Microwaves - Lecture Notes - v.1.3.

Dr. Serkan Aksoy - 2009

Microwaves
Lecture Notes
Dr. Serkan Aksoy

v.1.3.4

2009

These lecture notes are heavily based on the book of Microwave Engineering by David M. Pozar. For future versions or any proposals, please contact with Dr. Serkan Aksoy (saksoy@gyte.edu.tr).

Microwaves - Lecture Notes - v.1.3.4

Dr. Serkan Aksoy 2009

Content
1. LUMPED CIRCUIT MODEL -----------------------------------------1 2. FIELD ANALYSIS of LINES -----------------------------------------1
2.1. Wave Propagation along the Line --------------------------------------------------------------------- 1
2.1.1. 2.1.2. Lossless Transmission Lines ------------------------------------------------------------------------------ 2 Lossy Transmission Lines --------------------------------------------------------------------------------- 3

2.2. Smith Chart-------------------------------------------------------------------------------------------------- 3 2.3. Slotted Line-------------------------------------------------------------------------------------------------- 4 2.4. Generator & Load Mismatches ------------------------------------------------------------------------- 4

3. MICROWAVE NETWORKS -----------------------------------------5

3.1. Voltage, Current and Impedance ---------------------------------------------------------------------- 5 3.2. Impedance & Admittance Matrices ------------------------------------------------------------------- 5
3.2.1. 3.2.2. Scattering Matrix -------------------------------------------------------------------------------------------- 5 Transmission (ABCD) Matrix ---------------------------------------------------------------------------- 6

3.3. Equivalent Circuits for 2 Port Networks ------------------------------------------------------------- 6 3.4. Signal Flow Graphs --------------------------------------------------------------------------------------- 6

4. IMPEDANCE MATCHING ------------------------------------------8

4.1. L Networks Matching------------------------------------------------------------------------------------- 8 4.2. Quarter Wave Transformer ------------------------------------------------------------------------------ 8 4.3. Single Stub Tuning---------------------------------------------------------------------------------------- 8 4.4. Double Stub Tuning -------------------------------------------------------------------------------------- 9 4.5. Tapered Lines ----------------------------------------------------------------------------------------------- 9

5. POWER DIVIDERS & COUPLERS ------------------------------ 10

5.1. Three Port Networks (T Junction) --------------------------------------------------------------------10
5.1.1. 5.2.1. 5.2.2. 5.2.3. 5.2.4. 5.2.5. 5.2.6. T Junction Power Divider ------------------------------------------------------------------------------- 10 Waveguide Directional Coupler ----------------------------------------------------------------------- 10 Wilkinson Power Divider ------------------------------------------------------------------------------- 11 Hybrid Coupler -------------------------------------------------------------------------------------------- 11 Hybrid ------------------------------------------------------------------------------------------------ 11 Coupled Line Directional Coupler -------------------------------------------------------------------- 11 Lange Coupler --------------------------------------------------------------------------------------------- 11

5.2. Four Port Networks (Directional Coupler) ---------------------------------------------------------10

5.3. Other Couplers --------------------------------------------------------------------------------------------11

6. NOISE & ACTIVE COMPONENTS ----------------------------- 12

6.1. Noise Figure, --------------------------------------------------------------------------------------------12 6.2. Dynamic Range & Intermodulation Distortion---------------------------------------------------13 6.3. RF Diode Characteristics --------------------------------------------------------------------------------14

7. MICROWAVE AMPLIFIER DESIGN --------------------------- 15

7.1. Two-Port Power Gains ----------------------------------------------------------------------------------15 7.2. Stability -----------------------------------------------------------------------------------------------------15 7.3. Single Stage Amplifier Design ------------------------------------------------------------------------16 7.4. Broadband Amplifier Design--------------------------------------------------------------------------16 7.5. Power Amplifiers -----------------------------------------------------------------------------------------16
7.5.1. Large Signal Characterization -------------------------------------------------------------------------- 16

These lecture notes are heavily based on the book of Microwave Engineering by David M. Pozar. For future versions or any proposals, please contact with Dr. Serkan Aksoy (saksoy@gyte.edu.tr).

