ELEC 4505

Telecommunications Circuits
Preselect

Image
Filter

Amplifier
LNA

Mixer

Intermediate
Frequency

Backend

Speaker

Frequency
Synthesizers

Antenna
Switch
LPF

BPF

Power
Amplifier

Up converter
(Mixer)

Baseband
Stuff

2016 Course Notes, Labs

Plus: Handouts, Exams, Etc
2016 Calvin Plett

Source

Course Outline, Telecommunication Circuits, ELEC 4505, Fall 2016

Instructor: Prof. Calvin Plett, Office: 5146 ME
email: cp{at}doe.carleton.ca, web: www.doe.carleton.ca/~cp/telecom/telecom.html
Lectures: Mon, Wed 16:00-17:30, 4342 ME, (Term is Sept 7 Dec 9, 2015, break Oct 24-28)
Labs:

Wed 8:30-11:30, Thu 11:30-17:30, Odd Weeks room 4135 ME, (or computer lab for Lab 1)

Marks:

Lab 35%, Assignments 13%, Quizzes 4%, Final 50% (must get at least 35/100 in final exam)
Academic accommodation for any reason must be sought as soon as possible, preferably
early in the term. Verification will be required.

Course Description: Design of circuits making up the blocks in a communication system. Examples of
such blocks are tuned amplifiers, mixers, oscillators, phase shifters and detectors. Communications systems considered are wireless transceivers, AM, FM and TV. Use of the PLL will be discussed.
Course Outline:
Page
1. Introduction to Telecommunications
1
Components of a radio systems; noise, distortion, impedance matching.
2. Amplifier Design
17
Tuned amplifiers, class C amplifiers, extension to frequency multipliers.
3. Mixers and Modulators
39
4. Phase-Locked Loop and Applications
47
Introduction and applications such as: tracking filters, synthesizers, and FM demodulation.
5. Oscillators (For 2012, done after Amplifiers, before Mixers)
72
6. Amplitude-Modulated Radio
83
7. Frequency Modulators and Demodulators
89
8. Television Systems
102
Transmission and reception of video and audio;
Labs:
Group size is 2 for all labs, one write-up per group, labs are submitted electronically on cuLearn and are
due one week after the scheduled lab day. Lab 3 has two parts, done in weeks 7 and 9, the write-op
combines both parts and is due in week 10. (Note that unlike labs, assignments are done individually.)
1. Tuned Amplifiers (Simulation Lab) (and Assignment 1) Week 1 Sept. 14, 15, Week 3, Sept 28,
29)
115
Design and simulation (in ADS) of a 1.2 MHz tuned amplifier, built with a bipolar transistor and
passive components. You will learn about the use of transistor parameters, tuned circuits, noise figure and impedance matching.
2. Mixers and Modulators (Hardware Lab) (Week 5, Oct. 12, 13)
127
Use of an analog multiplier on an IC to build frequency changers. DSBSC, filtering for single sideband.
3 Phase-Locked Loops (Hardware Lab) (Week 7, Nov. 2, 3 and Week 9, Nov. 16, 17) 144
Use of a commercially available package to build a tracking filter, a synthesizer and an FM demodulator. The IC contains a voltage-controlled oscillator and phase detectors. In this lab, the VCO and
phase detectors will be characterized, then a complete phased-locked loop will be built. The main
external components will consist of a simple active loop filter and a divider to realize the synthesizer.
Assignments: Assignment 1 is part of Lab 1, Assignment 2 is on PLLs and is related to Lab 3 and done
in parallel with it, Assignment 3 in Week 11 is an oscillator simulation. Note that unlike labs, assignments are done individually.

