國
立
交
通
大
學
電機學院微電子奈米科技產業研發碩士班
碩
士
論
文
電流重複用之超寬頻金氧半低雜訊放大器應用於
3.1-10.6GH
Z
An Ultra-Wideband CMOS LNA with Current-Reused Technique
for 3.1 to 10.6GH
Z研 究 生:
李富國
指導教授 :
荊鳳德
教授
電流重複用之超寬頻金氧半低雜訊放大器應用於3.1-10.6GH
ZAn Ultra-Wideband CMOS LNA with Current-Reused Technique for
3.1 to 10.6GH
Z研 究 生:李富國 Student: Fu-Kuo Li
指導教授:荊鳳德 博士 Advisor: Dr. Albert Chin
國 立 交 通 大 學
電機學院微電子奈米科技產業研發碩士班
碩 士 論 文
A Thesis
Submitted to College of Electrical and Computer Engineering National Chiao Tung University
in partial Fulfillment of the Requirements for the Degree of
Master In
Industrial Technology R & D Master Program on Microelectronics and Nano Sciences
June 2008
Hsinchu, Taiwan, Republic of China
電流重複用之超寬頻金氧半低雜訊放大器
應用於
3.1-10.6GH
Z學生: 李富國
指導教授: 荊鳳德 博士
國立交通大學電機學院產業研發碩士班
摘
要
本論文研製之一個電流重複使用低雜訊放大器,此低雜訊放大器具有高
增益、低功率、低雜訊且小面積之優點。我們採用三級方式有別於一般電流重複
使用低雜訊放大器,來擴充高頻的頻寬,使增益更為平坦,在第一級部分,我們
是利用帶通慮波器來做input matching,接著第二級部分,利用疊接方式來達到
電流重複使用,以降低功率消耗,最後我們採用一個LC並聯共振電路方式,來增
強高頻部份的增益並增加頻寬作為第三級,在輸出部分則是利用current buffer
的方式來達到output matching。供應電壓V
DD為 1.8 伏特時,整個電路功率消耗
約為 10.3mW,及包含pad的情況下整個電路大小約為 0.89 mm
2。本研究的低雜訊
放大器所量測的規格,平均順向增益(S
21)在 3.1~10.6GHz時為 10dB,逆向隔離(S
12)
為-33dB以下,S11 為-9dB以下,S22 約為-11dB以下,而平均雜訊指數約為 3.6dB。
An Ultra-Wideband CMOS LNA
with Current-Reused Technique for 3.1 to 10.6GH
ZStudent: Fu-Kuo Li Advisor: Dr. Albert Chin
Industrial Technology R &
D Master Program of
Electrical and Computer Engineering College
National Chiao Tung University
ABSTRACT
A procedure
is for an Ultra-Wideband CMOS LNA with Current-Reused
Technique for 3.1 to 10.6GH
Z. It has an advantage of high gain, low power
consumption, low noise performance and small size. We utilize three-stage amplifier
to get a flat gain at high frequency. The first stage introduces the band pass filter for
input matching, the second stage introduces a current-reused cascaded
common-sources structure to lower power consumption. The last stage introduces
shunt-LC resonance to improve gain and the bandwidth. The current buffer
configuration is used for output matching. The total power dissipation of the chip is
about 10.3 mW at power supply 1.8 volt. The chip size included pad is 0.89 mm
2. The
measurement result of this study expect that the average forward S
21is 10dB at
3.1~10.6GHz, the reverse isolation S
12is under -33dB, the magnitude of S
11is under
誌
謝
本篇碩士論文順利完成,首先感謝指導老師荊鳳德教
授,老師熱誠的態度與對研究的敏銳度感染了實驗室每位同學,在老
師的指導下順利完成學業,讓我學習到很多。在此致上我最深的尊敬
與感謝。
感謝交通大學 ED633 實驗室的全體同學,感謝高瑄苓愽士,張
慈愽士,李佩諭學姐 給予我的建議與關心,提供我許多寶貴經驗。
感謝同屆同學陳膺任,劉思麟 在學業上一起經歷這兩年,一起修課,
一起寫作業,一起接受嚴厲考試的試煉;在生活互相砥礪,在研究中
有討論的同伴,一起經歷了美好的日子。還有感謝實驗室所有學長姐
和學弟妹們所給予我實驗室的新鮮有活力的研究風氣。還要感謝我的
家人,我的努力的動力,我不敢忘記家裡的需要與父母的辛苦,是你
們讓我能有美好的環境,有機會挑戰自我的理想。
最後感謝新竹,感謝交大,感謝那些幫助我關心我的人。
Contents
Abstract (in Chinese)
………IAbstract (in English)
………...IIContents
……….IVFigure Captions
……….VIChapter 1 Introduction
1.1 UWB CMOS Receivers
………..………….….11.2 Motivation
………3Chapter 2 Basic Concepts in RFIC Design
2.1 Noise Sources in MOSFETs
………42.1.1Drain Current Noise ……….………4
2.1.2Substrate Thermal Noise………...5
2.1.3 Drain Induced Gate Noise……….6
2.2 Noise Analysis
……….………92.2.1 The Concept of Noise Figure………9
2.2.2 Linearity in RF Circuits………...11
2.2.3 Third-Order Intercept point and The 1-dB Compression Point…………...13
Chapter 3 General Consideration in LNA Circuit Design
3.1 Low Noise Amplifier Basic
………...193.1.1 Impedance Matching Network………20
3.1.2 Stability………...25
3.2 Low Noise Amplifier Architecture Analysis
………..26Chapter 4 UWB CMOS LNA Design
4.1
Design Procedures
………304.1.1 Analysis of Input Matching ………32
4.1.2 Analysis of Output Matching………..34
4.1.3 Gain Analyzing………35
4.1.3 Bandwidth Analyzing………..36
4.2 Simulation Results
………374.3 Measurements and Conclusions
………42Chapter 5 Summary
………...48References
……….49Figure Captions
Chapter 1 Introduction
Chapter 2 Basic Concepts in RFIC Design
Figure 2-1 Substrate thermal noise………5
Figure 2.2 Drain induced gate noise………..6
Figure 2.3 Equivalent circuits………7
Figure 2-4 Definition of the 1-dB compression point ……….17
Figure 2-5 Plot of input output power of fundamental and IM3 versus input power..17
Figure 2-6 Cascaded nonlinear stages………..26
Chapter 3 Basic LNA Design
Figure 3-1 Common-source input stage with inductive source degeneration………..19Figure 3-2 Circuit embedded in a 50 ohm system………22
Figure 3-3 Circuit embedded in a 50 ohm system with matching circuit………22
Figure 3-4 Example of a very sample matching circuit network……….22
Figure 3-5 The eight possible impedance-matching networks with two reactive components……….23
Figure 3-7 Equivalent noise model of Figure 3-1………26
Chapter 4 UWB CMOS LNA Design
Figure 4-1 Circuits diagram……….31Figure 4-2 Chip layout……….32
Figure 4-3 simplification of the first stage………...33
Figure 4-4 the input impedance Zin………..33
Figure 4-5 the current buffer………34
