Documenti di Didattica
Documenti di Professioni
Documenti di Cultura
PHYS-328
Brigitte Vachon
<Brigitte.Vachon@McGill.ca>
Department of Physics, McGill University
December 7, 2015
2015 Vachon
Contents
1 A brief history of electronics
.
.
.
.
.
.
.
.
.
.
.
.
.
.
.
.
.
.
.
.
.
.
.
.
.
.
.
.
.
.
.
.
.
.
.
.
1
.
.
.
.
.
.
.
.
.
.
.
.
.
.
.
.
.
.
.
.
.
.
.
.
.
.
.
.
.
.
.
.
.
.
.
.
.
.
.
.
.
.
.
.
.
.
.
.
.
.
.
.
.
.
.
.
.
.
.
.
.
.
.
.
.
.
.
.
.
.
.
.
.
.
.
.
.
.
.
.
.
.
.
.
.
.
.
.
.
.
.
.
.
.
.
.
.
.
.
.
.
.
.
.
.
.
.
.
.
.
.
.
.
.
.
.
.
.
.
.
.
.
.
.
.
.
.
.
.
.
.
.
.
.
.
.
.
.
.
.
.
.
.
.
.
.
.
.
.
.
.
.
.
.
.
.
.
.
.
.
.
.
.
.
.
.
.
.
.
.
.
.
.
.
.
.
.
.
.
.
.
.
.
.
.
.
.
.
.
.
.
.
.
.
.
.
.
.
.
.
.
.
.
.
.
.
.
.
.
.
.
.
.
.
.
.
17
24
24
25
26
27
28
30
31
34
CONTENTS
3.2.1
3.2.2
3.2.3
CONTENTS
4 Transistors
4.1 Bipolar junction transistors (BJTs) . . . . . . . . . .
4.1.1 Theory . . . . . . . . . . . . . . . . . . . . .
4.1.2 Transistor biasing . . . . . . . . . . . . . . .
4.1.3 Intrinsic resistance . . . . . . . . . . . . . . .
4.1.4 Common-collector (emitter-follower) amplifier
4.1.5 Common-emitter amplifier . . . . . . . . . . .
4.1.6 Differential amplifier . . . . . . . . . . . . . .
4.2 Field effect transistors (FET) . . . . . . . . . . . . .
4.2.1 JFET . . . . . . . . . . . . . . . . . . . . . .
4.2.2 MOSFET . . . . . . . . . . . . . . . . . . . .
4.2.3 Theory . . . . . . . . . . . . . . . . . . . . .
4.2.4 Transconductance . . . . . . . . . . . . . . .
4.2.5 Current source circuit . . . . . . . . . . . . .
4.2.6 Transistor biasing . . . . . . . . . . . . . . .
4.2.7 Common-source amplifier . . . . . . . . . . .
4.2.8 Source-follower amplifier . . . . . . . . . . . .
.
.
.
.
.
.
.
.
.
.
.
.
.
.
.
.
.
.
.
.
.
.
.
.
.
.
.
.
.
.
.
.
.
.
.
.
.
.
.
.
.
.
.
.
.
.
.
.
.
.
.
.
.
.
.
.
.
.
.
.
.
.
.
.
.
.
.
.
.
.
.
.
.
.
.
.
.
.
.
.
.
.
.
.
.
.
.
.
.
.
.
.
.
.
.
.
.
.
.
.
.
.
.
.
.
.
.
.
.
.
.
.
.
.
.
.
.
.
.
.
.
.
.
.
.
.
.
.
.
.
.
.
.
.
.
.
.
.
.
.
.
.
.
.
.
.
.
.
.
.
.
.
.
.
.
.
.
.
.
.
.
.
.
.
.
.
.
.
.
.
.
.
.
.
.
.
.
.
.
.
.
.
.
.
.
.
.
.
.
.
.
.
.
.
.
.
.
.
.
.
.
.
.
.
.
.
.
.
.
.
.
.
.
.
.
.
.
.
.
.
.
.
.
.
.
.
.
.
.
.
.
.
.
.
.
.
.
.
.
.
.
.
.
.
.
.
.
.
.
.
.
.
.
.
.
.
.
.
.
.
.
.
.
.
.
.
.
.
.
.
.
