Design and Performance Analysis of Low-Noise Amplifier with Band-Pass Filter for 2.4-2.5 GHz

Department of Science and Technology Institutionen för teknik och naturvetenskap

Design and Performance

Analysis of Low-Noise

Amplifier with Band-Pass

Filter for 2.4-2.5 GHz

Muneeb Mehmood Abbasi

Mohammad Abdul Jabbar

Design and Performance

Analysis of Low-Noise

Amplifier with Band-Pass

Filter for 2.4-2.5 GHz

Examensarbete utfört i Elektroteknik

vid Tekniska högskolan vid

Linköpings universitet

Muneeb Mehmood Abbasi

Mohammad Abdul Jabbar

Handledare Adriana Serban

Examinator Magnus Karlsson

Detta dokument hålls tillgängligt på Internet – eller dess framtida ersättare –

under en längre tid från publiceringsdatum under förutsättning att inga

extra-ordinära omständigheter uppstår.

Tillgång till dokumentet innebär tillstånd för var och en att läsa, ladda ner,

skriva ut enstaka kopior för enskilt bruk och att använda det oförändrat för

ickekommersiell forskning och för undervisning. Överföring av upphovsrätten

vid en senare tidpunkt kan inte upphäva detta tillstånd. All annan användning av

dokumentet kräver upphovsmannens medgivande. För att garantera äktheten,

säkerheten och tillgängligheten finns det lösningar av teknisk och administrativ

art.

Upphovsmannens ideella rätt innefattar rätt att bli nämnd som upphovsman i

den omfattning som god sed kräver vid användning av dokumentet på ovan

beskrivna sätt samt skydd mot att dokumentet ändras eller presenteras i sådan

form eller i sådant sammanhang som är kränkande för upphovsmannens litterära

eller konstnärliga anseende eller egenart.

För ytterligare information om Linköping University Electronic Press se

förlagets hemsida

http://www.ep.liu.se/

Copyright

The publishers will keep this document online on the Internet - or its possible

replacement - for a considerable time from the date of publication barring

exceptional circumstances.

The online availability of the document implies a permanent permission for

anyone to read, to download, to print out single copies for your own use and to

use it unchanged for any non-commercial research and educational purpose.

Subsequent transfers of copyright cannot revoke this permission. All other uses

of the document are conditional on the consent of the copyright owner. The

publisher has taken technical and administrative measures to assure authenticity,

security and accessibility.

According to intellectual property law the author has the right to be

mentioned when his/her work is accessed as described above and to be protected

against infringement.

For additional information about the Linköping University Electronic Press

and its procedures for publication and for assurance of document integrity,

Design and Performance Analysis of

Low-Noise Amplifier with Band-Pass

Filter for 2.4-2.5 GHz

Mohammad Abdul Jabbar

Muneeb Mehmood Abbasi

Supervisor: Dr. Adriana Serban

Examiner: Dr. Magnus Karlsson

Department of Science and Technology

Abstract

Low power wireless electronics is becoming more popular due to durability, portability and small dimension. Especially, electronic devices in instruments, scientific and medical (ISM) band is convenient from the spectrum regulations and technology availability point of view. In the communication engineering society, to make a robust transceiver is always a matter of challenges for the better performance.

However, in this thesis work, a new approach of design and performance analysis of Low-Noise Amplifier with Band-Pass filter is performed at 2.45 GHz under the communication electronics research group of Institute of Science and Technology (ITN). Band-Pass Filtered Low-Noise Amplifier is designed with lumped components and transmission lines. Performances of different designs are compared with respect to noise figure, gain, input and output reflection coefficient. In the design process, a single stage LNA is designed with amplifier, ATF-58143. Maximally flat band-pass (BPF) filters were designed with lumped components and distributed elements. Afterwards, BPF is integrated with the LNA at the front side of LNA to get a compact Band-Pass Filtered Low-Noise Amplifier with good performance.

Advanced Design System (ADS) tool was used for design and simulation, and each design was tuned to get the optimum value for noise figure, gain and input reflection coefficient. LNA stand-alone gives acceptable value of noise figure and gain but the bandwidth was too wide compared to specification. Band-Pass Filtered Low-Noise Amplifier with lumped components gives also considerable values of noise and gain. But the gain was not so flat and the bandwidth was also wide. Then, Band-Pass Filtered Low-Noise Amplifier was designed with transmission lines where the optimum value of noise figure and gain was found. The gain was almost flat over the whole band, i.e., 2.4-2.5 GHz compared to LNA stand-alone and Band-Pass Filtered Low-Noise Amplifier designed with lumped components. It is observed that deviations of results from schematic to layout level are considerable, i.e., electromagnetic simulation is needed to predict the Band-Pass Filtered Low-Noise Amplifier performance.

Prototype of LNA, Band-Pass Filtered Low-Noise Amplifier with lumped and transmission lines are made at ITN’s PCB laboratory. Due to unavailability of exact values of Murata components and for some other technical reasons, the measured values of Band-Pass Filtered Low-Noise Amplifier with lumped components and transmission lines are deviated compared to predicted values from simulation.

Acknowledgement

With all praises to the almighty and by His blessings we have finally completed this thesis. We would like to express our gratitude to Dr. Magnus Karlsson who has graciously provided us his valuable time whenever we required his assistance. His counseling, supervision and suggestions were always encouraging and it motivated us to complete the job at hand. He will always be regarded as a great mentor for us.

We would also like to thank Dr. Adriana Serban for her valuable comments and suggestions. Finally the unwavering support from our loving families was an inspiration for us and we are extremely grateful to them.