Course Notes, Lab Notes, Assignments, extra info (past exams, lab hints, marking schemes): Will
all be available on a password protected course web page
Course Goals
The objective of this course is to learn about the design of communications circuits with topics as described in the above outline of the Lectures, Labs, and Assignments.
Student Learning Objectives
Upon successful completion of this course, students will be able to:
1. Understand the need in a communications system for amplification, filtering, mixing, signal generation, modulation, detection, and data conversion.
2. Understand how performance of telecommunications circuits is limited by parasitic capacitance, unwanted resistance, nonlinearity, and noise, and how to apply design principles to deal with such nonidealities.
3. Design tuned amplifiers and apply them in a communications system.
4. Design mixers and apply them in a communications system.
5. Design oscillators and apply them in a communications system.
6. Design PLL-based circuits such as a tracking filters, FM demodulators and frequency synthesizers
References:
Smith, Modern Communication Circuits 2nd Ed., McGraw-Hill 1998, TK6553.S5595
Krauss, Bostonian, Raab, Solid State Radio Engineering, Wiley 1980, TK6553.K73
Plett, Rogers, Radio Frequency Integrated Circuit Design, 2nd Ed. Artech House, 2010, TK7874
Sedra, Smith, Microelectronic Circuits, (for introduction to tuned amplifiers, oscillators)
Any introductory communications texts
Various Data books, Data Sheets

Communications Circuits - 1 of 16

Introduction to Communications Circuits

The objective is to transmit a signal from source (input) to destination (output). A transducer is
used to convert from one form of energy to another, e.g., a microphone converts sound waves
(sound pressure levels) in the air to electrical energy in wires or electrical circuits. To convert
between electrical signals and sound pressure levels, one uses a speaker or headphones. Another
example of a transducer is an antenna which converts between electromagnetic waves in the air
and electrical signals. Unlike the microphone, this can be used to convert in either direction, that
is between electrical and electromagnetic or from electromagnetic to electrical. An example communications system is shown below.
Input
Transducer

Output
Signal
Signal
Transmission
Transducer
Processing
Processing

Input
(e.g., voice)

Output
(e.g. Audio)
Transmitter (TX)

Receiver (RX)

This example is shown with an audio source (for example voice) feeding into a microphone, while
at the output the audio output is provided by a speaker. Transmission is done through the air and
the antenna behaves as a transducer converting between an electrical signal and the electromagnetic signal which goes through the air. Between the input and output transducers and the transmission transducers is the signal processing circuitry. In this course, we will spend most of our
time looking at examples of this signal processing circuitry.
Note that other types of inputs and outputs, transmission methods and transducers are possible in
communications systems. Table 1 below shows some of the more common types.
Table 1: Communications Systems: Inputs, Outputs, Transducers, Transmission
Application

Input

Transducer

Transmission
(transducers)

Output

Transducer

Telephone

voice

microphone

twisted pair

voice

speaker

Cell Phone
Radio Station

voice,
audio

cd player,
microphone

air (antennas)

audio

speaker
headphones

TV Station

video

camera

air (antennas), cable

video

TV Screen

Telemetry
(e.g., F1 race)

pressure,
temperature,

sensors

air (antennas)

readout,
numbers

video display,
files

keyboard,
storage

fiber optics (laser, detectors)

twisted pair or cable
(modem), air (antennas)

video,
audio
numbers

video display,
speaker,

data (e.g., inter- keystrokes,

net)
files

Such transmission should be accomplished with minimal degradation and noise added. Another
objective is to minimize power dissipation which is important for long battery life for hand-held
General

Communications Circuits - 2 of 16
personal communications devices. Other objectives might be low weight, small size, extra features, and low cost.
Modulation is used very often for transmission through the air, or for use in modems. In fact,
transmission of a plain telephone signal from the telephone to the switching office is one of the
few remaining examples where modulation is not necessarily used.
Higher Level (covered in other courses)
At the higher level, there are a lot of details we will not get into in this course. For example, there
is signaling, call setup and control, identification, billing, testing, tracking, coding and error correction, that goes on in many communications systems for example, in a landline phone, or a cellular phone. Some of this is handled by DSP in the backend of the communications transceiver,
but a lot of it is done by, or controlled by, the central office for landlines or the base station in cellular communications.
As an example, when a landline phone is picked up, the user hears a dial tone, the number is
entered, the user hears the phone ring, and the other party picks up the phone and says Oui
hello?. This seems straightforward, but actually involves a complex set of signals and behindthe-scene actions, some of which will now be briefly described.
Taking the phone off hook closes a switch that starts current flowing. The central office detects
that the phone is off hook, and connects a dial tone generator to the phone, so the user hears a dial
tone, indicating the system is ready.
Entering a number on a touch tone system sends a number of tones back to the central office.
These tones are converted back into numbers which are used to route the call, starting with country code, area code, then to the particular central office which is connected to the destination
phone. Communication between central offices is generally all digital and at much higher data
rates, so there is a whole host of conversions that take place as the signal is digitized and interleaved (multiplexed) with other messages being sent, converted to higher frequencies, possibly
converted to be sent over different media (for example, fiber, microwave link, etc).
In any case, with very little delay, the central office on the far side (or the base station if the destination is a cell phone) checks the status of the phone on the far side. If it is available, it sends a
ringing signal. At the same time, a ringing signal is sent back to the phone on the near side so the
original person knows the phone on the far side is being called. If the far-side phone is being used,
a busy signal is sent back to the first phone.
When the far side phone is picked up in response to the ringing, the far-side central office detects
the off-hook condition and stops sending a ringing signal. Instead the Oui Hello? voice signal in
the central office is digitized, coding and signalling information is added and the bits are interleaved (multiplexed) along with potentially multiple other conversations and sent along high
speed links back to their various destinations. At the near side, all conversations intended for the
near side are decoded, demultiplexed, and turned back into an analog electrical signal and connected to the correct line so the originating person hears the response. At the same time, the central office starts timing the call for billing information. And all of this takes a few tenths of a
second.
Note that in a plain landline phone, conversions between the analog voice signal and digital words

