Post on 13-Oct-2015
description
transcript
DESIGN OF AN ULTRA-WIDE-BAND CMOS LOW NOISE AMPLIFIER
A Thesis Submitted to the Department of Electrical and Electronic Engineering
Of AHSANULLAH UNIVERSITY OF SCIENCE AND TECHNOLOGY
By Md. Sadirul Islam, Student ID: 09.01.05.152 Farzana Ishtiaq, Student ID: 09.01.05.160 Afsana Akhtar, Student ID: 09.01.05.202 Shakib Aman, Student ID: 09.01.05.208
Tazreen Ahmed, Student ID: 09.01.05.028
In Partial Fulfillment of the Requirements for the Degree Of
Bachelor of Science in Electrical and Electronic Engineering July 2013
AHSANULLAH UNIVERSITY OF SCIENCE AND TECHNOLOGY Dhaka, Bangladesh
i
DECLARATION
This is to certify that this thesis is the outcome of the original work of the undersigned. No part of this work has been submitted elsewhere, partially and fully, for the award of any other degree or diploma. Materials of work found by other researcher are mentioned by references. Any material reproduced in this thesis has been properly acknowledged.
(Dr. Pran Kanai Saha) (Md. Sadirul Islam) Professor, Department of Electrical and Electronic Engineering BUET, Bangladesh
(Farzana Ishtiaq)
Signature of Supervisor
(Afsana Akhtar)
(Shakib Aman)
(Tazreen Ahmed)
Signature of Authors
ii
ACKNOWLEDGMENTS
We want to express our sincere gratitude to our supervisor, Professor Dr.
Pran Kanai Saha, for his continuous guidance and support throughout our
thesis at Ahsanullah University of Science And Technology. His wisdom and
kindness had been invaluable and made the whole thesis experience
enjoyable.
We specially thank Mr. Md. Sariful Islam (Lecturer, Dept. of EEE, AUST) for his
valuable suggestion and concise comments on some of the research papers of
the thesis. He has been very kind enough to extend his help at various phases
of this research.
This thesis would not have been possible without the guidance and the help of
several individuals who in one way or another contributed and extended their
valuable assistance in the preparation and completion of this thesis.
Last but not the least, to our family and the one above all of us, the
omnipresent God, for giving us the strength to complete this thesis within
given resource and time.
iii
ABSTRACT
The objective of this paper is to investigate a ultra-wideband (UWB) low noise
amplifier (LNA) by utilizing a simple inductively degenerated LNA with
resistive feedback into differential topology, which can improve the
bandwidth. The Proposed LNA has been simulated by using HSpice
considering BSIM3 model in 0.18 m RF CMOS technology. Based on those
technologies, this proposed UWB LNA obtained a flat gain 8dB bandwidth of
4.7GHz, the constant gain of 8dB, noise figure lower than 2.2dB, and the return
loss better than 30dB. HSpice simulation of the proposed LNA shows wider
bandwidth of 1.2GHz, low power consumption with gain of 8dB in the
frequency band of 4.2 GHz to 8.9GHz. Finally, LNA characterization exhibits
ultra-wide bandwidth characteristic and low noise margin. This proposed
LNA can be used in UWB and microwave communication system.
iv
Contents
Chapter 1 Introduction1
1.1: Ultra-wideband Technology.....................1
1.2: CMOS Technology.3
1.3: VLSI Technology3
1.4: Basic concept of Ultra-Wide-Band Technology4
1.5: Ultra-Wide-Band Technologies..6
1.5.1: Carrier Free Direct Sequence UWB Technology.6
1.5.2: Multi-Band OFDM (MB-OFDM) UWB Technology........6
1.6: Advantages of using Ultra-Wide-Band.6
1.7: Potential Applications of Ultra-Wide-Band.10
1.7.1: Communications and Sensors10
1.7.2: Position Location and Tracking.12
1.7.3: Radar Applications12
1.8: Ultra-Wide-Band Design Challenges12
1.9: Objective of this thesis..13
1.10: Organization of the thesis...13
Chapter 2 UWB Low Noise Amplifier (LNA)...14
2.1: Parameters in LNA Design.16
2.1.1: Noise Figure.16
2.1.2: Input Impedance Match.18
2.1.3: Gain.19
2.1.4: Linearity.20
2.2: Low Noise Amplifier (LNA) Topologies.22
2.2.1: Amplifier Distributed (DA).23
2.2.2: Common Gate (CG)24
2.2.3: Inductively Degenerated Common Source Topology (IDCS)...26
2.2.4: Cascode Topology..27
v
2.2.5: Folded Cascode Topology...28
2.2.6: Cascade Topology29
2.2.7: Current Reuse Topology..29
2.2.8: Differential Topology.31
2.2.9: Novel Topology.32
2.3: Proposed Topology.32
Chapter 3 Architecture of UWB LNA and Simulation Results..33
3.1 Architecture of UWB LNA...36
3.1.1 Single Ended LNA.37
3.1.2 Full Differential LNA..41
3.1.3 Balun Circuit43
3.2 Final Simulation Results..44
Chapter 4 Conclusion..52
4.1 Summary of Contribution...52
4.2 Future Work52
vi
List of Figures
Fig 1.1: Ultra wideband Technology ..2
Fig 1.2: Block Diagram of UWB System.2
Fig 1.3: Illustration of the Definition of UWB frequency range..4
Fig 1.4: UWB spectral mask for indoor communication systems..5
Fig 2.1: (a) Block Diagram of LNA at receiving end14
(b) Signal processing.15
Fig 2.2: Combination of power amplifier circuit and LNA circuit..15
Fig 2.3: Equivalent small signal circuit...17
Fig 2.4: More hardware in the input matching network..19
Fig 2.5: Equivalent circuit with inductive source degeneration.19
Fig 2.6: CP1 and 3rd order intercept point20
Fig 2.7: Two-tones with the IDM3 product...21
Fig 2.8: Generalized structure of distributed LNA...23
Fig 2.9: Conventional CG- LNA and low noise techniques employing feedback ..24
Fig 2.10: (a) Common source configuration for Narrow band 26
(b) Inductive degenerated CS for UWB..26
Fig 2.11: Cascode configuration..27
Fig 2.12: Folded cascade configuration..28
Fig 2.13: Conventional current reuse configuration..30
Fig 2.14: Differential capacitive cross coupled configuration..31
Fig 2.15: Noise cancelling principle of differential topology.31
Fig 3.1: Block Diagram of LNA.33
vii
Fig 3.2: Common LNA Architectures: Resistive termination34
Fig 3.3: Common LNA Architectures: 1/gm termination34
Fig 3.4: Common LNA Architectures: Shunt-series feedback...35
Fig 3.5: Common LNA Architectures: Inductive degeneration35
Fig 3.6: Simplified Block Diagram of proposed LNA..36
Fig 3.7: LNA topology for UWB single ended LNA...37
Fig 3.8: LNA topology for UWB LNA Small-signal equivalent circuit of the input of LNA...38
Fig 3.9: S21of Single Ended LNA...39
Fig 3.10: S11of Single Ended LNA39
Fig 3.11: AC output Noise Figure of Single Ended LNA.40
Fig 3.12: Maximum Power Gain of Single Ended LNA...40
Fig 3.13: Full-Differential LNA.41
Fig 3.14: Balun Circuit...43
Fig 3.15: S21 (dB) of Differential LNA...44
Fig 3.16: S11 (dB) of Differential LNA...44
Fig3.17: Comparison between S21 and S11 ...45
Fig 3.18: AC output Noise figure of Differential LNA.46
Fig 3.19: Minimum AC output Noise figure of Differential LNA..46
Fig 3.20: S22 of Differential LNA..47
Fig 3.21: Comparison between S11 and S22 of Differential LNA...47
Fig 3.22: S12 of Differential LNA..48
Fig 3.23: Output Noise of Differential LNA...48
Fig 3.24: Input Noise (dB) of Differential LNA...49
Fig 3.25: Comparison between input noise (dB) and output noise (dB)...49
Fig 3.27: Maximum Power Gain of Differential LNA..50
viii
List of Tables
Table 1.1: FCC allocations for each UWB category.5
Table3.1: Comparison of performance between Single ended LNA and Differential LNA........50
Table3.2: Performance of LNA51
Table3.3: Comparison with other LNA from the bibliography.51
ix
List of Abbreviations
BPSK Binary Phase Shift Keying
CGLNA Common-Gate Low-Noise Amplifier
CSLNA Common-Source Low-Noise Amplifier
CCC Capacitive Cross-Couple
CDMA Code Division Multiple Access
DS Direct Sequence
EVDO Enhanced Voice-Data Only
EIRP Effective Isotropic Radiated Power
FH Frequency Hopping
FCC Federal Communications Commission
IP3 3rd Order Intercept Point
IMD3 3rd Order Intermodulation
IDCS Inductively Degenerated Common Source
IIP3 3rd Order Input Power Level
LPI Low Probability of Intercept
LPD Low Probability of Detection
LOS Line Of Sight
MBOFDM Multi-Band Offset Frequency Division Modulation
MMIC Microwave Monolithic Integrated Circuit
NLOS Non-Line Of Sight
OFDM Offset Frequency Division Modulation
OIP3 3rd Order Output Power Level
x
PG Processing Gain
PDA Personal Digital Assistant
PSD Power Spectral Density
SNR Signal To Noise Ratio
SpO2 Oxygen Saturation Sensor
SFDR Spurious Free Dynamic Range
UWB Ultra-wideband
VSWR Voltage Standing Wave Ratio
WWAN Wireless Wide Area Network
WLAN Wireless Local Area Network
WPAN Wireless Personal Area Network
WBAN Wireless Body Area Network
1
Chapter 1
Introduction
The late 20th and early 21st century can be referred to as the wireless era, in which we
witnessed the rapid growth of wireless communication technologies. It is not uncommon
nowadays for an average home to have several digital cameras, portable music players,
cellular phones and external hard drives. As the number of device increase, so does the
number of wires connecting these devices together. It is usually frustrating to see a mess of
cables around the home or office desk and can also be inconvenient to have to carry these
cables when travelling.
