Simulation of GaAs phemt Ultra-Wideband Low Noise Amplifier using Cascaded, Balanced and Feedback Amplifier Techniques

Transcription

1 2011 International Conference on Circuits, System and Simulation IPCSIT vol.7 (2011) (2011) IACSIT Press, Singapore Simulation of GaAs phemt Ultra-Wideband Low Noise Amplifier using Cascaded, Balanced and Feedback Amplifier Techniques A. Salleh 1, A. S. Ja afar 2, M. Z. A. Abd. Aziz 3, N. Azran 4 and R. Affendi 5 1 Faculty of Electronic and Computer Engineering, Universiti Teknikal Malaysia Melaka (UTeM), Melaka, Malaysia Abstract. This paper presents the design and simulation of Ultra-Wideband (UWB) Low Noise Amplifier (LNA) using transistor GaAs phemt Mitsubishi MGF4941AL in AWR Microwave Office simulator. LNA should be designed to trade-off noise figure, gain, bandwidth, and return loss. The aims of design are to provide enough gain along with minimum noise figure. The design using three techniques, which is cascaded amplifier, balanced amplifier and feedback amplifier are used to replace the conventional distributed amplifier that suffers from relatively high power consumption. The LNA designed operate in UWB spectrum (3.1 GHz to 10.6 GHz), Noise Figure (NF) < 3 db, constant gain > 15 db and matched over the bandwidth. By using lumped element matching, the 3-stage cascaded give lowest NF about 0.5 db to 1.04 db while feedback amplifier give highest S 21 about db with fluctuation only 0.71 db. Keywords: LNA, UWB, phemt, Noise Figure 1. Introduction The implementation of LNA at front end of receiver system is one of the challenging aspects in emerging UWB radio frequency (RF) systems. Federal Communication Commission (FCC) has licensing on the use of UWB frequency (ranging from 3.1 GHz to 10.6 GHz) due to its benefits that are transmits information using very low power, short impulses thinly spreading over a wide bandwidth, high data rate and less multipath fading [1]. The design of broadband amplifiers introduces new difficulties which require careful considerations. Basically, the design of amplifier over a broad frequency range is matter of properly designing the maximum conjugate matched circuit, balance amplifier and feedback amplifier in order to compensate for variations of frequency [2]. The techniques that mentioned above, had been employed for the broadband systems as it is being firmly establish, reliable and robust devices that can be realized in Monolithic Microwave Integrated Circuit (MMIC) and UWB technologies. In this work, this method of broadband amplifier had been choose and design using Microwave Office due to the excellent bandwidth performance, low noise figure and also this devices become very popular. This is because the input and output capacitances of the active devices are absorbed in the distributed structures. As a result, the amplifier can exhibits very low sensitivities in process variations when designing and simulate. At the first stage of the receiver, LNAs are required to have high gain and low NF. But designing of the only single stage amplifier, the high gain, low noise figure and the stability of the amplifier cannot be achieved as we needed. As for the different biasing circuit, active biasing does not offer much advantage over the passive biasing circuit. The only improvement recorded is the noise figure performance of the LNA with active biasing circuit. The matching networks can be changed to lump elements for space reduction and cost saving. LNA usually implies RF or wireless applications. But noise is also a critical consideration for 1 A. Salleh. Tel.: ; fax: address: azahari@utem.edu.my. 277

