LiU-ITN-TEK-A--12/085--SE
Design and Performance
Analysis of Low-Noise
Amplifier with Band-Pass
Filter for 2.4-2.5 GHz
Muneeb Mehmood Abbasi
Mohammad Abdul Jabbar
2012-12-12
Department of Science and Technology
Linköping University
SE- 6 0 1 7 4 No r r köping , Sw ed en
Institutionen för teknik och naturvetenskap
Linköpings universitet
6 0 1 7 4 No r r köping
LiU-ITN-TEK-A--12/085--SE
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
Norrköping 2012-12-12
Upphovsrätt
Detta dokument hålls tillgängligt på Internet – eller dess framtida ersättare –
under en längre tid från publiceringsdatum under förutsättning att inga extraordinära 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,
please refer to its WWW home page: http://www.ep.liu.se/
© Muneeb Mehmood Abbasi, Mohammad Abdul Jabbar
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
Linköping University, SE-601 74 Norrköping, Sweden
Norrköping 2012
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.
i
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.
ii
Dedicated…
To parents, sisters and brothers
iii
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
3.5.4 Using Ideal Components with Biasing Network .................... 26
3.5.5 Using non-Ideal Components with Biasing Network ............. 29
iv
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
7 References...................................................................................... 68
v
List of Abbreviations
ISM
WLAN
LNA
BPF
IMN
OMN
LPD433
PMR446
CEPT
ETSI
ITU
ITU-R
FCC
WDCT
RFID
HiperLAN
Wi-Fi
PCB
PCS
WCDMA
ADS
WLL
ITN
SMD
Instruments Scientific and Medical
Wireless Local Area Network
Low-Noise-Amplifier
Band-Pass Filter
Input Matching Network
Output Matching Network
Low Power Device, 433 MHz
Private Mobile Radio, 446 MHz
European Conference of Postal and Telecommunications Administrations
European Telecommunications Standards Institute
International Telecommunication Union
The ITU Radio-communication Sector
Federal Communications Commission
Digital Cordless Telecommunications
Radio Frequency Identification
High Performance Radio LAN
Wireless Fidelity
Printed Circuit Board
Personal Communications Service
Wideband CDMA
Advanced Design System
Wireless Local Loop
Department of Science and Technology
Surface Mounted Device
vi
CHAPTER 1
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 lowpower signals without adding much noise. Therefore, to filter out the unwanted signal, a BandPass 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
1
CHAPTER 1
this thesis work, it is supposed to design and analysis the performance of band-pass filtered lownoise 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). ATF58143 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 3 Design of LNA with ATF-58143is described in details
Chapter 4 Design of BPF-LNA with the maximally flat BPF is depicted elaborately.
2
CHAPTER 1
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
3
CHAPTER 2
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].
According to European commission’s short range device regulations, the use of the ISM band is
being governed in Europe [5]. In most of the European zones, for license-free voice
communication, LPD433 band is allowed using analog frequency modulation [6].
4
CHAPTER 2
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 bandpass filter which provides some selectivity by filtering out received signals with unwanted
5
CHAPTER 2
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 downconvert 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:
=
+
(1)
6
CHAPTER 2
=
Where,
−
(2)
= 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].
7
CHAPTER 2
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
3456
7
(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].
8
CHAPTER 2
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 +
BKL )
MN
+
BKO )
MN ML
BKQ )
N ML …M@SN
+⋯+M
(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-noiseamplifier (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.
9
CHAPTER 2
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]
Z
< = 10 VWX Y Z@[ ] = −10 VWX 1 − ^_, ^
\
(12)
Where,
Pa = Power delivered to the load
Pb = Input power from the source
^ b ^ = Reflection coefficient looking towards the filter [10]
10
CHAPTER 2
•
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 = 0e* 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 fg hi
d4 O hi
=
3jfg hi )3\fg hi
3jO 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
11
CHAPTER 2
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]
Figure 2-5 Band-pass filter with Pi networks with order 3 [19]
12
CHAPTER 2
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
m
o
) transmission lines which is connected to the quarter wave ( ) 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
13
CHAPTER 2
principles, short-circuit quarter-wavelength stub works as shunt LC anti-resonators and opencircuit 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.
14
CHAPTER 2
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 ^L p^∆^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 (27) shows a general transistor amplifier circuit where IMN and OMN are designed with the
transistor.
15
CHAPTER 2
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
placed. If the impedance of BPF, , and impedance of LNA, 5 are known, these two values
can be used to calculate the characteristics impedance, l of the of quarter-wave transmission
line.
16
CHAPTER 2
m
o
Figure.2-10.Input and load impedance matched through 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
height and width of the transmission line can be found as well.
17
CHAPTER 3
3 Design of LNA
In this chapter, the design procedure of LNA is described step by step in the following subsections. 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
3.2.2 Applications
Applications of AVAGO Technologies’ ATF-58143 are following [28]:
18
CHAPTER 3
•
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 Vst
IDS
SRC4
Vdc=VDS
SRC3
Vdc=VGS
G
S2
S1
D
ATF58143_ADS_model
X2
Figure. 3-1 I-V characteristics simulation setup in ADS
60
VGS=0.550
Ids (mA)
VGS=0.538
40
VGS=0.516
VGS=0.494
20
VGS=0.472
VGS=0.450
0
0
1
2
3
4
5
Vds (V)
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
19
CHAPTER 3
seen that, at this Q point (Vut = 3 V, Iut = 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 Vst = 0.516 V.
VAR
VAR1
VDD =3.3 V
Var
Eqn
DC
-30.4 mA
3.30 V
3.30 V
30.4 mA I_Probe
V_DC
IDS
SRC1
Vdc=VDD
DC
DC1
R
R3
R=10 Ohm
3.00 V
95.4 uA
R
R1
R=5.4 kOhm
95.4 uA
R
R2
R=26 kOhm
515 mV
Drain
G
Gate
0A
-15.2 mA
Source
-30.4 mA
S2
-15.2 mA
30.3 mA
ATF58143_ADS_model
X1
S1
D
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 Iut = 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, Vut = 3 V, so here, by the above setup, exactly it is found 3 V.
In order to achieve these specification (i.e. drain current, Iut = 30 mA, and Vut = 3 V at Vst =
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, Vst = 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
without biasing networks.
20
CHAPTER 3
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
Gate
1
1
1
Term1
Z=Z0 Ohm
2
2
Drain
Ref
1
SNP1
2
Term2
Z=Z0 Ohm
1
3 S2P
R1
R=100 Ohm
Source
2
2
1
1
1
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.