Microwaves - Lecture Notes - v.1.3.4

Dr. Serkan Aksoy 2009

1. LUMPED CIRCUIT MODEL

Fields : Real size of the circuit, Circuit Transmission Line and will vary in magnitude and phase over its length. Circuits : Wavelength

2. FIELD ANALYSIS of LINES

The idea is to find relations between and with .

where , averaged values calculated for cross-sectional area.

and , , , length line and

are timeis the line

2.1. Wave Propagation along the Line

Using the frequency domain Telegrapher equation

: Resistance per unit length ( : Inductance per unit length ( : Capacitance per unit length ( : Conductance per unit length (

): Finite conductivity ): Self inductance of two wires where is propagation constant. ): Proximity conductors The solution of Telegrapher equation ): Dielectric losses

Kirchoff' Voltage Law

Then, taking the

derivation of

, the calculation of

Kirchoff' Current Law

Taking the limit as

, Telegraph equations Then

Time domain

Converting to time domain by using

Frequency domain

where

is the phase.

1 These lecture notes are heavily based on the book of Microwave Engineering by David M. Pozar. For future versions or any proposals, please contact with Dr. Serkan Aksoy (saksoy@gyte.edu.tr).

Microwaves - Lecture Notes - v.1.3.4 2.1.1. Lossless Transmission Lines

Dr. Serkan Aksoy 2009

The maximum value occurs when . The minimum value occurs when . When increases, increases as a measure of mismatch. Then Standing Wave Ratio ( ) is

When the lossless line is terminated by a load when means matched line. In that case at reflection coefficient and input impedance , the

Reflected waves occur.

Using the definition of , more useful form known as Transmission Line Impedance Equation as

Reflection coefficient at the load,

- Transmission Coefficient:Some part of EM wave is also transmitted to second region as

where and consists of a superposition of an incident and reflected waves called Standing Waves ( ).

Time Average Power Flow:

Insertion Loss:

Short Circuit: Open Circuit:

This shows is constant at anywhere on the line.

When the line is matched

is constant.

The proper length of open or short circuited transmission line Return can provide any desired reactance or susceptance.

When the line is mismatched ( Loss

(No regard to impedance is observed at the input. Transform) (Short circuit) (Open circuit)

): The same

Matched load (No reflected power, maximum power is delivered). Total reflection, (All power reflected).

(Quarter Wave Open circuit Short circuit

is not constant.
2 These lecture notes are heavily based on the book of Microwave Engineering by David M. Pozar. For future versions or any proposals, please contact with Dr. Serkan Aksoy (saksoy@gyte.edu.tr).

Microwaves - Lecture Notes - v.1.3.4 2.1.2. Lossy Transmission Lines

Power flow along lossy line: In practice, finite conductivity (or lossy dielectrics) lines can be evaluated as a Lossy Line. Power loss per unit length: Attenuation constant:

Dr. Serkan Aksoy 2009

Wheeler Incremental Inductance Rule:

Low-loss line ignoring the last term of , . Then

: Incremental inductance : Loss per unit length

: Power at the Then, Attenuation constant: .

In the lossy line; lossless line.

&

can be approximated to
where is skin depth and .

Distortionless Line: For the lossy line, in fact the exact is not a linear function of frequency means dispersive. But specifically if the following condition holds

Taylor series Inductance Rule:

where

then , mean that the lossy line 2.2. Smith Chart behaves as lossless (distortionless) line. Terminated Lossy Line: Loss is assumed small that where
is Resistance, is Reactance, is Conductance and is Susceptance. Whenever is normalized impedance

The apsis and ordinate of Smith chart are

and

Power lost in the line: Rearranging them

Perturbation Method for Calculating Attenuation

3 These lecture notes are heavily based on the book of Microwave Engineering by David M. Pozar. For future versions or any proposals, please contact with Dr. Serkan Aksoy (saksoy@gyte.edu.tr).

Microwaves - Lecture Notes - v.1.3.4

These are two families of circles as and . Superposition of Smith Chart and its 180o ( ) rotated version is known as Combined Impedance-Admittance Smith Chart.

Dr. Serkan Aksoy 2009

where and fixed and three cases are considered as

. Generally

is

Load Matched to Line: Then .

Generator Matched to Line:

Conjugate Latching:

is the complete revolution of Smith chart. is the half of Smith chart (180o). The images of is in Smith chart.

2.3. Slotted Line

This device is used to find - Measurement of - Calculate - Using and - Calculate , write at . as first . Maximum power transfer . If one directly chose , it does not mean that the best efficiency due to the phase differences. The efficiency can be improved only by making as small as possible.

on the

distance from the line.