One could see how wireless technologies could erase the life of user by removing some or
all of these wires. The current wireless data technologies can be divided into 3 categories:
WWAN (Wireless Wide Area Network),
WLAN (Wireless Local Area Network)
WPAN (Wireless Personal Area Network).
WWAN includes direct satellite link, CDMA based EVDO and WIMAX that covers regions of
several miles wide. WLAN system is Wi-Fi (802.11a/b/g). Wi-Fi has the potential for high
data rates. WPAN has a coverage range of a few feet. Two notable examples of WPAN are
infra-red and Bluetooth communication.
Among all these wireless technologies, Ultra-Wideband (UWB) is the most promising
technology for next generation WPANs, especially because of its low cost, low power and
high data rates for applications.
1.1 Ultra-wideband Technology
The recent surge in the demand for low power portable wireless electronics that can offer
extremely high data rates has resulted in much active research in Ultra-wideband (UWB)
systems. UWB is widely recognized as a promising technology for high data rate, short-
range applications with precise time resolution and high energy efficiency. Ultra wideband
(also known as UWB or as digital pulse wireless) is a wireless technology for transmitting
2
large amounts of digital data over a wide spectrum of frequency bands with very low
power for a short distance.
Figure 1.1: Ultra wideband Technology
All these benefits originate from the wideband current technology, UWB can offer data
rates up to 480 Mbps and its operational frequency spectrum is between 3.1 and 10.6GHz.
However, the wideband operation of UWB systems imposes many design challenges that
have not been explored before in the traditional narrowband ones.
A basic UWB system has a signal pulse generator which generates a Gaussian pulse to
transmit the encoded signal. The pulses are amplified and transmitted via antenna to the
receiver. When receiving antenna receives the transmitted signal, LNA amplify the signal
before it continues on into the receiver.
Figure 1.2: Block Diagram of UWB System.[1]
3
Here LNA is the subsystem and in this thesis our focus is the LNA design. We will discuss
about LNA in chapter 2 and LNA design in chapter 3.
1.2 CMOS Technology
Complementary metaloxidesemiconductor (CMOS) is a technology for constructing
integrated circuits. In CMOS technology, both N-type and P-type transistors are used to
realize logic functions.
CMOS was originally not considered a good technology for analog and RF applications
[Compared to Silicon-Germanium (SiGe) and Gallium-Arsenide (GaAs)]. However, the rapid
growth of the digital industry due to the continuous scaling of the CMOS technology has
motivated designers to create analog and RF CMOS circuits that can be integrated easily
with digital circuitry. The speed of CMOS is more than adequate for a common wireless
system which has a maximum operating frequency at 10 GHz. Despite the inferior
performance of RF CMOS circuits compared to SiGe and GaAs counterparts, the feasibility
of integrating analog/digital/RF circuits on chip, the potential low cost and low power
consumption and the dominance of CMOS in digital circuitry has provided reasonable
motives to adopt CMOS over other technologies.
1.3 VLSI Technology
Very-large-scale integration (VLSI) is design/manufacturing of extremely small, complex
circuitry using modified semiconductor material. It may contain millions of transistors,
each a few mm in size which is called Integrated circuit (IC).It is the process of creating
integrated circuits by combining thousands of transistor-based circuits into on a single chip
and has wide range of applications mostly based on electronic logic devices. As a side effect
of advances in the world of computers, there has been a dramatic proliferation of tools that
can be used to design VLSI circuits. CMOS having the ability to allow a high density of logic
functions on a chip, it became the most used technology to be implemented in VLSI chips.
4
1.4 Basic concept of Ultra-Wide-Band Technology
Ultra-Wideband Communications was first employed by Guglielmo Marconi in 1901 to
transmit Morse code sequences across Atlantic Ocean using spark gap radio transmitters.
Approximately fifty years after Marconi, UWB technology was applied to impulse radars in
military applications and this technology was restricted to military from 1960s to 1990s.
However, ultra-wideband is now ready for commercial applications because of recent
advancements in microprocessors stemming from the rapid development of semiconductor
technology.
Federal Communications Commission (FCC) in the United States defined UWB signal as any
emitting signal that has a fractional bandwidth greater than or equal to 0.20 or has a
bandwidth greater than or equal to 500MHz. The fractional bandwidth is calculated as 2
( fH - fL)/( fH + fL) and can be expressed as:
Bf= B
100% =
100%
Where B is the absolute frequency bandwidth, fc is the center frequency, fH is the upper -10
dB corner frequency and fL is the lower -10 dB corner frequency as shown in the figure.
Figure 1.3: Illustration of the Definition of UWB frequency range.
The early applications of UWB technology were primarily radar related, driven by the
promise of fine-range resolution that comes with large bandwidth. But the recent 3.1-
5
10.6GHz allocation extends the UWB use to a larger application area for which the specific
frequency ranges are reported in the table 1.4.1 FCC allocations for each UWB category.
Table 1.1: FCC allocations for each UWB category
Class/Application Frequency Band for operation
Communication and measurement systems 3.1 to 10.6 GHz (different out-of-band
emission Limits for indoor and outdoor
devices)
Imaging : ground penetrating radar, wall,
medical imaging
960 MHz or 3.1 to 10.6 GHz
Imaging: through wall 960 MHz or 1.99 to 10.6 GHz
Imaging: surveillance 1.99 to 10.6 GHz
UWB operational frequency bands, authorized by the FCC are, below 960 MHz, 3.1-10.6
GHz and 22-29 GHz. The FCC has mandated that UWB radio transmissions can legally
operate in the range 3.1 to 10.6 GHz, with the power spectral density (PSD) satisfying a
specific spectral mask assigned by FCC. Figure illustrates the UWB spectral mask for indoor
communications. According to the spectral mask the allowed effective isotropic radiated
power (EIRP) is below 41.3 dBm /MHz .
Figure 1.4: UWB spectral mask for indoor communication systems.
6
1.5 Ultra-Wide-Band Technologies
UWB communications transmit in which it does not interfere with conventional
narrowband and carrier wave used in the same frequency band.