Dedicated…

Table of Contents

1

Introduction... 1

1.1

Background and Motivation ... 1

1.2

Objectives ... 2

1.3

Outline of the Thesis ... 2

2

Theoretical Background ... 4

2.1

ISM Band ... 4

2.1.1 ISM Band Operation ... 4

2.1.2 Application ... 5

2.2

Radio Receiver Basics ... 5

2.3

Network Analysis ... 6

2.3.1 Two-Port Network ... 6

2.3.2 S-Parameter ... 6

2.4

Types of Noises ... 7

2.4.1 Thermal Noise ... 7

2.4.2 Shot Noise ... 8

2.4.3 Flicker Noise ... 8

2.5

Noise Figure ... 8

2.6

Active Device: FET ... 9

2.7

Design Process of BFP-LNA ... 9

2.7.1 Band-Pass Filter ... 10

2.7.2 Low-Noise-Amplifier (LNA) ... 14

2.7.3 Matching Network between BPF and LN A ... 16

3

Design of LNA ... 18

3.1

Design Specification ... 18

3.2

Transistor Selection ... 18

3.2.1 Features ... 18

3.2.2 Applications ... 18

3.3

Q-Point Determination ... 19

3.4

DC Biasing Network ... 20

3.5

Design of LNA with S2P File ... 20

3.5.1 Stability ... 21

3.5.2 Using Ideal Components without Biasing Network ... 22

3.5.3 Using non-Ideal Components without Biasing Network ... 24

3.6

Design of LNA with Electrical Model ... 31

3.6.1 Design with Ideal Components... 31

3.6.2 Design with non-Ideal Components ... 34

3.7

Layout Design of LNA ... 36

3.7.1 Design with non-Ideal Components ... 37

4

Design of BPF-LNA ... 40

4.1

Design Specifications of BPF ... 40

4.2

Design of Maximally Flat BPF ... 40

4.2.1 Design with Lumped Components ... 40

4.2.2 Design with Distributed Elements ... 42

4.3

Design of BPF-LNA with Lumped Components ... 47

4.3.1 Schematic Design with Ideal Components ... 47

4.3.1 Layout Design ... 49

4.4

Design of BPF-LNA with Distributed Elements ... 52

4.4.1 Design of Schematic ... 52

4.4.2 Design of Layout ... 54

5

Prototypes & Measurements ... 58

5.1

Prototype of LNA ... 58

5.1.1 Measurement Results ... 59

5.2

Prototype of BPF-LNA with Lumped Elements... 60

5.2.1 Measurement Results ... 61

5.3

Prototype of BPF-LNA with Distributed Elements ... 63

5.3.1 Measurement Results ... 64

5.4

Comparison of Layouts and Measured Results ... 65

6

Conclusion and Future Works ... 66

6.1

Conclusion ... 66

6.2

Future Works... 66

List of Abbreviations

ISM Instruments Scientific and Medical

WLAN Wireless Local Area Network

LNA Low-Noise-Amplifier

BPF Band-Pass Filter

IMN Input Matching Network

OMN Output Matching Network

LPD433 Low Power Device, 433 MHz

PMR446 Private Mobile Radio, 446 MHz

CEPT European Conference of Postal and Telecommunications Administrations

ETSI European Telecommunications Standards Institute

ITU International Telecommunication Union

ITU-R The ITU Radio-communication Sector

FCC Federal Communications Commission

WDCT Digital Cordless Telecommunications

RFID Radio Frequency Identification

HiperLAN High Performance Radio LAN

Wi-Fi Wireless Fidelity

PCB Printed Circuit Board

PCS Personal Communications Service

WCDMA Wideband CDMA

ADS Advanced Design System

WLL Wireless Local Loop

ITN Department of Science and Technology

1

Introduction

Demand of wireless communication systems with robust transmitting and receiving performance is growing tremendously due to the modern technology intense society. Frequency spectrum is a natural resource as well as limited and need to be used very keenly with high attention of distribution. Instruments Scientific and Medical (ISM) band is unlicensed and becomes most popular because of its free uses. The engineering community is giving high attention as well to design devices which is compatible with this band. Cordless phone, Wireless LAN, Bluetooth, Wi-Fi all are operated in the 2.4 to 2.5 GHz.

In wireless communications, receivers need to be able to detect and amplify incoming low-power signals without adding much noise. Therefore, to filter out the unwanted signal, a Band-Pass filter (BPF) is placed before low noise amplifier (LNA)’s placement. A low noise amplifier (LNA) is often used as the first stage of these receivers. To design an LNA integrated with Band-Pass Filter (BPF), with trade-off or suitable compromise between gain and noise is always a matter of challenge.

1.1

Background and Motivation

Thesis work is a partial requirement of Master of Science in Wireless Networks and Electronics at Department of Science and Technology (ITN), Linköping University. In this thesis work, integration of Band-Pass filter with LNA will be performed where; BPF will be designed by both lumped and distributed elements. While BPF is designed with lumped components, no need to design an input matching network (IMN) in the front side of LNA, matching network between BPF and LNA will be fixed as IMN in front of LNA. In figure 1-1 a typical receiver block diagram is shown where, the BPF and LNA are put in the same block i.e. BPF will be integrated with LNA and this integrated block will be acting as a single block.

Electronic devices such as microwave oven to Bluetooth all are operated in ISM band. To keep in mind the scarcity of electromagnetic spectrum, design of equipments in ISM band is convenient for the engineering and technological entrepreneur as it is free of cost. However, in

this thesis work, it is supposed to design and analysis the performance of band-pass filtered low-noise amplifier (BPF-LNA) at 2.45 GHz with lumped and distributed elements.

Figure. 1-1 Block diagram of super-heterodyne receiver with combined BPF-LNA [1]

In general, BPF and LNA are different blocks in a receiver. Here it is tried to compact BPF with LNA in a single block which would be cost effective and have less circuit complexity and the dimension of the receiver will be reduced as well. Making a larger antenna is not cost effective rather putting an LNA to boost up the antenna signal to compensate for the feedline losses going from the antenna (outdoor) to the receiver (indoor). To design BPF-LNA, at first, it was needed to choose such a transistor which gives maximum gain and minimum noise figure (NF). ATF-58143 is selected for the whole design process.

It is highly expected that the outcome of the thesis would be highly appreciated by the industry people due to its robustness and cost effectiveness. LNA is being used in many applications such as ISM radio, cellular handset, GPS receiver, cordless phone, satellite communication and wireless LAN etc.

1.2

Objectives

The main objectives of this thesis work are following:

• Literature review on BPF and LNA

• Selection of suitable substrate for BPF-LNA

• Design and simulation of all the design in Advanced Design Tools (ADS)

• Optimization of LNA and BPF-LNA

• Fabrication of prototype of LNA and BPF-LNAs and performance analysis

• Evaluation of noise figure, gain, input and output reflection coefficient

1.3

Outline of the Thesis

Chapter 1 Describes a brief idea about the thesis background and motivation Chapter 2 Theoretical background consists of literature review

Chapter 5 Fabrication process and comparison of results of BPF-LNAs are shown Chapter 6 Concludes the thesis works and expectation of future works within this topic

2

Theoretical Background

To have a better understanding and supporting of the thesis work, a theoretical background literature is included in this part. Relevant theories are described briefly.

2.1

ISM Band

The ISM radio band is radio band (a small portion of radio spectrum) which is reserved internationally for the use of radio frequency (RF) energy for the purpose of industrial, scientific and medical equipments other than communications [2]. In general, communications equipment operating in these bands must have to tolerate any interference generated by the ISM equipments and for the case of ISM device operation, users have no regulatory protection. In spite of the intention of the original allocation, the uses of these bands become very popular for short-range communication and low power communication electronics systems.

2.1.1

ISM Band Operation

ITU-R has defined the ISM bands in 5.138, 5.150, and 5.280 of the radio regulations [3]. Due to the national radio regulations of spectrum management, individual countries' use of the bands designated in these sections may differ. Some communication devices which are using the ISM bands, it must tolerate any interference from ISM equipments. Normally unlicensed operations are allowed to use these bands, because the unlicensed operations are supposed to tolerate any external or internal interference from other devices. However, the ISM bands do have the licensed operations. Because of high possibilities of harmful interferences, licensed use of the ISM bands is not high. By the part 18 of the Federal Communications Commission (FCC), uses of ISM bands are being governed in USA, at the same time, part 15 contains the rules and regulations for unlicensed communication devices even though those use the ISM frequency bands [4].

2.1.2

Application

Microwave oven is one of most common examples of ISM device which operates at 2.45 GHz. Lately ISM bands have been shared with license-free communications applications for example 915 MHz and 2.450 GHz are for wireless sensor networks. 915 MHz, 2.450 GHz and 5.800 GHz are for wireless LNA and cordless phones respectively [3]. In radio frequency identification (RFID) applications such as biometric and contactless smart cards, ISM bands are being used widely [3].

Some low power remote control toys, gas powered cars and miniature aircraft use 2.4 GHz band range. Worldwide Digital Cordless Telecommunications (WDCT) is an ISM band technology which uses the 2.4 GHz radio spectrum. Wireless LAN devices use the following bands [3]:

• Bluetooth 2450 MHz band

• HIPERLAN 5800 MHz band

• IEEE 802.11/Wi-Fi 2450 MHz and 5800 MHz bands

2.2

Radio Receiver Basics

The super-heterodyne receiver is one of the most popular forms of receiver which is widely used today in a variety of applications from broadcast receivers to two way radio communications links as well as many mobile radio communications systems [1]. At the early stage of radio communication technology development, the super-heterodyne receiver offers many advantages in many applications.

Figure. 2-1 Block diagram of super-heterodyne receiver [1]

In this section, a typical block diagram (figure 2-1) of wireless receiver is drawn. According to this figure, the typical functionalities will be described shortly. The basic function of receiver is to recover the transmitted baseband signal by the reversing the functions of transmitter. An important component of receiver is antenna which receives the radiated electromagnetic waves from some other sources of broad frequency ranges [1]. Then the signal passes through a band-pass filter which provides some selectivity by filtering out received signals with unwanted

frequencies and passing some signals of desired frequency band. The desired signal from BPF will pass through a low-noise-amplifier (LNA). The basic function of LNA is to amplify the very weak received signal at the same time to minimize the noise power which is added to the received signals [1]. By putting a BPF in before LNA reduces the possibilities to add other interfering signals to the desired signal, this is how, the amplifier cannot be overloaded with other high power signals. The output from LNA is feed to a mixer which is used to down-convert the received radio signal to a lower frequency signal. A local oscillator (LO) is set at the level of the frequency which is near to the RF input and the output of the mixer will be relatively low and it could be filtered out by the IF band-pass filter [1]. The high gain IF amplifier raises the power level of the filtered signal thus the baseband information can be recovered without distortion [1].