.
39
42
45
50
54
57
61
66
74
75
76
77
82
84
87
88
91
5 Operational amplifiers
93
5.1 Characteristics of ideal op-amps . . . . . . . . . . . . . . . . . . . . . . . . . . . . . 96
Brigitte VACHON
December 7, 2015
CONTENTS
5.2 Comparator . . . . . . . . . . .
5.3 Feedback . . . . . . . . . . . . .
5.3.1 Effect of negative feedback
5.3.2 Effect of negative feedback
5.3.3 Effect of negative feedback
5.4 Op-amp Golden rules . . . . .
5.5 Buffer . . . . . . . . . . . . . . .
5.6 Inverting amplifier . . . . . . . .
5.7 Non-inverting amplifier . . . . .
5.8 Summing amplifier . . . . . . . .
5.9 Integrator . . . . . . . . . . . .
5.10 Current source . . . . . . . . . .
5.11 Characteristics of real op-amps .
5.12 Oscillators . . . . . . . . . . . .
. . . . . . . . . . . .
. . . . . . . . . . . .
on voltage gain . . .
on input impedance .
on output impedance
. . . . . . . . . . . .
. . . . . . . . . . . .
. . . . . . . . . . . .
. . . . . . . . . . . .
. . . . . . . . . . . .
. . . . . . . . . . . .
. . . . . . . . . . . .
. . . . . . . . . . . .
. . . . . . . . . . . .
6 Digital electronics
6.1 Number codes in digital electronics .
6.1.1 Binary . . . . . . . . . . . .
6.1.2 Hexadecimal . . . . . . . . .
6.1.3 Binary Coded Decimal (BCD)
6.1.4 Gray code . . . . . . . . . .
6.1.5 ASCII . . . . . . . . . . . . .
6.2 Parallel and Serial Bit Transmission .
6.3 Logic functions . . . . . . . . . . . .
Brigitte VACHON
CONTENTS
.
.
.
.
.
.
.
.
.
.
.
.
.
.
.
.
.
.
.
.
.
.
.
.
.
.
.
.
.
.
.
.
.
.
.
.
.
.
.
.
December 7, 2015
.
.
.
.
.
.
.
.
.
.
.
.
.
.
.
.
.
.
.
.
.
.
.
.
.
.
.
.
.
.
.
.
.
.
.
.
.
.
.
.
.
.
.
.
.
.
.
.
.
.
.
.
.
.
.
.
.
.
.
.
.
.
.
.
.
.
.
.
.
.
.
.
.
.
.
.
.
.
.
.
.
.
.
.
.
.
.
.
.
.
.
.
.
.
.
.
.
.
.
.
.
.
.
.
.
.
.
.
.
.
.
.
.
.
.
.
.
.
.
.
.
.
.
.
.
.
.
.
.
.
.
.
.
.
.
.
.
.
.
.
.
.
.
.
.
.
.
.
.
.
.
.
.
.
.
.
.
.
.
.
.
.
.
.
.
.
.
.
.
.
.
.
.
.
.
.
.
.
.
.
.
.
.
.
.
.
.
.
.
.
.
.
.
.
.
.
.
.
.
.
.
.
.
.
.
.
.
.
.
.
.
.
.
.
.
.
.
.
.
.
.
.
.
.
.
.
.
.
.
.
.
.
.
.
.
.
.
.
.
.
.
.
.
.
.
.
.
.
.
.
.
.
.
.
.
.
.
.
.
.
.
.
.
.
.
.
.
.
.
.
.
.
.
.
.
.
.
.
.
.
.
.
.
.
.
.
.
.
.
.
.
.
.
.
.
.
.
.
.
.
.
.
.
.
.
.
.
.
.
.
.
.
.
.
.
.
.
.
.
.
.
.
.
.
.
.
.
.
.
.
.
.
.
.
.
.
.
.
.
.
.
.
.
.
.
.
.
.
.
.
.
.
.
.
.
.
.
.
.
.
.
.
.
.
.
.
.
.
.
.
.
.
.
.
.
.
97
98
99
100
101
102
102
103
104
105
106
107
108
111
.
.
.
.
.
.
.
.
113
. 116
. 116
. 117
. 118
. 121
. 121
. 123
. 124
3
CONTENTS
CONTENTS
.