Despite the single named used for the ultra wideband (UWB) transmissions, there are two
very different technologies being developed:
Carrier free direct sequence ultra wideband technology
MBOFDM, Multi-Band OFDM ultra wideband technology
1.5.1 Carrier Free Direct Sequence UWB Technology
Direct sequence (DS) format for ultra wideband is often referred to as an impulse,
baseband or zero carrier technology. DS based UWB increases system performance over
traditional impulse based UWB by using direct sequence spread technology, while
maintaining a very simple system structure. This form of UWB technology transmits a
series of low power Gaussian shaped pulses which are coherently received at the receiver.
Each of the DS UWB pulses has an extremely short duration typically between 10 and 1000
picoseconds. As a result it is shorter than the duration of a single bit of the data to be
transmitted.
1.5.2 Multi-Band OFDM (MB-OFDM) UWB Technology
This form of ultra wideband technology uses a wide band or multiband orthogonal
frequency division multiplex (MBOFDM) signal that is effectively a 500 MHz wide OFDM
signal. This is 500 MHz signal is then hopped in frequency to enable it to occupy a
sufficiently high bandwidth.
1.6 Advantages of using Ultra-Wide-Band
The main advantages of UWB technology are high data rates, low cost and low power. This
means that many types of devices in the home and business environments, such as laptops,
mobile phones, TVs, DVDs, memory sticks, digital cameras and MP3 players could be
networked wirelessly at a data rate that would effectively be impossible using classical
7
radio technology. This is because of UWB's very large data rate at very low power and at
low cost.
Some forms of UWB allow distances to be measured accurately. The nature of the short-
duration pulses used in UWB technology offers several advantages over narrowband
communications systems.
1. Large channel capacity
Main advantage of UWB technology is perceived from the channel capacity. Channel
capacity is the theoretical maximum possible number of bits per second of information
which may be conveyed through one or more links in an area. According to the Shannon
Hartley theorem channel capacity can be expressed as:
C= B log2 (1+SNR)
Where C represents the channel capacity in bits per second, B is the channel bandwidth in
Hz and SNR is the signal-to-noise ratio.
Thus channel capacity increases linearly with increasing channel's bandwidth to the
maximum value available, or by increasing the signal power exponentially where
bandwidth is fixed.[2] By virtue of the large bandwidths inherent in UWB systems, large
channel capacities could be achieved in principle without invoking higher-order
modulations requiring a very high SNR. Ideally, the receiver signal detector should match
the transmitted signal in bandwidth, signal shape and time. A mismatch results in loss of
margin for the UWB radio link. Several gigahertz of bandwidth available for UWB signals, a
data rate of gigabits per second (Gbps) can be expected.
2. Ability to Work with Low Signal to Noise Ratios
The Hartley-Shannon formula for maximum capacity also indicates that the channel
capacity is only logarithmically dependent on signal-to-noise ratio (SNR). Therefore, UWB
communications systems are capable of working in harsh communication channels with
low SNRs and still offer a large channel capacity as a result of their large BW.
8
3. Ability to Share the Frequency Spectrum
The FCCs requirement of 41.3dBm/MHz, 5 equal to 75 nano watts/MHz for UWB systems,
puts them in the category of unintentional radiators, such as TVs and computer monitors.
Such power restriction allows UWB systems to reside below the noise floor of a typical
narrow band receiver and enables UWB signals to coexist with current radio services with
minimal or no interference. However, this all depends on the type of modulation used for
data transfer in a UWB system.
4. Low Probability of Intercept (LPI) and Low Probability of Detection (LPD)
Because of low average transmission power, UWB communications systems have an
inherent immunity to detection and intercept. The extremely narrow pulse-width and low
duty cycle because the UWB signals to be spread thinly over a wide bandwidth resulting in
extremely low power spectral densities over GHz bandwidths. With such low transmission
power, the eavesdropper has to be very close to the transmitter (about 1 meter) to be able
to detect the transmitted information. In addition, UWB pulses are time modulated with
codes unique to each transmitter/receiver pair. The time modulation of extremely narrow
pulses adds more security to UWB transmission, because detecting picoseconds pulses
without knowing when they will arrive is next to impossible. Therefore, UWB systems hold
significant promise of achieving highly secure, low probability of intercept and detection
communications that is a critical need for military operations.
5. Resistance to Jamming
Unlike the well-defined narrowband frequency spectrum, the UWB spectrum covers a vast
range of frequencies from near DC to several gigahertz and offers high processing gain for
UWB signals. Processing gain (PG) is a measure of a radio system's resistance to jamming
and is defined as the ratio of the RF bandwidth to the information bandwidth of a signal:
PG = RF Bandwidth
Information Bandwidth
The frequency diversity caused by high processing gain makes UWB signals relatively
resistant to intentional and unintentional jamming, because no jammer can jam every
frequency in the UWB spectrum at once. Therefore, if some of the frequencies are jammed,
9
there is still a large range of frequencies that remains untouched. However, this resistance
to jamming is only in comparison to narrowband and wideband systems.
6. Multipath Propagation Performance
The phenomenon known as multipath is unavoidable in wireless communications
channels. It is caused by multiple reflections of the transmitted signal from various surfaces
such as buildings, trees, and people. The straight line between a transmitter and a receiver
is the line of sight (LOS); the reflected signals from surfaces are non-line of sight (NLOS).
The effect of multipath is rather severe for narrowband signals; it can cause signal
degradation up to 40 dB due to the out-of-phase addition of LOS and NLOS continuous
waveforms. On the other hand, the very short duration of UWB pulses makes them less
sensitive to the multipath effect. Because the transmission duration of a UWB pulse is
shorter than a nanosecond in most cases, the reflected pulse has an extremely short
window of opportunity to collide with the LOS pulse and cause signal-degradation.
7. Superior Penetration Properties
Unlike narrowband technology, UWB systems can penetrate effectively through different
materials. The low frequencies included in the broad range of the UWB frequency spectrum
have long wavelengths, which allows UWB signals to penetrate a variety of materials,
including walls. This property makes UWB technology viable for through-the-wall
communications and ground-penetrating radars. However, the material penetration
capability of UWB signals is useful only when they are allowed to occupy the low-frequency
portion of the radio spectrum.
8. Simple Transceiver Architecture
UWB transmission is carrier-less, meaning that data is not modulated on a continuous
waveform with a specific carrier frequency, as in narrowband and wideband technologies.
Carrier-less transmission requires fewer RF components than carrier-based transmission.
For this reason UWB transceiver architecture is significantly simpler and thus cheaper to
build.
10
1.7 Potential Applications of Ultra-Wide-Band
There have been tremendous research efforts to apply ultra wide band (UWB) technology
to the military and government sectors. Some of them are already accomplished and some
are intended for future.
UWB technology can enable a wide variety of applications in wireless communications,
networking, radar imaging, and localizing systems. The potential applications UWB are
categorized into three parts:
Communications and sensors
Position location and tracking
Radar applications
1.7.1 Communications and Sensors
For wireless communication the use of UWB technology offers significant potential for the
deployment of 3 communication systems:
Low Data Rate for Sensor Network:
Under the low-rate operation mode, UWB technology could be beneficial and potentially
useful in sensor, positioning and identification networks. A sensor network comprises a
large number of nodes spread over a geographical area to be monitored. Depending on the
specific application, the sensor nodes can be static or mobile. With its unique properties of
low complexity, low cost and extremely low power spectral density (PSD), UWB technology
is well suited to sensor networks. The innovative communication method of UWB at low
data rate gives numerous benefits to different sectors. For instance, the wireless
connection of computer peripherals such as mouse, monitor, keyboard, joystick and printer
can utilize UWB technology. UWB allows the operation of multiples devices without
interference at the same time in the same space. It can be used as a communication link in a
sensor network. It can also create a security bubble around a specific area to ensure
security. It is the best candidate to support a variety of WBAN applications. A network of
UWB sensors such as electrocardiogram (ECG), oxygen saturation sensor (SpO2) and
electromyography (EMG) can be used to develop a proactive and a smart healthcare
11
system. This can benefit the patient in chronic condition and provides long term health
monitoring. In UWB system, the transmitter is often kept simpler and most of the
complexity is shifted towards receiver, which permits extremely low energy consumption
and thus extends battery life.