2 lower frequency analog applications. In order to avoid this, the developing of amplifier based several techniques will solve the problem above. By cascading the LNA, the signal power injected at matched input port is coupled and amplified by trans-conductance, G m of each device before at the end of the matched output will be terminate. From the perspective of a basic two-port, it can be shown that the optimum driving source susceptance for minimum NF is inductive in character, but has a capacitive variation. Furthermore, the optimum driving conductance should vary linearly with frequency. Achieving such a noise match, together with a good source impedance termination is especially difficult for wide-band systems as it involves synthesizing a network that provides these characteristics over a large frequency range [3]. Frederic Stubbe et al. [4] found few phenomenon s in their 1 GHz double stage LNA design. The LNA gain is decrease due to leakage of transconductance current of first stage transistor through internal capacitor. Noise source of second stage transistor also increase and contribute to gain decreasing. At higher frequencies, the internal gate-drain capacitor can perform as shunt-shunt feedback resulting reduces gain. Kyung Heon Koo et al. [5] had introduced another method to obtain desirable gain by combining optimum series and parallel impedances for multistage amplifier design. This technique considered the lossy matching network and its relation to the reflection coefficients. It is a common practice to use a cascode stage to obtain a gain that equal the product of the intrinsic gains of two stages. This paper focuses on the design and simulation of an UWB LNA using three different methods which the basic review of the amplifier design procedure and design issues are presented in section 2. Then section 3, discusses the simulation performance in term of S-parameter and NF. 2. Design Methodology The GaAs phemt MGF4941AL transistor from Mitsubishi Semiconductor is used in the design. This product has lower noise figure (ranging from 0.35 db to 0.5 db), high associated gain (12 db to 13.5 db), lower insertion loss in switch and lower power consumption. This advantage offers higher density and greater saturated electron velocity in GaAs channel compared with GaAs in MESFET [6]. Designing LNA consist on different steps. First of all, S-parameter simulations is done to find exact value of the S-parameters and to evaluate the stability of the model at the operating point. After that the input and output matching network are designed to make sure that maximum power is delivered when the load is matched to the line and power loss in fed is minimized [7]. The design is based on three matching techniques; lumped element, quarter wave and single stub matching. Finally the whole system is optimized to achieve the better gain and NF Stages Cascaded Amplifier The main way of determining the stability of a device is to calculate the Rollett s stability factor (K), which is calculated using a set of S-parameters for the device at the frequency of operation which started at 3.1GHz [7]. Thus, the value of K is that is less than 1, so the transistor is conditionally stable. A Smith chart plotted to determine the stable and unstable region which shown in Fig. 1. The alternative way is using bilateral decision by calculate radius and distance from the center of the Smith chart in order to find reflection coefficient at the input and output [7]. The stars indicate the point chosen for both input and output impedance. SCIR1 is input stability circle where the region inside is unstable. SCIR2 is output stability circle and the region inside is unstable region. After the stability consideration of the transistor is set, the input and output impedance can be determine which both values are transferred into appropriate matching technique variable such as resistor, capacitor, inductor, length of stub etc. The maximum gain or conjugate matching method is in the single stage amplifier design. This method will be realized when the overall gain given by transistor, G 0 provide a conjugate match between the amplifier source or load impedance and the transistor. The same procedures applied to amplifier at 8 GHz and 10.6 GHz. In order to improve the design, the optimization done by varying the variable values (in this case, the variables are value of inductor and capacitor at both input and output in every section). All the procedures also applied to design the cascaded amplifier using others matching techniques (single stub and quarter wave). 278

3 Fig. 1: Stability circle Stages Balanced Amplifier A fairly flat gain response can be obtained if the amplifier is designed for less than maximum gain, but the input and output matching will be poor. The balanced amplifier circuit solves this problem by using two 90º couplers to cancel input and output reflections from two identical amplifiers. The basic circuit of a balanced amplifier is shown in Fig. 2. The first 90º hybrid coupler divides the input signal into two equalamplitude components with a 90º phase difference, which drives the two amplifiers. The second coupler recombines the amplifier outputs. Due to the phasing properties of the hybrid coupler, reflections from the amplifier inputs cancel at the input to the hybrid, resulting in an improved impedance match; a similar effect occurs at the output of the balanced amplifier. The gain bandwidth is not improved over that of the single amplifier sections. This type of circuit is more complex than a single-stage amplifier since it requires two hybrid couplers and two separate amplifier sections, but it has a number of interesting advantages [7]. First, the individual amplifiers stages can be optimized for gain flatness or NF, without concern for input and output matching. Second, reflections are absorbed in the coupler terminations, improving input and output matching as well as the stability of the individual amplifiers. The circuit also provides a graceful degradation of a 6 db in gain if a single amplifier section fails. Lastly, the bandwidth can be octave or more, primarily limited by the bandwidth of the coupler. Fig. 2: A balanced amplifier using 90º hybrid couplers Stages Feedback Amplifier Negative feedback can be used in broadband amplifiers to provide a flat gain response and to reduce the input and output VSWR. It also controls the amplifier performance due to variations in the S-parameter from transistor to transistor, as a bandwidth requirement of the amplifier approach a decade of frequency, gain compensation based on matching network only is very difficult, and negative feedback technique are used. In fact, a microwave transistor amplifier using negative feedback can be designed to have very wide bandwidth with small gain variation. On the other side, negative feedback will degrade the noise figure and reduce the maximum power gain available from a transistor. The basic circuit configuration for single stage feedback amplifier is shown in Fig

