IETE Journal of Research
Home | About us | Search | Current Issue | Past Issues | Guidelines | Subscribe | ContactLogin 
IETE Journal of Research
  Users Online: 83 Print this page  Email this page Small font size Default font size Increase font size


 
ARTICLE
Year : 2009  |  Volume : 55  |  Issue : 6  |  Page : 309-314 Table of Contents   

A 1.5 V, 0.85-13.35 GHZ MMIC Low Noise Amplifier Design Using Optimization Technique


1 School of Electrical and Electronic Engineering, Universiti Sains Malaysia, 14300 Nibong Tebal, Penang, Malaysia
2 School of Computer Engineering, University Malaysia Perlis, Complex UniMAP, Jejawi, Perlis - 02600, Malaysia
3 School of Microelectronic Engineering, University Malaysia Perlis, Complex UniMAP, Jejawi, Perlis - 02600, Malaysia

Date of Web Publication18-Jan-2010

Correspondence Address:
Arjuna Marzuki
School of Electrical and Electronic Engineering, Universiti Sains Malaysia, 14300 Nibong Tebal, Penang
Malaysia
Login to access the Email id

DOI: 10.4103/0377-2063.59172

Get Permissions

   Abstract 

This paper describes how a broadband, 1.5 V, 0.85 - 13.35 GHz low noise amplifier in 0.15 mm 85 GHz PHEMT ­process is synthesized to simultaneously meet multiple design specifications such as bandwidth, noise f0 igure, power gain and power consumption. Power-constrained synthesis technique is used to design the broadband amplifier. The ­simulated peak S21 is 19.8 dB, maximum noise Figure is 2.5 dB, 3-dB bandwidth is 12.5 GHz and power consumption is 73.5 mW. The calculated Figure of merit (FOM) is better than many reported broadband low noise amplifier (LNA).

Keywords: Broadband amplifier, Integrated circuit, Monolithic microwave integrated circuit, optimization, pseudomorphic high electron mobility transistor, power-constrained.


How to cite this article:
Marzuki A, Md. Shakaff AY, Sauli Z. A 1.5 V, 0.85-13.35 GHZ MMIC Low Noise Amplifier Design Using Optimization Technique. IETE J Res 2009;55:309-14

How to cite this URL:
Marzuki A, Md. Shakaff AY, Sauli Z. A 1.5 V, 0.85-13.35 GHZ MMIC Low Noise Amplifier Design Using Optimization Technique. IETE J Res [serial online] 2009 [cited 2013 Jun 19];55:309-14. Available from: http://www.jr.ietejournals.org/text.asp?2009/55/6/309/59172


   1.Introduction Top


A broadband amplifier is in demand for ­broadband ­wireless applications such as ultra wideband ­application [1] . The broadband ­amplifier can also be applied to multiple-band ­applications such as personal communication service (PCS), international mobile telecomunication-2000 (IMT-2000) and wireless local area network (WLAN) [2] .

There are many types of broadband topology or ­architecture that can be applied for the design. Parallel architecture, which employs many dedicated signals together with band pass filters and narrow band low noise amplifiers (LNAs), is popular for broadband. However, this topology leads to bigger die size. Distributed amplifier topology can be used for the design but it also increases die size and has poor performance for LNA. A cascode topology with wideband matching which can also be used consumes die area. Other than larger die size these topologies are prone to oscillation.

Common source amplifier with resistor and ­capacitor (RC) feedback from drain to gate would provide a good solution for broadband application. This ­topology can offer stability and therefore reduce any ­possibility of oscillation. The drawback of this topology is ­undoubtedly high noise.

Most of these broadband amplifiers employ noise ­cancellation technique [3] . This technique uses auxiliary amplifier as a feedback to reduce the noise which increases the die size. Low noise ­technology such as pseudomorphic high electron mobility transistor (PHEMT) can instead be used in the design. The PHEMT technology is cheaper solution for low volume market segment. Hence, RC feedback ­topology is chosen as the topology of the broadband amplifier. The process technology of 0.15 mm, 85 GHz PHEMT is used for this work.

Simulation-based optimization technique is used to design the amplifier. This technique will reduce the ­number of design cycles. Agilent ADS software is employed for the optimization.

Target specification for the amplifier is power gain, S21 = 18 dB, noise Figure, NF = 3 dB, ­bandwidth = 12 GHz and power consumption is 70 mW. The amplifier must be stable at all frequencies. This paper is in four sections: This section introduces the challenges and the decision to use RC feedback. The cascade RC feedback amplifier section explains the set up of the optimization. This ­section also explains the cascade RC feedback ­amplifier. The third section includes results and discussion on observations. Finally, the conclusion is in the last section of this paper.

1.1 Basic Theory of Cascade Resistor and Capacitor Feedback Amplifier

[Figure 1] shows the RC feedback amplifier schematic which consists of two stages of RC feedback amplifiers. The unit gate width (UGW) of M1 and M2 is 50 mm.