Figure 4-6 Current-reused two stage cascade amplifier with series inter-stage resonance………..35
Figure 4-7 Small signal equivalent representation of the circuit from node X to Y…35 Figure 4-8 the gain of the circuit………..37
Figure 4-9 Simulated S11……….38
Figure 4-10 Simulated S22………...39
Figure 4-11 Simulated S21………...39
Figure 4-12 Simulated S12………...40
Figure 4-13 Simulated NF………40
Figure 4-14 Simulated stability………41
Figure 4-15 simulated IIP3………...42
Figure 4-17 Measured S11………..44
Figure 4-18 Measured S22………...45
Figure 4-19 Measured S12………..45
Figure 4-20 Measured noise figure………..46
Figure 4-21 Measured linearity………46
Chapter 1
Introduction
---
1.1 UWB CMOS Receivers
UWB (Ultra Wideband) is a new wireless technology capable of
transmitting data over a wide frequency bands with low power and high data rates. It
communicates with short pulses on the order of nanoseconds, thus spreading the
energy of the radio signal over a very wide bandwidth. Compared to traditional
narrow band communication systems, UWB technology has the promising ability to
provide high data rate at low cost with relatively low power consumption.
The FCC has allocated 7.5 GHZ of spectrum for unlicensed use of UWB devices in
the 3.1 to 10.6 GHZ frequency band. The low noise amplifier needs to amplify the
received UWB signal with sufficient gain and as little as possible. From Shannon’s
equation C = B log (1+
BNo BSo
= ) for the channel capacity, we know that a UWB
wireless network, the bandwidth will likely be much higher than the data rate, so that
the system can operate at very low signal to noise ratio[1].
The majority implementation of the RF integrated circuit used for wireless
devices are encounter with various possibilities: CMOS, Bi CMOS, and GaAs
We just focus on the CMOS technology, CMOS process reduce the minimum channel
length from the present years, so the unity gain cut off frequency (ft ) is increasing.
For example, a deep sub-micron prototype CMOS technology has realized devices
with ft exceeding 100 GHz [2] and minimum noise figures less than 0.5-dB at 2 GHz.
The more commercially available sub-micron CMOS technologies have display ft’sof
20GHz and minimum noise figures of 1.6-dB at 2 GHZ [3]. The VLSI capabilities of
CMOS make it proper to very high levels of mixed signal radio integration while
increasing the functionality of a single chip radio to cover multiple RF standards [4].
Due to the advancement of circuit design technology, circuit size is small and cost
down consideration. With low cost and low power devices of RF front-end system
implemented by CMOS technology, the prospect of a single chip CMOS system has
received considerable interest. Even the SOC is difficult and hard to implement at this
time, but a set of separate chips in the same CMOS technology may bring significant
1.2 motivation
For portable wireless communication devices has given great push to the
development of a next generation of low power radio frequency integrated circuits
(RFIC) product. Such as wireless phones, cordless and cellular, global positioning
satellite (GPS), pagers, wireless modems, wireless local area network (LAN), and RF
ID tags, etc., require more low cost, low noise and high power efficiency solutions to
supply the demand for low-price product [6].
Chapter 2 discusses the basic concepts in RF design. Chapter 3 presents
the basic low-noise amplifiers design for UWB. Chapter 4 discusses the design
procedures of this circuit by using of the band pass filter , the current buffer
configuration, the shunt-LC resonance, Current-Reused Technique to get a good
input and output matching, broadband, a low power consumption, and also reveals
Chapter 2
Basic Concepts in RFIC Design
---2.1 Noise Sources in MOSFETs
2.1.1 Drain Current Noise
There are three main sources which contribute the thermal noise of
MOSFETs [7]. And the dominate noise source of RF MOSFETs is the drain current
noise which is expressed as:
i
nd=
KT
g
d0Δ
f
2
4
γ
(2.1) where gd0 is the drain-source conductance at zero VDS. The coefficient γ has a vale of
unity at zero VDS and, in long channel devices, decrease toward a value of 2/3 in
saturation [8]. Some measurements show that short-channel devices exhibit noise
considerably in excess of values predicted by long-channel theory, sometimes by an
order of magnitude in extreme cases. Some of the literature attributes this excess noise
to carrier heating by the large electric fields commonly encountered in such devices.
In this view, the high fields produce carriers with abnormally high energies. No longer
in quasi-thermal equilibrium with the lattice, these hot carriers produce abnormal
amount of noise. But in contrast to other groups, we find only a moderate
measurements.