2.3

Network Analysis

In this section, two-port network and S-parameter will be discussed briefly.

2.3.1

Two-Port Network

A two-port network is an electrical circuit which consists of four terminals to be connected with other external network or circuit [7]. It is represented by four variables such as at the input port

voltage, current, and at the output port voltage, and current, [8]. Figure 2-3 shows a

two-port network which has four terminals.

Figure. 2- 2 Two-port scattering network with source and load [9]

2.3.2

S-Parameter

Scattering parameters or S-parameters have significant role in RF system design. RF engineers use S-parameter to define the relationship between input-output of an electrical network in terms of incident and reflected power waves [10]. According to figure 2-2, an incident

normalized power wave, and a reflected normalized power wave,

The mathematical expression for incident and reflected normalized power wave can be written as:

= − (2) Where,

= Port 1or 2

= Characteristics impedance of the connectinglines [10]

Four waves such as , , and are related through following equations (3) and (4) where

, , and are the S-parameters of the above network [10]

= + (3)

= + (4)

Combining equation (3) and (4), the matrix form is as follows:

= (5)

Where,

= Input reflection coefficient = Input reflection coefficient = Forward voltage gain = Reversed voltage gain

2.4

Types of Noises

Noise is an undesired random disturbance in the communication systems which can degrade the useful signal [f]. It comes from natural or man-made sources. For wireless system performance evaluation, noise is an important factor to be taken into account. Normally, noise exists in all radio frequency (RF) and microwave systems. Receiver performances can be limited by the noises effect [1]. There are several parameters such as signal-to-noise ratio; dynamic range, bit error rates and minimum detectable signal level all are directly dependent on the noise effect [1]. In the following sections, some noises of electronics devices are discussed briefly:

2.4.1

Thermal Noise

Due to random thermal motions of electrons inside electronics devices generate some noises which are called thermal noise. Thermal noise is also called as Johnson–Nyquist noise [11]. Throughout the whole spectrum, the power spectral density is almost equal. The amplitude of the signal is very close to the Gaussian probability density function [11].

The electrons in a resistor are in a random motion, with a kinetic energy which is proportional to the temperature; T. Due to these random motions of these electrons, small random voltage fluctuations is produced across the terminal of the resistor. Calculations shows, the mean value of this produced voltage is zero but r.m.s. value is not zero, which can be calculated using the following equation through a narrow frequency bandwidth, B [1].

V = 4kTBR (6) Where,

k = 1.380x10) * _{J/K (Boltzmann’s constant)}

T = Temperature, degree Kelvin (°K) B = Bandwidth, Hz

R = Resistance,

2.4.2

Shot Noise

Due to thermal fluctuations of stationary charge carriers, a different type of noise is generated which is called shot noise. In case of higher frequencies and low level temperature, shot noise behaves as the dominant source of electronic noise [12]. Shot noise follows Poisson distribution, and the r.m.s. value of current fluctuations can be modeled by the following equation [13]:

+,= 2. ∆0 (7) Where, . = Charge of an electron = DC current flowing ∆0 = Bandwidth

2.4.3

Flicker Noise

Flicker noise or pink noise is inversely proportional to the frequency. At higher frequencies the noise is not considerable but at low frequency, it is troublesome [14]. Because of the imperfect contacts between conductors and semiconductor, this type of noise is generated inside electronics devices [15, 16].

This noise can be expressed by the following mathematical equation: [17]

1 = 2

34567 (8)

Where,

8 9= Oxide capacitance per unit length

: = Process dependent constant ; = Channel width

< = Channel length 0 = Frequency

2.5

Noise Figure

Noise figure (NF) is one of the most important parameters to evaluate the radio performance of communication system. It is a measurement of degradation of signal-to-noise ratio (SNR) between the input and output of the component [1].

When the network is noisy, the output noise power is greater than the output signal power; this is how, output SNR will be decreased because of high output noise power. Once the noise and desired signal are applied to the input of a noiseless network, may be both the noise and signal will be amplified or attenuated by the same degree, that’s why, SNR will not be changed [1]. The noise figure (NF) can be calculated using the following mathematical equation:

=> = ?@/B@ ? /B = ?BC@ ?BC ≥ 1 (9) => EF = 10 log => (10) Where,

,= Input signal power

=,= Input noise power

= Output signal power = = Output noise power

Using the following Friis equation, noise figure (NF) of LNA of a receiver can be obtained:

=>J J= 1 + => − 1 + BK_ML_N) + BK_MNO_M)_L + ⋯ +_MNBK_MLQ_…M)_@SN (11)

Where,

T, = Gain of each stage

=>U= Noise figure of each stage

From Friis equation (equation no. 11), it is understandable that the total noise figure =>_{J J} is

dominated by the noise figure of first stage, => which is the noise figure of the

low-noise-amplifier (LNA). Simultaneously the gain of the first stage T reduces the noise in the

consecutive stages [18].

2.6

Active Device: FET

Amplification is one of the most critical functions in all the wireless receivers and transmitters. Engineers pay high attention for designing the semiconductor transistor to get the acceptable value of amplification. Today, microwave and RF amplifiers commonly use three-terminal solid–state devices such as silicon or silicon germanium (SiGe) bipolar transistors, gallium arsenide (GaAs) field effect transistors (FETs) and high electron mobility transistors (HEMTs) etc [1]. RF and Microwave transistors are used as amplifiers which are low-cost, reliable and can be easily integrated due to high gain and low noise figure in the millimeter wave range [1].

2.7

Design Process of BFP-LNA

Band-pass filter and low-noise-amplifier have to be designed individually. Once these two blocks are designed, integration of these two blocks make a single module named BPF-LNA.

The following figure 2-4 shows the complete block of BPF-LNA where two blocks (BPF and LNA) are connected through a matching network of lumped or transmission lines.

Figure. 2-3 Complete BPF-LNA block diagram

2.7.1

Band-Pass Filter

In RF transmitter and receiver, filters are key components which is used to selectivity pass or reject signals based on frequency. Generally, there are four types of filters such as low-pass, high-pass, band-pass and band-stop filter. Combination of high-pass filter and low-pass filter make a band-pass filter (BPF) which is used to reject unwanted frequency bands and pass a narrow pass-band [1].

Normally, a pre-select BPF is setup in front of the first RF amplifier to the RF tuning range of the receiver (see figure 2-1). To make noise figure as less as possible, the filter should have low insertion loss (IL) as a result the cut-off characteristics of the filter will not be very sharp [1].

There are several classes of band-pass filter such as Butterworth or maximally flat, Chebyshev and elliptical BPF. BPF can be designed in some ways like using lumped components and distributed components. In this thesis work, maximally flat BPF is considered to design with lumped and distributed components. More details can be found in the chapter-4. Some parameters need to keep in mind during design of filters such as:

• Insertion Loss: An ideal filter has zero insertion loss (IL) when it is integrated in to the

RF circuitry as it does not introduce any power loss in the pass-band. But in practical a filter has some power loss in the pass-band. 0 dB line shows how much power is deviated which is quantified as insertion loss .It can be stated as the following mathematical equation: [10]

< = 10 VWX YZ@[

Z\] = −10 VWX 1 − ^_, ^ (12)

Where,

Pa= Power delivered to the load

P_b = Input power from the source

• Ripple: In a band-pass filter, flatness is highly desired and it can be achieved by controlling the ripple. The less difference between maximum and minimum of the amplitude of the pass-band will provide more flat band filter. Design of Chebyshev is a better way to control the magnitude of the ripple in the pass band [10].

