Chapter 1 Introduction
1.3 Thesis Organization
This thesis discuss about the receiver front-end circuits (low noise amplifier and mixer) design and implementation for the UWB frequency bands. The contents consist of two major topics: “A 0.18µm CMOS ultra-wideband low noise amplifier circuit” and “A 0.18µm CMOS 3~8GHz direct conversion broadband mixer circuit”.
In Chapter 2, we will introduce the fundamentals of conventional low noise amplifiers and mixer. And some theoretical MOSFET noise model and noise theory are presented. LNA noise analysis also introduce in this chapter. Furthermore, many important design parameters and direct conversion receiver front end would be presented in this chapter.
In Chapter 3, we start to design an ultra-wideband low noise amplifier. The matching method of ultra-wideband amplifiers design is negative resistor feedback plus the inductive degeneration method to enhance bandwidth. And the feedback resistors are the main components to achieve the ultra-wideband matching.
In Chapter 4, we design a 3~8 GHz direct convention broadband mixer based on gilbert cell structure. In this work, we will use chebyshev bandpass filter method to satisfy input return loss less than -10dB in our interest band.
In Chapter 5, conclusions and future work. Some issues that should be noted for future works on this topic are also summarize.
Chapter 2
The Fundamentals of Low Noise Amplifier and Mixer
In this chapter, we will introduce the principles of the low noise amplifier and mixer. Furthermore, this chapter also shows what kinds of parameters and consideration are important in LNA and mixer design. In section 2.1 illustrates the noise sources type and MOSFET noise model [2,3,4]. In section 2.2 shows principle of low noise amplifier and its noise analysis [3,4]. In section 2.3 shows the principle of mixer and some of its important characteristic [5,6]. In section 2.4, we will introduce the advantage and disadvantage of the direct conversion receivers [6].
2.1 Noise Analysis
The LNA is the first stage of the receiver front end. The noise performance is first consideration in design low noise amplifiers. Therefore, the study of noise is very important because it represents a lower limit to the size of electrical signal that can be amplified by a circuit without significant deterioration in signal quality. In this section, the various sources of electronic noise are considered separately. And MOSFET’s noise model will be described here.
2.1.1 Source of Noise Type
In the integrated circuit, there are many kinds of noise sources, such as shot noise, thermal noise and flicker noise (1/f noise).
Shot noise is always associated with a direct current flow and is present in diodes, MOS and bipolar transistors. The origin of shot noise can be seen by considering the diode and the carrier concentrations in the device in the forward-bias region. The
passage of each carrier across the junction, which can be modeled as a random event, is dependent on the carrier having sufficient energy and a velocity directed toward the junction. The shot noise current can be represented as i2 =2qID∆f , q is the electronic charge (1.6 10 C× −19 ), ∆ is the bandwidth in hertz. f
Thermal noise is generated from the conventional resistors. It is due to the random thermal motion of the electrons and is unaffected by the presence or absence of direct current, since typical electron drift velocities in a conductor are much less than electron thermal velocities. Since this source of noise is due to the thermal motion of electrons, we expect that it is related to absolute temperature T. The thermal noise can be represented by a series voltage generator v or shunt current generator 2 i . These representations are 2 v2 =4kTR f∆ ,i2 =4kT(1/ )R ∆f where k is Boltzmann’s constant. At room temperature:4kT =1.66 10× −20V− . C
Flicker noise is a type of noise found in all active devices, as well as in some discrete passive elements such as carbon resistor. The origins of flicker noise are varied, but it is caused mainly by traps associated with contamination and crystal defects. These traps capture and release carriers in a random fashion and the time constants associated with the process give rise to a noise signal with energy concentrated at low frequencies. Flicker noise, which is always associated with a flow of direct current, displays a spectral density of the form i2 =K I1( /a fb)∆f , where
∆ is small bandwidth at frequency f , I is direct current, K1 is constant for a f particular device, a is constant in the range 0.5 to 2, b is constant of about unity. It is apparent that flicker noise is most significant at low frequencies, although in devices exhibiting high flicker noise levels, this noise source may dominate the device noise at frequencies well into the megahertz range.
2.1.2 Noise Model of MOSFET
MOSFET’s noise source mainly comes from gate current noise, drain current noise (channel thermal noise) and flicker noise.