2.1.2 Substrate Thermal Noise
substrate
cbC
subR
drain
source
gate
substrate
cbC
subR
drain
source
gate
Figure 2-1 Substrate thermal noise
Figure 2-1 shows a simplified picture of how the thermal noise associated
with the substrate resistance can produce measurable effect at the main terminals of
the devices. At frequencies low enough that we may ignore Ccb (open), the thermal
noise of Rsub modulates the potential of the back gate, contributing some noisy drain
current:
i
nd
2
,
sub
=
4
KTR
sub
g
mb
2
Δ
f
(2.2) Depending on bias conditions – and also on the magnitude of theeffective substrate resistance and size of the back-gate transconductance – the noise
the ordinary channel charge. In this regime, layout strategies that reduce the substrate
resistance have a noticeable and beneficial effect on noise.
At frequencies well above the pole formed by Ccb and Rsub, however,
the substrate thermal noise becomes unimportant, as is readily apparent from
inspection of the physical structure and the corresponding frequency-dependent
expression for the substrate noise contribution[8]:
f
C
R
g
KTR
i
cb sub mb sub sub nd=
+
2Δ
2 2 ,)
(
1
4
ω
(2.3)The characteristics of many IC processes are such that this pole is often around 1 GHz.
Excess noise produced by this mechanism consequently will be most noticeable
below about 1 GHz.
2.1.3 Drain Induced Gate Noise
2 ng
i
drain
source
gate
2 ngi
drain
source
gate
2 ng
i
g
gC
gs 2 ngi
g
gC
gsFigure 2.3 Equivalent circuits
In addition to drain noise, the thermal agitation of channel charge has
another important consequence: gate noise. The fluctuating channel potential couples
capacitively into the gate terminal, leading to a noisy gate current (see Figure 2-2).
Noisy gate current may also be produced by thermally noisy resistive gate material.
But this noise source will be separately discussed later, even though it is more and
more important in nano-scale devices. Although the drain-induced-gate-noise is
negligible at low frequencies, it can dominate at radio frequencies. Van der Ziel has
shown that the drain-induced-gate-noise may be expressed as:
i
ng2=
4
KT
δ
g
gΔ
f
(2.4) where the parameter gg is:0 2 2
5
d gs gg
C
g
=
ω
(2.5) Van der Ziel gives a value of 4/3 (twice γ) for the gate noise coefficient, δ, in longchannel devices [8].
connected between gate and source, shunted by a noise current source (see Figure 2-3).
This noise current clearly has a spectral density that is not constant. In fact, it
increases with frequency, so perhaps it ought to be called “blue noise” to continue the
optical analogy. Because the drain thermal current noise and the
drain-induced-gate-noise do share a common origin, they are correlated. That is, there
is a component of the gate noise current that is proportional to the drain noise current
on an instantaneous basis.
Although the noise behavior of long-channel devices is fairly well
understood, the precise behavior of δ and γ in the short-channel regime is still
unknown at present. That’s why we have to do more research on the thermal noise of
MOSFETs. Thermal noise of deep sub-micrometer MOSFETs has received
considerable attention lately, which is mainly triggered by publications that report a
severe enhancement of the thermal noise with respect to long-channel theory [9]-[10].
In the earliest of these publications [9], thermal noise was found to be enhanced by a
factor up to 12 in n-channel devices with 0.7μm gate length and hot electrons were
proposed to explain these results. Evidently, the reported noise enhancements would
seriously limit the viability of RF CMOS and a detailed study is called for. Therefore,
in this paper, we perform an extensive study of the RF noise in 0.18μm RF CMOS
2.2 Noise Analysis
2.2.1 The Concept of Noise Figure
Noise is usually generated by the random motions of charges or charge carriers
in devices and materials. Because the noise process is random, one cannot identify a
specific value of voltage at a particular time, and the only recourse is to characterize
the noise with statistical measures, such as the mean-square or root-mean-square
values. Because of having various noise sources in the circuit, we need to simplify
calculation of the total noise at the output [11]. Obviously, the output-referred noise
does not allow a fair comparison of the performance of different circuits because it
depends on the gain. According the circuit theory, we can use the input-referred noise
of circuits to represent the noise of behavior in the circuits.
The signal-to-noise ratio (SNR), defined as the ratio of the signal power to the
total noise power, is an important parameter. In RF circuit, most of the front-end
receiver blocks are characterized in terms of their “noise figure” rather than the
input-referred noise. Noise figure has many different definitions. The most commonly
accepted definition is noise figure
out in SNR
SNR
= , (2.6)
Noise figure is a measure of how much the SNR degrades as the signal passes through
a circuit. If a circuit has no noise source, the SNRout = SNRin, regardless of the gain.
input. Thus, for reliable detection, the previously calculated minimum detectable
signal level must be modified to include the noise from the active circuitry. Noise
from the electronics is described by noise factor F, which is a measure of how much
the signal-to-noise ratio is degraded through the system. We note that
S
G
S
o = ⋅ i (2.7)where is the input signal power, is the output signal power, and is the
power gain
i
S So G
i o S
S . We derive the following equation for the noise factor:
) ( ) ( 0 ) ( ) ( source i total total o o source i i o i
N
G
N
N
S
N
S
SNR
SNR
F
⋅
=
=
=
(2.8)where is the total noise at the output. If is the noise at the output
originating at the source, and is the noise at the output added by electronic
circuitry, then we can write:
) (total o N No(source) ) (added o N ) ( ) ( )
(total o source o added
o
N
N
N
=
+
(2.9)Noise factor can be written in several useful alternative forms:
) ( ) ( ) ( ) ( ) ( ) (
1
source o added o source o total o source i total oN
N
N
N
N
G
N
F
=
=
+
⋅
=
(2.10)This shows that the minimum possible noise factor, which occurs if the
electronics add no noise, is equal to 1.Noise figure NF is related to noise factor F by
F
NF
=
10
log
10 (2.11)an electronic system that adds no noise has a noise figure of 0 dB.