• Bandwidth: In case of a band-pass filter, the difference between upper and lower

frequencies is defined as the bandwidth which is measured at the 3 dB attenuation. The value of the bandwidth can be written by the following expression [10]:

F;* cd _{= 0} e * cd− 05 * cd (13) Where, F;* cd _{= Bandwidth} 0e * cd = Upper Frequency 05 * cd = Lower Frequency

• Shape Factor: Sharpness is a highly expected factor in the filter design. The following

factor depicts the sharpness of the band-pass filter which is calculated using the ratio of bandwidths at 60 dB and 3 dB [10]. > =d4_d4 fg hi O hi =3j fg hi)3\ fg hi 3j O hi)3\ O hi (14) Where, > = Shape factor

F; kl cd _{= Bandwidth at 60 dB attenuation and}

F; * cd _{= Bandwidth at 3 dB attenuation}

• Rejection: Infinite number of components makes filter ideal, but its circuit becomes

more complex which is not practically convenient. That is why, in practical, finite number of components are used to design filters which is mostly specified 60 dB as the rejection rate [10].

However, it is not practically possible to make high performance band-pass filter in the integrated circuit form. Due to inherent losses of RF and microwave integrated circuits, filter experiences high insertion losses and low attenuation rates in out-band. Now-a-days, in most of the devices, off-chip filter is being used which is optimized for better performance but at the same time it is costly [1].

2.7.1.1

Lumped-Components Filter

Generally, the filters which are designed by lumped components (inductor, capacitor) are called lumped components filters. Lumped components are considered to design the filters when it is needed to reduce the dimension of the filter and if the assigned frequency band is low [10]. At high frequency, filter design with lumped components become less ideal [19]. There are some problems to design filter at higher frequencies, for example, the wavelengths become equal to the dimensions of the lumped components which causes of different types of losses and degradation of performances [10]. The terms "tee" and "pi" are used to describe lumped element

filters, and other networks. A tee element starts with a series element, while a pi network starts with a shunt element as shown below [19].

(a) (b)

Figure. 2-4 Network topology a. Pi Network low-pass filter b. Tee network high-pass filter [19] The following figures represent band-pass filters of Tee and Pi Networks of order 3

Figure. 2-3 Band-pass filter with Tee networks with order 3 [19]

2.7.1.2

Distributed-Elements Filter

A distributed element filter is an electronic filter which contains capacitance, inductance and resistance interns of transmission lines instead of conventional discrete circuit elements. The functionalities of this distributed element filters are same as conventional one [20]. To design RF and microwave circuit at higher frequency using distributed elements is convenient rather lumped elements. At high frequency, to design with lumped components have some losses because of deviation in behaviour [21].

There are two ways to convert lumped components to distributed components such as Richard

transform and Kuroda’s identity [22]. These two methods consider m

n transmission lines. To

form a lumped component from a transmission line, the width of microstrip line ( _l ) is used

[23]. There are several ways to design distributed elements filters such stub filters and coupled lines.

In this thesis work, stub filter is designed and implemented. Stub filter is implemented by using

quarter wave ( m

o ) transmission lines which is connected to the quarter wave (

m

o ) stubs [24]. A

stub behaves like a capacitor or an inductor over a narrow band and in case of wide range of frequencies it shows resonance properties. The impedance of the stub can be found by its length [25].

(a)

(b)

Figure. 2-6 Quarter wave stub resonator [22] (a) equivalent circuit of short-circuit (b) equivalent circuit of open-circuit

In figure 2-6 (a), quarter wave stub resonator of equivalent short-circuit and in figure 2-6 (b) quarter wave stub resonator of open-circuit are designed respectively. According to RF

principles, short-circuit quarter-wavelength stub works as shunt LC anti-resonators and open-circuit quarter-wavelength stub works as series LC resonator. To build complex filters, stubs can be used in combination with impedance transformers which could be most useful in case of band-pass applications [25]

Figure. 2-7 Band-pass filter using quarter wave transmission lines and short-circuit stubs [24]

In figure 2-7, a band-pass filter is shown using ( m_o ) transmission lines and quarter wave ( m_o )

short-circuit stubs. The short circuit stubs are used to pass the required frequency signal through the transmission lines by behaving as an open circuit at the joint of transmission line and stub. And short-circuit stubs behave as a short circuit at the joint of the transmission line and stubs for all other out of band frequencies

2.7.2

Low-Noise-Amplifier (LNA)

Low-noise-amplifier (LNA) is one of the most important key components of the communication system. It is used in the input stage of the receiver. It deals with two important parameters such as gain (in dB) and the noise figure [1]. In a few words, the purpose of the LNA is to amplify the received signal to acceptable levels while minimizing the noise which is added from the channel.

According to Friss equation (equation no. 11), it is very important for RF and microwave engineers to design RF receiver with low noise at the input stage. Once the signal is received by the antenna, passing through the BPF and LNA, it is not possible to get the high gain and low noise at the same time. That’s why, it is important to consider a trade-off between gain and noise figure [26].

2.7.2.1

Design Specification

Before going to design BPF-LNA, design specification should be made properly. The following things should be given attention such as bandwidth and central frequency for measurements, noise figure (NF), gain, transistor model, Q-point, source impedance, load impedance, matching network.

2.7.2.2

Transistor

In order to make an LNA, the choice of transistor is critical. This is one of the most important steps in designing a low-noise-amplifier (LNA). Different types of transistors are available for LNA applications. According to specifications, appropriate transistor should be selected for low-noise-amplifier due to its low noise figure and high gain [27]. The numbers of transistors are limited at the interested frequency. In this thesis work, ATF58413 is chosen.

2.7.2.3

Stability Analysis

Stability test is one of the most important tasks to verify whether the amplifier is stabled or not. Due to improper stability, an RF circuit approaches to be oscillated. To verify the stability of a transistor, Rollet’s conditions are used such as [10]:

: = )^?NN^L)^?LL^Lp^∆^L

^?NL^^?LN^ > 1 (15)

∆ = ^ ^^ ^ − ^ ^^ ^ (16)

If : > 1 and ^∆^ < 1 then the amplifier is stabled throughout the selected frequency band and

bias conditions.

By putting a shunt conductance or a series resistance either at input port or output port, an amplifier can be stabilized. It is recommended not to put a resistive element at the input side as it causes additional noises to be amplified. After stabilization through adding resistors, may be gain can be low or noise figure increases so it’s a trade off [10].

2.7.2.4

Q-Point Selection

The operating point of a device is known as Q-point, which is the steady-state operating condition of an active device without applying any input signal. Here, at first a suitable Q-point needs to be found for correct biasing of the transistor throughout the entire bandwidth.

2.7.2.5

DC Biasing Network

Biasing is a process of setting up the bias point at the middle of the DC load line applying drain voltage and current [27]. In a field-effect transistor (FET), bias is the DC voltage supplied from a battery which is applied at the drain. According to the selected Q-point, the biasing circuit is designed to operate the transistor at that Q-point.

2.7.2.6

Input and Output Matching Networks

Matching networks is one of the important steps to design LNA. Impedance matching is used to minimize the reflections and obtain an acceptable amount of noise figure and maximum gain by making the load impedance equal to the source impedance [22].To get an optimal value of input reflection coefficient, gain and noise figure (NF); input matching network is tuned and for output reflection coefficient; output matching network (OMN) is tuned. The following figure (2-7) shows a general transistor amplifier circuit where IMN and OMN are designed with the transistor.

Figure. 2-8 A general transistor amplifier circuit [1]

Generally it is not possible to obtain both minimum noise figure and maximum gain for an amplifier. So, some sort of compromise must be made. This can be done by using constant gain circles and circle of constant noise figure to select a usable trade-off (check it from book, trade off: up-down or compromise: linear) between noise figure and gain [1]. IMN and OMN can be designed by lumped and distributed components. More details will be discussed in Chapter: 3.

Four parameters are considered to check the design of LNA such as gain ( ), noise figure

(NF) and input reflection coefficient ( )

2.7.3

Matching Network between BPF and LNA

Integration of band-pass filter (BPF) and low-noise-amplifier (LNA) can be performed using matching network which is shown in the figure 2-4. This matching network can be designed using lumped elements or quarter-wave transmission lines.