General

Communications Circuits - 3 of 16
usually takes place in the central office, while in the modern cell phone such conversions are done
in the cell phone. The digital bits are used to modulate the cell phone carrier frequency. In a cell
phone, there may be further signalling, for example, once call set-up is complete, a particular
channel will be assigned, to allow multiple cell phones in the same area to be in use simultaneously yet independently. Other interventions by the base station is for hand-offs as a user
crosses cell boundaries, or there may be the need for power control to keep the signal levels constant as the user moves with respect to the base station. In any case, it should be apparent that
there is a lot that goes on in communications systems besides simply designing amplifiers, mixers, oscillators, etc.
Why modulation?
1. Higher frequency is important since transmission efficiency is best if the antenna size is of the
order of a quarter wavelength.
e.g. voice 200 Hz - 4 kHz, at 4 kHz, wavelength = 75 km
CHEZ 106 (at 106.1 MHz): = 2.83 m
2 GHz = 15 cm
2. -each conversation can use different carrier frequencies. This is known as multiplexing and is
why there can be multiple radio stations or telephone conversations, TV signals etc. without
interfering with each other.
Modulation can be analog or digital.
Analog
Modulation can be of amplitude (AM) or frequency (FM). FM is less prone to noise because it is
known to have constant amplitude. The first example below show AM where the variation of
amplitude represents the modulating signal. The second example shows FM where the frequency
is varying according to the modulating signal.
Digital
This can be Frequency, Phase, or Amplitude Shift Keying (FSK, PSK, or ASK), or some combination, e.g., Quadrature Amplitude Modulation (QAM) discussed much later. The third and fourth
example below show FSK and PSK being used to represent a bit stream. Various filtering schemes
can be used to smooth sudden phase or amplitude shifts.
AM
FM
FSK
PSK

1 0

0 1

0 1 1

Digital advantages includes less problems with attenuation and noise, since all that is important is
that you can tell if a 1 or a 0 has been sent. Noise will affect bit error rate BER. A bit error occurs
when a 1 is seen as a zero or a zero as a 1. Depending on application and amount of error correcting, typical bit error rate might be 1 error out of a hundred to one error out of a million.
Analog - good because of high-frequency capability, simplicity.
Digital often requires extra circuitry such as clock and/or carrier recovery, A/D, D/A convertors,
DSP blocks.
General

Communications Circuits - 4 of 16
In transmission, there can be many causes of signal impairment, for example, for transmission
through air, there can be problems with attenuation, distortion, multipath, blocking, noise added...
multipath
Tx
Rx
The above shows an example of multipath where signals take different paths between the source
and destination. Often, this is because of reflections off other objects (e.g., buildings if outdoors,
or walls if indoors). These signals will arrive at different times, and if they arrive out of phase will
result in destructive interference causing the received signal to disappear. This is called fading.