Drain current noise id2 (channel thermal noise) is the dominant noise source of MOS devices. Since MOSFETs are essentially voltage-controlled resistors, they exhibit thermal noise. In the triode region of operation particularly, one would expect noise commensurate with the resistance value. Indeed, detailed theoretical considerations lead to the following expression for the drain current noise of FETs.
d2
i =4kT gγ d0∆f (2-1) where gd0 is the drain-source conductance at zero VDS. The parameter g has a value of unity at zero VDS and, in long devices, decreases toward a value of 2/3 in saturation. Note that the drain current noise at zero VDS is precisely that of an ordinary conductance of value gd0.
Another kind of thermal noise is gate current noise ig2. The fluctuating channel potential couples capacitively into the gate terminal, leading to a noisy gate current.
Although this noise is negligible at low frequencies, it can dominate at radio frequencies. The gate current noise may be expressed as
g2
where δis the coefficient of gate noise, classically equal to 4/3 for long-channel devices. Equation (2-2) is valid when the device is operated in saturation.
The gate noise is partially correlated with the drain noise, with a correlation coefficient given by
≡ × ≈
where the value of 0.395j is exact for long-channel devices. The correlation can be treated by expressing the gate noise as the sum of two components, the first of which is fully correlated with the drain noise, and the second of which is uncorrelated with the drain noise. Hence, the gate noise is re-expressed as
g2 2
Because of the correlation, special attention must be paid to the reference polarity of the correlated component. The value of c is positive for the polarity.
In electronic devices, 1/f noise (flicker noise) arises from a number of different mechanisms, and is most prominent in devices that are sensitive to surface phenomena.
Charge trapping phenomena are usually invoked to explain 1/f noise in transistors.
Some types of defects and certain impurities can randomly trap and release charge.
The trapping times are distributed in a way that can lead to a 1/f noise spectrum in both MOS and bipolar transistors. Larger MOSFETs exhibit less 1/f noise because their larger gate capacitance smooths the fluctuation in the channel charge. Here, if good 1/f noise performance is to be obtained from MOSFETs, the largest practical device sizes must be used (for a given gm). The mean-square 1/f drain noise current is given by where A (=WL) is the area of the gate and K is a device-specific constant. Thus, for a fixed transconductance, a larger gate area and a thinner dielectric reduce this noise term.
2.2 Principle of Low Noise Amplifiers
In the design of low noise amplifiers, there are several common goals. These include minimizing the noise figure of the amplifier, providing gain with sufficient linearity – typically measured in terms of the third order intercept point, IP3 and providing a stable 50Ω input impedance to terminate an unknown length of transmission line which delivers signal from the antenna to the amplifier. A good input matching is even more critical when a preselect filter precedes the LNA because such filters are often sensitive to the quality of their terminating impedances. The additional constraint of low power consumption which is imposed in portable systems further complicates the design process.
2.2.1 Basic Topologies of Low Noise Amplifier
Fig 2.1 Common LNA architectures (a) Resistive termination (b) 1/gm termination (c) Shunt - series feedback (d) Inductive degeneration [3]
Here, there are four basic topologies of low noise amplifiers as shown in Fig. 2.1.
These techniques provide good stable input impedance matching. The first technique of low noise amplifiers is shown in Fig. 2.1(a). It uses resistive termination of the input port to provide 50Ω impedance. Unfortunately, the use of real resistors in this fashion has a deleterious effect on the amplifier’s noise figure.
The second architecture, shown in Fig. 2.1(b), uses the source or emitter of a common gate or common base stage as the input termination. It’s also called 1/gm termination architecture. Assuming matched conditions, yields the following lower bounds on noise factor for the cases of bipolar and CMOS amplifiers:
Bipolar: 3
The bipolar representation neglects the effect of base resistance in bipolar devices. In CMOS expressions, γ is the coefficient of channel thermal noise, gm is the device transconductance, and gd0 is the zero bias drain conductance. For long channel devices, γ=2/3, α=1. But in short channel MOS devices, γ can be greater than one, and α can be much less than one. Accordingly, the minimum theoretically achievable noise figures tend to be around 3dB or greater in practice.