RIN, CIN, ROUT, COUT, ROUT2, COUT2 are ­primarily used to stabilize the circuit. These components value will be finalized during the optimization. RS, CS, RS2, CS2 and RIN2 are used to bias M1 and M2. This ­configuration must ensure voltage at gate M1 and M2 is 0 V. To define RS and RS2, usually a current value must be decided first. One can calculate the value of these resistors using the Ids vs. Vgs curve [Figure 2], CS and CS2 are used as bypass capacitors so that AC gm will be equal to DC gm. In other words, these capacitors will short the source of the transistors to the ground, the value of reactance of this capacitor must be small at the interested frequency range. RFB, CFB, LFB and RFB2, CFB2 are used as feedback networks for M1 and M2 respectively. These components can, undoubtedly, affect performance of the whole circuit. It can affect the gain, input and output return loss. CCC is coupling capacitor. L, CC, RS, RL are external components with RS and RL equal to 50 V, CC equal to 100 pF and L equal to 0.1 μH. The core circuit of ­single-stage closed-loop gain, AV



where GF, GS, AOL is conductance of feedback ­resistance, conductance of source resistance and open-loop gain respectively. The closed-loop gain will be equal to an open-loop gain if GF approaches zero. In this case, the open loop gain is referring to transistor gain without feedback topology.

1.2 Implementation of Simulation-Based Optimization

Core-based design with parasitic-aware approach ­Marzuki et al., [4] was used in designing the monolithic microwave integrated circuit (MMIC). The flow is shown in [Figure 3]. This paper presents the activity in step circuit optimization.

To reduce input port sensitivity with respect to the load, transistor size must be determined. ­[Figure 4] and [Figure 5] are simulated result using transistors m­easurement-based model with standard s-parameter test bench as shown in [Figure 6]. [Figure 4] shows the range of ­transistor finger for S12 <-20 dB. Low S12 could reduce the sensitivity [5] .

[Figure 5] shows S21 of transistor with different ­finger values. As discussed in the previous section, the required power gain of the amplifier is 18 dB. Hence, it is decided to choose device finger which can deliver minimum gain of 14 dB through out the operating frequency.

From [Figure 4] and [Figure 5], it is clear that finger value from 2 to 10 is chosen for optimization. [Table 1] shows the design variables which will be optimized in order to achieve the required specifications. These design variables are the same for power-constrained and non-power-constrained. The current budget is 50 mA for the power constrained while no current budget for non-power-constrained.


   2.Results and Discussion Top


[Table 2] shows the optimized design variable of ideal ­components. The current consumption for non-power-constrained is 60 mA. Optimization with gradient type is used for all conditions. These ­idea­components are replaced with measurement-based modeled component from foundry. Self-resonant ­frequency for all passive components is higher than 14 GHZ.

[Figure 7], [Figure 8] and [Figure 9] show final the simulated results of modelled components from foundry. The 3-dB ­bandwidth of 12.5 GHz is shown in [Figure 7].

[Figure 8] shows noise Figure vs. frequency. Maximum noise Figure is 2.5 dB. The broadband amplifier is ­unconditionally stable; this is shown in [Figure 9]. The mu_load and mu_source are more than 1.

Monte-Carlo analysis is done for Noise Figure and S21. The analysis is done with standard deviation of 2 for RS, RS2, RFB and RFB2. The results are shown in [Figure 10] and [Figure 11]. Both results show that the parameters are still within the specification.

[Table 3] summarizes most of broadband amplifier works. Amplifiers are often compared using Figure of merit (FOM) Linten et al. [6] , the better the FOM the better is the design. The formula of FOM is shown in equation (2).



The works [7],[8],[9],[10] do not use any ­simulation-based optimization technique and non-power constrained designs. These are more on ways to meet the ­bandwidth requirement. Work [1],[12] use current as one of the ­constraints for the broadband amplifier design, but there are no reports on the performance of these works on ­non-power-constrained.

This work has shown that power constraint is ­important to get better performance for broadband amplifier; this can be seen from the FOM of power-constrained is 0.4 ­better than non-power constrained. The layout [Figure 12] is manually drawn and die size is 1.162 mm 2 . One typical GaAs wafer price is ­approximately USD 4,000. ­Therefore the die cost for 6 inch wafer is approximately 25 cents. The design is targeted to be tested using ­standard ­microwave on-wafer measurement method, thus there is no study on the design performance due to ­package ­parasitic. The layout uses back via process which ­eliminates the ground pad.


   3.Conclusion Top


Amplifier with performance of 12.5 GHz for ­3dB-­bandwidth, 73.5 mW of power consumption, a noise Figure of 2.5 dB and power gain of 19.8 dB is designed and fabricated. A simulation-based ­optimization ­technique is used to design the broadband amplifier. For this work, common source with resistor and capacitor feedback from drain to gate has been employed. Two amplifiers are arranged in cascade to achieve sufficient power gain, sensitivity of the transistor size reviewed and investigated. This step has been introduced to broadband amplifier design methodology. It is clear that for RC feedback, topology, power constraint and design technique would give ­better performance. Since the mask cost for GaAs is less than the RF CMOS, this work offers a good bargain for a ­commercialized product.