Block Diagram of Communications Transceiver

An example communications transceiver is shown below. Many aspects of this receiver are common to all receivers.
Amplifier
Preselect
LNA

Image
Filter
BPF

BPF
Rx

Audio
Intermediate (Limiter)
Amp
Mixer Frequency Detector Processing
RF
IF
LO

BPF
demod
Frequency
Synthesizers

Antenna
Switch

Tx

TLO

BPF

LPF

TRF
Power
Amplifier

speaker

A/D
I/Q
2nd IF

TIF

up converter processing
(Mixer)

Amp

Source

This transceiver has a transmit side (Tx) and a receive side (Rx). These are connected alternately
to the antenna, using the antenna switch. Note, in some systems where different transmit and
receive frequencies are used, band-pass filters could instead be used to separate Tx from Rx. In
the following discussion, first the receiver then the transmitter blocks will be discussed at a systems level. More detailed design of these blocks will be discussed in later parts of the course.
Input Preselection Filter
This filter may be required to prevent overloading of the LNA. For example, in a cable system, 60
or more channels may be present simultaneously, and filtering is required to limit the overall signal level.
LNA (Low Noise Amplifier)
Input signal needs to be amplified without adding much noise. Then at its output both signal and
noise are higher but the signal-to-noise ratio is nearly the same i.e. SNR OUT SNR In . As a result,
noise is of less importance in further stages.
General

Communications Circuits - 5 of 16
Mixer
The following multiplier symbol represents the mixer.
v1

vo
v2

multiplier action is:

v o = kv 1 v 2
where k is some constant.
if v 1 = A 1 cos 1 t , and v 2 = A 2 cos 2 t
kA 1 A 2
then: v o = --------------- cos 1 + 2 t + cos 1 2 t
2
sum

difference

Here, let 1 be the RF signal and 2 be the LO signal. As shown in the example below, with the
RF frequency at 900 MHz and the IF at 100 MHz, the LO could be at 1000 MHz if it is a highside LO, or it could be at 800 MHz if it is a low-side LO. Note also that there is a sum frequency
formed at f RF + f LO which occurs at 1900 MHz or at 1700 MHz for the two cases shown below.
These are easily removed by the IF filter.
.
Low Side Oscillator

High Side Oscillator

1
2 1

100
IF

2
1 + 2

900 1000
RF LO

1900

1 2

100
IF

2 1
1 + 2

800 900
LO RF

1700

On what basis are high-side or low-side LO chosen? Low side LO being at a lower frequency may
be easier to design. However, given a certain range of frequencies, the high-side LO will have a
smaller fractional range of frequencies. For example, with the frequencies above, if the input RF
frequency range is from 850 to 950 MHz, then the LO will have to have a range from 950 to 1050
MHz for high-side, and 750-850 for low side. This is 5 % for high-side and 6.25 % for lowside LO.
There are many other (probably even more important) considerations in frequency planning. For
example, harmonics and intermodulation of various frequencies, images and other known signals
must be avoided. Some of these topics will be discussed later.

General

Communications Circuits - 6 of 16
The IF Filter
Typically, there will be a whole bunch of channels. It is the job of the local oscillator to mix the
desired channel (shown darker below) down to the IF. That is, the LO is tunable and is tuned to
f RF + f IF for a high side oscillator. Then it is the job of the IF filter to select only one channel, as
shown below. The channel separation shown here is about 200 kHz. A 200 kHz channel selection
filter is easier to design at a fixed IF of 100 MHz compared to making a tunable filter of the same
bandwidth at 900 MHz.

Channels mixed down to IF

IF

Channels at RF
LO

RF

IF Filter

100

900.2

1000.2

Image Reject Filter

In the above example, if there were any signal, or noise, present at 1100.2 MHz, it would also mix
with the 1000.2 MHz LO down to the 100 MHz IF. This can be prevented by placing an image
reject filter around the desired RF band, as shown below.
Channels mixed down to IF
IF

100

IF Filter

Channels at RF
RF

900.2

Image Reject Filter

LO

Image Region
Noise and
Other Channels

1000.2

1100.2

Note that wherever there is a mixer, there will be a desired frequency and an image frequency.
The image frequency and desired frequency will be separated by twice the intermediate frequency, in this case they are separated by 200 MHz. Thus, while mixing down to the IF has simplified the channel select filter, there is now the need for an image reject filter. Fortunately, a filter
which removes components more than 200 MHz from the carrier at 900 MHz is quite achievable,
although still not easy.
Detector
Following the down converter will be the detector which extracts the modulated signal. In some
cases, there will be a limiter preceding this stage (for example in FM, or in many of the digital
transceivers). For a digital system, the output will be digital bits.
Processing
May be as simple as filtering for an analog transceiver. In a digital system, the system may need
General