High Data Rate:
The unique applications of UWB systems in different scenarios have initially drawn much
attention, since many applications of UWB spans around existing market needs for high
data rate applications. Demand for high density multimedia applications is increasing,
which needs innovative methods to better utilize the available bandwidth. UWB system has
the property to fill the available bandwidth as demand increases. The problem of designing
receiver and robustness against jamming are main challenges for high-rate applications.
The large high-resolution video screens can benefit from UWB. These devices stream video
content wirelessly from video source to a wall-mounted screen. Various high data rata
applications include internet access and multimedia services, wireless peripheral
interfaces and location based services. Regardless of the environment, very high data rate
applications (>1 Gbps) have to be provided. The use of very large bandwidth at lower
spectral efficiency has designated UWB system as a suitable candidate for high internet
access and multimedia applications. Standardized wireless interconnection is highly
desirable to replace cables and propriety plugs. The interconnectivity of various numbers
of devices such as laptops and mobile phones is increasingly important for battery-
powered devices.
Home Network Applications:
Home network application is a crucial factor to make pervasive home network
environment. The wireless connectivity of different home electronic systems removes
wiring clutter in living room. This is particularly important when we consider the bit rate
needed for high definition television that is in excess of 30 Mbps over a distance of at least
few meters. In IEEE 1394, an attempt has been made to integrate entertainment, consumer
electronics and computing within a home environment. It provides isochronous mode
12
where data delivery is guaranteed with constant transmission speed. It is important for
real time applications such as video broadcasts.
1.7.2 Position Location and Tracking
Position location and tracking have wide range of benefits such as locating patient in case
of critical condition, hikers injured in remote area, tracking cars, and managing a variety of
goods in a big shopping mall. For active RF tracking and positioning applications, the short-
pulse UWB techniques offer distinct advantages in precision time-of-flight measurement,
multipath immunity for leading edge detection, and low prime power requirements for
extended-operation RF identification. The reason of supporting human-space intervention
is to identify the persons and the objects the user aims at, and identifying the target task of
the user. Knowing where a person is, we can figure out near to what or who this person is
and finally make a hypothesis what the user is aiming at. This human-space intervention
could improve quality of life when used in a WBAN. In a WBAN, a number of intelligent
sensors are used to gather patients data and forwards it to a PDA which is further
forwarded to a remote server. In case of critical condition such as arrhythmic disturbances,
the correct identification of patients location could assist medical experts in treatment.
1.7.3 Radar Applications
Short-pulse UWB techniques have several radar applications such as higher range
measurement accuracy and range resolution, enhanced target recognition, increased
immunity to co-located radar transmissions, increased detection probability for certain
classes of targets and ability to detect very slowly moving or stationary targets. UWB is a
leading technology candidate for micro air vehicles (MAV) applications. The nature of
creating millions of ultra-wideband pulses per second has the capability of high
penetration in a wide range of materials such as building materials, concrete block, plastic
and wood.
1.8 Ultra-Wide-Band Design Challenges
Designing UWB systems have several challenges, some of which have not been reached
before in traditional narrowband systems. One particular design techniques have evolved
to handle narrowband signals extremely efficiently, but these techniques do not apply UWB
13
systems. Furthermore, since the entire system power consumption is no longer dominated
by the radiation power, achieving low power consumption in the LNA and mixer is
extremely important.
The requirement for maximum total power consumption set by the 802.15.3a specification
at 110 Mb/s and 200 Mb/s is 100 mW and 250 mW, respectively. Which implies that, in
case of designing a UWB transceiver, a power save mode must be supported. To meet these
constraints, a transceiver must either target the lowest power for all data rates or use an
architecture that scales power with data rate. It is also advantageous to have an
architecture that scales power consumption under optimal channel conditions.
Another challenge is narrowband jammers. Due to the wide operational band of UWB,
interferences from existing radio systems are unavoidable. The FCC spectral mask for UWB
EIRP is -41.3 dBm/MHz, Which is low enough to not cause interference to other wireless
system sharing the same bandwidth. However, this does not guarantee that in-band
narrowband transmitters do not saturate UWB receivers. This presents a serious challenge
for UWB since other systems operating in the same band usually have much higher
transmitted powers. For example, the transmitted power of cellular radios can be up to +30
dBm, which are several orders of magnitude higher than UWB transmitters are permitted.
1.9 Objective of this thesis
The objective of this thesis is to design a ultra-wideband (UWB) low noise amplifier (LNA)
with minimum noise figure, Maximum Signal Voltage gain transfer, minimum power
dissipation
1.10 Organization of this thesis
This thesis begins with an overview of UWB signals and systems followed by a discussion
about UWB LNA and then proceeds this research in two sections: a simple single ended
inductively degenerated LNA with resistive shunt feedback and a fully differential LNA
based on the single ended LNA proposed before. Both sections analyze the performance of
the LNA and systems in detail, and numerous measurements accompany the design choices
to show the improvements and tradeoffs. The final chapter concludes and discusses
exciting new research directions that expand this body of work.
14
Chapter 2
UWB Low Noise Amplifier
Low Noise Amplifier (LNA) is an electronic amplifier, located close to the detection device
which can reduce the losses in the feedline. It is placed at the front of a radio receiver
circuit. The receivers front end is dominated by the first few stages. The effect of noise
from these stages of the receiver chain is reduced by the gain of LNA, while the noise is
injected directly into the receiver signal. So for a LNA, to boost the desired signal power is
necessary. Its main function is to amplify extremely low signals without adding noise, thus
preserving the required Signal-to-Noise Ratio (SNR) of the system at extremely low power
levels.[3] Additionally, for large signal levels, the LNA amplifies the received signal without
introducing any distortions, which eliminates channel interference.
Figure 2.1: (a) Block Diagram of LNA at receiving end
15
Figure 2.1: (b) Signal processing.
A good LNA has-
1. A low NF (like 1 db)
2. A large enough gain (like 20 db)
3. A large enough intermodulation (IP3) and compression point (P1 db)
Criteria are- 1.Operating bandwidth
2. Gain flatness
3. Stability
4. Input-output voltage standing wave ratio (VSWR)
Figure 2.2: Combination of power amplifier circuit and LNA circuit.
16
2.1 Parameters in LNA Design
For any communication system LNAs are basic building blocks. There are 4 important
parameters are in LNA design:
1. Noise figure
2. Input impedance match
3. Gain
4. Linearity
2.1.1 Noise Figure
The ratio between the total output noise power due to all noise sources and the output
noise power generated by the source internal resistance is known as noise figure (NF).[4]
Measuring noise figure in low noise elements become particularly important to the
development of next generation communications system. This application note examines
the process of making practical noise figure measurements of low noise amplifiers. Keeping
a low NF is a crucial process when designing the LNA.
According to Friis for n-stage network equation is:
NFt = 1+ (NF1-1) + (NF2-1)/Ap1 + (NF3-1)/Ap2Ap3 +..+ (NFn-1)/Ap1 Ap(n-1)
Where, Ap =gain of the stage
Overall NF is dominated by NF1.
We can discard the noise contribution of the later stages on overall noise figure
calculation[5], which means we could only focus on the noise situation, which is given
below:
17
Figure 2.3: Equivalent small signal circuit.
It contains current noise in drain ), current noise in gate ,
resistance from Rg and Lg
). The noise figure of such LNA can be expressed as:
.. (1)
Where,
.. (2)
Here,
= Gate noise in the channel
= Thermal noise in the channel
gm1/gdo1 ( =1 in long channel devices)
QL= 1/ 0RsCgs1 = Quality factor of input match network
Cgs1= WLCox
c=0.395 (even consider the short-channel effect [6])
0=work frequency
r=resonance frequency gm1/Cgs1
18
From (1) and (2) we get that,
The value of QL.opt = 3.5 5.5
Finally F reaches its minimun value and that is[7]:
The channel bandwidth,
.(3)
2.1.2 Input Impedance Match
It is mainly conjugate pair of load and source impedances. The input impedance of the
amplifier (Zin) which is now tuned to the source impedance (Z0) using two inductors LS and
Lg. From figure 2.3 input impedance of LNA is expressed as:
Zin=sLg+RLg+Rg+Z1
Where,
Z1= sLs+1/sCgs1+ TLs
We can adjust Ls to make the input impedance that equals Rs (50) when working
frequency is close to 0 ,which means Zin=Re(Zin) = TLs=Rs ; (neglecting RLg & Rg)[8]
Ls =Rs/ T
& Lgs=1/ 02Cgs1-Ls
Ls does not affect the real part of Zin.