.
.
.
.
.
.
.
.
.
.
.
.
.
.
.
.
.
.
.
December 7, 2015
.
.
.
.
.
.
.
.
. .
. .
. .
. .
. .
. .
. .
. .
SR
. .
. .
. .
. .
. .
. .
. .
. .
. .
.
.
.
.
.
.
.
.
. . . . .
. . . . .
. . . . .
. . . . .
. . . . .
. . . . .
. . . . .
. . . . .
flip-flop)
. . . . .
. . . . .
. . . . .
. . . . .
. . . . .
. . . . .
. . . . .
. . . . .
. . . . .
.
.
.
.
.
.
.
.
.
.
.
.
.
.
.
.
.
.
.
.
.
.
.
.
.
.
.
.
.
.
.
.
.
.
.
.
.
.
.
.
.
.
.
.
.
.
.
.
.
.
.
.
.
.
.
.
.
.
.
.
.
.
.
.
.
.
.
.
.
.
.
.
.
.
.
.
.
.
.
.
.
.
.
.
.
.
.
.
.
.
.
.
.
.
.
.
.
.
.
.
.
.
.
.
.
.
.
.
.
.
.
.
.
.
.
.
.
.
.
.
.
.
.
.
.
.
.
.
.
.
.
.
.
.
.
.
.
.
.
.
.
.
.
.
.
.
.
.
.
.
.
.
.
.
.
.
.
.
.
.
.
.
.
.
.
.
.
.
.
.
.
.
.
.
.
.
.
.
.
.
.
.
.
.
.
.
.
.
.
.
.
.
.
.
.
.
.
.
.
.
.
.
.
.
.
.
.
.
.
.
.
.
.
.
.
.
.
.
.
.
.
.
.
.
.
.
.
.
.
.
.
.
.
.
.
.
.
.
.
.
.
.
.
.
.
.
.
.
.
.
.
.
.
.
.
.
.
.
.
.
.
.
.
.
.
.
.
.
.
.
.
.
.
.
.
.
126
126
130
131
137
138
145
146
148
149
149
153
156
157
159
163
165
166
.
.
.
.
169
. 170
. 171
. 172
. 175
4
CONTENTS
7.4
7.5
7.6
7.7
Brigitte VACHON
CONTENTS
.
.
.
.
.
.
.
.
.
.
.
.
.
.
.
.
.
.
.
.
.
.
.
.
.
.
.
.
.
.
.
.
.
.
.
.
.
.
.
.
.
.
.
.
.
.
.
.
.
.
.
.
.
.
.
.
.
.
.
.
.
.
.
.
.
.
.
191
. 192
. 196
. 197
. 198
. 199
. 200
. 201
microprocessor...
. . . . . . . . . . . . . . . . . . . . .
. . . . . . . . . . . . . . . . . . . . .
. . . . . . . . . . . . . . . . . . . . .
203
. 203
. 212
. 216
. . . . . .
. . . . . .
. . . . . .
. . . . . .
(EPROM)
. . . . . .
. . . . . .
ASIC, FPGA,
. . . . . . . . . .
. . . . . . . . . .
. . . . . . . . . .
December 7, 2015
.
.
.
.
.
.
.
.
.
.
.
.
.
.
.
.
.
.
.
.
.
.
.
.
.
.
.
.
.
.
.
.
.
.
.
.
.
.
.
.
.
.
.
.
.
.
.
.
.
.
.
.
.
.
.
.
.
.
.
.
.
.
.
.
.
.
.
.
.
.
.
.
.
.
.
.
.
.
.
.
.
.
.
.
.
.
.
.
.
.
.
.
.
.
.
.
.
.
.
.
.
.
.
.
.
.
.
.
.
.
.
.
.
.
.
.
.
.
.
.
.
.
.
.
.
.
.
.
.
.
.
.
.
.
.
.
.
.
.
.
.
.
.
.
.
.
.
.
.
.
.
.
.
.
.
.
.
.
.
.
.
.
.
.
.
.
.
.
.
.
.
.
.
.
.
.
.
.
.
.
.
.
.
.
.
.
.
.
.
.
.
.
.
.
.
.
.
.
.
.
.
.
.
.
.
.
.