Communications Circuits - 7 of 16
digital filters, D/A converters, clock recovery circuits (carrier recovery may also be required). For
either or both systems, an automatic gain control system and a receive signal strength indicator
may also be included.
Audio Amplifier and Speaker
This amplifies the electrical signal and converts it to an audio signal.

Transmit Side
Source and Amplifier
These convert a low-frequency signal (if the source were a microphone, this would be an audio
signal) to an electrical signal and amplify it.
Processing
This might be filtering, A/D conversion (for a digital system) and modulation onto an appropriate
intermediate frequency carrier. As described before, modulation can be AM, FM, PSK, FSK, etc.
Up Converting Mixer
This mixes the transmit IF (or TIF) to the Transmit RF (or TRF) signal by using the Transmit LO
(or TLO). As discussed for the receive side down converter, this can be a high-side LO where
f RF = f LO f IF , or low side LO where f RF = f LO + f IF . For example, if the incoming frequency was at 70 MHz and the desired output RF was at 900 MHz, then the transmit LO could be
either at 830 MHz or 970 MHz. The diagram below shows the latter case.

Filter

TLO

TIF

70

900

970

1040

Filter
For the above example, the up converting action produces two sidebands. In the case of the highside LO, the desired output is at 900 MHz, but another band is created at 1040 MHz ( f LO + f IF ).
This is filtered out with a band pass filter centered at 900 MHz.
Note the frequency at 1040 MHz is not the image frequency. Recall, the image is the frequency
which would mix to the same output frequency. In this case the image is at 1870 MHz, since this
frequency mixed with the LO at 970 MHz will result in an output at 900 MHz.
Note: in some AM transmitters the upconverting mixer is directly used as the modulator. In this
case, both sidebands are desired.

General

Communications Circuits - 8 of 16
Power Amplifier
Typically, high efficiency is the most important feature of the power amplifier. To achieve this, it
may operate as class B or class C and thus may have a fair bit of harmonic distortion.
Filter After PA
This filter cleans up some of the additional frequency componenets generated due to distortion in
the power amplifier.
Antenna Switch
With the switch connected to the transmit side, there should be high isolation to the LNA, to protect the LNA from the large signal the power amplifier is producing. In this position, the switch
must be able to handle the power from the PA with low loss and minimal distortion. When the
switch is connected to the receive side, since the receive signal can be very small, there should be
very little insertion loss in the switch. As well, in this position, the PA should be disconnected to
prevent it from loading the input to the LNA.

Important Concerns: Noise and Distortion

Note that noise is what limits how small a signal can be detected. The limit to how large a signal
can be handled is determined by distortion.

Noise
Sources of noise:
External to circuit: noise from the input, power supply noise, noise from other circuits on the
same printed circuit board or on the same substrate for an integrated circuit.
inside of circuits: shot noise, thermal noise, flicker noise, etc.
2

Resistors: thermal noise 4kTR V /Hz due to thermal motion of conduction electrons
- where k is Boltzmanns constant, T is temperature in Kelvin, R is resistance.
- this is random (Gaussian amplitude spectrum, white frequency spectrum),
Bipolar transistors: the most important noise sources are thermal noise due to base resistance and
base and collector shot noise as shown below. In addition, all transistors have flicker noise.
Resistor
Model

vn =

4kTR

at R = 50 , 4kTR 8 10
Transistor Model
vn =

4kTr b rb

21

W Hz

19

V Hz

at T = 297 K, kT 4.15 10

and

4kTR 0.9 nV Hz

Collector Shot Noise where IB and IC are base and collector

bias currents and q is the charge on an
i nc = 2qI C

Base Shot Noise

19
electron, q 1.6 10
i nb = 2qI B
in class: show small-signal transistor model with noise, gm, r, re, principles of design.