2.7.3.1

Matching Network with Lumped Components

There are different topologies of matching networks which can be designed by lumped elements such as T-networks, Pi-network and L-network. In this thesis work, T-network is used as connector between BPF and LNA.

(a) (b)

Figure. 2-9 Topology a. Pi network low-pass filter b. Tee network high-pass filter [19]

2.7.3.2

Matching Network with Distributed Elements

The connection between BPF and LNA can be made using quarter transmission line as well. In this case, IMN of LNA is removed and matching network by quarter-wave transmission line is

Figure.2-10.Input and load impedance matched through m

o line [10]

l can be determined using the following equation [10]:

l= 5 , (17)

Where,

l = Characteristic impedance of the line

, = Impedance from BPF

5 = Impedance from LNA

Once, _l is calculated, afterwards using Agilent ADS’s line calculation option, corresponding

3

Design of LNA

In this chapter, the design procedure of LNA is described step by step in the following sub-sections. The operating frequency of the design is 2.45 GHz. The design is simulated and optimized in Advanced Design System (ADS)

3.1

Design Specification

The design specifications for the low noise amplifier are as follows:

• Gain > 15.5 dB

• Noise Figure < 0.55 dB

• Used lumped components for – matching networks

• Bandwidth: 100 MHz from 2.4 GHz to 2.5 GHz

3.2

Transistor Selection

The AVAGO Technologies’ ATF58143 is chosen for designing BPF-LNA due to its following features.

3.2.1

Features

There are some mentionable features of ATF58143 such as [28]

• Low noise and high linearity performance

• Enhancement Mode Technology

• Excellent uniformity in product specifications

• Low cost surface mount small plastic package SOT-343 in Tape-and-Reel packing

option available

• Lead-free option available

• Cellular /PCS/WCDMA base stations

• Pre-driver amplifier for 3-4 GHz WLL

• Low noise and high linearity application at 450 MHz to 6 GHz.

3.3

Q-Point Determination

The following circuit (figure 3-1) is setup for I-V characteristics simulation in Advanced Design

System (ADS). Figure 3-2 shows different I-V curves with respect to different V_st

Figure. 3-1 I-V characteristics simulation setup in ADS

Figure. 3-2 I-V curves of ATF-58143

For this thesis work, such a Q point is chosen according to data sheet in which it is possible to get the minimum noise figure at 2.45 GHz which is the central frequency. In the data sheet, it is

S1 G D S2 ATF58143_ADS_model IDS SRC3 X2 SRC4 Vdc=VDS Vdc=VGS 1 2 3 4 0 5 20 40 0 60 VGS=0.450 VGS=0.472 VGS=0.494 VGS=0.516 VGS=0.538 VGS=0.550 Vds (V) Id s ( m A )

seen that, at this Q point (V_ut= 3 V, I_ut= 30 mA), the minimum Noise Figure NF_{b and the Gain are 0.55 dB and 16.5 dB respectively.

3.4

DC Biasing Network

Figure 3-3 shows the setup for the desired Q point at V_st = 0.516 V.

Figure. 3-3 DC biasing network setup in ADS

In order to get the Q point, the above circuit is designed. At drain, it is needed to have I_ut = 30

mA. In the above circuit setup, it is 30.3 mA which is very close to desired one. The drain

voltage, according to Q point, is, V_ut = 3 V, so here, by the above setup, exactly it is found 3 V.

In order to achieve these specification (i.e. drain current, I_ut = 30 mA, and V_ut = 3 V at V_st =

0.5 V). There are three resistors used such as R1, R2 and R3. R1 and R2 are used for voltage

divider and by changing their values, getting the gate voltage, V_st = 0.51 V

3.5

Design of LNA with S2P File

In this section, LNA is designed using S2P file with ideal and no-ideal components with and

515 mV 3.30 V 3.30 V 3.00 V Gate Source Drain 95.4 uA DC S1 G D S2 -15.2 mA 0 A 30.3 mA -15.2 mA Var Eqn 30.4 mA -30.4 mA -30.4 mA 95.4 uA VAR DC R I_Probe ATF58143_ADS_model R V_DC R VAR1 DC1 R2 IDS X1 R1 R3 SRC1 R=26 kOhm Vdc=VDD R=10 Ohm VDD =3.3 V R=5.4 kOhm

3.5.1

Stability

S2P file is used to check the stability of the transistor. First the S2P file is run alone and it is found that in the whole bandwidth (BW) i.e. from 2.4 GHz to 2.5 GHz, the transistor is

unstable, because the value of stability factor is : < 1

Figure. 3-4 Schematic for stability test

In order to make it stabled, a series resistor is connected in front of the drain and by changing its

different values it was found that stability factor becomes, : > 1only in the central frequency

(at 2.45 GHz). To get stability factor : > 1 in the whole bandwidth, a shunt resistor (R1) is

connected to the drain as shown in the figure 3-4.

Figure. 3-5 Transistor stability test

When a 100 shunt resistor is connected to the drain, the value of stability factor, : > 1

through the whole bandwidth as shown in figure 3-5. Along to x-axis frequency and along to

y-axis stability factor are plotted.

Drain Source Gate 2 1 1 1 1 1 1 2 1 2 Ref 1 2 1 2 3_S2P SNP1 Term2 Term1 R1 R=100 Ohm Z=Z0 Ohm Z=Z0 Ohm

2.2

2.4

2.6

2.8

2.0

3.0

Frequency (GHz)

3.5.2

Using Ideal Components without Biasing Network

In the following circuit, input matching network (IMN) and output matching network (OMN) are designed by using Smith chart tool in ADS.

Figure. 3-6 Schematic with ideal components without biasing network

According to this figure, several parameters will be discussed such as forward voltage gain

, noise figure (NF), input reflection coefficient, . The circuit is optimized in order to

get required noise figure (NF) and power gain.

Figure. 3-7 Simulation result of noise figure (NF)

In figure 3-7, along to x-axis frequency and along to y-axis noise figure are plotted. Noise figure is found as 0.57 dB and minimum noise figure (NF) can be achieved 0.55 dB at the central frequency 2.45 GHz. But if this amount of noise figure is achieved, by changing input matching

Gate Source Drain 2 1 1 1 1 1 1 1 1 2 1 2 Re f 1 2 1 2 3 2 1 2 1 1 2 1 2 S2P L1 L2 C2 C1 SNP1 Term1 Term2 R1 C=1.216 pF R= L=2.1 nH L=1 nH C=1.08 pF Z=Z0 Ohm Z=Z0 Ohm R=100 Ohm 1.5 2.0 2.5 3.0 3.5 4.0 4.5 1.0 5.0 1 2 3 4 5 0 6 Frequency (GHz) N o is e F ig u re (N F ) d B

network, the value of input reflection coefficient, goes higher than -6 dB which is undesirable.

Figure. 3-8 Simulation result of forward voltage gain

In figures3-8 and 3-9 along to x-axis frequency and along to y-axis input reflection coefficient and forward voltage gain are plotted respectively. Here, forward voltage gain is 14.6 dB and it can be achieved 17 dB at the central frequency 2.45 GHz but if forward voltage gain is increased by changing the IMN and OMN, noise figure also increases. To design LNA, main concern is to get the acceptable value of noise figure and forward voltage gain,

Figure. 3-9 Simulation result of input reflection coefficient

1.5 2.0 2.5 3.0 3.5 4.0 4.5 1.0 5.0 0 5 10 15 -5 20 Frequency (GHz) F o rw a rd V o lt a g e G a in ( S 2 1 ) d B 1.5 2.0 2.5 3.0 3.5 4.0 4.5 1.0 5.0 -8 -6 -4 -2 -10 0 Frequency (GHz) In p u t R e fl e c ti o n C o e ff ic ie n t (S 1 1 ) d B

The value of input reflection coefficient can be changed by changing the value of IMN’s components. The value of input reflection coefficient of figure 3-9 can be decreased but if the value decreases, then noise figure increases. So, there is see a trade-off between noise figure and input reflection coefficient. Practically, its value should be less than -6 dB, so here, the value is achieved which is less than -6 dB.