Stability Factor (K)
1.2
1.1
1.0
0.9
2.0
2.2
2.4
2.6
2.8
3.0
Frequency (GHz)
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 yaxis stability factor are plotted.
21
CHAPTER 3
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.
1
Gate
2 1
1
Term1
Z=Z0 Ohm
1
1
L1
L=2.1 nH
R=
2
Re f
3
2
Drain
1
S2P
SNP1
2
Source
2
C1
2 C=1.216 pF
1
1
2
L2
L=1 nH
1
C2
C=1.08 pF
R1
R=100 Ohm
1
1
1
Term2
Z=Z0 Ohm
1
2
2
1
1
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.
6
Noise Figure (NF) dB
5
4
3
2
1
0
1.0
1.5
2.0
2.5
3.0
3.5
4.0
4.5
5.0
Frequency (GHz)
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
22
CHAPTER 3
network, the value of input reflection coefficient,
goes higher than -6 dB which is
undesirable.
Forward Voltage Gain (S21) dB
20
15
10
5
0
-5
1.0
1.5
2.0
2.5
3.0
3.5
4.0
4.5
5.0
Frequency (GHz)
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,
Input Reflection Coefficient (S11) dB
0
-2
-4
-6
-8
-10
1.0
1.5
2.0
2.5
3.0
3.5
4.0
4.5
5.0
Frequency (GHz)
Figure. 3-9 Simulation result of input reflection coefficient
23
CHAPTER 3
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.
1
1
2
Gate
1
1
L1
Term1
Z=Z0 Ohm
1
2
3
C1
2
1
1
2
Re f
2
Drain
S2P
SNP1
1
2
1
L2
2
1
Term2
Z=Z0 Ohm
R1
Source
C2
1
1
1
1
2
2
1
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
C2 = 1.1
L1 = 2.2
L2 = 1.2
24
CHAPTER 3
12
Noise Figure (NF) dB
10
8
6
4
2
0
1.0
1.5
2.0
2.5
3.0
3.5
4.0
4.5
5.0
Frequency (GHz)
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.
Forward Voltage Gain (S21) dB
20
15
10
5
0
-5
-10
-15
1.0
1.5
2.0
2.5
3.0
3.5
4.0
4.5
5.0
Frequency (GHz)
Figure. 3-12 Simulation result of forward voltage gain
25
CHAPTER 3
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 nonideal components.
Input Reflection Coefficient (S11) dB
0
-2
-4
-6
-8
-10
1.0
1.5
2.0
2.5
3.0
3.5
4.0
4.5
5.0
Frequency (GHz)
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.
26
CHAPTER 3
1
1
R3
R=10 Ohm
SRC1
Vdc=3.3 V
1
2
R2
R=26 kOhm
2
1
R1
R=5.4 kOhm
1 2
1
1
2
1
L4
L=5.6 nH
R=
L3
L=2.525 nH
R=
2
1
1
2
C1
C=8.06 pF
2
Term1
Z=Z0 Ohm
1
1
1
2
L
L1
L=2.1 nH
R=
C
2 C2
1 C=1.244 pF
1
Gate
2
1
2
Re f
3
2
1
Drain
S2P
SNP1
2
Source
1
1
1
2
L2
L=1 nH
R=
2
1
1
1
R4
R=100 Ohm
C4
C=8.2 pF
C3
C=1.08 pF
1
2
2
1
Term2
Z=Z0 Ohm
Figure. 3-14 Schematic with ideal components with biasing network
6
Noise Figure (NF) dB
5
4
3
2
1
0
1.0
1.5
2.0
2.5
3.0
3.5
4.0
4.5
5.0
Frequency (GHz)
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.
27
CHAPTER 3
Forward Voltage Gain (S21) dB
20
15
10
5
0
-5
1.0
1.5
2.0
2.5
3.0
3.5
4.0
4.5
5.0
Frequency (GHz)
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.
Input Reflection Coefficient (S11) dB
0
-2
-4
-6
-8
-10
1.0
1.5
2.0
2.5
3.0
3.5
4.0
4.5
5.0
Frequency (GHz)
Figure. 3-17 Simulation result of input reflection coefficient
28
CHAPTER 3
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.
1
SRC1
Vdc=3.3 V
1
R3
2
1
2
1
2
1
R1
2
1
R2
1
1
L2
L3
2
1
1
2
C1
1
2
2
L1
1
1
Gate
1
Term1
2 Z=Z0 Ohm
2
Re f
3
2
S2P
SNP1
1
2 2
Drain
1
L4
1
2
R4
C2
C3
Source
2
1
1
1
2
1
1
1
1
C4
2 Term2
Z=Z0 Ohm
1
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
R2 = 26 k
C1 = 8.0
C2 = 1.2
L1 = 2.2
L2 = 2.7
R3 = 10
C3 = 1.1
L3 = 5.6
R4 = 100
C4 = 8.2
L4 = 1.2
29
CHAPTER 3
10
Noise Figure (NF) dB
8
6
4
2
0
1.0
1.5
2.0
2.5
3.0
3.5
4.0
4.5
5.0
Frequency (GHz)
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.
Forward Voltage Gain (S21) dB
20
15
10
5
0
-5
-10
-15
1.0
1.5
2.0
2.5
3.0
3.5
4.0
4.5
5.0
Frequency (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
increases.
30
CHAPTER 3
Input Reflection Coefficient (S11) dB
0
-2
-4
-6
-8
-10
1.0
1.5
2.0
2.5
3.0
3.5
4.0
4.5
5.0
Frequency (GHz)
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 (
gain (
), noise figure (NF) and forward voltage
) are observed.
31
CHAPTER 3
1
1
SRC1
Vdc=3.3 V
R1
R=5.4 kOhm
1 2
1
1
2
1
R2
2
R=26 kOhm
2
1
1
L2
L=2.775 nH
R=
1
1
2 1 1
2
2
L1
L=1.552 nH
R=
C1
C=8.2 pF
Term1
2 Z=50 Ohm
C2
2 C=1.71 pF
1
1
1
2
1
G
S2
S1
D
R3
R=10 Ohm
1
4
2
3
1
L3
L=5.6 nH {t}
R=
2
1
2 12
1
L4
L=1.136 nH
R=
C3
ATF58143_ADS_model
C=3 pF
X1
R4
2 R=100 Ohm
1
1
1
C5
C=8.2 pF
C4
2 C=1.2 pF
1
Term2
Z=50 Ohm
2
1
Figure. 3-22 Schematic with ideal components
Noise Figure (NF) dB
8
6
4
2
0
1.0
1.5
2.0
2.5
3.0
3.5
4.0
4.5
5.0
Frequency (GHz)
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
32
CHAPTER 3
Forward Voltage Gain (S21) dB
20
15
10
5
0
-5
1.0
1.5
2.0
2.5
3.0
3.5
4.0
4.5
5.0
Frequency (GHz)
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,
Input Reflection Coefficient (S11) dB
0
-2
-4
-6
-8
-10
-12
-14
1.0
1.5
2.0
2.5
3.0
3.5
4.0
4.5
5.0
Frequency (GHz)
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.