General

Communications Circuits - 9 of 16
- Shot noise is due to the randomness of current (discrete events) as it crosses a junction. Like
resistive thermal noise, this is Gaussian in amplitude and white in frequency.
-Flicker noise has a 1/f frequency spectrum, that is, there is more noise at low frequency.
Flicker noise is significantly more important for MOS compared to bipolar.
Noise Factor is given by:

S N IN
S IN N OUT
F = ------------------------- = ------------ ------------ S N OUT
S OUT N IN

Where Signal S and Noise N have units of power. (Note, we will often use shorter forms, So, Si,
S OUT
No, Ni for output and input signal and noise.) But ------------ = G = Gain
S IN
N OUT
GN IN + N add
N add
F = ------------- = ------------------------------- = 1 + ------------GN IN
GN IN
GN IN
Thus, the minimum value of F is 1.
Noise Figure NF or F dB :

N o Tot
Total Output Noise
F = --------------- -------------------------------------------------------------------------------N o Src Output Noise due to Source

N o Tot
N OUT
F dB = 10 log ------------ = 10 log --------------GN IN
N o Src

Thus the minimum value of F dB is 0dB.

NIN usually taken as Noise Reference, the available noise from a 50 source.
R
vn =

vo =

kTR

4kTR
R

v
P = ----- = kT
R

Resistor Model
Here v n is a voltage spectral density with units of V Hz . If bandwidth is taken into account as
4kTRB , then voltage spectral density has been converted into rms voltage.
Now consider a cascade of components:
RS

N IN

G LNA
N LNA

G MIX
N MIX

G IF
N IF

N OUT
N IN G LNA G MIX G IF + N LNA G MIX G IF + N MIX G IF + N IF
F = ------------- = --------------------------------------------------------------------------------------------------------------------------------------------- (simplify in class)
N IN G LNA G MIX G IF
GN IN

General

Communications Circuits - 10 of 16

Distortion
Distortion occurs due to nonlinear transfer function e.g.
vo

i
saturation
vi
v
diode or vbe

amplifier

output, clipping

input

output, bias wrong

The result is harmonic distortion and intermodulation, as will be shown shortly.

Distortion: input at f1, output at f1, 2f1, 3f1, 4f1,..
intermodulation: input at f 1 and f 2 , for example at 7 and 8 MHz.
Mathematics of Nonlinearity
2

let the two tone test be:

v i = cos 1 t + cos 2 t = x 1 + x 2
2

v o = a 1 v i + a 2 v i + a 3 v i + a 4 v i + ....

model nonlinearity as:

second order

third order

v0 = a0 + a1 x1 + x2 + a2 x1 + x2 + a3 x1 + x2

HD2

MIX

2
2
2
x1 + x2 = x1 + 2 x1 x2 + x2 ;

HD2

3
2
2
3
3
x 1 + x 2 = x 1 + 3x 1 x 2 + 3x 1 x 2 + x 2

HD3

IM3

desired

then

IM3

HD3

General

Communications Circuits - 11 of 16

Table 2: Outputs From Nonlinear Circuits with Inputs at f 1 = 7 , f 2 = 8 MHz.

First Order

Second Order

symbolic
frequency

example
frequency

name

Comment

f1 , f2

7, 8

fundamental

Desired Output

2f 1 , 2f 2

14, 16

HD2 (Harmonics)

Can filter

f2 f1

1
15

IM2 (Mixing)
IM2 (Mixing)

Can filter all of these

3f 1 , 3f 2

21, 24

HD3 (Harmonic)

Can filter harmonics

2f 1 f 2

6
9

IM3 (Intermod)
IM3 (Intermod)

close to fundamental, hard

to filter 3rd order intermods

f2 + f1
Third Order

2f 2 f 1

These numbers are plotted in the following graph. It is apparent that harmonics can be filtered out
easily, while the third order intermodulation terms, being close to the desired tones, may be difficult to filter.
f1
f2 f1

f2

2f 1 f 2

f2 + f1

2f 2 f 1

10

2f 1

12

14

3f 1

3f 2

2f 2
16

18

20

22

24

Distortion Versus Power Level

Above, we have looked at frequency components. However, the relative levels of these components is dependent on power level. Power is measured as:
2

v rms
v peak
vp p
P = ---------- = ------------ = ----------- ,
R
2R
8R

P out
P dBm = 10 Log 10 ---------------
1 mW

1 mW = 0 dBm , 10 mW = 10 dBm , 100 mW = 20 dBm , 1 Watt = 30 dBm

commercial circuits often have compression points listed

1 dB compression: occurs where gain is down 1 dB compared to ideal gain

Third-Order Intercept Point IP3 is a measure of distortion. This is the power level at which the
extrapolation of the Pout vs. Pin curve and IM3 vs. Pin curve intercept. Note, that one would
General