2.4. Generator & Load Mismatches

Then, using this

where

and

.
4

These lecture notes are heavily based on the book of Microwave Engineering by David M. Pozar. For future versions or any proposals, please contact with Dr. Serkan Aksoy (saksoy@gyte.edu.tr).

Microwaves - Lecture Notes - v.1.3.4

Dr. Serkan Aksoy 2009

3. MICROWAVE NETWORKS

It can be shown that the real parts of , and are even in , but the imaginery parts of them are odd in . are the even function in . At low frequencies for electrically small circuits, lumped active and and passive circuit elements are enough for analyzing the circuit leading a type of a Quasi-Static solution (assumption of 3.2. Impedance & Admittance Matrices negligible phase change in any where of the circuit) of Maxwell's equations and to Kirchoff Current and Voltage Laws The and of a port microwave network having 'th terminal with impedance concept. Moreover fields are considered as TEM type. But this way is not possible to analyze microwave circuits. The circuit concept should modify and apply to microwave network theory developed in MIT in 1940. The The impedance matrix is in the form of reasons of using it are as follow

Much easier than field theory, Calculations are performed at terminals, not everywhere, Easy to modify and combine different problems, The field solution of Maxwell's equation gives more Similarly the admittance matrix is in the form of information at the every time and place of the network, but difficult. At microwave frequencies, although the definition of the terminal pairs for line is relatively easy, the terminal pair for line does not strictly exist.

3.1. Voltage, Current and Impedance

The measurements of and at microwave frequencies are difficult due to not easily defined terminals for non-TEM waves. Because of that the fields are measured and used as

Clear that

. It can be shown that

Then and are known as Input Impedance and Transfer than, the impedance can be defined as . Because the Impedance, respectively. fields depend on the coordinates (like in waveguide), special attenuation should be given for extraction of and . The way is If the network is reciprocal (no ferrites, plasmas and to do that active devices inside), and are the right

relations. If the network is lossless, imaginary. 3.2.1. Scattering Matrix

and

are purely

The form of the scattering matrix gives the port networks with the incident can be defined as complete description of the and reflected waves as . The impedance concept first used by O. Heaviside, and then after application to transmission lines, to electromagnetics by Schelkunoff. In this manner, types of it then, the impedance

Intrinsic Impedance, : It depends on only the material parameters. Wave Impedance, : TM, TE, TEM Any element of the scattering matrix types are present which may depend on type of guide, material and frequency. Characteristic Impedance, : For TEM, it is unique, but for TE and TM, not unique because and can not be determined, uniquely.
5 These lecture notes are heavily based on the book of Microwave Engineering by David M. Pozar. For future versions or any proposals, please contact with Dr. Serkan Aksoy (saksoy@gyte.edu.tr).

Microwaves - Lecture Notes - v.1.3.4

and are Reflection and Transmission Coefficient (from port to port ), respectively. Network analyzer is used to measure parameters of a network. If network is reciprocal, is symmetric. If network is lossless, is unitary means that . Using the relations as and with between and matrixes

Dr. Serkan Aksoy 2009

3.3. Equivalent Circuits for 2 Port Networks

Transition between a coaxial line and microstrip line can be chosen as an example of two port networks. Because of discontinuity in the transition region, EM energy is stored in the vicinity of the transition leading to reactive effects mean that the transition region should be modeled as black box (There is an unlimited way of equivalent circuits, but choosing matrix equivalence), then

In the , all the and are defined as a reference point at the end of every lines. If the reference point is shifted, then

where

is the electrical length of the outward shift.

3.2.1.1. Generalized Scattering Matrix In the previous chapter, is defined for networks with same characteristic impedance for all ports. Generally for not same impedance for all ports, a new set of wave amplitudes as , Then, generalized matrix

A nonreciprocal network can not be represented by passive equivalent circuit using reciprocal elements. If the network is reciprocal (there are six degrees of freedom, six independent parameters), the presentations lead naturally to and equivalent circuits as

- In reciprocal networks, is symmetric, - In lossless networks, is unitary and satisfies the equation

If the network is lossless, the impedance and admittance matrix elements are purely imaginary, the degrees of 3.2.2. Transmission (ABCD) Matrix freedom reduces to three and the elements of and Many microwave networks consisting cascade connection and equivalent circuits should be constructed from purely need building block fashion in practice. ABCD matrix is defined reactive elements.
to satisfy this as

3.4. Signal Flow Graphs

Signal flow graph is an additional technique to analyse microwave networks in terms of reflected and transmitted where port 1 and port 2 are completely isolated in the equation waves. Three different forms of it are given below with nodes means that cascade multiplication is possible. The current and branches. direction is also specially designed for ABCD matrix. The relation between and matrix parameters are as

If the network is reciprocal (

), then

6 These lecture notes are heavily based on the book of Microwave Engineering by David M. Pozar. For future versions or any proposals, please contact with Dr. Serkan Aksoy (saksoy@gyte.edu.tr).