33
CHAPTER 3
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.
1
SRC1
Vdc=3.3 V
1
R3
1
2
2
2
1
1
R1
2
1
R2
1
1
L2
L3
2
1
1
2
1
1
L1
C1
2
Term1
Z=50 Ohm
C2
2
1
2
2
1
2
1
G
S1
S2
D
1
4
3
1
ATF58143_ADS_model
X1
2
1 1
2
C3
2
1
1
C5
R4
Term2
Z=50 Ohm
C4
2
1
1
1
L4
1
2
2
1
Figure. 3-26 Schematic with non-ideal components
Table 3 List of components
Resistor
R1 = 5.4 k
R2 = 26 k
Capacitor (pF)
C1 = 8.2
C2 = 1.6
Inductor (nH)
L1 = 1.5
L2 = 2.7
R3 = 10
C3 = 3.0
L3 = 5.6
R4 = 100
C4 = 1.2
L4 = 1.2
--
C5 = 8.2
--
34
CHAPTER 3
14
Noise Figure (NF) dB
12
10
8
6
4
2
0
1.0
1.5
2.0
2.5
3.0
3.5
4.0
4.5
5.0
Frequency (GHz)
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
Forward Voltage Gain (S21) dB
20
15
10
5
0
-5
-10
-15
1.0
1.5
2.0
2.5
3.0
3.5
4.0
4.5
5.0
Frequency (GHz)
Figure. 3-28 Simulation result of forward voltage gain
35
CHAPTER 3
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.
Input Reflection Coefficient (S11) dB
0
-2
-4
-6
-8
-10
-12
-14
1.0
1.5
2.0
2.5
3.0
3.5
4.0
4.5
5.0
Frequency (GHz)
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~~
36
CHAPTER 3
3.7.1 Design with non-Ideal Components
The following layout is designed with non-ideal components using S2P file. Figure 3-30 and 331 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
1 8
1 3
2
1
S RC 1
V dc =3. 3
6
V
2
1
1 7
R3
R=1 0
O
h m
2
1 6
22
1
1
R1
R= 5 . 4
24
1
1 1
k O
R2
R=2 6
hm
1
15
k O
h m
10
L 2
P ar t Num
1
Te r m
1
Z= 50
O
1
2
1
2
2
7
3
1
1 0
2
be r = LQ
G
1 8 HN 2 N 7 S0 0
L 3
P ar t Num
2
2
9
1 9
be r = LQ
G
1 8 HN 5 N 6 S0 0
29
8
1
2 0
3
11
3
2
2 3
3
32
1
34
3 6
1
T er m
Z =5 0
hm
C1
P a r t Nu m
2
C2
be r = G
Q
M
1 8 7 5 C2 E8 R2 CB 1 2
Pa r t Nu m
C3
Pa r t N u m
b er =G
Q
M
b er = G
Q
M
C4
1 8 7 5 C 2 E 3 R 0 B B1
P2
a r t Nu m
ber =G
Q
M
1 8 7 5 C2 E1 2 0 G
C5
P ar t Num
B 12
be r = G
Q
M
1 8 7 5 C 2 E 8 R 2 CB 1 2
2
O
hm
1 8 7 5 C2 E1 R3 B B1 2
2
1
1 6
2
25
24
1
1
4
L1
Pa r t Nu m
1
b er =L Q
G
2
17
2
R1 0
R= 1 0 0
1 8 H N1 N8 S0 0
3
S2 P
SNP 1
3
15
L4
O
hm
Pa r t Nu m
2
2
37
28
ber =L Q
G
1 8 H N2 N2 S0 0
2
5
21
22
1 3
c e_
l 2
I __ 48
Figure. 3-31 Layout symbol of LNA with lumped components
37
CHAPTER 3
Noise Figure (NB) dB
20
15
10
5
0
1.0
1.5
2.0
2.5
3.0
3.5
4.0
4.5
5.0
Frequency (GHz)
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.
Forward Voltage Gain (S21) dB
15
10
5
0
-5
-10
-15
1.0
1.5
2.0
2.5
3.0
3.5
4.0
4.5
5.0
Frequency (GHz)
Figure. 3-33 Layout result of forward voltage gain
38
CHAPTER 3
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.
Input Reflection Coefficient (S11) dB
0
-2
-4
-6
-8
-10
-12
-14
-16
1.0
1.5
2.0
2.5
3.0
3.5
4.0
4.5
5.0
Frequency (GHz)
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.
39
CHAPTER 4
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 stopband frequencies, pass-band frequencies, stop-band attenuation, pass-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 set-up at 40 dB and pass-band attenuation is setup 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.
40
CHAPTER 4
L2
L=48.984245 nH
R=1e-12 Ohm
L1
L=521.913291 pH
Term1
R=1e-12 Ohm
Z=50 Ohm
L4
L=20.289939 nH
R=1e-12 Ohm
C2
C=86.473425 f F
C1
C=8.115975 pF
L3
L=216.183563 pH
R=1e-12 Ohm
C
C4
C=208.765314 f F
Term2
Z=50 Ohm
C3
C=19.593698 pF
Figure. 4-1 Schematic of band-pass filter using lumped components
Input Reflection Coefficient (S11) dB
0
-20
-40
-60
-80
-100
-120
-140
1.0
1.5
2.0
2.5
3.0
3.5
4.0
4.5
5.0
Frequency (GHz)
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.
41
CHAPTER 4
Forward Transmission (S21) dB
0
-20
-40
-60
-80
-100
1.0
1.5
2.0
2.5
3.0
3.5
4.0
4.5
5.0
Frequency (GHz)
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.