.
.
.
.
.
.
.
.
.
.
.
.
.
.
.
.
.
.
.
.
.
.
.
.
.
.
.
.
.
.
.
.
.
.
.
.
.
.
.
.
.
.
.
.
.
.
.
.
.
.
.
.
.
.
.
.
.
.
.
.
.
.
.
.
.
.
.
.
.
.
.
.
.
.
.
.
.
.
.
.
.
.
.
.
.
.
.
.
.
.
.
.
.
.
.
.
.
.
.
176
177
179
180
180
182
183
185
187
188
CONTENTS
CONTENTS
A Grounding issues
218
A.1 Ground loops . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . 220
A.2 AC ground . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . 221
B Semiconductor technology
222
C Current in semiconductors
230
240
253
Brigitte VACHON
December 7, 2015
Chapter 1
A brief history of electronics
1870-1910: Cathode rays, X rays and the electron (Hittorf, Crookes, Braun[Nobel],
Roentgen[Nobel], JJ Thomson[Nobel], Millikan[Nobel])
1885-1910:Wireless telegraphy (Hertz, Branly, Lodge, Marconi[Nobel])
1904-1932: Vacuum Tubes (Edison, Fleming, Coolidge, Schottky, Langmuir[Nobel])
1900-1927: Radio (Fessenden, Armstrong, Hazeltine, Black)
1920-1930: Television (Baird, Farnsworth, Zworykin)
1920-1940: Radar (Watson-Watt, Loomis, Rabi[Nobel], Alvarez[Nobel], Purcell[Nobel], Dicke)
1900-1930: Electrons, waves and the birth of solid state physics (Boltzmann,
Planck[Nobel], de Broglie, Schrodinger, Fermi[Nobel], Dirac[Nobel], Bloch[Nobel], Peierls, Wilson)
1934-1948: Transistors (Southworth, Ohl, Lark-Horovitz, Shockley[Nobel], Brattain[Nobel],
Bardeen [2Nobel])
7
CHAPTER 1. HISTORY
Brigitte VACHON
December 7, 2015
CHAPTER 1. HISTORY
(source: http://adoxoblog.wordpress.com/2013/02/27/colossus/)
(source: http://www.maximumpc.com/article/features/12_things_you_didn%E2%80%99t_know_about_eniac)
(source: http://www.computersciencelab.com/ComputerHistory/HistoryPt4.htm)
The die from an Intel 8742, an 8-bit microcontroller that includes a CPU running at 12 MHz, 128
bytes of RAM, 2048 bytes of EPROM, and I/O in the same chip.(source: wikipedia)
Brigitte VACHON
December 7, 2015
CHAPTER 1. HISTORY
Your typical electronic devices rely largely on integrated circuits and semi-conductor devices.
(source: http://www.sonyclassics.com/itmightgetloud/)
(source: Apple)
What is a semiconductor?
Material that has electrical conductivity between that of a conductor and that of an insulator.
What is a semiconductor device?
An electrical component that relies on the properties of semiconductor (ex. silicon).
Brigitte VACHON
December 7, 2015
10
CHAPTER 1. HISTORY
In the context of scientific research, electronics often custom designed and built for specific applications and needs.
Typical Data AcQuisition (DAQ) system
sensor
Signal conditioning
(ex. filter, pre-amp)
Processing
stage (ex. ADC,
microprocessor)
Thin Gap Chamber readout for the ATLAS detector at the CERN LHC.
Detector testing facility at McGill University.(source: B.Vachon)
Brigitte VACHON
December 7, 2015
11
Chapter 2
A few useful reminders about impedance
Impedance: Measure of the opposition that a circuit presents to a current when a voltage is applied.
An extension of the concept of resistance to AC circuits. (source: Wikipedia)
Since the concept of impedance in an AC circuit takes into account effects of inductance and capacitance, impedance is a frequency dependent quantity. Effects of inductance and capacitance
generally more pronounced at high frequency.
12
CHAPTER 2. IMPEDANCE
2.1
Input impedance
2.2
Output impedance
Large output impedance Small current drawn out of device before voltage at the output drops
(Ex. Electric guitar, high-Z microphone)
Brigitte VACHON
December 7, 2015
13
CHAPTER 2. IMPEDANCE
2.3
Electronics 2.3.
for Scientists
SIGNAL
Device A
Device B
Zout
Zin
Videal
Rule of thumb: Input impedance should be at least 10 times larger than output impedance in order
to avoid the Device B overloading Device A.