Communications Circuits - 12 of 16
normally not operate at this power level, since the resulting distortion levels are very high.
However knowing the intercept point allows one to work backwards to find distortion at other
more useful power levels.
Pout
(dBm)

1 dB compression

OIP3 +20
+10

dB

/1

dB

op
e

-50
-60
-70
-80
P03
-90
-100
-110

sl

P01

l
enta
m
a
B
d
Fun
/1d
B
d
1
e=
slop
n
tio
a
ul
od
m
er
t
n
rI
e
rd
O
d
3r

0
-10
-20
-30
-40

-40

-30

-20

-10

Pi

+10

Pin (dBm)

IIP3

In class example: Relate coefficients a1, a2, a3 to IP3 and show a numerical calculation.

Impedance Matching
For optimum power transfer, prevention of ringing and radiation, good noise behaviour we need
the load matched to the source.

RS
50

N IN

50 Transmission
Line

Circuit
Would like 50 here

50 Transmission
Line

RL
50

If the circuit input impedance, Z IN is not the desired impedance, may need a matching circuit as

General

Communications Circuits - 13 of 16
follows:
RS
50

Matching
Circuit

50 Transmission
Line
50

Circuit

Z IN

An example of matching a transistor amplifier with a capacitive input is shown below. The series
inductance adds an impedance of jL to cancel the input capacitive impedance. Note that the
matching circuit is lossless.

Matching
+j30
50

Z in
50 j30

A more general matching circuit is required if the real part is not 50 . For example, most input
impedances can be matched by one of the two circuits shown on the next page. For a capacitive
input impedance, as with most transistor amplifiers, the first is used if the real part of Z in is less
than 50 , while the second circuit is used if the real part of Z in is bigger than 50 . Note that
the matching is still lossless (if L and C are ideal reactances).
Later (lab1) we will discuss the use of a transformer for impedance matching. Note: not all circuits are matched to 50 . For example, video circuits are often matched to 75 .
--------------------------------------------------------------------------------------------------------------------Next page gives systematic way to solve for series L, parallel C type of matching network.
Although it will not be done in this course, in practice, most experienced designers tend to use the
Smith Chart, or one of the widely available programs, as an aid to matching. However, it is good
to know how to use the techniques shown on the next page to understand the matching process,
and because it is the basis for the programs to do matching.
Note that after solving for L and C, sometimes upon simulation, it will turn out that the impedance
has not been matched well enough to 50 Ohms. In such a case, the component values can be
adjusted by iteration to bring the impedance arbitrarily close to 50 Ohms. The first step is to adjust
the component closest to the load first to bring the real part close to 50 ohms. Then the component
close to the source is adjusted to set the imaninary part of the impedance to 0.

General

Communications Circuits - 14 of 16
Impedance Matching Examples (Capacitive Zin. left: Real{Zin}<50, right: Real{Zin}>50)
Matching

Matching

C
Z3 = Z0
Y3 = Y0

C
Z 2 or Y 2

Z in or Y in

Z3 = Z0
Y3 = Y0

Z 2 or Y 2

Z in or Y in

Z 2 = Z in + jL (1),

Y 2 = Y in + jC (1),

1
1
Y 2 = ----- = ----------------------- = Y 0 jC (2),
Z2
Z in + jL

1
1
Z 2 = ----- = ------------------------ = Z 0 jL (2),
Y2
Y in + jC

since

since

1
Y 3 = Y 2 + sC = ----- = Y 0 (3),
Z0

Z 3 = Z 2 + sL = Z 0 (3),

1
40 jX
Y 2 = ----- = -------------------- = 0.02 jC since,
2
2
Z2
40 + X
Y 3 = Y 2 + jC = 0.02 Real part is

