized, the full integration of the antenna and device is completed. It should be noted that, because a short-circuited ring is used, a small coupling capacitor has been placed over the patch to enable the feed for the active device. The gate voltage, depending on the device pinch-off voltage, [V.sub.p] and the saturated drain current, [I.sub.dss], is then adjusted for the desired value of drain current. [I.sub.d] is calculated from
[V.sub.gs] = [V.sub.p](1-[square root][I.sub.d]/[I.sub.dss]) (1)
The fourth step consists of the computer-aided design of the complete HEAP antenna using commercial software such as Touchstone[R] or Spice.[R] The use of Spice requires the knowledge of the active device model, while the microstrip antenna can be simulated according to the model previously described. The use of Touchstone requires the scattering parameters of both the transistor and microstrip antenna. Then the prototype can be mounted to prove the design.
The fifth step is the prototype construction to conclude all the previous tasks. Both the amplifier and microstrip antenna are mounted in a single structure, as shown in Figures 3 and 4. Different measurements are made to show the results obtained. The radiation pattern, [S.sub.21], [S.sub.11] and efficiency as a function of input power are shown in Figures 5, 6, 7 and 8, respectively ([S.sub.21] has been measured using an auxiliary test antenna that receives the radiated signal from the HEAP antenna and from a conventional short-circuited ring antenna). The radiation pattern is nearly the same for both antennas. The main differences between the two antennas are the gain ([S.sub.21] parameter) and efficiency. From the displayed data, it can be seen that the [S.sub.21] of the HEAP antenna is 10 dB higher than the [S.sub.21] of the conventional one at the operating frequency. Also, the efficiency is not as high as the one presented previously.  This is a logical result due to the fact that BAR mode amplifie rs have lower efficiency than mode E amplifiers. However, there is a great advantage of using BAR mode amplifiers because of the lower input power required.
Finally, Table 4 shows the main results of the BAR mode HEAP antenna.
In this article the principles of design of HEAP antennas have been presented. Two main results have been shown that are displayed in Tables 1 and 2 -- the optimal impedance that an antenna should present to become a part of a high efficiency active device, and the relation between possible realizations and the topology of the patches, respectively.
The HEAP antenna concept is first intended for communication applications and is primarily based on the integration of high efficiency amplifiers and a patch antenna. Its principle of operation is intended for the use of both BAR and E modes, but the initial requirement of low supply power restricts its use to the BAR mode at the cost of reduced efficiency. The supplied power of the BAR mode HEAP antenna is very low according to the initial requirements.
V. Gonzalez received his BS degree in physics in 1995 from the U.N.E.D. and his PhD in physics in 2001 from the Universidad Carlos III de Madrid. He is currently an assistant professor at the Technical Telecommunication School at the Universidad Politecnica de Madrid. His interests are related to microstrip antennas and microwave technology.
J.M. Rodriguez received his BS degree in physics in 1986 and his PhD in physics in 1991, both from the Universidad Complutense de Madrid. He is currently an assistant professor at the Technical Telecommunication School at the Universidad Politecnica de Madrid. His interests are related to the optical properties of semiconductors, optical communications and microwave technology.
J.E. Gonzalez received his ingeniero tecnico en equipos electronicos degree and his ingeniero de telecomunicacion degree from the Universidad Palitecnica de Madrid in 1985 and 1996, respectively. From 1986 to 1989 he worked at the Meteorological Spanish Institute in Madrid. Since 1989 he has worked as an assistant professor in the department of engineering, audio visual and communications at the Universided Politenica de Madrid. His research interests include telecommunications systems and digital signal processing.
C. Rueda received his degree in ingeniero de telecomunicacion in 1990. He is currently an assistant professor at the Technical Telecomnmnication School at the Universidad Politecnica de Madrid. His interests are related to mobile communications.