42
CHAPTER 4
TL1
W=0.73 mm
L=5.1 mm
T ee1
W1=0.65 mm
W2=0.86 mm
W3=0.613 mm
T L4
W=3.574 mm
L=16.91 mm
T L7
W=3.574 mm
L=16.91 mm
Cros1
W1=0.86 mm
W2=3.574 mm
W3=1.007 mm
W4=3.574 mm
TL3
W=0.86 mm
L=19.36 mm
T erm1
Z=50 Ohm
Cros2
W1=1.007 mm
W2=3.574 mm
W3=0.86 mm
W4=3.574 mm
TL6
W=1.007 mm
L=19.4678 mm
T L2
W=0.613 mm
L=16.91 mm
Tee2
W1=0.86 mm
TL11
W2=0.577 mm
W=0.73 mm
W3=0.613 mm L=4.75 mm
Term2
Z=50 Ohm
T L9
W=0.86 mm
L=19.36 mm
T L5
W=3.574 mm
L=16.91 mm
T L8
W=3.574 mm
L=16.91 mm
T L10
W=0.613 mm
L=16.91 mm
Figure. 4-4 Schematic of band-pass filter using distributed elements
Input Reflection Coefficient (S11) dB
0
-5
-10
-15
-20
-25
-30
-35
1.0
1.5
2.0
2.5
3.0
3.5
4.0
4.5
5.0
Frequency (GHz)
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.
43
CHAPTER 4
Forward Transmission (S21) dB
0
-10
-20
-30
-40
-50
-60
-70
1.0
1.5
2.0
2.5
3.0
3.5
4.0
4.5
5.0
Frequency (GHz)
Figure. 4-6 Layout result of forward transmission
In figure 4-6, the forward transmission,
is showing that the signal passing at -3 dB
attenuation is from 2.26 GHz to 3.09 GHz. But the desired bandwidth is from 2.4 GHz to 2.5
GHz. As a margin the bandwidth is selected from 2.3 GHz to 3.09 GHz because at layout it will
be left shifted.
In this part, layout of BPF is designed with the previous specifications of section 3.7. The
following layout is designed for stub filter which is shown in figure 4-4. Figure 4-7 and 4-8
represent the layout of the stub filter and layout symbol of stub-filter respectively which
dimension is 56.90 mm x 38.63 mm. A number of vias are created for better grounding. After
generation the symbol, it was then simulated in the schematic window and got the following
responses of input reflection coefficient,
and forward transmission,
44
CHAPTER 4
Figure. 4-7 Layout of BPF
Ter m 2
Z=50
O hm
Ter m 1
Z=50 O hm
BPF Layout TX af t er m eet n
i g 17oct
I __1
Figure. 4-8 Layout symbol of BPF
45
CHAPTER 4
Forward Transmission (S21) dB
0
-20
-40
-60
-80
-100
1.0
1.5
2.0
2.5
3.0
3.5
4.0
4.5
5.0
Frequency (GHz)
Figure. 4-9 Layout result of forward transmission
In figure 4-9, the forward transmission,
is showing that the signal passing at -3 dB
attenuation is from 2.0 GHz to 2.67 GHz. As already mentioned in the schematic of this filter
that the signal at -3 dB will be shifted towards left side, so in this case it is seen that it moves
from (2.3-3.09) GHz to (2.0-2.67) GHz. The required bandwidth which is (2.4-2.5) GHz is still
in the range of -3 dB attenuation.
Input Reflection Coefficient (S11) dB
0
-10
-20
-30
-40
-50
1.0
1.5
2.0
2.5
3.0
3.5
4.0
4.5
5.0
Frequency (GHz)
Figure. 4-10 Simulation result of input reflection coefficient
46
CHAPTER 4
In figure 4-10, the value of
is -19.35 dB at the central frequency which is 2.45 GHz. Bandas compared to schematic level which is -
pass filter is showing very acceptable results for
10.80 dB
4.3 Design of BPF-LNA with Lumped Components
In this section, BPF-LNA is designed using lumped components. Schematic and layout designs
results are compared as well.
4.3.1 Schematic Design with Ideal Components
Order 4 lumped-filter with maximally flat response is connected with LNA through -matching
network in figure 4-11. The schematic is simulated and found the following responses of noise
figure (NF), input reflection coefficient, and forward voltage gain.
SRC1
Vdc=3.3 V
R3
R1
R2
L6
L7
1
L2
Term1
Z=50 Ohm
L1
C1
L4
C2
L3
C3
L5
C4
C5
2
Ref
L8
S2P C7
SNP1
C6
R4
C9
C8
Term2
Z=50 Ohm
Figure. 4-11 Schematic of BPF-LNA using ideal lumped components
Table 4 List of components
Resistor
Capacitor (pF)
Inductor (nH)
R1 = 5.4 k
R2 = 26 k
C1 = 8.11
C2 = 0.086
L1 = 0.52
L2 = 48.98
R3 = 10
C3 = 19.95
L3 = 0.21
R4 = 100
C4 = 0.20
L4 = 20.21
--
C5 = 2.0
L5= 1.80
--
C6 = 1.0
L6= 2.77
--
C7= 3.0
L7= 5.60
--
C8= 1.28
L8= 1.50
--
C9= 8.20
--
47
CHAPTER 4
Noise Figure (NF) dB
100
80
60
40
20
0
1.0
1.5
2.0
2.5
3.0
3.5
4.0
4.5
5.0
Frequency (GHz)
Figure. 4-12 Simulation result of noise figure (NF)
In figure 4-12, the noise figure (NF) is 0.761 dB and minimum noise figure is 0.611 dB at the
central frequency 2.45 GHz
Forward Voltage Gain (S21) dB
20
0
-20
-40
-60
-80
-100
1.0
1.5
2.0
2.5
3.0
3.5
4.0
4.5
5.0
Frequency (GHz)
Figure. 4-13 Simulation result of forward voltage gain
In figure 4-13, forward voltage gain,
is 14.28 dB at the central frequency 2.45 GHz and it is
giving almost flat gain from 2.31 GHz to 2.54 GHz. If this forward voltage gain is changed by
changing the IMN and OMN, noise figure will be also increased. The main reason for
connecting the BPF with LNA is that it gives flat gain only in the desired bandwidth.
48
CHAPTER 4
Input Reflection Coefficient (S11) dB
0
-10
-20
-30
-40
1.0
1.5
2.0
2.5
3.0
3.5
4.0
4.5
5.0
Frequency (GHz)
Figure. 4-14 Simulation result of input reflection coefficient
In figure 4-14, the value of
is -35 dB at the central frequency which is 2.45 GHz. Band-pass
is -120 dB and after connecting with LNA the overall
filter stand alone has the value of
increases but it is still acceptable because it is less than -6 dB.