3.5.3

Using non-Ideal Components without Biasing Network

The following circuit is designed with non-ideal components without biasing network. The schematic was simulated and found the following responses of input reflection coefficient, noise figure and forward voltage gain. Table 1 shows the list of components used in figure 3-10.

Figure. 3-10 Schematic with non-ideal components without biasing network

Table 1 List of components

Resistor ( ) Capacitor (pF) Inductor (nH) R1 = 100 C1 = 1.2 L1 = 2.2 -- C2 = 1.1 L2 = 1.2 Gate Source Drain 2 1 1 1 1 1 1 1 1 2 1 2 Re f 1 2 1 2 3 1 2 1 2 2 1 2 1 S2P C2 C1 L1 L2 SNP1 Term1 Term2 R1 Z=Z0 Ohm Z=Z0 Ohm

Figure. 3-11 Simulation result of noise figure (NF)

Non-ideal components deviates the result from the ideal components because there are some parasitic effects involved in non-ideal components. In figure 3-11, the noise figure is found as (NF) 0.578 dB and it can be achieved as 0.56 dB which is the minimum noise figure at central frequency 2.45 GHz. If the desired amount of noise figure is achieved by changing the input matching network, the input reflection coefficient can be high.

Figure. 3-12 Simulation result of forward voltage gain

1.5 2.0 2.5 3.0 3.5 4.0 4.5 1.0 5.0 2 4 6 8 10 0 12 Frequency (GHz) N o is e F ig u re (N F ) d B 1.5 2.0 2.5 3.0 3.5 4.0 4.5 1.0 5.0 -10 -5 0 5 10 15 -15 20 Frequency (GHz) F o rw a rd V o lt a g e G a in ( S 2 1 ) d B

In figure 3-12 it is found that forward voltage gain is 14.58 dB and it can be achieved 16.70 dB at the central frequency 2.45 GHz but if this forward voltage gain is increased by changing the IMN and OMN, noise figure (NF) will be increased also. The non-ideal components gain slightly decreases as compared to ideal components and the reason is parasitic effects in non-ideal components.

Figure. 3-13. Simulation result of input reflection coefficient

Input reflection coefficient can be varied with the change of the value of IMN’s components. There is a trade-off between noise figure (NF) and input reflection coefficient. In practical,

value of should be less than 6 dB. Here, it is found that the input reflection coefficient is

-9.27 dB which is acceptable.

3.5.4

Using Ideal Components with Biasing Network

The following circuit of LNA is designed with ideal components and biasing network is added as well. After simulation of the schematic, the following responses of input reflection coefficient, noise figure and forward voltage gain are found which are described briefly.

1.5 2.0 2.5 3.0 3.5 4.0 4.5 1.0 5.0 -8 -6 -4 -2 -10 0 Frequency (GHz) In p u t R e fl e c ti o n C o e ff ic ie n t (S 1 1 ) d B

Figure. 3-14 Schematic with ideal components with biasing network

Figure. 3-15 Simulation results of noise figure (NF)

It is found that there is a very small difference in results between with and without biasing design. Because S2P file has already saved data for AC signal and applying DC voltage cannot change its results. In figure 3-15, the noise figure is 0.62 dB and it can be achieved 0.59 dB which is the minimum noise figure (NF) at the central frequency 2.45 GHz.

Drain Source Gate 2 1 1 1 1 1 2 1 2 Re f 1 2 1 2 3 1 1 2 1 1 2 2 1 1 1 1 2 1 1 2 1 2 1 2 2 1 2 1 2 1 1 2 2 1 C L S2P _C4 C2 C1 R2 L4 R3 R1 L3 L1 SRC1 L2 SNP1 Term2 Term1 R4 C3 C=8.2 pF C=8.06 pF L=1 nH R=26 kOhm L=5.6 nH R= R=10 Ohm R=5.4 kOhm L=2.525 nH R= R= C=1.244 pF Vdc=3.3 V R= Z=Z0 Ohm Z=Z0 Ohm R=100 Ohm C=1.08 pF L=2.1 nH 1.5 2.0 2.5 3.0 3.5 4.0 4.5 1.0 5.0 1 2 3 4 5 0 6 Frequency (GHz) N o is e F ig u re (N F ) d B

Figure. 3-16 Simulation result of forward voltage gain

In figure 3-16 the forward voltage gain, is 14.99 dB and it can be achieved 16.72 dB at

central frequency 2.45 GHz but once the forward voltage gain is increased, the noise figure (NF) will also be increased. Gain has also no effect of DC biasing.

Figure. 3-17 Simulation result of input reflection coefficient

1.5 2.0 2.5 3.0 3.5 4.0 4.5 1.0 5.0 0 5 10 15 -5 20 Frequency (GHz) F o rw a rd V o lt a g e G a in ( S 2 1 ) d B 1.5 2.0 2.5 3.0 3.5 4.0 4.5 1.0 5.0 -8 -6 -4 -2 -10 0 Frequency (GHz) In p u t R e fl e c ti o n C o e ff ic ie n t (S 1 1 ) d B

There is a trade-off between noise figure (NF) and input reflection coefficient. In practical, value of S (11) should be less than -6 dB. In figure 3-17, the value of input reflection coefficient is -9.687 dB

3.5.5

Using non-Ideal Components with Biasing Network

The following circuit is designed with non-ideal components with biasing network. After simulation the schematic, the following responses of input reflection coefficient, noise figure and forward voltage gain are found.

Figure. 3-18 Schematic with non-ideal components with biasing network

Table 2 List of components

Resistor Capacitor (pF) Inductor (nH) R1 = 5.4 k C1 = 8.0 L1 = 2.2 R2 = 26 k C2 = 1.2 L2 = 2.7 R3 = 10 C3 = 1.1 L3 = 5.6 R4 = 100 C4 = 8.2 L4 = 1.2 Gate Source Drain 2 1 2 1 2 1 2 1 1 1 1 1 1 1 1 1 1 2 1 2 Re f 1 2 1 2 3 1 2 1 2 1 2 1 2 2 1 1 2 2 1 2 1 2 1 S2P C3 C4 C2 L1 C1 L4 SRC1 L3 L2 SNP1 Term1 Term2 R3 R1 R2 R4 Vdc=3.3 V Z=Z0 Ohm Z=Z0 Ohm

Figure. 3-19 Simulation result of noise figure (NF)

In figure 3-19, along to x-axis frequency and along to y-axis noise figures are plotted. Noise figure (NF) is 0.618 dB and minimum noise figure is 0.602 at the central frequency 2.45 GHz.

Figure. 3-20 Simulation result of forward voltage gain

In figure 3-20 along to x-axis frequency and along to y-axis forward voltage gain are plotted.

Here, forward voltage gain, is 14.813 dB and it can be achieved up-to 16.42 dB at central

frequency 2.45 GHz but when the forward voltage gain increases, noise figure (NF) also

1.5 2.0 2.5 3.0 3.5 4.0 4.5 1.0 5.0 2 4 6 8 0 10 Frequency (GHz) N o is e F ig u re (N F ) d B 1.5 2.0 2.5 3.0 3.5 4.0 4.5 1.0 5.0 -10 -5 0 5 10 15 -15 20 Frequency (GHz) F o rw a rd V o lt a g e G a in ( S 2 1 ) d B

Figure. 3-21 Simulation result of input reflection coefficient

When noise figure (NF) increases, input reflection coefficient decreases, so there is a trade-off between noise figure (NF) and input reflection coefficient. In the case of LNA design, value of should be less than -6 dB. In figure, 3-21, the value of input reflection coefficient is -8.706 dB.

3.6

Design of LNA with Electrical Model

In this section, the LNA is designed with electrical model; but Electrical model does not explain the results in all the frequencies as compared to S2P file. So for layout design S2P file was used to design LNA. Biasing network is designed by using electrical model.