Small output impedance typically a good thing since it allows to source a greater current while
avoiding excessive power lost to resistive heating.
We will see that one of the useful characteristics of operational amplifiers (Op amps) is their large
input impedance and low output impedance.
Brigitte VACHON
December 7, 2015
14
CHAPTER 2. IMPEDANCE
Z0 < ZL
Zo > ZL
Z0 = ZL
Brigitte VACHON
December 7, 2015
15
CHAPTER 2. IMPEDANCE
Twisted-pair cable
Ribbon cable
(source: wikipedia)
(source: wikipedia)
(source: wikipedia)
Brigitte VACHON
December 7, 2015
16
Chapter 3
Diode
A diode is a two-terminal, passive (i.e. no built-in power source), non-linear 1 device.
A diode is essentially an electrically controlled one-way gate.
Usually made from semiconductors but can also be a vacuum tube device.
We will only consider here semiconductor diodes.
Thevenin and Norton laws do not apply to circuits containing non-linear devices like diodes!
17
CHAPTER 3. DIODE
Diodes are useful devices that can be used for many applications such as:
power conversion (rectifiers)
ex. automotive alternator, ACDC conversion
Over-voltage protection
ex. diodes on connection pins of an IC
logic gates
temperature measuring
ionizing radiation detectors
voltage dropper
etc.
(source: http://en.wikipedia.org/wiki/Diode)
Circuit Symbol:
Brigitte VACHON
anode
December 7, 2015
cathode
18
CHAPTER 3. DIODE
n-type
p-type
December 7, 2015
19
CHAPTER 3. DIODE
The external battery voltage attracts electrons and holes away from the p-n junction,
thereby creating a depletion region. No current can flow through the diode.
December 7, 2015
20
CHAPTER 3. DIODE
Ideal diode model: Diode acts as short-circuit (wire) in one direction and as open circuit in other
direction. In other words, it only conducts current in one direction, with a very small forward voltage
drop.
(source: http://www.engineersgarage.com/tutorials/diodes)
Brigitte VACHON
December 7, 2015
21
CHAPTER 3. DIODE
Practical diode model: For many applications and circuit analysis a simple working model of a
diode is sufficient. In this working model, the diode is assumed to be a perfect diode and the average
forward voltage drop is approximately VD 0.6 V.
Brigitte VACHON
December 7, 2015
22
CHAPTER 3. DIODE
(source: http://www.engineersgarage.com/tutorials/diodes)
Diode (Ebers-Moll or Shockley) equation for voltage greater than breakdown voltage.
I = Io(eV /VT 1) where,
VT = kT /e (thermal voltage) is a characteristic of each diode (e.g. VT 50 mV for Silicon.)
I0 is characteristic of each diode and temperature DEPENDENT (e.g. near 200 C, I0 increases by
9% by degree).
Brigitte VACHON
December 7, 2015
23
CHAPTER 3. DIODE
3.1
Diode circuits
3.1.1
Voltage divider
Example: You have a 9 V battery but need lower voltage to power a device.
V0
0.6 V
0.6 V
0.6 V
Similar result could be achieved with resistor(s), but the use of diode(s) typically reduces the amount
of power wasted due to resistive heating and will supply a regulated voltage that is sensitive to current
variations.
Brigitte VACHON
December 7, 2015
24
CHAPTER 3. DIODE
3.1.2
Voltage regulator
Brigitte VACHON
December 7, 2015
25
CHAPTER 3. DIODE
3.1.3
Type of circuits used to cap voltages to preset limits. Useful as voltage protection circuitry or
signal conditioning.
Brigitte VACHON
December 7, 2015
26
CHAPTER 3. DIODE
3.1.4
December 7, 2015
27
CHAPTER 3. DIODE
3.1.5
Clamp circuit
When Vin < 0.6 V, diode conducts and charge capacitor to VC = Vpeak 0.6.