Microwaves - Lecture Notes - v.1.3.4

Dr. Serkan Aksoy 2009

The decomposition rules are also given below

7 These lecture notes are heavily based on the book of Microwave Engineering by David M. Pozar. For future versions or any proposals, please contact with Dr. Serkan Aksoy (saksoy@gyte.edu.tr).

Microwaves - Lecture Notes - v.1.3.4

Dr. Serkan Aksoy 2009

4. IMPEDANCE MATCHING
Impedance matching (or tuning) is an important issue for - Maximum power is delivered when load is matched to line (assuming the generator is matched) - Power loss is minimized. ratio of receiver components is increased. - Amplitude and phase errors are reduced. Whenever has nonzero real part, impedance matching is possible with the factors such as - Complexity: Simpler one is prefable. - Wide Bandwidth: Match a load over a band. - Implementation: Easier one is prefable - Adjustability: Adjust to match a variable load impedance. For multiple reflection, total is

The condition of enough to make total reflection (multiple) is zero. If and , then and .

is also

4.1. L Networks Matching

For each frequency, has to be equal to length is possible only for one frequency. . Thus fixed line

Bandwidth performance of QWT for wide band matching: If one set the maximum value of reflection coefficient, The normalized should be converted to with adding that can be tolerated, then the fractional bandwidth is impedance, then adding the impedance matching will be successful.

4.2. Quarter Wave Transformer

This shows that as becomes closer to , the bandwidth increases. This result is valid only for TEM lines. The step changes of reactance effect can be compensated by making a small adjustment in the length of the matching section. Approximate behavior of reflection coefficients is shown at below.

It is used for only real load impedance. Complex load impedance can always be transformed to real impedance by appropriate length of transmission line. But this generally alters the frequency dependency of the equivalent load reducing the bandwidth of the matching. The following relation has to be The QWT can be extended as a multisection form for matching of broader bandwidth. satisfied as

4.3. Single Stub Tuning

where A single open-circuited (or short-circuited) transmission line is connected either in parallel or series with the feed line at a certain distance from the load. Single Sub Tuning can match any load impedance to line, but suffer from disadvantage of requiring a variable length between the load and stub.
8 These lecture notes are heavily based on the book of Microwave Engineering by David M. Pozar. For future versions or any proposals, please contact with Dr. Serkan Aksoy (saksoy@gyte.edu.tr).

Microwaves - Lecture Notes - v.1.3.4

Dr. Serkan Aksoy 2009

ripple) multisection matching transformers can be used. In first one: the derivatives of is settled to zero, in second one: is equated to Chebyshev polynomial.

4.5. Tapered Lines

The line can be continuously tapered for decreasing the effect of the step changes in characteristic impedance between the discrete sections. The incremental reflection coefficient

Since lumped elements are not required, single stub is convenient and easy to fabricate in microstrip form. Two adjustable parameters and susceptance or reactance provided , by using theory of by stub. Although for microstrip lines open circuit is easy to Since small reflections fabricate since a via-hole is enough, for coax or waveguide short circuit is preferred since open circuit line may be large for radiation. If the impedance has the form of at the distance . Then stub reactance can be chosen as , resulting for matching condition. For a given susceptance or reactance, the difference in lengths of open and short-circuited stub is . where from can be found. Chancing type of taper (Exponential, , Triangular, Klopfenstein), different 4.4. Double Stub Tuning band pass characteristic may be applied. Klopfenstein yields the The double stub tuner can not match all load impedances, but shortest matching section. The Bode-Fano criterion for certain type of canonical load impedances will help us to define load may be arbitrary distance from the first stub. theoretical limit on the minimum reflection with the upper limit of matching performance and provide a benchmark against which a practical design can be compared.