The third topology is shown in Fig. 2.1(c). This architecture uses resistive shunt and series feedback to set the input and output impedances of the LNA. Amplifiers using shunt-series feedback often have extraordinarily high power dissipation compared to others with similar noise performance. Intuitively, the higher power is partially due to the fact that shunt series amplifiers of this type are naturally broadband, and hence techniques which reduce the power consumption through LC tuning are not applicable. For GPS applications, a broadband front end is not required,
and it is desirable to make use of narrowband techniques to reduce power.
The fourth topology is shown in Fig. 2.1(d). It is desirable to have a narrowband RF signal processing, to get rid of out of band blockers. It employs inductive source or emitter degeneration to generate a real term in the input impedance. It offers the possibility of achieving the best noise performance of any architecture.
2.2.2 L-degeneration LNA Noise Analysis
Fig. 2.2 Cascode LNA architecture
Fig. 2.3 Small-signal model for LNA noise calculations [3]
Fig. 2.2 shows the basic cascode LNA architecture. A simple analysis of the input impedance shows that
in s g m1 s T s shown in Fig. 2.3. In this circuit, AR represents the series resistance of the inductor Lg, Rg is the gate resistance of the NMOS device, id2 represents the channel thermal noise of the device. ig2,c represents the portion of the total gate noise that is correlated with the drain noise. ig2,u represents the portion that is uncorrelated with the drain noise. Analysis based on this circuit neglects the contribution of subsequent stages to the amplifier noise figure. Recall that the noise factor for an amplifier is defined as
Total_output_noise
F=Total_output_noise_due_to_the_source (2-7)
To evaluate the output noise when the amplifier is driven by a 50Ω source, we first evaluate the transconductance of the input stage. With the output current proportional to the voltage on Cgs, and noting that the input circuit takes the form of a series-resonant network Qin is the effective Q of the amplifier input circuit. From the equation (2-8), the output noise power density due to the 50Ω source is
T2
In a similar way, the output noise power density due to
R
Aand Rg can be expressed asg
Next, the noise power density associated with the correlated portion of the gate noise and drain noise can be expressed as
d g d
The last noise term is the contribution of the uncorrelated portion of the gate noise.
This contributor has the following power spectral density:
g d
We observe that all of the noise terms contributed by the first device M1 are proportional to
S
a i, d( )ω
o , the contribution of the drain noise. Hence, it is convenient to define the contribution of M1 as a whole asd
where, after some slight simplification
2 2 with (2-14) and (2-15), it is clear that the effect of induced gate noise is to modify the
noise contribution of the device in proportion to
χ
. It follows directly thatTo understand the implications of this new expression for F, we observe that
χ
includes terms which are constant, proportional to QL, and proportional to Q2L. It follows that (2-18) will contain terms which are proportional to QL as well as inversely proportional to QL. Therefore, a minimum F exists for a particular QL.
2.3 Principle of Mixers
The mixer is an ubiquitous component of wireless systems. An ideal mixer multiplies the signal at the radio frequency (RF) port with a signal at the local-oscillator (LO) port to create the intermediate-frequency (IF) signal. If the RF and LO signals are sinusoids, it is clear that the IF signal has components at two frequencies. There is a high frequency component at the sum of the RF and LO frequencies, and a low frequency signal at the difference of the RF and LO signals.
Therefore, a mixer can effect up conversion or down conversion [5]. In design of RF CMOS mixers is a very challenging task involving trade offs between gain, linearity, noise figure, voltage supply and power consumption.
2.3.1 Conversion Gain
One important mixer characteristic is conversion gain, which is defined as the desired IF output
Conversion Gain=
the value of the RF input (2-19) The gain of mixers must be carefully defined to avoid confusion. The “voltage conversion gain” of a mixer is defined as the ratio of the rms voltage of the IF signal to the rms voltage of the RF signal. Note that these two signals are centered around two different frequencies. The voltage conversion gain can be measured by applying a sinusoid at ω and examining the amplitude of the downconverted component at RF ω . IF
The “power conversion gain” of a mixer is defined as the IF power delivered to the load divided by the available RF power from the source. If the input impedance and the load impedance of the mixer are both equal to the source impedance, for example, 50Ω, then the voltage conversion gain and power conversion gain of the mixer are equal when expressed in decibels.
Conjugate matching at the input of the mixers is necessary in the first downconversion stage of heterodyne receivers that employ image reject filters. The transfer function of these filters is usually characterized for only a standard termination impedance and may exhibit ripples if other impedance levels are used.