4 RES ID=R1 R=576 Ohm IND ID=L3 L=5.73 nh PORT P=1 Z=50 Ohm 1 2 SUBCKT ID=S1 NET="mgf4941al" PORT P=2 Z=50 Ohm 3. Result and Analysis Fig. 3: Single stage feedback amplifier In this section, the comparison between three techniques in several parameters like S-parameter and NF are presented. Figure 4 shows the comparison of transducer gain, S 21 for three different design techniques. The gain for the maximum conjugate matching (3-stage cascaded amplifier is db in average with fluctuation of 3.19 db. For the 3-Stage Balance Amplifier, the value is db in average with fluctuation of 3.05 db; while 3-Stage Feedback Amplifier gives the most uniform value of gain, db with fluctuation only 0.71 db. Thus, in this analysis, the 3-Stage Feedback Amplifier is the best in most uniform gain (minimum fluctuation). The insertion loss and return loss also important in the simulation. Both values should be as minimum possible because the power fed into the LNA cannot afford to be losing as the signal fed into antenna is sufficiently small enough, maximum conjugate matching shows value of S 11 at db to db and S 22 at db to db. 3-stage balance amplifier shows a little decrease in performance compared to maximum conjugate matching, which S 11 at db to db and S 22 at -1.6 db to db. The best is 3-stage feedback amplifier that shows S 11 at db to db and S 22 at db to db. When the S 11 is above -3 db, it means that the signal fed will have loss more than half of its power, also, will produce a poor performance over system. Fig. 5 shows the NF for each design technique. For 3-stage cascaded amplifier, the value is ranging from 0.5 db to 1.04 db, better compared to other techniques. 3-stage balance amplifier gives value of NF at 0.89 db to 1.46 db, while 3-stage feedback amplifier gives value at db to db. Fig. 4: Comparison of S 21 using cascaded, balanced and feedback amplifier. 280

5 Fig. 5: Comparison of NF using cascaded, balanced and feedback amplifier. 4. Conclusion In this paper, UWB LNA was successfully simulated. For the first technique, the maximum conjugate matching, or 3- stage cascaded amplifier give a better performance in term of noise figure. However the 3- stage feedback amplifier gives the best uniform gain compared to other techniques. In term of matching techniques, lumped element matching performs better compared to quarter wave and single stub matching. All the techniques achieved the target for uniform gain > 15 db and NF < 3 db. 5. Acknowledgements The authors would like to acknowledge the funding from Short Term Grant of Universiti Teknikal Malaysia Melaka (UTeM). 6. References [1] Ultra-Wide-Band (UWB) First Report and Order Transl.: Federal Communications Commission (FCC), Feb [2] B.Y.Banyamin, M.Berwick: Analysis of the Performance of Four-Cascaded Single-Stage Distributed Amplifiers, IEEE Trans Microw. Theory Tech., Vol.48, 2000, pp [3] Dennis Ma, RF Receiver Systems and Circuits, ECE1371 Term Paper. [4] Frederic Stubbe, S. V. Kishore, C. Hull, and V. Della Torre, A CMOS RF-Receiver Front-End for 1 GHz Applications, Symposium on VLSl Circuits Digest of Technical Papers, [5] Kyung Heon Koo and Choong Woong Lee, Optimum Combination of Series and Parallel Immittances for Broadband Lossy Match Amplifier,International Conference on Circuits and Systems, June 1991, Shenzhen, China. [6] S. N. Prasad, S. Ponnala, S. Moghe, Z. M. Li, Cascaded transistor cell distributed amplifiers, Microwave and Optical Technology Letters, Vol. 12, No. 3, June , pp [7] D. M. Pozar, Microwave and RF Design of Wireless Systems, John Wiley & Sons, Inc.,