19
To get more degrees in the design, we can add additional hardware to the input impedance
matching network. It is possible for more area on the chip.[9]
Figure 2.4: More hardware in the input matching network.
2.1.3 Gain
LNA must have high gain for the processing of signal into post circuit. High gain LNA is
designed for application by manufacturer. If the LNA doesn't have high gain, the signal will
be affected in by noise in LNA circuit itself. So high gain of LNA is a important parameter of
LNA. An inductive load is used to maximize amplifier gain and linearity. The power gain is
given by when LNA is matched at input and output.
Figure 2.5: Equivalent circuit with inductive source degeneration.
20
From this equivalent circuit,we can express the transconductance (Gm1) of the first stage
(common source):
T/2 0Rs
For a certain technology we considered that gm= nCox(W/L)(VGS-VTS),all parameters are
fixed except channel bandwidth. For large gain,we can increase W for the weak signal.
According to equation (3), increased width could elevate noise figure and introduce to
more noise to LNA.
2.1.4 Linearity
Linearity is an important parameter of LNA. There are many ways to describe the linearity
of LNA. High linearity is necessary for low adjacent channel leakage power. For linearity,
third order (IP3) intercept point is the commonly used method. The IIP3 is plotted
graphically by plotting the output power verses the input power both on dB scales. Here
two curves are drawn:
1. One for the linearly amplified signal at an input tone frequency
2. Another for a non-linear product
Figure 2.6: CP1 and 3rd order intercept point.
21
From the figure 2.6, it shows 1 dB compression point (P1 dB) and 3rd order intercept point
(IP3). IP3 is a figure of merit for linearity. For the derivation of IP3, two-tone test is
typically used. IP3 is an important parameter for system designers to estimate the spurious
free dynamic range (SFDR).
The P1 dB is defined as the point where the gain has dropped by 1 dB on the logarithmic
scale of gain as a function of input power. The point where the curves of the fundamental
signal and third order distortion product signal intercept, is known as IP3.The input power
level is known as IIP3 and the output power is known as OIP3.Figure 2.6 shows the output
power vs. input power of the fundamental frequency and the 3rd order intermodulation
(IMD3) product.[10]
The method of measuring two-tone IP3, begins by applying two sinusoids to the circuits
input at frequencies f1 and f2. The frequencies at which the IMD3 products appear for the
signals:
2f1 f2
2f2 f1
Where,
2f1 = 2nd harmonic of f1
2f2 = 2nd harmonic of f2
Figure 2.7: Two-tones with the IDM3 product.[11]
22
IP3 can be calculated by two equations:
OIP3=P1st+(P1st-P3rd)/2 dBm
IIP3=OPI3-G dBm
Where,
P1st = Power of the fundamental in dbm
P3rd= Power of the 3rd order intermodutlation product in dbm[12]
To increase the current density or current draw of the LNA is the easiest way to improve
the IP3 performance.
2.2 Low Noise Amplifier (LNA) Topologies
There are various types of LNA topologies. The two well-known topology structure found
are common source (CS) topology and common gate topology (CG). There are various other
possible topologies, such as:
1. Distributed Amplifier (DA) topology
2. Common Gate (CG) Topology
3. Inductively degenerated common source topology (IDCS)
4. Cascode Topology
5. Folded Cascode Topology
6. Cascade Topology
7. Current reuse topology
8. Differential topology
9. Novel topology
23
2.2.1 Amplifier Distributed (DA)
Distributed amplifiers are widely used in wideband LNA design. This type of structure
which utilizes several stages of common-source amplifiers is well-known for large power
and area consumption (0.8mm2 in and 1.4mm2 in ). It combines the parasitic gate-source
capacitance with on-chip inductors to build transmission lines which have intrinsic
broadband frequency response that goes all the way down to DC.
Figure 2.8: Generalized structure of distributed LNA.[10]
Features of DA topology are:
The capability of designing on-chip T-lines, and high-Q inductors
Integrated circuits incorporating on-chip T-lines
In frequency domain, the transistors parasitic capacitances are absorbed
into the constants of the T-line
Until the cut-off frequency of the line itself is reached, the circuits bandwidth
remains approximately constant
Capable of providing a wideband input/output matching
Advantages of DA topology are:
Very broad band performance
Good Input/output matching
Power combining characteristics
Partially isolated port for dc bias injection[13]
24
Drawbacks of DA topology are:
Large silicon area, which is due to the presence of several on-chip inductors
and/or transmission lines
High power consumption that is related to the number of stages
required to enhance the gain
2.2.2 Common Gate (CG)
The conventional common-gate low-noise amplifier (CGLNA) exhibits a relatively high
noise figure (NF) at low operating frequencies relative to the MOSFET fT, which has limited
its adoption even though its superior linearity, input matching, and stability compared to
the inductively degenerated (CS-LNA). The design of a low-power inductorless LNA using a
capacitive cross-coupled (CCC) gm -boosting and a positive feedback schemes is described
that improves the NF and retains the advantages of the CGLNA. The LNA designed in 0.13
mum CMOS shows more than 15 dB input return loss, maximum gain of 18.2 dB, NF of 2.86
dB, and the third-order input intercept point (IIP3) is 2 dB while dissipating 1.8 mA from
1.2 V supply.
Figure 2.9: Conventional CG- LNA and low noise techniques employing feedback. [10]
25
Features of CG topology are:
It is a parallel resonance as opposed to the series resonance of the
inductor-degenerated LNA
Low Q (quality factor)
Wider bandwidth
Advantages of CG topology are:
Wide operation BW
Accurate ranging
Extremely high-speed data communication
Drawbacks of CG topology are:
Number of on-chip inductors
Becomes more troublesome when designing a balanced or differential LNA
Large group-delay variation due to several resonances in the input-
matching network
It requires moderate to large silicon area and introduces a high insertion-
loss that degrades the amplifier NF and gain
26
2.2.3 Inductively Degenerated Common Source Topology (IDCS)
The inductive source degeneration design is widely used in ultra-wideband low noise
amplifier (3.1-10.6 GHz band). It is a widely used topology in narrow-band RF applications.
The cascode device decouples the input and output matching.
Figure 2.10: (a) Conventional inductively degenerated common source
configuration for narrow band applications. [10]
(b) Inductive degenerated CS for UWB.
Features of IDCS topology are:
Increases the voltage gain of the LNA & the output impedance
Additional cascode device provides large active load to improve voltage
gain at high frequency
Advantages of IDCS topology are:
High gain, low noise, high linearity over a wide frequency range
Doesnt use any off-chip components, it can be easily integrated as one
part of a complete low-voltage transceiver
Drawback of IDCS topology is:
Comparatively large in size.
27
2.2.4 Cascode Topology
The cascode is a two-stage amplifier composed of a transconductance amplifier followed by
a current buffer. Compared to a single amplifier stage, this combination may have one or
more of the following characteristics: higher input-output isolation, higher input
impedance, high output impedance, higher gain or higher bandwidth. In modern circuits,
the cascode is often constructed from two transistors (BJTs or FETs), with one operating as
a common emitter or common source and the other as a common base or common gate.
The cascode improves input-output isolation (or reverse transmission) as there is no direct
coupling from the output to input.
Figure 2.11: Cascode configuration. [10]
Features of Cascode topology are:
Increases the voltage gain of the LNA & the output impedance
Additional cascode device provides large active load to improve voltage
gain at high frequency
Advantages of Cascode topology are:
High gain, low noise, high linearity over a wide frequency range
Doesnt use any off-chip components, it can be easily integrated as one
part of a complete low-voltage transceiver
Drawback of Cascode topology is:
Comparatively large in size
28
2.2.5 Folded Cascode Topology
The basic folded cascode consists of three fundamental circuit elements: a common-base
transistor, an associated emitter resistor and a voltage reference, which is connected to the
base of the cascode transistor. The input of such a stage is in the form of a current, which is
applied to the emitter of the common-base transistor. The output is also in the form of a
current, available at the collector of the common-base transistor.