Y 2 = 0.01 + j0.005 + jC = 0.01 + jY

The two unknowns C and L can be solved

The two unknowns C and L can be solved for by solving the real and imaginary parts
for by solving the real and imaginary parts of (2).
of (2) (as equations or with numbers).
Example: Z in = 80 j40 , Z 0 = 50 .
Example: Z in = 40 j30 , Z 0 = 50 ,
We will find L and C by going through the
1
Y 0 = ----- = 0.02 . We will find L and C by
1
1
Z0
above equations. Y in = ------- = -------------------- ,
Z in
80 j40
going through the above equations.
multiply top and bottom by 80 + j40 , result:
Z 2 = 40 j30 + jL = 40 + jX
80 + j40
- = 0.01 + j0.005
Y in = ---------------------2
2
where X = L 30
80 + 40

40
2
-------------------- = 0.02 or X = 2000 40 = 20 ,
2
2
40 + X
thus L = 50 . Imaginary part
X
is: C = -------------------- , substitute X = 20 from
2
2
40 + X
20
previous part, C = ---------------------- = 0.01
2
2
40 + 20

where Y = 0.005 + C
0.01 jY
1
Z 2 = ----- = ------------------------ = 50 jL . Real:
2
2
Y2
0.01 + Y
0.01
0.01
2
-----------------------=
50
Y
=
--------
0.01
= 0.01
2
2
50
0.01 + Y
thus C = 0.005 . Imaginary part:
Y
L = ------------------------ , substitute Y = 0.01 from
2
2
0.01 + Y
0.01
- = 50
previous part, L = ------------------------------2
2
0.01 + 0.01
General

Communications Circuits - 15 of 16

Amplifiers
common emitter, simple
V CC
R1
RS

RC

CC

vo

CB

vS

RL
R2

RE

CE

V EE
:
RS
vS

v1

r v

vo

gm v

ro

RC

RL

Gain Estimate
vo
----- = g m R L R C , assumptions: R 1 and R 2 are large compared to R S and r , and that
vi
r
R s r , otherwise, need a factor of ----------------- in the gain equation.
RS + r
Low Frequency Poles
1
The coupling and bypass capacitors C E , C B and C C are selected so that ---------- R seen , where
jC
R seen is the equivalent resistance seen by the capacitor in question.
e.g., C B sees r Z C R 1 R 2 + R S ;

C C sees R C r o + R L ;

C E sees r e R E

Note at high frequency the impedance of C is important in the calculation of CB

High-Frequency poles:
1
-------------------------------------------------------------------------------------------- ;

R S R 1 R 2 r C + C gain

1
output: --------------------------------------- r o R C R L C

input:

Miller
General

Communications Circuits - 16 of 16
Cascode Amplifier
V CC
RC

CC

vo

Q2

vS

RS

RL

Q1

CB
RE

CE

V EE
Because of the cascode transistor, Q 2 , the effect of C is decreased allowing higher frequency.
As well, the output impedance is higher. However, the operating range is reduced because there
are now two transistors in series. Note in earlier days, ELEC 3509, 15 V power supplies have
been used, so the reduced range was not important. However, on an integrated circuit, the power
supply voltages are steadily decreasing year after year, and in the latest integrated circuit processes power supply voltages are now down to 1.2 V or even less.
Extension to Amplifier with Parallel RLC load (discussed again more fully in a few pages):
v
With resistor RC, gain is given by: ----o- = g m R C R L , with a more general load impedance of
vi
gm
v
v
1
ZC , gain is given by: ----o- = g m Z C , or noting that Y = ------ , ----o- = ------ .
Y
ZC
v
v
i

For a parallel resonant circuit, with resistor R, capacitor C, inductance L, admittance adds, so this
final form is an excellent way to express gain. The resulting gain is:
gm
gm
gm
gm
vo
v
------------------------------- = ----------------------------- . Let s = j : ----o- = ---------------------------------- = ---------------------------------------- .
----- = Y
1 1
1
j
1
1
vi
vi
C + YR + YL
sC + --- + -----jC + --- --------- + j C -------
R sL
R L

R
L
1
This is a bandpass filter with centre frequency 0 = ----------- . At resonance, the term in the bracket
LC
disappears and peak gain is g m R , same as before. The -3dB bandwidth of the filter can be shown
1
to be B = -------- .
RC

General