Capacitor cannot discharge
Vout = Vin + VC
Brigitte VACHON
December 7, 2015
28
CHAPTER 3. DIODE
Brigitte VACHON
December 7, 2015
29
CHAPTER 3. DIODE
3.1.6
Rectifier circuits: Making alternating voltage uni-directional (first step in building a circuit that
transforms ACDC.)
This is called a half-wave rectifier circuit since it only provides a positive voltage for half the period
of the wave.
Brigitte VACHON
December 7, 2015
30
CHAPTER 3. DIODE
3.1.7
Brigitte VACHON
December 7, 2015
31
CHAPTER 3. DIODE
(source: http://www.electronics-tutorials.ws/diode/diode_6.html)
Brigitte VACHON
December 7, 2015
32
CHAPTER 3. DIODE
If you add a capacitor in parallel with the load, the capacitor smooths out the ripples of the full-wave
rectifier output thereby providing essentially a DC voltage from an AC input. Voila!
(Vinpeak
T
...
1.2) 1
2RLC
(source: http://www.electronics-tutorials.ws/diode/diode_6.html)
December 7, 2015
33
CHAPTER 3. DIODE
3.2
In addition to the ordinary diode that we have been looking at until now, there exists other types
of diodes, here are a few you may come across as some point or another.
3.2.1
Circuit symbol:
(source: http://en.wikipedia.org/wiki/Light-emitting_diode)
Brigitte VACHON
December 7, 2015
34
CHAPTER 3. DIODE
3.2.2
Photodiode
Circuit symbol:
Here, the inverse process occurs. Incoming photons are absorbed in the p-n junction and the energy
deposited create additional electron-hole pairs in the depletion region. Think of it as a photoelectric effect taking place in the depletion region of the p-n junction.
The number of electron-hole pairs created is directly proportional to the intensity of light (i.e. number of photons).
Note that all p-n junction are sensitive to EM radiaton (UV, x-rays, -rays). In order to avoid
the malfunctioning of semiconductor devices due to induced photocurrent, electronic devices (ex.
transistors, ICs) are encapsulated in an opaque casing.
There are two principal modes of operation of a photodiode:
Photovoltaic mode: Photodiode is used with zero bias. Light creating electron-hole pairs
produces a current. (ex. solar cell)
Photoconductive mode: Photodiode reversed biased creating a larger depletion region. Incoming photons create electron-hole pairs which results in a reverse current flow which is directly
proportional to the illuminance2. (ex. optocoupler)
2
Brigitte VACHON
December 7, 2015
35
CHAPTER 3. DIODE
December 7, 2015
36
CHAPTER 3. DIODE
Optical fibers can transport data over much longer distances with minimal losses
Much higher bandwidth of data can be sent over optical fibers
Provide electrical isolation between two different electrical circuits
low noise
(source: atlas.ch)
Brigitte VACHON
December 7, 2015
37
CHAPTER 3. DIODE
3.2.3
Zener diode
Circuit symbol:
Zener diodes are similar to normal diodes but have smaller breakdown voltage, Vb, which is meant
to be attained (i.e. not bad or dammaging point to exceed in the operation of the diode).
These diodes are intended to be used in reverse bias mode.
Zener diodes are manufactured to have extremely sharp I-V curve breakdown. They are particularly
well-suited for the production of a reference voltage point and as voltage stabilizer removing ripple
from a low quality DC voltage.
Brigitte VACHON
December 7, 2015
38
Chapter 4
Transistors
Transistors are essential ingredients of virtually every electronic circuit!
Invented by Shockley, Bardeen and Brattain in 1947. Nobel prize in Physics in 1956.
39
CHAPTER 4. TRANSISTORS
Transistors are three-terminal active components (i.e. need to supply voltage/current to make
them function).
(source: http://www.mikroe.com/old/books/keu/04.htm)
They can be thought of as being similar to a water faucet. While a knob on a faucet controls the
flow of water, a voltage and/or current applied to a control lead on a transistor controls the current flowing through the other two leads. In order words, the amount of current flowing through a
transistor is proportional to a control signal.
Ok, not very exciting up to now, is it? We are just more or less talking about an electrically controlled current faucet where the output current is proportional to the controlling input signal. But
heres what makes transistors so famous: The output power of a transistor can be higher
than the controlling input power, which means that you can use a transistor to
amplify a signal!