In the receiver chain, for components with loss (such as switches and filters),
the noise figure is equal to attenuation of the signal. For example, a filter with 3 dB of
loss has a noise figure of 3 dB. This is explained by noting that output noise is
approximately equal to input noise, but signal is attenuated by 3 dB. Thus, there has
degradation of SNR by 3 dB [12].
2.2.2 Linearity in RF Circuits
Mathematically, any nonlinear transfer function can be written as series
expansion of power terms unless the system contains memory. While many RF
circuits can be approximated with a linear model to obtain their response to small
signals, nonlinearities often lead to interesting and important phenomena. For
simplicity, we assume that:
...
3 3 2 2 1 0+
+
+
+
=
in in in outk
k
v
k
v
k
v
v
(2.12)One common way of characterizing the linearity of a circuit is called the
two-tone test. In this test, an input consisting of two sine waves is applied to the
circuit. 2 1 2 2 1 1
cos
t
v
cos
t
X
X
v
v
in=
ω
+
ω
=
+
(2.13) When this tone is applied to the transfer function given in (2.12), the result is anumber of terms:
(
)
(
)
(
)
order third order ond desiredX
X
k
X
X
k
X
X
k
k
v
3 1 3 3 sec 2 2 1 2 2 1 1 0 0=
+
+
+
+
+
+
(2.14)(
)
(
)
(
2)
1 2 2 1 2 2 1 3 1 3 2 2 2 1 2 1 2 2 1 1 0 03
3
2
X
X
X
X
X
X
k
X
X
X
X
k
X
X
k
k
v
+
+
+
+
+
+
+
+
+
=
(2.15)These terms can be further broken down into various frequency components.
For instance, the term has a zero frequency (dc) component and another at the
second harmonic of the input:
2 X1
(
v
t
)
v
(
t
X
1 2 1 2 1 1 2 11
cos
2
2
cos
ω
=
+
ω
=
)
N
(2.16)The second-order terms can be expands as follows:
(
)
N
2 2 2 2 2 1 2 2 1 2 2 12
HD dc IM HD dcX
X
X
X
X
X
+ ++
+
=
+
(2.17)
where second-order terms are composed of second harmonics HD2, and mixing
components, here labeled IM2 for second-order intermodulation. The mixing
components will appear at the sum and difference frequencies of the two input signals.
Note also that second-order terms cause an additional dc term to appear.
The third-order terms can be expanded as follows:
(
)
N
N
3 3 2 3 2 2 1 3 2 2 1 3 3 1 3 2 13
3
HD FUND FUND IM FUND IM HD FUNDX
X
X
X
X
X
X
X
+ + + ++
+
+
=
+
(2.18)
Third-order nonlinearity results in third harmonics HD3 and third-order
intermodulation IM3. Expansion of both the HD3 and IM3 terms shows output signals
appearing at the input frequencies. The effect is that third-order nonlinearity can
change the gain, which is seen as gain compression. This is summarized in Table 2.1.
Note that in the case of an amplifier, only the terms at the input frequency are
desired. Of all the unwanted terms, the last two at frequencies 2ω1−ω2 and
1 2
2ω −ω are the most troublesome, since they can fall in the band of desired output if
1
ω is close in frequency to ω and therefore cannot be easily filtered out. These two 2
tone are usually referred to as third-order intermodulation terms (IM3 products)
2.2.3 Third-Order Intercept point and The 1-dB Compression Point
One of the most common ways to test the linearity of a circuit is to apply
two signals at the input, having equal amplitude and offset by some frequency, and
plot fundamental output and intermodulation output power as function of input power
as show in Figure 2-5. From the plot, the third-order intercept point (IP3) is
determined. The third-order intercept point is a theoretical point where the amplitudes
of the fundamental tones at 2ω1−ω2 and 2ω2−ω1 are equal to the amplitudes of
the fundamental tones at ω and 1 ω . 2
fund= 1 3 3
4
9
i ik
v
v
k
+
(2.19)The linear component of (2.19) given by
fund=
k
1v
i (2.20) can be compared to the third-order intermodulation term given byIM3= 3 3
4
3
iv
k
(2.21)The small , the fundamental rise linearity (20dB/decade) and that the
IM3 terms rise as the cube of the input (60dB/decade). A theoretical voltage at
which these two tones will be equal can be defined:
i v
1
4
3
3 1 3 3 3=
IP IPv
k
v
k
(2.22)This can be solved for vIP3 :
3 1 3
3
2
k
k
v
IP=
(2.23)That (2.23) gives the input voltage at the third-order intercept point. The
input power at this point is called the input third-order intercept point (IIP3). If IP3 is
specified at the output, it is called the output third-order intercept point (OIP3).
The third-order intercept point cannot actually be measured directly, since
it is useful to describe a quick way to extrapolate it at a given power level. Assume
that a device with power gain G has been measured to have an output power of at
the fundamental frequency and a power of at the IM3 frequency for a given input
power of , as illustrated in Figure 2-5. On a log plot of and versus , the
IM3 terms have a slope of 3 and the fundamental terms have a slope of 1. Therefore,
1 P 3 P i P P3 P1 Pi
1
3
3
1=
−
−
iP
IIP
P
OIP
(2.24)3
3
3
3=
−
−
iP
IIP
P
OIP
(2.25)since subtracation on a log scale amounts to division of power.
Also note that
i
P
P
IIp
OIP
G
=
3
−
3
=
1−
(2.26)These equations can be solved to given
[
1 3]
[
1 3]
12
1
2
1
3
P
P
P
G
P
P
P
IIP
=
+
−
−
=
i+
−
(2.27)In addition to measuring the IP3 of a circuit, the 1-dB compression point
(Figure 2-4) is another common way to measure linearity. This point is more directly
measurable than IP3 and requires only one tone rather than two. The 1-dB
compression point is simply the power level, specified at either the input or the output,
where the output power is 1dB less than it would have been in an ideally linear device.