C. Martin Pascual received the PhD in Physics (1976)from the University of Madrid, and the degree of Maitre en Sciences Aeronantiques (1970) from ENSA (Paris). Between 1968 and 1970 he was with the Centre d' Etudes et Recherches en Microondes (ONEBA, Paris). Since 1970 he has been with the lEC, in relation to teledetection, telecommunications, microwave links, devices and measurements, antennas and mode matching techniques. Since 1982 he has acted as a project manager for the Microwaves Department in contracts with ESTEC. In 1984 he became head of the department. Today he is the director of the TeDeCe and assistant professor at the Universidad Carlos III de Madrid.
(1.) V. Radisic, S.T. Chew, Y. Qian and T. Itoh, "High Efficiency Power Amplifier Integrated with Antenna," IEEE Microwave and Guided Letters, February 1997.
(2.) V. Gonzalez, J.M. Rodriguez, C. Rueda, I. Gomez, J.L. Jimenez and C. Martin Pascual, "Development of High Efficiency Active Patch Antennas," Microwave and Optical Technology Letters, October 2000.
(3.) V Gonzalez, J.M. Rodriguez, C. Rueda, D. Segovia, E. Rajo and C. Martin Pascual, "Low Bias BAR Modes HEAP Transmitting Antennas," Microwave and Optical Technology Letters, May 2001.
(4.) S.C. Gripps, "High Efficiency RF Power Amplifier," USA Patent 39119656, 1976.
(5.) H.L. Krauss, G.W Bostian and F.H. Raab, Solid State Radio Engineering, John Wiley & Sons, New York, 1990.
TABLE I OUTPUT IMPEDANCE OF DIFFERENT AMPLIFICATION CLASSES Maximum Fundamental Amplification Output Impedance Class [[eta].sub.out] (%) [Z.sub.fo] A saturated 63.5 Real A saturated 80 Real and overexcited B 78.5 Real C 100 Real C saturated 100 Real C_E 80 Complex D 100 Real E 100 Complex F 88 Real BAR 80 Real Even Harmonic Odd Harmonic Amplification Impedance Impedance Class [Z.sub.even harmonic] [Z.sub.odd harmonic] A saturated Short Circuit Open Circuit (SC) (OC) A saturated SC Real and overexcited B SC SC C SC SC C saturated SC SC C_E HIGH Z (OC) HIGH Z (OC) D SC OC E HIGH Z (OC) HIGH Z (OC) F SC Real BAR HIGH Z (OC) HIGH Z (OC) TABLE II AVAILABILITY OF DIFFERENT TOPOLOGIES IN PATCH ANTENNAS Amplification Rectangular Circular Open-circuit Short-circuited Class Patch Patch Ring Patch Ring Patch A saturated yes yes yes yes A saturated yes possible possible possible and overexcited B possible possible possible possible C hard difficult difficult difficult C saturated hard difficult difficult difficult C-E hard yes yes yes D hard difficult difficult difficult E hard yes yes yes F possible difficult difficult difficult BAR possible yes yes yes TABLE III HEAP ANTENNA SPECIFICATIONS Frequency (MHz) 1650 EIRP (W) =3 Efficiency (%) 60 Bias Voltage (V) 3 Antenna's shape Short-circuited ring Omnidirectional Radiation pattern in azimuth, maximum at the zenith Polarization Circular TABLE IV BAR MODE HEAP ANTENNA PERFORMANCE Maximum EIRP (W) 2.5 Maximum added efficiency (%) 58 Gain (dB) 14
|Printer friendly Cite/link Email Feedback|
|Title Annotation:||antenna power amplifiers for mobile communication systems, using a HEAP system|
|Comment:||DESIGN METHODOLOGY FOR HIGH EFFICIENCY ACTIVE RADIATORS.(antenna power amplifiers for mobile communication systems, using a HEAP system)|
|Author:||GONZALEZ, V.; RODRIGUEZ, J. M.; GONZALEZ, J. E.; RUEDA, C.; PASCUAL, C. MARTIN|
|Article Type:||Brief Article|
|Date:||Sep 1, 2001|
|Previous Article:||SPECTRUM MONITORED ADAPTIVE FEEDFORWARD LINEARIZATION.|
|Next Article:||A SIMPLE VVA RFIC DESIGN WITH A FOCUS ON REPEATABILITY AND STABILITY.|