Figure 2.12: Folded cascade configuration. [10]
Features of Folded Cascode topology are:
Composed of one NMOS transistor & one PMOS transistor
Requires a large supply voltage
Total transconductance is increased with the same current consumption
Advantages of Folded Cascode topology are:
Possesses exclusive advantages in terms of amplifier linearity noise figure,
and bias stability
One single transistor exists in each DC path which increases the voltage
swing and consequently improves the circuit linearity and power gain
Drawback of Folded Cascode topology is:
The inherently low gain is one of the major concerns as the current
consumption is limited
29
2.2.6 Cascade Topology
A cascade amplifier or topology is any diode constructed from a series of amplifiers. It is
basically a differential amplifier with one input grounded and the side with the real input
has no load. It can also be seen as a common collector (emitter follower) followed by a
common base. Since the input side has no load there is no gain on that side and the Miller
effect does not come into play.
Features of Cascade topology are:
Found in multi section or multi stage circuit topology
Two or more stages connected to achieve the high gain
Advantages of Cascade topology are:
Two or more LNA blocks of CS and CG can be connected in cascade to get
desired gain
Drawbacks of Cascade topology are:
Supply voltage requirement is more in cascade topology
Consumes large silicon area as compared to cascode topology
2.2.7 Current Reuse Topology
Low power low voltage operation is found to be the bottleneck of future CMOS system
implementations. To comply with these constrains, a current reuse configuration is here
reported to design UWB Low Noise Amplifiers (LNA).
30
Figure 2.13: Conventional current reuse configuration. [10]
Features of Current Reuse topology are:
Reduce the power consumption of LNA
Preserve high gain
The power consumption, noise and the IIP3 can be improved
Advantages of Current Reuse topology are:
Suitable configuration for LNA implementation because of its low DC
power consumption, high and flat gain, low NF and high reverse
isolation
Can be used with any circuit configuration like cascode, CS,CG and
feedback topologies or even with multi stage cascaded structures to
reduce the DC power consumption
Drawbacks of Current Reuse topology are:
Since the second stage of the traditional current-reused LNA topology
requires a DC bias as well as biasing resistor, they will result in extra
noise and signal leakage
High input and output impedances, thus requiring external impedance
matching networks
31
2.2.8 Differential Topology
Differential topology considers the properties and structures that require only a smooth
structure on a manifold to be defined. Smooth manifolds are 'softer' than manifolds with
extra geometric structures, which can act as obstructions to certain types of equivalences
and deformations that exist in differential topology.
Figure 2.14: Differential capacitive cross coupled configuration.
Figure 2.15: Noise cancelling principle of differential topology. [10]
Features of Differential topology are:
Can Highly diminish the 2nd order nonlinearity
Shunt series triple resonance peaking technique is adopted to achieve
wideband flat gain
32
Advantages of Differential topology are:
Provide the input matching and low noise figure, but the power gain of
6.1-8.5dB (or 5.2-8.2dB) is insufficient to suppress the noise of the
subsequent components
Input matching, high power gain, and low noise figure under the
wideband condition
Better immunity to environmental noise, improved linearity, low power
consumption
Drawbacks of Differential topology are:
Not suitable for low-power applications
Single-stage LNAs consume low power but their output bandwidth
cannot provide the flat-high-gain response
2.2.9 Novel Topology
Few novel topologies are also possible other than the topologies discussed above. Novel
topology is nothing but the combination of above basic topology with either bandwidth
extension techniques or noise cancelling techniques. Even in few papers topology is
invented to optimize BW, gain and NF simultaneously at the cost of other parameters.
2.3 Proposed Topology
After taking all the benefits and drawbacks of different topologies, Inductively Degenerated
Common Source (IDCS) and Cascode topology is chosen for higher gain, better input
impedence matching and Differential LNA is chosen for linearity and stability along with
reverse isolation and immunity to noise. Easement of designing as well as further
improvement of the device is also given priority.
33
Chapter 3
Architecture of proposed UWB LNA and Simulation
Results
For the UWB receiver using direct-conversion architecture, Low Noise Amplifier (LNA) is
an important system block. It is the first gain stage after antenna in the receive path, so the
LNA noise figure is critical since it directly adds to the total noise figure of whole system. It
is also expected for LNA to have good gain, linearity and stability. Furthermore, good
reverse isolation is desirable for LNA to reduce the amount of LO signal that leaks from the
mixer back to antenna.
Figure 3.1: Block Diagram of LNA
Based on the way input matching is provided LNA can be classified into four
architectures [13] : Resistive termination, 1/gm termination, shunt-series feedback and
inductive degeneration.
34
Resistive termination technique is a straightforward matching method which uses a
resistor at the input port to provide a termination of 50 impedance. It suffers from poor
noise performance due to the deleterious effect of real resistor at input port.
Figure 3.2: Common LNA Architectures: Resistive termination
The 1/gm termination method uses the source of common gate stage as the termination
point and the impedance is set by the 1/gm of the transistor in the common gate stage. It is
more suitable for bipolar transistor than CMOS to achieve low noise figure performance.
Figure 3.3: Common LNA Architectures: 1/gm termination
35
Shunt-series feedback uses shunt series feedback to set the input and output impedances of
the system. The bias point of the input is fixed with the output voltage. This method can
increase the power dissipation compared to others with similar noise performance.
Figure 3.4: Common LNA Architectures: Shunt-series feedback
The inductive source degeneration method employs inductor at source terminal to
generate a real term for the input impedance.
Figure 3.5: Common LNA Architectures: Inductive degeneration
This method often works with cascode transistor and has been widely used for LNA design
due to its good noise performance, reverse isolation and high feasible gain.
In this paper, an inductively degenerated LNA with resistive shunt feedback is adopted to
achieve wide bandwidth and good noise performance.
36
3.1 Architecture of the proposed LNA
Figure 3.6 shows the simplified block diagram of the proposed LNA architecture, where
two simple inductively degenerated LNA with resistive shunt feedback is paired to make a
Differential LNA to get higher linearity, better input matching, higher power gain and lower
noise figure for wide band application.
Figure 3.6: Simplified Block Diagram of proposed LNA
37
3.1.1 Single Ended LNA
As shown in Fig. 3.5, this LNA employs cascode transistor (M1 and M2) to reduce miller
effect of gate-drain capacitor (Cgd) and enhance gain. Inductor Ls is for simultaneous noise
and input matching; inductor Lg is for the impedance matching between the source
resistance Rs and the input of the LNA.
Figure 3.7: LNA topology for UWB single ended LNA [13]
Capacitor Cgs in Fig. 3.3 represents the gate-source capacitance of the input transistor M1.
Resistor Rf is a shunt feedback resistor added to the conventional cascode narrowband
LNA. The capacitor Cf is for ac signal coupling purpose. The inductor Lload is a shunt peaking
inductor at the output port. The large amplitude range of signals, received in antenna in
wireless communications systems, requires variable gain stages to enhance the signal to
the noise ratio (SNR). The variable gain feature is realized by including an additional n-type
MOSFET in parallel with the cascode device in the traditional cascode LNA.
38
Figure 3.8: LNA topology for UWB LNA Small-signal equivalent circuit of the input of LNA
While in high-gain mode, the original cascode device steers current signal to LC load and
the additional NMOS bypasses the signal directly to VDD to lower the gain in low gain mode.
Note that in Fig.3.5, Ms1 is used to bypass the signal in lower gain mode. As shown in the
small-signal equivalent circuit (Fig. 3.6) for input part of this LNA, a series combination of
reactive elements is chosen to have a serial resonance at the frequencies of interest. The
symbol t means the cutoff frequency of transistor M1. The -3-dB bandwidth of a typical
RLC resonant circuit is inversely proportional to its Q-factor (BW-3db=0/QWB), where QWB
is approximately determined by following equation.
