Precalculo_-_Matematicas_para_el_calculo.pdf

Design and Implementation of a LNA in UHF band using Microstrip Rodríguez, Sebastián. Correa, Andrés. Fajardo, Arturo. Paez, Carlos Department of Electronics Pontificia Universidad Javeriana (PUJ) Bogotá, Colombia { jrodriguezp, jairo.correa, fajardoa, paez.carlos }@javeriana.edu.co Abstract— This article presents a design methodology for low noise amplifiers (LNA) in the UHF band, optimized for narrowband operation, based on a proper choice of the bias operating point, modeling of the active element through S parameters and the appropriate design of matching networks for minimum noise figure and maximum gain under stable conditions. Likewise, presents a case study related with the design of an LNA with this methodology, together with the experimental results and the results simulated with the Ansoft Designer® platform. Keywords-component; Design methodology, Amplifier, Noise Figure, Amplifier Stability.

Low

Noise

I. INTRODUCTION Most electronic applications of modern telecommunications are focused on the ability to use the wireless environment as a clear channel of communication. The reasons for the use of such technologies are transparency to the user, portability and ubiquity. The main application of low noise amplifiers (LNA) is to amplify the signal received at the Radio Frequency (RF) front end [1-5] in the receiver systems of the equipment, like cell phones where the signal received is very low (typically in the range of -110 dBm and -70 dBm [6]) and must be amplified before its demodulation. A block diagram of a typical receiver of such systems is shown in Fig. 1. During the integration of the LNA in the receiver system, the designer has to deal with various compromises between noise figure (NF), gain, linearity, impedance matching and power dissipation [6, 7]. In general, for consider the amplifier as an LNA, it must at least meet certain thresholds in the noise figure (NF), gain, stability and matching coupler [6-9].

With the development of new applications involving broadband technologies, usually implemented with OFDM modulation, has emerged a proposal of LNA design in which the main objective is to increase the bandwidth of the amplifier and its linearity, sacrificing performance in NF and gain [3, 4, 9]. Finally, due to the migration of wireless communications portable devices to the WLAN band, there have been some studies where the design of the LNA and the antenna are considered as a constitutive element called ALNA (Active Integrated Antenna with Low Noise Amplifier) [13]. In the literature there is a considerable number of designs and implementations of different LNA topologies in different technologies such as SiGe, Bipolar, GaAs, MMIC, SOI CMOS and microstrip [1-6, 10-15]. As part of the project "Design and implementation of a linear amplification system for ASK modulation in the UHF band" led by the Telecommunications Research Group (SISCOM) of the Pontificia Universidad Javeriana, it was undertook a systematic study of design methodologies for LNAs. As a partial result of this study, it is proposed a design methodology for the UHF band (300MHz-3GHz). The first part of this document shows the general topology of a LNA and a proposed design methodology along with a study case and its simulated results using the Ansoft Designer® software platform. The last part presents the conclusions found by the researchers.

II. LNA DESIGN METHODOLOGY FOR UHF The main purpose of the methodology proposed in this paper, is to present a design methodology of a narrowband LNA for the UHF band using microstrip technology, which allows synthesize the circuit from a particular optimization criterion, defined from some specific requirements of NF, gain, impedance matching and stability. The general topology of the LNA is shown in the Fig. 2, the transistor is modeled by its scattering matrix. It consists of a bias stage, a gain element, an input matching network (IMN) and an output matching network (OMN).

Fig. 1. Block diagram of a wireless equipment receiver. Taken from [6] In this context, LNA design should include other additional variables such as low cost, small size, low weight and low power dissipation [1, 2, 9, 10]. Independent of the use context of a LNA, traditional design methodologies presented in the literature, basically consist of ensuring simultaneously fulfillment of the required NF and input matching. A comparative study of some design methodologies directed to achieve this objective is presented in [10]. Another less conventional methodology aimed at the design of LNA for wireless applications, proposes to reduce energy consumption by reducing current consumption or input voltage, minimizing the impact on its other merit parameters [11, 12].