4.3.1 Layout Design
With the previous substrate specifications of section 3.7, the following layout of BPF-LNA is
designed with non-ideal components. Figure 4-15 and 4-16 represent the layout and layout
symbol of BPF-LNA respectively which dimension is 51.3 mm x 14 mm. A number of vias are
created for better grounding. After generation the symbol, it was then simulated in the
schematic window and got the following responses of noise figure, input reflection coefficient,
and forward voltage gain,
Figure. 4-15 Layout of BPF-LNA
49
CHAPTER 4
SRC1
Vdc=3. 3 V
R3
R=10 O hm
R1
R=5. 4 kO hm
R2
R=26 kO hm
L6
Par t Num ber =LQ G 18HN2N7S00
C4
L4
Par t Num ber =G Q M 1875C2E1R0CB12
Par t Num ber =LQ G 18HN6N8J00
L2
C2
Par t Num ber =LQ G 18HN1N2S00
Par t Num ber =G Q M 1885C2A1R0BB01
L7
Par t Num ber =LQ G 18HN5N6S00
L8
Par t Num ber =LQ G 18HN1N2S00
C5
Par t Num ber =G Q M 1875C2E1R5CB12
C7
Par t Number =G Q M 1875C2E3R0CB12
C6
Par t Num ber =G Q M 1875C2E1R2BB12
C9
Par t Num ber =G Q M 1875C2E3R0BB12
Ter m 2
Z=50 O hm
Ter m 1
Z=50 O hm
C1
Par t Num ber =G Q M 1885C2A1R0BB01
L1
Par t Num ber =LQ G 18HN1N2S00
L5
Par t Num ber =LQ G 18HN1N2S00
L3
C3
Par t Num ber =LQ G 18HN1N
Par2S
t 00
Num ber =G Q M 1875C2E1R0CB12
R4
R=100 O hm
1
C8
Par t Number =G Q M 1875C2E1R0BB12
2
Ref
S2P
SNP1
BPFLNA lum ped new layout
I __40
Figure. 4-16 Layout symbol of BPF-LNA with lumped components
70
Noise Figure (NF) dB
60
50
40
30
20
10
0
1.0
1.5
2.0
2.5
3.0
3.5
4.0
4.5
5.0
Frequency (GHz)
Figure. 4-17 Layout result of noise figure (NF)
In figure 4-17, the noise figure (NF) is 1.37 dB and the minimum noise figure 1.051 dB can be
achieved at the central frequency 2.45 GHz. In order to achieve the desired amount of NF,
input matching network’s component should be changed which causes increase the value of
input reflection coefficient.
50
CHAPTER 4
Forward Voltage Gain (S21) dB
20
0
-20
-40
-60
-80
1.0
1.5
2.0
2.5
3.0
3.5
4.0
4.5
5.0
Frequency (GHz)
Figure. 4-18 Layout result of forward transmission
In figure 4-18, forward voltage gain is 13.26 dB at the central frequency 2.45 GHz and it is
providing almost flat gain from 2.35 GHz to 2.60 GHz. The flatness becomes more and the
bandwidth is expanded as compared to schematic level because non-ideal components are used
with very selective values for it from Murata library
Input Reflection Coefficient (S11) dB
0
-2
-4
-6
-8
-10
-12
-14
-16
1.0
1.5
2.0
2.5
3.0
3.5
4.0
4.5
5.0
Frequency (GHz)
Figure. 4-19 Simulation result of input reflection coefficient
The value of
is also changed very much from schematic level which is -10.80 dB but still
acceptable. The reason for this change is also the use of non-ideal components with very
selected components library from Murata. In this design, non-ideal components availability
51
CHAPTER 4
problem was faced in band-pass filter. That is why; a new filter is designed with transmission
line in order to solve this problem.
4.4 Design of BPF-LNA with Distributed Elements
In this section, BPF-LNA is designed using distributed elements. Schematic and layout designs
results are compared as well.
4.4.1 Design of Schematic
Order 4 stub filter with maximally flat response is connected with LNA through quarter wave
matching network in figure 4-20. The schematic is simulated and found the following responses
of noise figure (NF), input reflection coefficient and forward voltage gain.
SRC1
Vdc=3.3 V
TL7
W=3.574 mm
L=16.91 mm
TL1
W=0.73 mm
L=5.1 mm
TL4
Tee1
W1=0.65 mm W=3.574 mm
W2=0.86 mm L=16.91 mm
W3=0.613 mm
Cros1
W1=0.86 mm
W2=3.574 mm
W3=1.007 mm
W4=3.574 mm
R3
Cros2
W1=1.007 mm Tee2
W2=3.574 mm W1=0.86 mm TL11
W3=0.86 mm W2=0.577 mm W=0.73 mm
W4=3.574 mm W3=0.613 mm L=4.75 mm
R1
R2
L2
L1
1
Term1
Z=50 Ohm
TL3
W=0.86 mm
L=19.36 mm
TL2
W=0.613 mm
L=16.91 mm
TL6
W=1.007 mm
L=19.4678 mm
TL5
W=3.574 mm
L=16.91 mm
TL9
W=0.86 mm
L=19.36 mm
TL8
W=3.574 mm
L=16.91 mm
TL10
W=0.613 mm
L=16.91 mm
TL12
Subst="MSub1"
W=0.65 mm
L=18.56 mm
2
Ref
L3
S2P C1
SNP1
R4
C3
Term2
Z=50 Ohm
C2
Figure. 4-20 Schematic of BPF-LNA using distributed elements
Table 5 List of components
Resistor
Capacitor (pF)
Inductor (nH)
R1 = 5.4 k
R2 = 26 k
C1 = 3.0
C2 = 1.2
L1 = 2.7
L2 = 5.6
R3 = 10
C3 = 8.2
L3 = 1.5
R4 = 100
--
--
52
CHAPTER 4
Noise Figure (NF) dB
100
80
60
40
20
0
1.0
1.5
2.0
2.5
3.0
3.5
4.0
4.5
5.0
Frequency (GHz)
Figure. 4-21 Simulation result of noise figure (NF)
In figure 4-21, along to x-axis frequency and along to y-axis noise figure are plotted. The noise
figure (NF) is 1.193 dB and the minimum noise figure is 1.166 dB at central the frequency 2.45
GHz
Forward Voltage Gain (S21) dB
20
0
-20
-40
-60
-80
-100
1.0
1.5
2.0
2.5
3.0
3.5
4.0
4.5
5.0
Frequency (GHz)
Figure. 4-22 Simulation result of forward voltage gain
In figure 4-22, forward voltage gain is 12.93 dB at the central frequency 2.45 GHz and the gain
is almost flat from 2.24 GHz to 2.59 GHz. Once this forward voltage gain increases by changing
the IMN and OMN, noise figure increases also at the same time.