Distance between the stubs should be generally chosen as or to reduce the frequency sensitivity. Single Section Transformer: The reflection coefficient of single section transformer can be written when discontinuities between the impedances are small as

This shows that

is dominated by

and

Multisection Transformer: If the applications require more bandwidth, multisection transformers consists of equal length sections of the lines can be used. The reflection coefficient can be written as

The importance of this result is that the desired reflection coefficients response as a function of frequency can be synthesized by proper choosing of . To obtain passband responses, binominal (maximally flat) and Chebysev (equal
9 These lecture notes are heavily based on the book of Microwave Engineering by David M. Pozar. For future versions or any proposals, please contact with Dr. Serkan Aksoy (saksoy@gyte.edu.tr).

Microwaves - Lecture Notes - v.1.3.4

Dr. Serkan Aksoy 2009

5. POWER DIVIDERS & COUPLERS

between two output ports. The fringing fields and higher order modes at the discontinuity leading stored energy can be These are passive components used for power division or power accounted by a lumped susceptance, . The output line and can be selected to provide various power combining. In power division, an input signal is divided by impedances decision. coupler in two (or more) signals, equally or not. Resistive Divider: Possible to match all ports simultaneously, the resistive (lossy) divider is used, but no isolation between two output ports due to being not lossless. Half of the supplied power is dissipated in resistors.

5.1. Three Port Networks (T Junction)

It has two inputs with one output.

5.2. Four Port Networks (Directional Coupler)

It has two inputs and two outputs. After considering using the matrix for reciprocal, matched and lossless If component is passive (no anisotropic material), the network features of network, the possible solutions are means is reciprocal and when all ports are also matched Directional Coupler. Using different phase references, , considering lossless one Symmetrical or Anti-symmetrical Directional Coupler may be defined. The design parameters of directional coupler are

Using the features of

for lossless and reciprocal network

To satisfy, above equation at least two parameters have to be zero means that three port network can not be reciprocal, lossless and matched all ports. If the network is nonreciprocal with matching all port and satisfaction of energy conservation, such a device is known as Circulator relies on anisotropic materials.

The coupling factor shows the fraction of input power to the output. The directivity is a measure of isolation ability for forward and backward waves. The ideal coupler has infinite directivity and isolation and also lossless. The directional property of the all directional coupler is produced through the use of two separate waves or wave components, which add in phase at the coupled port, and cancel in phase at the isolated port.

5.2.1. Waveguide Directional Coupler Bethe Hole Coupler: Couple one waveguide to another through a single small hole in the common wall. Types of the parallel guides and skewed guides work properly only at the design if only two ports of the network are matched, a lossless and frequency (narrow bandwidth in reciprocal network can be physically realizable. terms of its directivity).
If the network is being lossy, network can be reciprocal and matched at all ports (Resistive Divider or Isolator). As an example, Ferrite Isolators are two-port device having unidirectional transmission characteristics. Because is not unitary, the isolator must be lossy. The isolators can be used between a high-power source and load to prevent possible reflections from damaging the source by absorbing reflected power.

Multi Hole Coupler: Series of coupling holes are used to increase bandwidth as similar design to multi section transformer. Making coupling coefficients proportional to binominal coefficients, maximally flat response can be obtained. Using Chebysev polynomial, different responses are possible.

5.1.1. T Junction Power Divider

This can be used for power division (or combining). Lossless Divider: This suffers from the problem of not being matched at all ports and in addition does not have any isolation
10 These lecture notes are heavily based on the book of Microwave Engineering by David M. Pozar. For future versions or any proposals, please contact with Dr. Serkan Aksoy (saksoy@gyte.edu.tr).