2.3 Cascaded Nonlinear Stages
Since in RF systems, signals are processed by cascaded stages, it is
important to know how the nonlinearity of each stage is referred to the input of the
cascade. Consider two nonlinear stages in cascade, as shown in Figure2-6. Assuming
that the input-output relationship is
(
)
(
)
(
)
3 3(
)
(2.28) 2 2 1 1t
x
t
x
t
x
t
y
=
α
+
α
+
α
(
)
(
)
(
)
3 13(
)
(2.29) 2 1 2 1 1 2t
y
t
y
t
y
t
y
=
β
+
β
+
β
Substitute (2.28) into (2.29) results in the relation
)
(
)
2
(
)
(
)
(
1 1 3 1 1 2 2 13 3 3 2t
x
t
x
t
y
=
α
β
+
α
β
+
α
α
β
+
α
β
(2.30)If we consider only the first- and third-order terms, then
3 3 1 2 2 1 1 3 1 1 3
2
3
4
β
α
β
α
α
β
α
β
α
+
+
=
IPA
(2.31)From equation (2.31) can be simplified if the two sides are inverted and squared:
2 , 3 2 2 1 2 2 1 , 3 2 3 2 1
2
3
1
1
IP IP IPA
A
A
α
β
β
α
+
+
=
, (2.32)where AIP3,1 and AIP3,2 represent the input IP3 points of the first and second stages,
respectively. From the above result, we note that as α increases, the overall IP1 3
decreases. This is because with higher gain in the first stage, the second stage senses
Figure 2-4 Definition of the 1-dB compression point
1 3,
IIP
IIP
3,2Figure 2-6 Cascaded nonlinear stages
Table 2.1
Frequency Component Amplitude
dc
(
2)
2 2 1 2 0 2 v v k k + + 1 ω ⎟ ⎠ ⎞ ⎜ ⎝ ⎛ + + 2 2 2 1 1 3 1 1 2 3 4 3 v v v k v k 2 ω ⎟ ⎠ ⎞ ⎜ ⎝ ⎛ + + 2 1 2 2 2 3 2 1 2 3 4 3 v v v k v k 1 2ω 2 2 1 2v k 2 2ω 2 2 2 2v k 2 1 ω ω ± k2v1v2 1 2 ω ω ± k2v1v2 1 3ω 4 3 1 3v k 2 3ω 4 3 2 3v k 2 1 2ω ±ω 2 2 1 3 4 3 v v k 1 2 2ω ±ω 2 2 1 3 4 3 v v kChapter 3
General Consideration in LNA Circuit Design
---
3.1 Low Noise Amplifier Basic
Low noise amplifier is the first gain stage in the receive path so its noise
figure directly adds to that of the system. Therefore, there are several common goals
in the design of LNA. These include minimizing noise figure of the amplifier,
providing enough gain with sufficient linearity and providing a stable 50 ohm input
impedance to terminate an unknown length of transmission line which delivers signal
from antenna to the amplifier [9]. Among LNA architectures, inductive source
degeneration is the most popular method since it can achieve noise and power
matching simultaneously, as shown in Figure 3-1. The following analysis in 3.2 is
based on this architecture. The LNA basic considerations are introduced as follows.
3.1.1 Impedance Matching Network
The need for impedance matching network becomes more important. In order to
deliver maximum power to a load, it must be properly terminated at both the input and
the output ports. The input impedance of a circuit can be any values in order to have
the best power transfer into the circuit. It is necessary to match this impedance to the
impedance of the source driving the circuit. The output impedance also must be
similarly matched in order to deliver maximum power to the 50 ohm load, it must
have the terminations ZS and ZL. The input matching network is designed to
transforms the generator impedance to the source impedance ZS, and the output
matching network transforms the 50 ohm termination to the load impedance ZL.
Consider the RF system shown in Figure 3-2. Here the source and load terminations
are 50ohm, as the transmission lines leading up to the circuit for optimum power
transfer, prevention of ringing and radiation, and good noise behavior. For example,
we needs the circuit input and output impedances matched to the system. In general,
some matching circuit must almost always be added to the circuit, as shown in Figure
3-3.Typically, reactive matching circuits are used because they are lossless, adding no
noise to the circuit, and will only be matched over a range of frequencies and not at
others. If a broadband matching is required, then other techniques may need to be
in Figure 3-4. The series inductance adds an impedance of jωL to cancel the input
capacitive impedance. Note that, in general, when the impedance is complex
(
)
, then to match it, the impedance must be driven from its complex conjugate .jX R+
(
R− jX)
The input, output impedance of a circuit is very common in using reactive
components to achieve impedance transformation, as they will not absorb any power
or add any noise. Thus, series or parallel inductance or capacitance can be added to
the circuit to provide an impedance transformation. Series components will move the
impedance along a constant resistance circle on the Smith Chart. Parallel components
will move the admittance along a constant conductance circle. Table 3.1 summarizes
the effect of each component.
With the proper choice of two reactive components, any impedance can be
moved to a desired point on the Smith Chart. There are eight possible
two-components matching networks, also known as Ell networks, as shown in Figure
3-5. Each will have a region in which a match is possible and a region in which a
Figure 3-2 Circuit embedded in a 50 ohm system
Figure 3-3 Circuit embedded in a 50 ohm system with matching circuit
Figure 3-5The eight possible impedance-matching networks with two
Table 3.1
Component Added Effect Description of Effect
Series inductor z→ z+ jωL Move clockwise along a resistance
circle
Series capacitor z → z− j ωC
Smaller capacitance increases impedance
(
− j ωC)
to movecounterclockwise along a conductance circle
Parallel inductor y → y− j ωL
Smaller inductance increases admittance
(
− j ωL)
to movecounterclockwise along a conductance circle
Parallel capacitor y→ y+ jωC
Move clockwise along a conductance circle
3.1.2 Stability
The stability of an amplifier is a very important consideration in a design and
can be determined from the S parameters, the matching networks, and the
terminations. A two-port network to be unconditionally stable can be derived from
(3.1) to (3.4) .