Where, Rfm=Rf/[1-Av] and Av represents the open-loop voltage gain of LNA. The key role of
the shunt-feedback resistor Rf is to reduce the Q-factor of the LNA input circuit. With low Q-
factor, LNA is capable of providing wide bandwidth (BW-3dB). Resistor Rf also helps in
flattening the gain over a wide bandwidth of frequencies with small noise figure
degradation.
39
Simulation Results of Single Ended LNA
Simulation results show that Single Ended LNA has S21 gain of 10dB with bandwidth 1GHz
between 8GHz and 9GHz with center Frequency 8.5GHz. Simulation results are shown in
Fig 3.9 to Fig 3.12.
Fig 3.9: S21of Single Ended LNA
Fig 3.10: S11of Single Ended LNA
40
Fig 3.11: AC output Noise Figure of Single Ended LNA
At the center frequency 8.5GHz Noise figure of the Single Ended LNA is 2dB. Maximum
Power gain of this LNA is 18.35dB at 7GHz, and 18dB at the center frequency.
Fig 3.12: Maximum Power Gain of Single Ended LNA
41
3.1.2 Differential LNA
LNA in our design, according to the specifications, is differential with variable gain mode.
Since the transceiver architecture is Zero-IF architecture, the output load is the capacitive
impedance seen by LNA when connected to the next mixer stage. A reasonable estimate of
the value of input capacitance of the mixer is 400 fF and is used in our design and
simulations in parallel with a 200 resistance. As shown in Fig. 3.4, a full-differential LNA
is designed based on the single-ended LNA in previous section.
Figure 3.13: Differential LNA [13]
The most important advantage of differential LNAs over single-end ones is that differential
LNA has much higher immunity to environmental noise. This is to make sure that an
adequate rejection of the noise and the interfering signals traveling through the common
substrate.
42
Design Procedure
The load inductor Lload= 2.664 nH is designed to resonate with the output capacitance (400
fF) and the parasitic drain capacitance of the cascode transistor at the center frequency of
the band. If we assume Cd,M2 200fF, then output resonance frequency is:
In order to reduce the Q-factor and improve the bandwidth of the output network, a small
resistor Load= 5 is added in series with the load inductor. The source input matching is
achieved by selecting the source degeneration inductor Ls= 0.17 nH to match the input
resistor 50 such that
The gate inductance Lg= 2.223 nH resonates with Cgs and parasitic source capacitance
(300 fF) of low-gain mode transistor at resonance frequency:
As mentioned earlier, the feedback resistance Rf= 1.14 k and Cf= 2 pF are chosen to
reduce the Q-factor of the input network and AC coupling respectively.
The bias resistance Rbias= 1.417 k is chosen high enough to minimize its noise power
contribution, which is given by:
43
While selecting the W/L ratio of the input transistors, trade-off between the noise power
contributed and the gain needs to be considered. The noise power of the n-channel
MOSFET is given by:
Whereas gain is:
If gm is increased to improve the gain, the noise contributed by the MOSFET increases
proportionally. The W/L ratio of the cascode transistor is selected such that the ratio of
gms of M1 and M2 is close to 550. Finally, the W/L of the low-gain mode transistor is
optimized to give the necessary gain in the low-gain mode.
3.1.3 Balun Circuit
A Balun is a device which converts unbalanced impedance to balanced and vice versa. A
Balun Circuit is used whenever a circuit design requires signals on two lines that are equal
in magnitude and 180 degrees out of phase. In our Design we have used Balun to create
differential input and achieve differential output.
Figure 3.14: Balun Circuit
44
3.2 Final Simulation Results
The Design was simulated with BSIM3 model provided for 0.18 m RF CMOS process. Some
of the important parameters for LNA such as S21, S11, S12, S22, Power gain and Noise Figure
are simulated in HSpice. Figures 3.13 to 3.24 show the corner simulations for each of
these parameters.
Fig 3.15: S21 (dB) of Differential LNA
Fig 3.16: S11 (dB) of Differential LNA
45
Fig3.17: Comparison between S21 and S11
Gain and input matching parameters (S21 and S11) are plotted in Fig 3.15 to Fig 3.17. S21 of
the proposed LNA is plotted in Fig 3.15. S21 is around 8dB from 4.8GHz to 9GHz. It can be
compared with S21 of the single ended LNA in Fig 3.10. S21 of Single ended LNA is 10dB
from 5.6GHz to 9GHz. Comparison shows that Differential LNA gets wider bandwidth but
lower gain.
Fig 3.16 shows S11 of the Differential LNA, which has a value of -30dB at 8.9GHz. Comparing
this plot with the S11 of single ended LNA of Fig 3.11, we get Single ended LNA has a S11 of -
12dB.
So, S21 and S11 of Fig 3.17 assure good input matching of the network.
46
Fig 3.18: AC output Noise figure of Differential LNA
Fig 3.18 shows the AC output noise figure and Fig 3.19 shows the stability factor of the
proposed LNA. AC output noise figure is less than 2.5dB.
Fig 3.19: Minimum AC output Noise figure of Differential LNA
47
Fig 3.20: S22 of Differential LNA
Fig 3.20 shows the S22 of the proposed LNA and the Comparison between S11 and S22 is
shown in Fig 3.21.
Fig 3.21: Comparison between S11 and S22 of Differential LNA
48
Fig 3.22: S12 of Differential LNA
Fig 3.22 shows the S12 of the LNA of this work. The output noise is shown in Fig 3.23.
Fig 3.23: Output Noise of Differential LNA
49
Fig 3.25: Input Noise (dB) of Differential LNA
The input Noise is shown in 3.25. Fig 3.26 shows the comparison between input noise and
output noise.
Fig 3.26: Comparison between input noise (dB) and output noise (dB)
50
Fig 3.24: Maximum Power Gain of Differential LNA
Maximum power gain of the proposed LNA is shown in Fig 3.24. At 7GHz we get the
maximum power gain with a value of 18.3dB. At center Frequency 8.9GHz Power Gain is
more than 18dB.
Comparison of performance between Single ended LNA and Differential LNA in this paper
is given in the table below.
Table3.1: Comparison of performance between Single ended LNA and Differential LNA
Parameter Single Ended Double Ended
Frequency 8.6GHz 8.9 GHz
Power Supply Voltage 1.8 V 1.8 V
Power Consumption 5.6 mW 2.7mW
S21 10 dB 8dB
S11 -12dB -30dB
Noise Figure 2dB 2.15dB
51
The performance of the proposed Differential LNA is summarized in the table below.
Table3.2: Performance of LNA
Figure Value
Maximum Gain 18 dB
BW-3dB 7.5-10.6GHz
Noise Figure < 2 dB
S11 -27 dB
S21 8 dB
Power Consumption 27mW
Bias Current 15mA
Comparison with other works in the reference is given in the table below.
Table3.3: Comparison with other LNA from the bibliography
Technology Bandwidth Gain NF Differential Input Match Power
[15] 0.5m 100 900 MHz
14.8 dB 3.3 dB no no 3.4 mA 3.3V
[11] 0.35m 50 900 MHz 11 dB 4.4 dB no yes 1.5 mA 3.3V
[14] 0.25m 2 1600 MHz 13.7 dB 2.5 dB no yes 14 mA 2.5V
[16] 0.18m .8 1 GHz 26 dB 4.1 dB yes yes 20 mA 1.8V
[17] 0.13m 100 930 MHz
13 dB 4 dB yes yes 0.6 mA 1.2V
This work
0.18 m 8.2-9.4GHz 18 dB < 2 dB yes yes 15mA 1.8V
52
Chapter 4
Conclusion
4.1 Summary of Contribution
A simple inductively degenerated single ended LNA proposed by [13] has been modified
to improve LNA gain, Bandwidth and Noise Figure. A differential topology of the single
ended LNA has also been implemented in this work. The aspect ratio of the MOS
transistor has been optimized in this work for increasing LNA gain. Resistance and
Inductance of the circuit has also been optimized for better matching of the circuit to
improve the reflection coefficient (S11). The circuit has been simulated in HSpice and
HSpice simulation results shows wider bandwidth of 8.2 9.4GHz and higher power gain
of 18dB, better matching performance with S11 around -30dB at 8.8GHz.