Microwaves - Lecture Notes - v.1.3.4 5.2.2. Wilkinson Power Divider

Dr. Serkan Aksoy 2009

Coupled lines of two (or more) transmission lines are closed together, power can be coupled between the lines. Generally TEM mode is assumed rigorously valid for striplines, but approximately valid for microstrips. Coupled Line Theory is based on types of excitations as even mode (strip currents are equal in amplitude with same directions) and odd mode (strip currents are equal in amplitude with opposite directions). Arbitrary excitation can be treated as a superposition of appropriate even and odd modes amplitudes. Moreover design It is a network with the useful property of being lossless when graphs are present for coupled lines. the output ports are matched, that is, only reflected power is Design Considerations: dissipated. It is known that a lossy three port network can be made having all ports are matched with isolation between the - Although a single section coupled line has limited bandwidth requirement, the bandwidth can be increased using output ports. Wilkinson Power Divider can be made in due to multiple sections coupled line having close relations to microstrip or stripline form with arbitrary power division of way Divider or Combiner. The even-odd mode technique is used multisection QWT. for analysis. - The assumption of the same velocity of propagation for even and odd modes in design, generally not satisfied for a coupled 5.2.3. Hybrid Coupler microstrip or non TEM lines. This gives poor directivity. By using more effective dielectric constant (smaller phase velocity) It has having types of the following. for even mode, phase differences should be minimized. This also produces problems as the mismatching phase velocities for 5.2.3.1. Quadrature Hybrid ( Hybrid) multisection case and degrades coupler directivity. Increasing This is a directional coupler (knows as Branch Line bandwidth can be obtained with low coupling limits. Hybrid) with a phase difference in outputs (2 3). Evenodd mode technique can be applied for analysis. matrix has a 5.2.6. Lange Coupler high degree of symmetry To increase coupling factor, means any port can be used Lange Coupler (several for input as given below lines) with phase difference between outputs is used as a 3 dB coupling ratio in an octave or more bandwidth can be achieved. The main disadvantage of it (a type of quadrature hybrid) is difficult to 5.2.4. Hybrid fabricate due to very It is a four port network with a phase shift (2 3) between narrow lines. Folded Lange two outputs (also may be in phase). It can be used as a combiner coupler is also used for more easily analysis to model equivalent and has unitary symmetric scattering matrix as circuit.

5.3. Other Couplers

Moreno Crossed Guide Coupler Schwinger Reversed Phase Coupler Riblet Short Slot Coupler Symmetric Tapered Coupled Line Coupler Coupler with Apertures in Planar Lines As an example of a device uses a directional coupler is Reflectometer isolate and sample the incident and reflected powers from a mismatch load as a heart of a scalar (or vectorial) network analyzer.

It may be produces as the form of ring hybrid (rate race), tapered matching lines and hybrid waveguide junction (Magic T, (Rate Race)) in which symmetrically (or antisymmetrical) placed tuning ports (or irises) can be used for matching.

5.2.5. Coupled Line Directional Coupler

11 These lecture notes are heavily based on the book of Microwave Engineering by David M. Pozar. For future versions or any proposals, please contact with Dr. Serkan Aksoy (saksoy@gyte.edu.tr).

Microwaves - Lecture Notes - v.1.3.4

Dr. Serkan Aksoy 2009

6. NOISE & ACTIVE COMPONENTS

where is output noise power and is amplifier gain. Excess Noise Ratio (ENR) is also used to characterize Noise Power of Noise is usually generated by random motions of charges (or active noise generator consisting of a diode or a tube as charge carriers in devices and materials). Such motions can be caused by the mechanism of Thermal Noise: Thermal vibrations of bound charges. Shot Noise: Random fluctuations of charge carriers. Flicker Noise: noise. Plasma Noise: Random motions of charges. Quantum Noise: Quantized nature of charge carriers. Noise is a random process and can be passed into a system from external sources or generated within the system itself. Noise level defining the system performance determines for minimum signal reliability detected by a receiver. where & generator. are Noise Power & Equivalent Temperature of

Measurement of Noise Power:

factor method is applied as

where

should be determined via power measurement. Then

Dynamic Range and Compression Point: The linearity and deterministic features of all components can be satisfied in a range called Dynamic Range. The floor level of noise dominates the output power at very low frequencies. Compression where & are temperature of hot & cold load, respectively. Point is defined as the input power for which the output is below that of an ideal amplifier. 6.1. Noise Figure, Noise Power and Equivalent Noise Temperature: Rayleigh- It is a measure of the degration in Jeans approximation results Voltage Fluctuations as and output as ratio between the input

where is Boltzmanns constant, is temperature, is bandwidth and is resistance. Because of frequency is defined for a matched input source and for a noise source independency, this is known as White Noise Source can be treated as Gaussian distributed variables. consist of a resistor temperature . A noisy resistor can be replaced with a noiseless resistor and a voltage source of RMS . Then connecting a load resistor Noise Figure of a Noisy Network: Having the parameters of results in maximum power transfer called Noise Power as with input noise and signal power , the output noise power and the output noise signal . Then Noise Figure is then then then : Cooler device, less noise power. : Smaller bandwidth, less noise power. : Use the exact definition of for . If the network is noiseless , .