1
<
Γ
s (3.1)1
<
Γ
L (3.2)1
1
22 1 2 2 1 11−
Γ
<
Γ
+
=
Γ
L L N IS
S
S
S
(3.3)1
1
11 1 2 2 1 22−
Γ
<
Γ
+
=
Γ
S s OUTS
S
S
S
(3.4)The two-port network is shown in Figure 3-6. For unconditional stability any
passive load or source in the network must produce a stable condition. The solution of
(3.1) to (3.4) gives the required conditions for the two-port network to be
unconditionally stable [4]. 21 2 1 2 2 2 2 2 11
2
1
S
S
S
S
k
=
−
−
+
Δ
(3.5) 21 12 22 11S
S
S
S
−
=
Δ
(3.6) A convenient way of expressing the necessary and sufficient conditions forunconditional stability is
k
>
1
(3.7)1
<
Δ
(3.8)E
SZ
sΓ
sΓ
I NZ
I NT w o - p o r t
n e tw o r k
O U TΓ
LΓ
Z
O U TZ
LFigure 3-6 Stability of two-port networks
3.2 Low Noise Amplifier Architecture Analysis
gs C 2 ngu i vgs gmvgs ind2 2 Rg v -+ 2 ngc i g R l R g L s L s R 2 Rl v 2 s v 2 out i
Figure 3-7 Equivalent noise model of Figure 3-1
In Figure 3-7, the input impedance can be expressed as
gs s g o s T s gs m s gs m gs s g in C L L at L L C g L C g sC L L s Z ) ( 1 1 ) ( + = = ≈ ⎟ ⎟ ⎠ ⎞ ⎜ ⎜ ⎝ ⎛ = ⎟ ⎟ ⎠ ⎞ ⎜ ⎜ ⎝ ⎛ + + + = ω ω ω (3.9)
frequency of the device and source inductance at resonant frequency. Therefore it can
be set to 50 ohm for input matching while resonant frequency is designed to be equal
to the operating frequency.
According to prior introduction, the equivalent noise model of common-source
LNA with inductive source degeneration can be expressed as Figure 3-7, where
is the parasitic resistance of the inductor and is the gate resistance of the device.
Note that the overlap capacitance C
l
R
g
R
gd has also been neglected in the interest of
simplicity. Then the noise figure can be obtained by computing the total output noise
power and output noise power due to input source. To find the output noise, we first
evaluate the trans-conductance of the input stage. With the output current proportional
to the voltage no Cgs and noting that the input circuit takes the form of series-resonant
network, the transconductance at the resonant frequency can be expressed as
s o T s T s gs o m in m m R L R C g Q g G ω ω ω ω ( + ) = 2 = = (3.10)
where Qin is the effective Q of the amplifier input circuit. So the output noise power
density due to the source can be expressed as
2 2 2 2 . , ) 1 ( 4 ) ( s s T s o T eff m Rs o Rs a R L R kT G S S ω ω ω ω + = = (3.11)
2 2 2 2 , , ) 1 ( ) ( 4 ) ( s s T s o T l g o R R a R L R R R kT S l g ω ω ω ω + + = (3.12)
Furthermore, channel current noise of the device is the dominant noise
contributor, and its noise power density associated with the correlated portion of the
gate noise can be expressed as
2 , , ) 1 ( 4 ) ( s s T do o i i a R L g kT S ngc nd ω γκ ω + = (3.13)
where γ is the coefficient of channel thermal noise, α =gm/gd0 and
2 2 2 5 2 1 5 ⎥⎦ ⎤ ⎢ ⎣ ⎡ + + = γ δα γ δα κ c cQL (3.14) gs s o L C R Q ω 1 = (3.15)
The last noise term is the contribution of the uncorrelated portion of the gate
noise, and its output noise power density can be expressed as
2 , ) 1 ( 4 ) ( s s T do o i a R L g kT S ngu ω γξ ω + = (3.16) where ) 1 )( 1 ( 5 2 2 2 L Q c + − = γ δα ξ (3.17)
According to equations (3.11), (3.12), (3.13) and (3.16), the noise figure at the
resonant frequency can be expressed as
⎟⎟ ⎠ ⎞ ⎜⎜ ⎝ ⎛ + + + = T o L s g s l Q R R R R F ω ω α γχ 1 (3.18)
where ) 1 ( 5 5 | | 2 1+ c QL 2 + 2 +QL2 = γ δα γ δα χ (3.19)
From equation (3.19), we observe that χ includes the terms which are
constant, proportional to , and proportional to . It follows that equation (3.19)
will contain terms which are proportional to as well as inversely proportional to
. A minimum noise figure, therefore, exits for a particular .
L Q QL2 L Q L L Q Q
Chapter 4
UWB CMOS LNA Design
---
4.1 Design Procedures
This architecture is for an Ultra-Wideband CMOS LNA with
Current-Reused Technique for 3.1 to 10.6 GHZ. It has an advantage of high gain, low
power consumption, low noise performance and small size. We utilize three-stage
amplifier to get a flat gain at high frequency. The first stage introduces the band pass
filter for input matching, it includes two inductances, one capacitor and one
inductance at gate. The second stage introduces a current-reuse cascaded
common-sources structure to lower power consumption. The last stage introduces LC
in shunt connection to improve gain and bandwidth for high frequency. The current
buffer configuration is used for output matching. Its circuit diagram is shown in
Figure 4-2 Chip layout
4.1.1 Analysis of Input Matching
In input part, we adopt a common-source amplifier, by introducing
matching components to let the frequency resonate at which we want. We also
1 1 1 1 1 1 1
1
s gs m s g gsL
C
g
)
L
s(L
sC
Z
=
+
+
+
(4.1)In addition, we put a shunt-LC resonance in input part, following with the
equivalent circuit Z1, to make it have a good resonance in the imagine part, and get
real part impedance-matching
(R
s=g
m1L
s1/C
gs1)
at the same time. Form figure 4-4we can get the input matching impedance Zin in equation (4.2).