The load impedance of the mixer, on the other hand, is typically not equal to 50Ω because most passive IF filters have an input impedance of 500 to 1000Ω. In architectures such as homodyne topologies, the load seen by the mixer may be even higher to maximize the voltage gain.
From the above observation, we note that the voltage and power conversion gains of a mixer may not be equal in decibels.
2.3.2 Noise Figure : SSB and DSB
Noise Figure is defined as
signal to noise ratio at RF port Noise Figure=
signal to noise ratio at IF port (2-20) In a typical mixer, there are actually two input frequencies that will generate a given intermediate frequency. One is the desired RF signal, and the other is called the image signal. The existence of an image frequency complicates noise figure computations because noise originating in both the desired and image frequencies therefore becomes IF noise, yet there is generally no desired signal at the image frequency. In the usual case where the desired signal exists at only one frequency, the noise figure that one measures is called the single-sideband noise figure (SSB NF); the rarer case, where both the “main” RF and image signals contain useful information, leads to a double-sideband (DSB) noise figure. Clearly, the SSB noise figure will be greater than for the DSB case, since both have the same IF noise but the former has signal power in only a single sideband. Hence, the SSB NF will normally be 3dB higher than the DSB NF.
2.3.3 Isolation
Isolation is one of another parameter in design mixer. It is generally desirable to minimize interaction among the RF, IF, and LO ports. For instance, since the LO signal power is generally quite large compared with that of the RF signal, any LO feedthrough to the IF output might cause problems at subsequent stages in the signal processing chain. This problem is exacerbated if the IF and LO frequencies are similar, so that filtering is ineffective. Even reverse isolation is important in many instances, since poor reverse isolation might permit the strong LO signal to work its way back to
the antenna, where it can radiate and cause interference to other receivers.
2.3.4 Single Balanced and Double Balanced Gilbert Mixer
Fig. 2.4 Single Balanced and Double Balanced Gilbert Mixer
The circuit of single balanced and double balanced Gilbert mixer is shown in Fig. 2.4.
The lower stage is operated as a transconductance amplifier which converts RF voltage into a current and then performs a multiplication in the current domain. VLO
is chosen large enough so that the transistors alternately switch all of the tail current from one side to the other at the LO frequency. In single balanced mixers, the output consists of sum and difference components, each the result of an odd harmonic of the LO mixing with the RF signal. It makes the linearity worse. So the double balanced mixers exploit symmetry to remove the undesired output LO component through cancellation. Double balanced mixers make the linearity better than single balanced mixers.
In the low voltage applications, the DC current source can be replaced by a parallel LC tank to create a zero headroom AC current source. The resonant frequency of the tank should be chosen to provide rejection of whatever common mode component is most objectionable. If several such components exist, one may use series combinations of parallel LC tanks.
2.4 Direct Conversion Receivers
Fig. 2.5 is a simple direct conversion receiver, where the LO frequency is equal to the input carrier frequency. The circuit of Fig. 2.5(a) operates properly only with double-sideband AM signals because it overlaps positive and negative parts of the input spectrum. For frequency- and phase-modulated signals, the down-conversion must provide quadrature outputs [Fig. 2.5(b)] so as to avoid loss of information. This is because the two sides of FM or QPSK spectra carry different information and must be separated into quadrature phases in translation to zero frequency.
The simplicity of the direct conversion architecture offers two important advantages. First, the problem of image is circumvented because ωIF=0. As a result, no image filter is required, and the LNA need not drive a 50Ω load. Second, the IF SAW filter and subsequent down-conversion stages are replaced with low pass filters and base-band amplifiers that are amenable to monolithic integration. But this architecture has a DC offset problem. We will discuss this issue in detail below.
DC Offsets problem:
Since in a direct conversion topology the downconverted band extends to zero frequency, extraneous offset voltages can corrupt the signal and, more importantly, saturate the following stages. To understand the origin and impact of offsets, consider the receiver shown in Fig. 2.6, where the LPF is followed by an amplifier and an A/D
converter. Let us make two observations. First, the isolation between the LO port and the inputs of the mixer and LNA is not infinite; that is, a finite amount of feedthrough
converter. Let us make two observations. First, the isolation between the LO port and the inputs of the mixer and LNA is not infinite; that is, a finite amount of feedthrough