If is not strong function of frequency (White Noise), an Equivalent Noise Temperature is defined as

Noise Figure of a Two-Port Passive and Lossy Network: Having such as attenuator (or lossy line) with a matched source resistor at , overall system temperature also at , noise factor

where is Noise Power delivered to load . A noisy amplifier with a source of resistor at a temperature of can be replaced with a noiseless amplifier and a resistor having Equivalent Noise Temperature as where temperature is lossy factor. The equivalent noise

12 These lecture notes are heavily based on the book of Microwave Engineering by David M. Pozar. For future versions or any proposals, please contact with Dr. Serkan Aksoy (saksoy@gyte.edu.tr).

Microwaves - Lecture Notes - v.1.3.4

Dr. Serkan Aksoy 2009

matched to its input and output circuits. Consider that line is mismatched to input circuit as

If the line is at the temperature

Noise Figure of Cascaded System: A cascaded system of two components ( ), ( ) and Then The output of first stage : The output of second stage:

When the line is matched the matched lossy line.

same as

6.2. Dynamic Range & Intermodulation Distortion

All realistic devices are nonlinear at very low power levels due to noise effects and also practical components became nonlinear at high power levels. A given component (or network) can operate as desired between minimum and maximum realistic power ranges known as Dynamic Range. In the sense of and . The first stage is intermodulation distortion, it is called as Spurious Free reduces to second stage. Generally Dynamic Range. The output response of a nonlinear device (diode, transistor) by using a Taylor series expansion

where dominant due to

where is DC output, is Lineer output, others are Nonlinear outputs. Different output response can be obtained relating to coefficients as Only Noise Figure of a Passive Two Port Network: Available Power Only Gain does not depend on . Only , Rectifier , Attenuator or Amplifier , Mixing or Frequency Conversion

Gain Compression:

where as

, then

In most practical system, , then gain tends to decrease is named Gain Compression or Saturation.

Intermodulation Distortion:

the output is the combination of two input frequencies are called Intermodulation Products. The definitions of and network is matched are close as lossy lines. When the , then Passive Intermodulation: Connectors, cables, antennas, and every metal-metal contact can cause passive intermodulation due to poor mechanical contact, oxidation, contamination etc., Noise Figure of a Mismatched Lossy Line: Previously, Noise and also thermal effects of high power source. This has Figure of a lossy line is calculated under assumptions of line is generally lower power levels.
13 These lecture notes are heavily based on the book of Microwave Engineering by David M. Pozar. For future versions or any proposals, please contact with Dr. Serkan Aksoy (saksoy@gyte.edu.tr).

Microwaves - Lecture Notes - v.1.3.4

Dr. Serkan Aksoy 2009

6.3. RF Diode Characteristics

Shottky Barrier Diode Detectors: This is a nonlinear device consisting of semiconductor-metal junction resulting lower junction capacitance can be used frequency conversion (rectification, detection, mixing). It has a

with a Small Signal Model

These diodes are used as rectifiers, detectors and demodulation of an AM modulated RF carrier. PIN Diode: This is used to construct an electronic switching for control circuits such as phase shifters and attenuators. These are preferable because of small size, high speed and inerrability with planar circuits. Especially single-pole PIN diode switches can be used in either a series or a shunt configuration to form a single pole RF switch. Insertion Loss of switches

where

is diode impedance as

Varactor Diode: Junction capacitance varies with bias voltage used for electronically frequency tuning. Impatt Diode: Similar to PIN diode, but based on avalanche effects exhibiting negative resistance over a broad frequency range, therefore used to directly convert DC to RF power. Gunn Diode: It exhibits a negative differential resistance based on Gunn effect and used to generate RF power to DC. Baritt Diode: Similar to junction transistor without a base contact and useful for detector and mixer applications with advantages of lower AM noise.

14 These lecture notes are heavily based on the book of Microwave Engineering by David M. Pozar. For future versions or any proposals, please contact with Dr. Serkan Aksoy (saksoy@gyte.edu.tr).