1 2 2 1 2 2 ) 1 // ( ) 1 // ( Z sC sL Z sC sL Zin + = (4.2)
We can also keep a very low noise to procure a band-pass amplifier for an entire
Ultra-Wideband.
A
Z1
Figure 4-3 simplification of the first stage
4.1.2 Analysis of Output Matching
The current buffer configuration is used for output matching as shown in
Figure4-5,and make the entire frequency matching to 50 Ω.In this circuit, M3 is the
source follower, and the steady current of M3 is provided by M4.We only need to
control the supply voltage and the size of the transistor, and then a good output
matching will be obtained as equation below.
3 4 3
1
||
1
m o m outg
r
g
Z
=
≈
(4.3)4.1.3 Gain Analyzing
Figure 4-6 Current-reused two stage cascade amplifier with series inter-stage resonance
Figure 4-7 Small signal equivalent representation of the circuit from node X to Y
In Figure 4-6 is the Current-reused two stage cascade amplifier with
series inter-stage resonance, and the Small signal equivalent representation of the
circuit from node X to Y is in Figure 4-7. If we divide id2 to id1, we can get the gain as in equation (4.4). (4.4) 2 2 2 2 / S R id +
so we can see that resonates with at high frequency
2 2 2 2 1 2 ) / 1 1 ( g gs LS gs m sL sC sC L L sC g id + + + × = g g sC 2 LS 1/sC 2 +1/ s2
and. If provides high impedance, then, we can get 1 2 id id as below. 2 L L2 >>RLS2 ω ωT gs m sC g id id ≅ ≅ 2 2 1 2 (4.5)
where is M2’s cutoff frequency, is the operation frequency. ωT ω
4.1.4 Bandwidth Analyzing:
In order to get a low power consumption low noise amplifier LNA, we
cascade the second amplifier M2 on the first amplifier M1 to use the same dc current
to achieve the goal of power saving. Besides we also utilize substrate bias at M1 to
enhance the gain, at the same time reduce the power consumption.
For gain and the bandwidth, we utilize three-stage amplifier to achieve a
high gain and the flat bandwidth design. In the circuit, the requirement of the
current-reused structure is Coupling capacitor (C2) which is connected to the output of
the first stage amplifier (at drain) and also connected to the input of the stage second
stage amplifier (at gate) .Therefore it can provide signal coupling between the two
stages and the Bypass capacitor (Cbypass) functions as an ac ground at the source of
transistor M2, for avoiding the coupling effect to the first stage. Aside from, we join
the shunt-LC resonance CB and LB, by this LC resonance the bandwidth and the gain
can be improve a lot in order to get a high gain Ultra-Wideband circuit. In Figure 4-8,
Figure 4-8 the gain of the circuit
4.2 Simulation Results
Figure 4-9 shows the Return loss Simulation. S11 is lower than -9.4dB
between 3.1 and 10.6GHz. The output buffer achieves excellent matching such that
S22 is lower than -11dB from 3.1GHz to 10.6 GHz in Figure 4-10. Figure 4-11 is the
power gain versus frequency, the gain at 3.1GHz is 13.567dB and at 10.6GHz is
13.319, so the different between the wideband is only 0.25dB.The reverse isolation
S21 is lower than -38dB shown in Figure 4-12. The noise figure (NF) of this UWB
LNA is shown in Figure 4-13.The NF at 3.1GHz is 2.686dB, at 10.6GHz is 2.842dB
and the minimum noise figure (NFmin) at 5.7GHz is 2.146dB. The stability factor
4-14.The third-order intermodulation distortion(IIP3) is -3dBm and shown in Figure
4-15 ,the test is performed at 4 GHz. These results imply excellent linearity of our
LNA. The proposed UWB LNA dissipates 10.3mW with a power supply of 1.8V.
Figure 4-10 Simulated S22
Figure 4-12 Simulated S12
Figure 4-15 simulated IIP3
4.3 Measurements and Conclusions
The following Figure4-16 ~ Figure4-21 are the measurement result which
are slightly different from simulation. Which imply good accuracy of simulation and
good circuit design. But the gain dropping at high frequency (in Figure 4-16) is due to
The bandwidth of this work with considering matching and gain is from
3.1 to 10.6 GHz, while the average gain is about 10dB. Figure 4-17 and Figure 4-18
show the measurement results of S11 and S22. Input and output matching are
achieved very well from 3.1 to 10.6 GHz, so S11 can bellow -9dB and the S22 can
bellow -11dB.S12 below up to -33dB in Figure 4-19. The noise figure is about 3.6
shown in Figure 4-20.The noise performance is very flat and the minimum noise
figure is 2.85dB at 5GHz. The noise figure can be better if we solve the resistor
parasitic. The linearity of the third-order intermodulation distortion(IIP3) is -3dBm
as measurement and shown in Figure 4-21. Figure 4-22 shows the die photo of this
circuit. Total power consumption is 10.3 mW which the Vdd is 1.8V.In the chip, the apply Vb is 0.1V, Vg1 is 0.61V and Vg2 is 0.61V. Table 4.1 is the measurement result summary. By the band pass filter, the current buffer configuration, the shunt-LC
resonance, Current-Reused Technique we proposed, a good input and output matching,
broadband, a low power consumption LNA is developed for UWB system
3 4 5 6 7 8 9 10 2 11 0 5 10 -5 15 freq, GHz d B (S (2 ,1 )) Figure 4-16 Measured S21 3 4 5 6 7 8 9 10 2 11 -25 -20 -15 -10 -30 -5 freq, GHz d B (S (1 ,1 )) Figure 4-17 Measured S11
3 4 5 6 7 8 9 10 2 11 -22 -20 -18 -16 -14 -24 -12 freq, GHz d B (S (2 ,2 )) Figure 4-18 Measured S22 3 4 5 6 7 8 9 10 2 11 -38 -36 -34 -40 -32 freq, GHz d B (S (1 ,2 )) Figure 4-19 Measured S12
0 5 10 15 2 3 4 5 6 7 8 9 10 11
Freq,GHZ
NF(dB)
Figure 4-20 Measured noise figure
Input Power (dB)
-40 -30 -20 -10 0Output Power (dB)
-100 -80 -60 -40 -20 0 OP3 OP1--
3
3
Figure 4-22 Die photo
B.W
(GHz)
Gain
(dB)
NF
(dB)
S11
(dB)
S22
(dB)
IIP3
(dBm)
Pdc
(mW)
3.1~10.6
10
3.6
< -9
< -13
-3
10.3
Chapter 5
Summary
---By the Current-Reused Technique for 3.1 to 10.6GHZ we proposed, a good
input and output matching, broadband, a low power consumption amplifier is
developed for UWB system applications.