Fig. 2. General diagram of a RF amplifier. Taken from [17]

The proposed methodology is evaluated through a case of study, which is the design of a LNA amplifier using the proposed methodology, with an operating frequency of 900MHz, specifications of maximum 2.5 for VSWRin and VSWRout, gain greater than 10dB and maximum noise figure of 3dB. Both are explained step by step in the next paragraphs:

1. Choice of transistor and choice of biasing. A first step for the choice of transistor is to define the amplifier topology that best suits the design requirements. As this is a high frequency design, the parameters found in the device data sheet typically are: parameters S, MAG (Maximum Available Gain), and Rollet factor (k). These parameters allow a first feasibility analysis of the design with a specific transistor. For the case of study, the chosen is the NE68033, which has a MAG between 20dB and 19dB, greater than de desired, and a k between 0.73 and 0.85 at the desired operating frequency, the problem of this choice is to be close to a region where the transistor is no longer unconditionally stable. Generally, manufacturers do not provide data for all frequencies, or for all possible biasing schemes that can be design, so that a great limiting factor of this type of design is the availability of data provided for the device. Manufacturers also often provide two types of models to work with simulators: an approach using a nonlinear model, which is modeled as an ideal component and parasitic lumped elements; and a type of model that consists of a table of experimental data with more specific measurements of the S parameters that the simulator must interpolate, these models often contain information of the noise figure. 2. Bias circuit design. Once selected the transistor and the bias operating point, the next step is the design of the bias stage. The bias circuit must simultaneously ensure a stable operating point and a certain isolation of the RF stage. A comparative analysis of different biasing schemes is found in [16, 17]. Subsequently, the circuit is designed to meet the chosen biasing, considering that very high values of resistance naturally add more noise to the system, and very low values of resistance increase the power consumption of the system. The designed bias circuit can be seen in Fig. 3, adding the transmission lines and DC blocking capacitances.

requirements, even though it is potentially unstable. In Fig. 4 it can be seen the areas where the transistor is potentially unstable, plotted on a Smith chart, using the Smith Tool®. |S |

|∆|

|S S

∆

S S |

0.9229 and |Δ|

1

(1) 1

0.5059

|S |

18.03dB

|S |

(2) (3) (4)

Fig. 4. Stability circles.

4. First evaluation of constant gain and noise circles. At this point begins a selecting process of the operating point according to the intersection between the constant gain circles and the constant noise figure circles, to find the points at which are best satisfied both specifications. In this step, the reflection coefficients (ΓS_M , ΓL_M ) are calculated looking for maximum power transfer and minimum NF, using the equations presented in [16, 17].

Γ

B B C C

Fig. 3. Bias circuit.

|

|

_

Γ

3. Evaluating stability zones and Maximum Available Gain. The next step following the bias design, is the analysis of the transistor stability [17] [18]. Conditions in which a transistor is unconditionally stable are (1) and (2). If one of these conditions is not fulfilled, the stability conditions of the transistor must be considered in the design of the matching networks, to avoid possible oscillations. A tool to assess the stability conditions are the so-called stability circles, which can be plotted on a Smith chart in accordance with the equations proposed by [17]. Through these circles, it is possible to observe the values of the reflection indexes (Γ) at the input Γin and output Γout of the system (Fig. 2) and find unstable zones of operation. Comparing the MAG value in the chosen conditions with the gain requirements of the design, it should be analyzed the feasibility of using the selected transistor. The data provided in the model given by the manufacturer, in a S2P file, are taken as design values. The S parameter information is imported into 2-port devices in order to use the LNA design tools provided by Ansoft Designer® (as Smith Tool®) and it is analyzed its stability. Applying (1) and (2), (3) it is obtained. Therefore, the circuit is potentially unstable because k