   Author Top



Arjuna Marzuki received B.Eng. in electronics from University of Sheffield, UK in 1997 and MSc from Univ. Sains Malaysia in 2004. After graduating with his first degree, he joined Hewlett-Packard as RFIC Design engineer. Currently he is a lecturer in Universiti Sains Malaysia.


Ali Yeon Md. Shakaff recieved his Ph.D. in Electronic Engineering from the University of Newcastle-upon-Tyne, England in 1987. Currently he is a Professor in Universiti Malaysia Perlis. His current research interests are in Embedded Systems Design and Sensors.


Zaliman Sauli is an associate professor in Universiti Malaysia Perlis. He was dean of School of microelectronic in Unimap for several years. His current research interests are in microelectronic.





 
   References Top

1.A. Bevilacqua, and A.M. Niknejad. "An Ultra-Wideband CMOS LNA for 3.1 to 10.6 GHz wireless receiver", Proceedings International Solid State circuit conference. 2004.  Back to cited text no. 1      
2.D. An, E.H. Rhee, J.K. Rhee, and S.D. Kim. " Design and fabrication of a wideband MMIC low-noise amplifier using Q-Matching", Journal of the Korean Physical Society, Vol. 37, No. 6, pp. 837-41, 2000.  Back to cited text no. 2      
3.E.A. Klumperink, F. Bruccoleri, P. Stroet, and B. Nauta. "Amplifiers exploiting thermal noise canceling: A Review", Proceedings 12th GaAs Symposium, pp. 371-4, 2004.  Back to cited text no. 3      
4.A. Marzuki, A. Rasmi, Z. Sauli, and A.Y. Shakaff. "Core-based design with parasitic-aware approach for medium power amplifier at 900 MHz, 2.4 GHz, 3.5 GHz and 5.85 GHz", Informacije MIDEM, Vol. 38, No. 2, pp. 131-9, 2008.  Back to cited text no. 4      
5.V. Antonocelli, R.G. Antonocelli, M. Rizzi, and B. Castagnolo. "The role of tunability in microwave synthesis", 2nd International ¬Conference on Microwave and Millimeter Wave Technology Proceedings, pp. 9-12. 2000.  Back to cited text no. 5      
6.D. Linten, S. Thijs, M.I. Natarajan, P. Wambacq, W. Jeamsaksiri, and J. Ramos, et al. "A 5GHz fully integrated ESD-protected low noise amplifier in 90 nm RFCMOS", Proceedings of the 30th ESSCC, Vol. 30, pp. 291-4, 2004.  Back to cited text no. 6      
7.M.V. Cherkashin, D. Eyllier, L.I. Babak, L. Billonnet, B. Jarry, D.A. Zait Sev, and A.V. Dyagilev. "Design of 2-10GHz feedback MMIC LNA using visual technique", European Microwave Conference ¬Proceedings, Vol. 2. 2005.  Back to cited text no. 7      
8.J. Janssens, M. Steyart, and H. Miyakawa. "A. 2.7 Volt CMOS broadband low noise amplifier", Symposium on VLSI Circuits Digest of Technical Paper, pp. 87-8, 1997.  Back to cited text no. 8      
9.A. Marzuki, T.Z. Zulkifli, N.M. Noh, and A.A. Aziz. "A broadband RF feedback amplifier design with simple feedback network", Proceedings of RF and Microwave Conference" pp. 1-4, 2004.  Back to cited text no. 9      
10.A. Ismail, and A.A. Abidi. "A 3-10GHz Low-noise amplifier with wideband lc0 -ladder matching network". IEEE Journal of solid state circuits, Vol. 39, No. 12, pp. 2269-77, 2004.  Back to cited text no. 10      
11.C.S. Wang, and C.K. Wang. "A 90 nm CMOS Low noise amplifier using noise neutralizing for 3.1-10.6 GHz UWB System", Proceedings of the 32nd European Solid State Circuits Conference, pp. 251-4. 2006.  Back to cited text no. 11      
12.A. Jajoo, M. Sperling, and T. Mukherjee. "Synthesis of a wideband low noise amplifier", Proceedings Great Lakes Symposium on VLSI, pp. 57-62, 2006.  Back to cited text no. 12      


    Figures

  [Figure 1], [Figure 2], [Figure 3], [Figure 4], [Figure 5], [Figure 6], [Figure 7], [Figure 8], [Figure 9], [Figure 10], [Figure 11], [Figure 12]
 
 
    Tables

  [Table 1], [Table 2], [Table 3]



 

Top
 
  Search
 
  
    Access Statistics
    Email Alert *
    Add to My List *
* Registration required (free)  

 
  In this article
    Abstract
    1.Introduction
    2.Results and Di...
    3.Conclusion
    Author
    References
    Article Figures
    Article Tables

 Article Access Statistics
    Viewed2196    
    Printed127    
    Emailed0    
    PDF Downloaded262    
    Comments [Add]    

Recommend this journal