Table 5.1 is the comparison of broadband LNA performance. We can find out
by this table, by using Current-Reused Technique and shunt-LC resonance, can pull to
being wide very big arrival 3.1~10.6GHz frequently. All the advantages are important
for UWB system considerations.
specifications This work Ref. [14] Ref. [15] Ref. [16] Ref. [17] Process 0.18 μm 0.18 μm 0.18 μm 0.18 μm 0.18 μm Frequency (GHz) 3.1~10.6 2.4~9.5 2~4.6 2~6.5 3.1~10.6 S11 (dB) <-9 <-9 <-9 <-7.8 <-8 S22 (dB) <-13 <-20 <-10 <-16 ---- Gain (dB) 10 9.3 9.8 11.9 13.5~16 NF (dB) 2.85~4.5 4~9 2.3~6 4.1~4.6 3.1-6 IIP3 (dBm) -3 -6.7 -7 4 -7 Pdiss (total) (mW) 10.3 18 12.6* 27 11.9 Die size (mm2) 0.89 1.1 0.9 0.88 1.2 Table 5.1 Comparison of broadband LNA performance (*Only core LNA)
Reference
[1] P.Heydari, “Design Consideration for Low-Power Ultra Wideband Receivers,”
IEEE Quality of Electronic Design, 2005. ISQED 2005. Sixth International
Symposium on 21-23 March 2005 Page(s):668-673.
[2] H. S. Momose, F. morifuji, T. Yoshittomi, T Ohguro, M. Saito, T. Morimoto, Y.
Katsuma, H. Iwai, “High frequency AC characteristics of 1.5nm gate oxide
MOSFET ” IEEE international Electron Device Meeting, December 1996.
[3] S. P. Voinigescu, S. W. Tarasewicz, T. MacElwee, and J. Ilowski, ”An assessment
of the state-of-the-art 0.5um bulk CMOS technology for RF applications” proc. IEEE
internationall Electron Devices Meeting, 1995.
[4] J.C. Rudell, J.J. Ou, R. S. Narayanaswami, et al. “Recent development in high
integration multi-standard CMOS transceivers for personal communication systems”
invited paper at the 1998 International Symposium on Low Power Electronics,1998.
[5] C.Yoo and Q.Huang, “A common-gate switched,0.9W class E power with 41%
HI),pp.56-57, June 2000.
[6] P. Miliozzi, K. Kundert , K. Lampaert , P. Good, and M. chian, “A design system
for RFIC: Challenges and solutions.” Proceedings of the IEEE, Oct.2000.
[7] T. H. Lee, “The Design of CMOS Radio-Frequency Integrated Circuits,”.
Cambridge University Press,1998.
[8] A. van der Ziel, Noise in Solid State Devices and Circuits, Wiley, New York, 1986.
[9] A. A. Abidi, “High-frequency noise measurements on FET’s with small
dimensions,” IEEE Trans. Electron Devices, vol. ED-33, pp. 1801–1805, 1986.
[10] Klein, P. “An analytical thermal noise model of deep submicron MOSFET’s,”
IEEE Electron Device Lett., vol. 20, pp. 399–401, 1999.
[11] T. H. Lee, The Design of CMOS Radio-Frequency Integrated Circuits, 1st ed.
NewYork: Cambridge Univ. Press, 1998.
Boston :Artech House,c2003.
[13] B. Razavi, RF Microelectronics, 1st ed. NJ, USA: Prentice-Hall PTR, 1998.A.
[14]Bevilacqua, and Ali M. Niknejad, “An Ultrawideband CMOS Low-Noise
Amplifier for 3.1-10.6 GHz Wireless Reveivers,” in IEEE Journal of Solid-State
Circuits, Vol. 39, No. 12, 2005, pp. 2259-2268.
[15]C. W. Kim, M.-S. Kang, P. T. Anh, H.-T. Kim and S.-G. Lee, “An Ultra-Wideband
CMOS Low Noise Amplifier for 3-5GHz UWB System,” in IEEE Journal of
Solid-State Circuits, Vol. 40, No. 2, 2005, pp. 544-547.
[16]J. Jung, K. Chung, T. Yun, J. Choi, and H. Kim, “Ultra-wideband low noise
amplifier using a cascade feedback topology,” in Silicon Monolithic Integrated
Circuits in RF Systems Dig., 2006, 202-205.
[17]Y.-J. Lin, S. S. H. Hsu, J.-D. Jin and C. Y. Chan, “A 3.1-10.6 GHz
Ultra-Wideband CMOS Low Noise Amplifier with Current-Reused Technique,” in