53
CHAPTER 4
Input Reflection Coefficient (S11) dB
0
-5
-10
-15
-20
-25
1.0
1.5
2.0
2.5
3.0
3.5
4.0
4.5
5.0
Frequency (GHz)
Figure. 4-23 Simulation result of input reflection coefficient
In figure 4-23, the value of
is -8.31 dB at the central frequency which is 2.45 GHz. Band-
pass filter stand-alone has the value of
overall value of
is -10.80 dB and after connecting with LNA the
increases also but it is still acceptable which is less than -6 dB
4.4.2 Design of Layout
With the previous substrate specifications of section 3.7, the following layout of BPF-LNA is
designed with distributed components. Figure 4-24 and 4-25 represent the layout and layout
symbol of BPF-LNA respectively which dimension is 97.10 mm x 39.75 mm. A number of vias
are created for better grounding. There are two big vias are created to tie-up Vuu . However,
after generation the symbol, it was then simulated in the schematic window and found the
following responses of noise figure, input reflection coefficient,
and forward voltage gain,
.
54
CHAPTER 4
Figure. 4-24 Layout of BPF-LNA with distributed components
V_ DC
SR C2
Vd c = 3 . 3
V
R3
R= 1 0
R2
R=2 6
k O
O
hm
hm
L 2
R1
R=5 . 4
L1
Pa r t Nu m
k O
ber =L Q
G
P ar t Num
1 8 H N2 N7 S0 0
be r = LQ
1
Z =5 0
O
1 8 HN 5 N 6 S 0 0
L 3
P ar t Num
C1
Pa r t N u m
T er m
G
h m
b er =G
Q
M
b er = LQ
G
1 8 HN 1 N 2 S 0 0
C3
P ar t Num
b er = G
Q
M
M
1 8 7 5 C 2 E 1 R 0 B B1 2
1 8 7 5 C 2 E 3 R 6 BB 1 2
1 8 7 5 C2 E 1 R 8 B B1 2
T er m
Z =5 0
hm
C2
Pa r t N u m
R4
R= 1 0 0
O
b er = G
Q
h m
S2 P
SN P1
BP F L NA
I _ _1
TX
a
l y ou t
af t e r
m
e et n
i g
o c t
17
Figure. 4-25 Layout symbol of BPF-LNA
55
2
O
h m
CHAPTER 4
Noise Figure (NF) dB
80
60
40
20
0
1.0
1.5
2.0
2.5
3.0
3.5
4.0
4.5
5.0
Frequency (GHz)
Figure. 4-26 Layout result of noise figure (NF)
In figure 4-26, the noise figure (NF) is 1.05 dB and it can be achieved the minimum noise figure
0.94 dB at the central frequency 2.45 GHz. In order to achieve the desired amount of NF, input
matching network’s component should be changed which causes increase the value of input
reflection coefficient.
Forward Voltage Gain (S21) dB
20
0
-20
-40
-60
-80
1.0
1.5
2.0
2.5
3.0
3.5
4.0
4.5
5.0
Frequency (GHz)
Figure. 4-27 Layout result of forward voltage gain
56
CHAPTER 4
In figure 4-27, the forward voltage gain 12 dB at the central frequency 2.45 GHz and the gain is
almost flat from 2.15-2.66 GHz. The flatness and the bandwidth are almost close to the
schematic level because now non-ideal components are not being used for filter design. The use
of transmission lines has solved the components unavailability problem.
Input Reflection Coefficient (S11) dB
0
-5
-10
-15
-20
1.0
1.5
2.0
2.5
3.0
3.5
4.0
4.5
5.0
Frequency (GHz)
Figure. 4-28 Layout result of input reflection coefficient
In figure 4-28, the value of
is also changed from the schematic level which is -10.36 dB and
it is a better value than schematic level. Using of transmission lines to design filter is the reason
for this improvement.
57
CHAPTER 5
5 Prototypes & Measurements
In this chapter, prototype of LNA, BPF-LNA with lumped components and BPF-LNA with
transmission lines are described which are fabricated at PCB laboratory of ITN. In all the
prototypes Roger’s substrate RO4350B is used. Once the prototype is fabricated, components
are soldered and parameters are measured using vector network analyzer. All the components
are used in the prototype are of standard size 0603 inch.
5.1 Prototype of LNA
The following figure shows the complete prototype of LNA stand-alone which is tested and
measured to see the performances. After getting the result from network analyzer, the generated
S2P file is run in ADS and found the following results of input reflection coefficient and
forward voltage gain.
Figure. 5-1 Photograph of the prototype of LNA stand-alone
58
CHAPTER 5
The following tables show the values of components which were used at layout level and
prototype respectively. Due to unavailability of components, the values of table 7 were used for
the prototype of LNA.
Table 6 Used values at layout
Resistor (k )
Capacitor (pF)
Inductor (nH)
26
1.3
1.8
5.4
3.0
2.7
Table 7Used values at prototype
Resistor (k )
Capacitor (pF)
27
5.6
Inductor (nH)
1.0
1.0
2.2
2.2
5.1.1 Measurement Results
In figure 5-2, along to x-axis and along to y-axis frequency and forward voltage gain are
plotted. The forward voltage gain is 7 dB at the central frequency 2.45 GHz and the gain is
almost flat from 1.8-4.1 GHz. The flatness is almost same to the layout level.
Forward Voltage Gain (S21) dB
10
5
0
-5
-10
-15
-20
-25
-30
1.0
1.5
2.0
2.5
3.0
3.5
4.0
4.5
5.0
Frequency (GHz)
Figure. 5-2 Measurement result of forward voltage gain
In figure 5-3, along to x-axis and along to y-axis frequency and forward voltage gain are plotted
is -14 dB which is almost close to the layout level gain. The
respectively. The value of
59
CHAPTER 5
values which are used at the layout level, in most of cases, those component values were not
found in the desired companies. That is why, the result of prototype level is deviated from the
layout level.