3.6.1

Design with Ideal Components

The following circuit is designed with ideal components. After simulation the schematic,

following responses of input reflection coefficient ( ), noise figure (NF) and forward voltage

gain ( ) are observed.

1.5 2.0 2.5 3.0 3.5 4.0 4.5 1.0 5.0 -8 -6 -4 -2 -10 0 Frequency (GHz) In p u t R e fl e c ti o n C o e ff ic ie n t (S 1 1 ) d B

Figure. 3-22 Schematic with ideal components

Figure. 3-23 Simulation result of noise figure (NF)

In figure 3-23, along to x-axis, frequency and along to y-axis noise figure are plotted. The noise figure (NF) is 0.567 dB and it can be obtained up-to 0.545 dB at central frequency 2.45 GHz. In order to achieve the desired amount of NF by changing input matching network, the value of input reflection coefficient will go high

S1 G D S2 2 1 3 4 1 2 1 2 1 1 1 1 1 2 1 2 1 2 1 1 1 2 1 2 2 1 1 1 2 2 1 1 2 1 1 2 2 1 2 1 2 1 1 2 1 ATF58143_ADS_modelC3 C1 R4 C2 L4 C5 R3 L2 L3 R1 R2 SRC1 Term2 Term1 X1 C4 L1 C=3 pF C=1.2 pF C=8.2 pF R=100 Ohm C=8.2 pF L=1.136 nH R= L=1.552 nH R= R=10 Ohm R= L=2.775 nH R= L=5.6 nH {t} R=5.4 kOhm R=26 kOhm Vdc=3.3 V Z=50 Ohm Z=50 Ohm C=1.71 pF 1.5 2.0 2.5 3.0 3.5 4.0 4.5 1.0 5.0 2 4 6 0 8 Frequency (GHz) N o is e F ig u re (N F ) d B

Figure. 3-24 Simulation result of forward voltage gain

In figure 3-24, forward voltage gain is found as 13.50 dB and it can be achieved 14.57 dB at central frequency 2.45 GHz but if this forward voltage gain increases by changing the IMN and OMN, noise figure also increases. As the aim is to design LNA, the main target is to get minimum noise figure and required forward voltage gain,

Figure. 3-25 Simulation result of input reflection coefficient

In figure 2-25, along to x-axis frequency and along to y-axis input reflection coefficient are plotted. If the value of input reflection coefficient decreases, then noise figure (NF) increases.

1.5 2.0 2.5 3.0 3.5 4.0 4.5 1.0 5.0 0 5 10 15 -5 20 Frequency (GHz) F o rw a rd V o lt a g e G a in ( S 2 1 ) d B 1.5 2.0 2.5 3.0 3.5 4.0 4.5 1.0 5.0 -12 -10 -8 -6 -4 -2 -14 0 Frequency (GHz) In p u t R e fl e c ti o n C o e ff ic ie n t (S 1 1 ) d B

So, we see a trade-off between noise figure and input reflection coefficient. Expected value of input reflection coefficient is less than -6 dB and here, it is achieved less than -6 dB.

3.6.2

Design with non-Ideal Components

The following circuit is designed with non-ideal components. After simulation of the schematic, the following responses of input reflection coefficient, noise figure and forward voltage gain are found which are described briefly.

Figure. 3-26 Schematic with non-ideal components

Table 3 List of components

Resistor Capacitor (pF) Inductor (nH) R1 = 5.4 k C1 = 8.2 L1 = 1.5 R2 = 26 k C2 = 1.6 L2 = 2.7 R3 = 10 C3 = 3.0 L3 = 5.6 R4 = 100 C4 = 1.2 L4 = 1.2 -- C5 = 8.2 -- S1 G D S2 2 1 3 4 1 2 1 2 1 1 1 1 1 2 1 2 1 2 1 1 2 1 2 1 1 1 2 1 1 2 2 1 1 2 1 2 1 2 1 2 1 2 1 2 1 ATF58143_ADS_model C5 C1 C4 C2 L3 L2 R4 L4 L1 R3 C3 R1 R2 SRC1 Term2 Term1 X1 Vdc=3.3 V Z=50 Ohm Z=50 Ohm

Figure. 2-27 Simulation result of noise figure (NF)

Non-ideal components deviates the result from the ideal components because there are some parasitic effects involved in non-ideal components. In figure 3-27, along to x-axis, frequency and along to y-axis noise figure are plotted. From the figure, the noise figure is found as (NF) 0.573 dB and the minimum noise figure is 0.556 dB at central frequency 2.45 GHz

Figure. 3-28 Simulation result of forward voltage gain

1.5 2.0 2.5 3.0 3.5 4.0 4.5 1.0 5.0 2 4 6 8 10 12 0 14 Frequency (GHz) N o is e F ig u re (N F ) d B 1.5 2.0 2.5 3.0 3.5 4.0 4.5 1.0 5.0 -10 -5 0 5 10 15 -15 20 Frequency (GHz) F o rw a rd V o lt a g e G a in ( S 2 1 ) d B

In figure 3-28, along to x-axis, frequency and along to y-axis, forward voltage gain are plotted. Here, the forward voltage gain is 13.32 dB and it can be obtained up-to 14.34 dB at the central frequency 2.45 GHz but if gain is increased, by changing the IMN and OMN, noise figure also

increases.

Figure. 3-29 Simulation result of input reflection coefficient

The value of input reflection coefficient can be changed by changing the value of input matching network (IMN) components. The expected value of input reflection coefficient is less than -6 dB and in figure 3-29, the value is achieved less than -6 dB

3.7

Layout Design of LNA

In the following sections, layout of LNA is designed with Roger’s substrate (Rogers 4350B) which specifications are [29]:

• Substratethickness, | = 0.254 ~~

• Relative dielectric constant, •€= 3.48

• Conductorthickness, • = 35 µ~

• Dielectric loss tangent, • ‚ = 0.0004

• Conductivity of conductor is 5.8 ∗ 10ll„ /~ • Conductorsurfaceroughness is 0.001~~ 1.5 2.0 2.5 3.0 3.5 4.0 4.5 1.0 5.0 -12 -10 -8 -6 -4 -2 -14 0 Frequency (GHz) In p u t R e fl e c ti o n C o e ff ic ie n t (S 1 1 ) d B

3.7.1

Design with non-Ideal Components

The following layout is designed with non-ideal components using S2P file. Figure 30 and 3-31 represent the layout and the layout symbol of LNA respectively with components which

dimension is 36.7 mm x 14 mm. A number of vias are created for better grounding. After

generation the symbol, it was simulated and the following responses of input reflection coefficient, forward voltage gain and noise figure are seen which are described briefly.

Figure. 3-30 Layout of LNA

Figure. 3-31 Layout symbol of LNA with lumped components

1 1 1 2 1 2 1 2 1 2 1 2 1 2 1 2 1 1 1 1 1 1 2 3 2 1 2 1 2 1 2 1 1 2 1 2 1 2 3 7 3 6 3 5 3 4 3 3 3 2 3 1 3 0 2 9 2 8 2 7 2 6 2 5 2 4 2 3 2 2 2 1 2 0 1 9 1 8 1 7 1 6 1 5 1 4 1 3 1 2 1 1 1 0 9 8 7 6 5 4 3 2 1 1 2 1 2 c e ll_ 2 S 2 P I _ _ 4 8 L 1 C 1 L 4 L 3 C 2 C 3 C 4 C 5 L 2 R 1 0 R 3 R 2 R 1 T e r m2 S R C 1 T e r m1 S N P 1 P a r t N u mb e r = L QG1 8 H N 1 N 8 S 0 0 P a r t N u mb e r = GQM1 8 7 5 C 2 E 8 R 2 C B 1 2 P a r t N u mb e r = L QG1 8 H N 2 N 2 S 0 0 P a r t N u mb e r = L QG1 8 H N 5 N 6 S 0 0 P a r t N u mb e r = GQM1 8 7 5 C 2 E 1 R 3 B B 1 2 P a r t N u mb e r = GQM1 8 7 5 C 2 E 3 R 0 B B 1 2P a r t N u mb e r = GQM1 8 7 5 C 2 E 1 2 0 GB 1 2 P a r t N u mb e r = GQM1 8 7 5 C 2 E 8 R 2 C B 1 2 P a r t N u mb e r = L QG1 8 H N 2 N 7 S 0 0 R = 1 0 0 Oh m R = 1 0 Oh m R = 2 6 k Oh m R = 5 . 4 k Oh m Z = 5 0 Oh m V d c = 3 . 3 V Z = 5 0 Oh m

Figure. 3-32 Layout result of noise figure (NF)

In figure 3-32, along to x-axis, frequency and along to y-axis noise figure are plotted. Here the noise figure (NF) is 0.92 dB and it can be achieved up-to 0.69 dB which is minimum noise figure at the central frequency 2.45 GHz. These results are deviated from schematic results (figure 3-19) because now transmission lines are used to connect the non-ideal components. All the parasitic effects are also considered. That is why, noise figure deviates from 0.618 dB to 0.92 dB.