Microwaves - Lecture Notes - v.1.3.4

Dr. Serkan Aksoy 2009

More generally, most useful power definition is Transducer Power Gain account for both source and load mismatch

7. MICROWAVE AMPLIFIER DESIGN

7.1. Two-Port Power Gains

The gain and stability of a general two-port amplifier in terms of parameters of transistor will be investigated for amplifier and where oscillator design. Three types of power gain can be derived as , If transistor is unilateral,
; ,

, then

where is independent of . is defined with an assumption Similar relation can be obtained by Equivalent circuit parameter. that conjugate matching of both source and load depend on but not . depends on both and . Whenever input and 7.2. Stability output are both conjugately matched, gain is maximized and . There are necessary conditions for a transistor amplifier to be stable based on the possible oscillation for input and output The average power delivered to network impedance has a negative real part as a two-sub group: - Conditional Stability: If and , network is stable for a range of passive source and load impedance. The power delivered to load - Unconditional Stability: If and for all passive sources and loads, network is unconditionally stable. The stability condition is usually frequency dependent since matchings generally depend on frequency (stability may be possible for a frequency but not possible for others). Rigorous treatment of stability requires parameters of network have no poles in the right-half complex plane in addition to and . If device is unilateral , more simply results and are enough for stability. Stability Circles: Applying the above requirement for unconditional stability, following conditions have to be satisfied

Then, the power gain

where

If If

, then , then

. , Unilateral Transducer Power Gain,

These conditions define a range for and where amplifier will be stable. Finding this range by using Smith chart, plotting the input and output Stability Circles are defined as loci in the (or ) plane for which (or ), then define boundaries between stable and unstable regions. The equations for input and output stability conditions can be extracted as

15 These lecture notes are heavily based on the book of Microwave Engineering by David M. Pozar. For future versions or any proposals, please contact with Dr. Serkan Aksoy (saksoy@gyte.edu.tr).

Microwaves - Lecture Notes - v.1.3.4

Dr. Serkan Aksoy 2009

7.5. Power Amplifiers

This is used to increase power level with consideration of efficiency, gain, intermodulation and thermal effect. Amplifier Efficiency is defined as

with the effect of input power, Power Added Efficiency, PAE If device is unconditionally stable, the stability circles must be completely outside (or totally enclose) the Smith chart. This can be stated mathematically as for for If or , the amplifier can not be unconditionally stable because a source (or load impedance) leading to (or ) can cause (or ). If transistor is only conditionally stable, and must be chosen in stable regions. Specially unconditionally stability can be tested with the methods of Rollets Condition or Parameter. where is small signal (linear) power gain. Class A amplifiers with theoretically maximum efficiency are inherently linear that transistor is biased to conduct over entire range of input signal cycle (low-noise amplifier). Class B amplifiers with theoretically maximum efficiency are biased to conduct only during one-half of input signal cycle (Push-Pull amplifier). Class C amplifiers with efficiency near are operated with transistor near cut-off for more than half of the input signal 7.3. Single Stage Amplifier Design cycle (in a resonant circuit, constant envelope modulation). Maximum gain with stability can be realized when input and Higher classes such as D, E, F and S are also used with high output sections provide a conjugate match between source and efficiency. load impedance, but generally as a narrowband. To perform this where is power gain. PAE drops quickly with frequency. Another parameter is Compressed Gain defined as the gain of amplifier at 1 dB compression gain as

7.5.1. Large Signal Characterization

conditions simultaneously have to be satisfied means that also by maximizing the transducer gain, first of all , then should be solved by considering stability conditions. It is also preferable to design for less than the maximum obtainable gain, to improve bandwidth (or to obtain a specific amplifier gain). To do that, Constant Gain Circles on the Smith chart to represent loci of and that give fixed values of gain are used. Besides stability and gain, the Noise Figure of the amplifier should be minimized by using Constant Noise Figure Circles. If input power is small enough, parameter is independent from input power and linear small signal model is suitable for modeling. But for high input powers, transistor behaves as a nonlinear device (Large-Signal Characterization) and more difficult to design. The following methods are possible for Large Signal Characterization - Measure the output power as a function of source and load impedances and produce tables, then determine large signal source and load reflection coefficients to maximize power gain for a particular output power.

7.4. Broadband Amplifier Design

- Plot contours (Load-Pull Contours) of constant power output The bandwidth can be improved with designing for less than on a Smith chart as a function of load reflection coefficient with maximum gain will improve bandwidth, but the input and conjugately matching at input and design for a specified gain. output ports will be poorly matched. Common approaches to - Use nonlinear equivalent of the transistor circuit. solve this problem are listed below - Compensated Matching Network, - Resistive Matching Network, - Negative Feedback, - Balanced Amplifiers, - Distributed Amplifiers. Especially for designing Class A amplifiers, the stability can be checked by using small signal model because instabilities begin at low signal levels.

16 These lecture notes are heavily based on the book of Microwave Engineering by David M. Pozar. For future versions or any proposals, please contact with Dr. Serkan Aksoy (saksoy@gyte.edu.tr).