Input Reflection Coefficient (S11) dB
2
0
-2
-4
-6
-8
-10
-12
-14
-16
1.0
1.5
2.0
2.5
3.0
3.5
4.0
4.5
5.0
Frequency (GHz)
Figure. 5-3 Measurement result of input reflection coefficient
5.2 Prototype of BPF-LNA with Lumped Elements
The following figure shows the complete prototype of BPF-LNA with lumped components
which is tested and measured to see the performances at the prototype. After getting the result
from network analyzer, the generated S2P file is run in ADS and found the following results of
input reflection coefficient and forward voltage gain.
Figure. 5-4 Photograph of the prototype of BPF-LNA with lumped elements
60
CHAPTER 5
The following tables show the values of components which were used at layout level and
prototype level respectively.
Table 8 Used values at layout
Resistor (k )
Capacitor (pF)
Inductor (nH)
26
5.4
1.0
1.2
1.2
2.7
--
3.0
--
--
1.5
--
--
3.0
--
Table 9 Used values in prototype
Resistor (k )
Capacitor (pF)
Inductor (nH)
27
5.6
2.2
1.0
2.2
2.2
--
2.2
--
--
2.2
--
--
3.3
--
5.2.1 Measurement Results
In figure 5-5, Forward voltage gain with respect to frequency is shown. The forward voltage
gain is 10 dB at the central frequency 2.45 GHz and the gain is not as flat as expected. But the
value of gain is satisfactory.
61
CHAPTER 5
Forward Voltage Gain (S21) dB
20
0
-20
-40
-60
-80
1.0
1.5
2.0
2.5
3.0
3.5
4.0
4.5
5.0
Frequency (GHz)
Figure. 5-5 Measurement result of forward voltage gain
In figure 5-6, along to x-axis and along to y-axis frequency and input reflection coefficient are
plotted respectively. The value of
is – 5.5 dB which is almost acceptable but still it is little
higher than – 6 dB.
Input Reflection Coefficient (S11) dB
2
0
-2
-4
-6
-8
-10
-12
-14
-16
1.0
1.5
2.0
2.5
3.0
3.5
4.0
4.5
5.0
Frequency (GHz)
Figure. 5-6 Measurement result of input reflection coefficient
62
CHAPTER 5
5.3 Prototype of BPF-LNA with Distributed Elements
The following figure shows the complete prototype of BPF-LNA with distributed components
which is tested and measured to see the performances at the prototype. After getting the result
from network analyzer, the generated S2P file is run in ADS and found the following results of
input reflection coefficient and forward voltage gain which are shown in figures 5-8 and 5-9.
Figure. 5-7 Photograph of the prototype of BPF-LNA with distributed element
The following tables show the values of components which were used at layout level and
prototype level respectively.
Table 10 Used values at layout
Resistor (k )
Capacitor (pF)
Inductor (nH)
26
5.4
1.8
3.6
1.2
2.7
Table 11 Used values at prototype
Resistor (k )
Capacitor (pF)
Inductor (nH)
27
5.6
1.0
3.3
2.2
2.2
63
CHAPTER 5
5.3.1 Measurement Results
In figure 5-8, forward voltage gain is shown where the value of forward voltage gain is 7 dB at
the central frequency 2.45 GHz and the gain is not as flat as layout level. The bandwidth is also
wide which is not expected.
Forward Voltage Gaint (S21) dB
20
0
-20
-40
-60
-80
1.0
1.5
2.0
2.5
3.0
3.5
4.0
4.5
5.0
Frequency (GHz)
Figure. 5-8 Measurement result of forward voltage gain
In figure 5-9, along to x-axis and along to y-axis frequency and input reflection coefficient are
is – 5.0 dB which is higher than -6.
plotted respectively. The value of
Input Reflection Coefficient (S11) dB
0
-2
-4
-6
-8
-10
1.0
1.5
2.0
2.5
3.0
3.5
4.0
4.5
5.0
Frequency (GHz)
Figure. 5-9 Measurement result of input reflection coefficient
64
CHAPTER 5
5.4 Comparison of Layouts and Measured Results
In this section, the comparison of layout and measured results are shown in tabular form. In case
of layout, the results are much better than the measurement results. The optimal values of
components were not found in the Murata and other companies, that is why, other closest
available values were used. Table-6, 8, 10 show the components values which are used in layout
design and simulation, and table-7, 9, 11 are used in the respective prototype. Since equivalent
values of layout components are not used in prototype, that is why, the results are much
deviated into the prototype.Table-12 shows the summary of layout results of LNA stand-alone,
BPF-LNA (Lumped) and BPF-LNA (T-Line). Table-13 shows the summary of measurement
results of LNA stand-alone, BPF-LNA (Lumped) and BPF-LNA (T-Line).
Other non-ideal effects also ought to be responsible for some of the unexpected deviation
between simulation and measurement results. The surface of the conductor layer is not ideally
smooth, then the signal might not propagate entirely as expected. In addition, may be, the
connectors and other SMD components are not placed properly which results a small gap
between the conductor layer and the respective component and these various situations can
cause a capacitive/inductive effect, which causes deviation as well. In the PCB, wires are used
for grounding and Vuu instead of using layer. Throughout the whole design stability was
checked and the operation point was steady. Finally, noise figure was not measured due to
unavailability of instrument.
Table 12 Comparison of layout parameters
Prototypes
S 11 (dB)
S21 (dB)
Noise Figure (NF) (dB)
LNA-Stand alone
-15.30
14.73
0.92
BPF-LNA (Lumped)
-10.80
13.26
1.37
BPF-LNA (T-Line)
-10.36
12
1.05
Table 13 Comparison of measured parameters
Prototypes
S 11 (dB)
S 21 (dB)
LNA-Stand alone
-14.0
7
BPF-LNA (Lumped)
-5.5
10
BPF-LNA (T-Line)
-5.0
7
65
CHAPTER 6
6 Conclusion and Future Works
6.1 Conclusion
According to the design specifications, LNA stand-alone, BPF-LNA with lumped components,
and transmission lines are designed withATF-58143,and their performance are compared in the
simulation (schematic and layout) and measurement level (prototype level).Optimization was
performed according to get the desired responses in all the designs. All the PCB prototypes
were fabricated using a standard PCB (etch based) process. In case of LNA stand-alone, the
optimum value input reflection coefficient and gain are -14 dB and 7 dB respectively but the
bandwidth is too wide compared to the specification and gain is not much flat. BPF-LNA with
lumped components has a input reflection coefficient and gain of – 5.5 dB and 10 dB,
respectively and the bandwidth is narrower than LNA-stand-alone but still it is wider than the
specified 100 MHz. BPF-LNA with transmissions lines was measured and the input reflection
coefficient is – 5.0 dB and the gain is 7 dB. As the required values of components were not used
to prototype due to unavailability of components that is why, the measurement results of PCB
level is not satisfactory. Once the required values are used, the bandwidth and gain ought be
narrow and almost flat respectively, over the whole bandwidth which is expected for the desired
signal with minimum noise.