Figure. 3-33 Layout result of forward voltage gain

1.5 2.0 2.5 3.0 3.5 4.0 4.5 1.0 5.0 5 10 15 0 20 Frequency (GHz) N o is e F ig u re ( N B ) d B 1.5 2.0 2.5 3.0 3.5 4.0 4.5 1.0 5.0 -10 -5 0 5 10 -15 15 Frequency (GHz) F o rw a rd V o lt a g e G a in ( S 2 1 ) d B

In figure 3-33, the forward voltage gain is 14.73 dB at the central frequency 2.45 GHz but if this forward voltage gain is increased by changing the IMN and OMN, noise figure also increases. Gain is very close to the schematic results which is 14.813 dB.

Figure. 3-34 Layout result of input reflection coefficient

When noise figure (NF) increases, input reflection coefficient decreases, so there is a trade-off

between noise figure (NF) and input reflection coefficient. As the value of is acceptable

up-to -6 dB, in figure, 3-34, the value is -15.30 dB which is less than -6 dB. Schematic results have

low noise figure as compared to layout results but at layout better value of is found as

compared to schematic level (-8.70 dB), so it’s a trade-off between noise figure and input

reflection coefficient. 1.5 2.0 2.5 3.0 3.5 4.0 4.5 1.0 5.0 -14 -12 -10 -8 -6 -4 -2 -16 0 Frequency (GHz) In p u t R e fl e c ti o n C o e ff ic ie n t (S 1 1 ) d B

4

Design of BPF-LNA

In this chapter, the procedure of design of BPF is described step by step. Afterwards, a matching network is designed to integrate LNA with BPF. Lastly, layout of BPF-LNA is designed.

4.1

Design Specifications of BPF

There are numbers of specifications have to consider to design maximally flat BPF such as stop-band frequencies, pass-stop-band frequencies, stop-stop-band attenuation, pass-stop-band attenuation and filter order. In this thesis work, stop band is set-up at 0.1 GHz and 3 GHz. Pass-band is set-up from 2.3 GHz to 2.6 GHz. Stop-band attenuation is up at 40 dB and pass-band attenuation is set-up at 3 dB. In addition, filter order is 4.

4.2

Design of Maximally Flat BPF

In this section, maximally flat band-pass filter is designed with lumped components and distributed elements.

4.2.1

Design with Lumped Components

The following circuit of maximally flat band-pass filter of order 4 is designed with ideal lumped components. The pass band is selected from 2.3-2.6 GHz and attenuation for pass-band is -3dB. Series resonators have very low impedance for the desired bandwidth which is 2.4-2.5 GHz. Parallel resonators have very high impedance for desired bandwidth to stop the signal from ground. Filter order 4 is used to design band-pass filter. Higher order of filters has higher loss due to more components but more sharp and flat response. Circuit complexity goes high as well with physical dimension. However, the schematic was simulated and the following responses of

input reflection coefficient and forward transmission are observed. This circuit is

designed alone on the required band and then it will be connected with LNA by matching network.

Figure. 4-1 Schematic of band-pass filter using lumped components

Figure. 4-2 Simulation result of input reflection coefficient

In figure 4-2 and 4-3, along to x-axis, frequency and along to y-axis input reflection coefficient

and forward transmission are plotted respectively. The value of is -120 dB at the central

frequency which is 2.45 GHz. Band-Pass filter is showing very appropriate results

for individually but when this BPF is attached with the LNA circuit using matching network

then it is needed to further optimization to get better results for whole BPF-LNA. C Term1 L4 L3 L2 L1 C4 C3 C2 C1 Term2 Z=50 Ohm R=1e-12 Ohm L=20.289939 nH R=1e-12 Ohm L=216.183563 pH R=1e-12 Ohm L=48.984245 nH R=1e-12 Ohm L=521.913291 pH C=208.765314 f F C=19.593698 pF C=86.473425 f F C=8.115975 pF Z=50 Ohm 1.5 2.0 2.5 3.0 3.5 4.0 4.5 1.0 5.0 -120 -100 -80 -60 -40 -20 -140 0 Frequency (GHz) In p u t R e fl e c ti o n C o e ff ic ie n t (S 1 1 ) d B

Figure. 4-3 Layout result of forward transmission

In figure 4-3, the forward transmission, is showing that the signal is passing at -3 dB

attenuation from 2.3 GHz to 2.6 GHz. But, the desired bandwidth is from 2.4 GHz to 2.5 GHz. As a margin, from 2.3 GHz to 2.6 GHz is selected.

4.2.2

Design with Distributed Elements

In this project work, all the measurements were performed at the central frequency 2.45 GHz. At the higher frequencies such as approximately at 1 GHz, lumped components behaves differently and that is why, use of transmission lines theory is a best option instead [26].

In this section, order 4 stub filter with maximally flat response is designed in figure 4-4. Series transmission lines have very low impedance for desired bandwidth which is from 2.4 GHz to 2.5 GHz. Parallel short circuit stubs have very high impedance for the desired bandwidth to stop the signal from ground. After simulation, the following responses of input reflection coefficient,

and forward transmission, are found. This circuit is designed alone on the required

band and then it will be connected with LNA by a matching network.

1.5 2.0 2.5 3.0 3.5 4.0 4.5 1.0 5.0 -80 -60 -40 -20 -100 0 Frequency (GHz) F o rw a rd T ra n s m is s io n ( S 2 1 ) d B

Figure. 4-4 Schematic of band-pass filter using distributed elements

Figure. 4-5 Simulation result of input reflection coefficient

In figure 4-5, the value of is -10.80 dB at the central frequency which is 2.45 GHz.

Band-Pass filter is showing very acceptable results for < -6 dB alone but when this BPF is attached

with LNA using matching network then it is needed further optimization to get better results for the overall BPF-LNA. In this case, quarter wave transmission line is used for matching between BPF and LNA. Cros1 _Cros2 TL3 TL6 T L9 TL1 TL11 T L2 T L5 T L4 T L8 T L7 T L10 T ee1 Tee2 T erm1 Term2 W4=3.574 mm W3=1.007 mm W2=3.574 mm W1=0.86 mm W4=3.574 mm W3=0.86 mm W2=3.574 mm W1=1.007 mm L=19.36 mm W=0.86 mm L=19.4678 mm W=1.007 mm L=19.36 mm W=0.86 mm L=5.1 mm W=0.73 mm L=4.75 mm W=0.73 mm L=16.91 mm W=0.613 mm L=16.91 mm W=3.574 mm L=16.91 mm W=3.574 mm L=16.91 mm W=3.574 mm L=16.91 mm W=3.574 mm L=16.91 mm W=0.613 mm W3=0.613 mm W2=0.86 mm W1=0.65 mm W3=0.613 mm W2=0.577 mm W1=0.86 mm Z=50 Ohm Z=50 Ohm 1.5 2.0 2.5 3.0 3.5 4.0 4.5 1.0 5.0 -30 -25 -20 -15 -10 -5 -35 0 Frequency (GHz) In p u t R e fl e c ti o n C o e ff ic ie n t (S 1 1 ) d B