4.2 Future Work
LNA in this paper can be extended to operate in other band groups of UWB standard
either by tuning the inductances and capacitances or placing the different matching
circuitry for each band in parallel. Combinations of these methods can also be used to
achieve the same. The simulations on layout are not available at the time of publishing
this paper due to certain limitations in the design kit, which have not been completely
resolved, can be solved in future.
The Noise Figure of the LNA can significantly be improved if a noise cancelling technique
is applied. In the future, noise canceling methods should be investigated while trying to
maintain the other performance measures at almost the same level.
Another issue to be addressed is that the input return loss is relatively high. This is
because the input matching network was not optimized for input return loss, as it was
designed to optimize gain, noise figure and power dissipation instead. However, this
53
problem could likely be solved by trading off increased power consumption for a better
input return loss.
At the circuit level, managing the tradeoff between supply voltage, clock speed, and
power consumption has been widely explored for digital circuits. However, analog
designers are faced with a few challenges without answers as transistors continue to
scale: creating new architectures to support low voltage operation and/or operating at
the optimal supply voltage. Instead of blindly following voltage scaling for digital
circuits, it is important to revisit and analyze tradeoffs made when important techniques
and architectures such as cascoding and op-amps cannot be used. Ultimately, to survive
with voltage scaling, analog designers must think in sub-systems of simple circuits that
functionally provide the same (or better!) performance as compared to well-known,
monolithic analog circuits.
In summary, the UWB LNA has room for improvement. However, it is important to pay
attention to the overall performance (the figure of merit) in order to decide whether a
particular modification is acceptable.
HSpice netlists
Single ended LNA .temp 27 .options post .param Vdd=1.8 .global gnd lload nvdd 1 2.664nH lg nin+ 7 2.223nH ls 4 gnd 0.17nH rload 1 2 5 rf 2 5 1.14K rbias 6 nvbias 1.417K cf 5 6 2pF c1 7 6 300fF c2 2 nvout 400fF Ms nvdd nvm 6 cmosn l=0.18u w=100u M2 2 nvdd 3 cmosn l=0.18u w=100u M1 3 6 4 cmosn l=0.18u w=100u vbias nvbias gnd dc=0.8v vm nvm gnd dc=0.8v vdd nvdd gnd dc=1.8v *vin nin+ gnd dc ac 500mv 0 .ac lin 50 3G 15G P1 nin+ gnd port=1 z0=50 dc = 0 P2 nvout gnd port=2 z0=50 .lin noisecalc=1 sparcalc=1 .print ac s11(db) s12(db) s21(db) s22(db) .print ac nf(db) nfmin(db) .print g_max .probe .end
Differential LNA .temp 27 .options post .param Vdd=1.8 .global gnd .subckt balun 1 2 3 4 E1 5 2 1 0 0.5 V1 3 5 F1 1 0 V1 -0.5 R1 1 0 1T E2 6 4 1 0 0.5 V2 2 7 F2 1 0 V2 -0.5 R2 7 6 1u .ends balun *part 1 lload1 nvdd 1 2.664nH lg1 nin+ 7 2.223nH ls1 4 gnd 0.17nH rload1 1 2 5 rf1 2 5 1.14K rbias1 6 nvbias 1.417K cf1 5 6 2pF c1 7 6 300fF c2 2 nvout1 400fF Ms1 nvdd nvm 6 cmosn l=0.18u w=100u M2 2 nvdd 3 cmosn l=0.18u w=100u M1 3 6 4 cmosn l=0.18u w=100u *part 2 lload2 nvdd 8 2.664nH lg2 nin- 14 2.223nH ls2 11 gnd 0.17nH rload2 8 9 5 rf2 12 9 1.14K rbias2 13 nvbias 1.417K
cf2 12 13 2pF c3 14 13 300fF c4 9 nvout2 400fF Ms2 nvdd nvm 13 cmosn l=0.18u w=100u M4 9 nvdd 10 cmosn l=0.18u w=100u M3 10 13 11 cmosn l=0.18u w=100u vbias nvbias gnd dc=0.8v vdd nvdd gnd dc=1.8v *vin nin gnd dc ac 500mv 0 xdi d c nin+ nin- balun xdo dd cd nvout1 nvout2 balun .ac lin 50 3G 12G P1 d gnd port=1 z0=50 dc = 0 P2 dd gnd port=2 z0=50 .print ac s11(db) s12(db) s21(db) s22(db) .print ac nf(db) nfmin(db) .print noise inoise(db) onoise(db) .print ac g_max .lin noisecalc=1 sparcalc=1
.end
Bibliography
[1] Ultra-Wideband Communications: Fundamentals and Applications By Faranak Nekoogar
[2] Ultra-Wideband Communications Systems: Multiband OFDM Approach By W. Pam
Siriwongpairat, K. J. Ray Liu
[3] Nejati, H.; Ragheb, T.; Nieuwoudt, A.; Massoud, Y., Modeling and Design of Ultrawideband
Low Noise Amplifiers with Generalized Impedance Matching Networks, ISCAS 2007. IEEE,
p.2622-2625, (2007)
[4] Nejati, H.; Ragheb, T.; Massoud, Y., Analytical Modeling of Common-Gate Low Noise
Amplifiers, ISCAS.2008, IEEE, p.888
[5] T. H. Lee, The Design of CMOS Radio-Frequency Integrated Circuits (second edition),
Publishing House of Electronics Industry, 2006
[6] Jiwei Chen, Bingxue Shi, Analysis and optimization of the impact of channel resistance on
CMOS LNA noise performance, Microelectronics Journal 33, p. 1027-1031, (2002)
[7] B. Chi, Z. Yu, B. Shi, Analysis and Design of CMOS RF Integrated Circuits, Tsinghua Press,
2006
[8] Massoud, Y.; Nieuwoudt, A.; Ragheb, T., Variability-Aware Synthesis for Wideband Low
Noise Amplifiers, ISCAS 2007, IEEE, p.3219-3222
[9] Design Considerations And Performance Requirements For High Speed Driver Amplifiers.
By: Nils Nazoa, Consultant Engineer (LA Techniques Ltd)
[10] Sunny Gyamlani, Sameena Zafar, Jigisa Sureja, Jigar Chaudhari, Comparative Study of
various LNA topologiesUsed for CMOS LNA Design,
[11] F. Bruccoleri, et al., Noise cancelling in wideband CMOS LNAs, IEEE ISSCC Dig. Tech.
Papers, pp. 406-407, 2002.
[12] C. Kim, M. Kang, P. T. Anh, H. Kim, and S. Lee, An ultra-wideband CMOS low noise
amplifier for 3-5 GHz UWB system, IEEE Journal of Solid-State Circuits, Vol.40, No. 2, Feb.
2005.
[13] C. Garuda, X. Cui, P.-C. Lin, S. J. Doo, P. Zhang, and M. Ismail, A 3-5 GHz Fully Differential
CMOS LNA with Dual-gain Mode for Wireless UWB Applications, 48th Midwest Symposium on
Circuits and Systems, Aug. 2005
[14] F. Bruccoleri, et al., Generating All Two-MOS-Transistor Amplifiers leads to new wide-
band LNAs, IEEE J. Solid-State Circuits, vol. 36, pp. 1032-1040, July 2001.
[15] J. Janssens, et al., A 10-mW Inductorless, Broadband CMOS Low Noise Amplifier for
900MHz Wireless Communications, Proc. of IEEE CICC, pp. 75-78, May 1998.
[16] Adiseno, et al., A 1.8-V Wide-band CMOS LNA for Multiband Multistandard Front-end
Receiver, Proc. of IEEE ESSCIRC, pp. 141-144,Sept. 2003.
[17] Stanley B. T. Wang, Ali M. Niknejad, and Robert W. Brodersen, A Sub-mW 960-MHz
Ultra-Wideband CMOS LNA, Berkeley Wireless Research Center, Dept. of EECS, UC Berkeley,
Berkeley, CA 94704, USA
[18] Microelectronic Circuits, by Prof. H. Rashid, pp. 554 580