Furthermore, throughout the whole design, transistor was stable. The level of satisfaction of this
thesis work is satisfactory. However, due to the parasitic effects and unavailability of required
Murata components, there are some deviations from expectations in the measured results. This
thesis work gives a closer and wide view of all the relevant background theories and design
technologies to the designer. PCB lab works gave a manufacturing hands-on experience which
implies expanding reality of theoretical knowledge.
6.2 Future Works
Though the responses are satisfactory, but still there are scopes to improve the performances.
Some of the circuits can be improved in design and with more proper optimization to have
66
CHAPTER 6
better responses. In future, different classes of BPFs such as elliptical, Chebyshev with different
orders can be designed with LNA which will provide more options to compare for the better
one. Furthermore, exact values of components which were used in the designs can be purchased
and made new prototypes, which may produce better responses of BPF-LNA (designed with
lumped components). However, the acquired knowledge from this thesis work can help to
design the whole RF receiver system in the ISM band.
67
REFERENCES
7 References
[1] David M. Pozar, Microwave and RF Wireless System, John Willey & Sons, Inc. Third
Edition, 2000. Chapter-10
[2]
International Telecommunication Union, "Industrial, Scientific and Medical (ISM)
applications of radio frequency energy in the field of telecommunications.", 19 October,
2009.
[3]
International Telecommunication Union, Visited date: 14 August, 2012 http://www.itu.int
[4]
Federal Communications Commission. “Authorization of Spread Spectrum Systems
Under Parts 15 and 90 of the FCC Rules and Regulations". 18 June, 1985. Retrieved
2007-08-31.
[5]
European Commission, Visited date: 16 August, 2012, http://ec.europa.eu/
[6]
Electronic
Communications
http://www.erodocdb.
[7]
Ghosh, Smarajit, “Network Theory: Analysis and Synthesis”, Prentice Hall of India ISBN
81-203-2638-5 pp. 353
[8]
Two-Port Networks, Visited date: 30 August, 2012 , http://fourier.eng.hmc.edu
[9]
Andres Moran Valerio, Alonso Perez Garrido, Thesis title “Design and Implementation of
6-8.5 GHz LNA”, Institute of Science & Technology, Linkoping University, 2008-11-07,
pp. 13
Committee,
“ERC
Recommendation
70-03”,
[10] Reinhold Ludwig, Pavel Bretchko, “RF Circuit Design”, Prentice-Hall, Inc. New Jersey
07458, ISBN: 0-13-095323-7, 2000. Chapter-2, 4
[11] C. D. Motchenbacher, J.A. Connelly, “Low-Noise Electronic System Design”. Wiley
Interscience. 1993.
[12] Dennis V. Perepelitsa, “Johnson Noise and Shot Noise”, MIT, Department of Physics, 27
November , 2006. pp. 1
[13] M. Blanter, M. Büttiker, Physics Reports on “Shot Noise in Mesoscopic Conductors”,
2000. DOI:10.1016/S0370-1573(99) 00123-4.
[14] Jimmin Chang, A.A. Abidi and C.R. Viswanathan, "Flicker Noise in CMOS Transistors
from Subthreshold to Strong Inversion at Various Temperatures". IEEE Transactions on
Electron Devices, Vol. 41, No. 11 November 1994. pp. 1965
68
REFERENCES
[15] Devendra K. Misra, “Radio-Frequency and Microwave Communication Circuits;
Analysis And Design”, John Wiley and Sons, 2001. Chapter-2
[16] Henry W.Ott, “Noise Reduction Techniques in Electronic Systems”, John Wiley and Sons
1988. Chapter-8
[17] Behzad Razavi, “Design of Analog CMOS Integrated Circuits”, McGraw-Hill, 2000,
Chapter 7: Noise.
[18] Adriana Serban Craciunescu, “Low-Noise Amplifier for Ultra-Wideband System”. LiUTEK-LIC-2006.
[19] Lumped Element Filters, Visited date: August, 2012
http://www.microwaves101.com/encyclopedia/Lumpedfilters.cfm
[20] F.R. Connor, “Wave Transmission”, Edward Arnold Ltd., 1972 ISBN 0-7131-3278-7, pp.
13-14
[21] M. Afzal, N. Ahmad , Thesis title, “Investigation of Different Diplexer Design
Techniques for 4G Mobile Communications”, Institute of Science & Technology,
Linkoping University, 2011, LiU-ITN-TEK-A-11/068-SE, pp. 7
[22] David M. Pozar, Microwave Engineering, John Willey & Sons, Inc. Second Edition,
1998, pp. 449-450
[23] C. Zhu, L Yao, J Zhao, “Novel Microstrip Diplexer based on a Dual Band Bandpass Filter
for WLAN Systems”, State Key Laboratory of Millimeter Waves, Southeast University,
Nanjing, 210096, China, Proceedings of Aisa-Pacific Microwave Conference-2010.
[24] Zverev, Matthaei, Young, Jones Dishall, “Handbook of Filter Synthesis”, Artech House,
ISBN 0-890-06099-1-5, Nov. 1951, Chaper-8.
[25] Matthaei, L. George, Young, Leo, Jones, “Microwave Filters, Impedance-Matching
Networks and Coupling Structures”, McGraw-Hill 1980, ISBN 0-89006-099-1
[26] Robin S. Johansson, Torbjörn E. Karlsson, "Low Noise Amplier 2.45 GHz", Lund
Institute of Technology, Lund University, Sweden. May 16, 2011
[27] Venkat Ramana. Aitha, Mohammad Kawsar Imam,Master’s Thesis title “Low Noise
Amplifier for Radio Telescope at 1.42 GHz”, Computer and Electrical Engineering,
Halmstad University, Sweden, IDE0747, May 2007 pp. 29
[28] Data Sheet, AVAGO Technologies, ATF-58143, Visited date: 17 August, 2012
http://www.avagotech.com/
[29] Roger R04350B data sheet, Visited date: August, 2012 www.rogerscorp.com
69