Brigitte VACHON
December 7, 2015
40
CHAPTER 4. TRANSISTORS
Brigitte VACHON
December 7, 2015
41
CHAPTER 4. TRANSISTORS
4.1
Brigitte VACHON
December 7, 2015
42
CHAPTER 4. TRANSISTORS
If a positive voltage of at least 0.6 V is applied at the base, the holes in the base are
pushed toward the middle of the p-n junction and electrons attracted. Recombination
of electron-holes can occur and a current can
flow from the base to the emitter.
December 7, 2015
43
CHAPTER 4. TRANSISTORS
Brigitte VACHON
December 7, 2015
44
CHAPTER 4. TRANSISTORS
4.1.1
Theory
Now lets define different voltages and currents of a bipolar junction transistor in oder to be able to
describe its operation more quantitatively.
Brigitte VACHON
December 7, 2015
45
CHAPTER 4. TRANSISTORS
Lets now go into more details of how a npn1 bipolar junction transistors functions in normal operation.
For current to flow through transistor, voltage at collector must be greater than voltage at the
emitter, i.e. VC > VE.
The base-to-emitter voltage exhibits a voltage drop of 0.6 V, therefore VB must be at least
0.6 V greater than VE. It also means that in this case, you get VBE 0.6 V
By Kirchoffs current law: IE = IC + IB .
By Kirchoffs voltage law: VCE = VBE + VCB .
The most important feature of a transistor is described by the following formula: IC = hF E IB = IB .
Here, refers to the current gain. Typical values of are around 100.
Using the previous formula, you can show that in fact IE = ( + 1)IB, and for typical values of
this means IE IC .
Note: Every transistor has its own unique value which depends on temperature and changes
in VCE. Therefore, do not design circuits where the behaviour directly depends on the precise
value of !
1
Brigitte VACHON
December 7, 2015
46
CHAPTER 4. TRANSISTORS
Exercise: Given the values below and = 100, find VE,IE, IB, IC and VC.
Brigitte VACHON
December 7, 2015
47
CHAPTER 4. TRANSISTORS
Characteristic curves of a transistor are used to identify the working points of a transistor.
Find IC
assuming
VCE=0
(source: http://www.electronics-tutorials.ws/transistor/tran_2.html)
Brigitte VACHON
December 7, 2015
48
CHAPTER 4. TRANSISTORS
(Linear) active region: Region of operation where the collector current is roughly proportional to
the base current. [Transistor fully ON, without output current control]
Saturation region: Region of operation where collector current is no longer proportional to the base
current, the transistor acts like a closed switch from collector to emitter.
cutoff region: Region where the base current approaches zero resulting in a zero collector current.
[Transistor fully OFF]
Note that it is important to remember that a transistor has maximum allowed values of IC, IB and
VCE not too be exceeded, otherwise you burn the transistor. Transistors also have a maximum power
rating given by the maximum product of VCE and IC.
The desired operation point of a transistor is set by biasing the transistor, i.e.
setting the DC voltages of the transistors terminals.
Brigitte VACHON
December 7, 2015
49
CHAPTER 4. TRANSISTORS
4.1.2
Transistor biasing
To keep the operating point of a transistor in the linear active region for use as an amplifier, one
needs to carefully set the DC voltages of the transistors terminals.
Brigitte VACHON
December 7, 2015
50
CHAPTER 4. TRANSISTORS
Brigitte VACHON
December 7, 2015
51
CHAPTER 4. TRANSISTORS
Here is the common method of setting the constant DC operating conditions of a transistor: the
quiescent point or Q-point.
Use a voltage-divider circuit to set the DC Base
voltage:
2
VB = VCC R1R+R
2
The following steps introduces some useful rules of
thumb:
1. Choose the DC value of your output in order to allow for the largest possible symmetric output signal
swing without clipping the signal. Use the rule of
thumb of setting the quiescent output voltage as
Vout(E or C) 1/2VCC.
2. Choose the quiescent current you want flowing
through your transistor. Typical values for IC
IE are of the order of mA (e.g. 1 to 10 mA). This
in turn will give you an approximate target value
for IB = IC /.
3. Pick a value of resistor(s) RE or C that satisfy the
quiescent output voltage and collector current you
choose. [Relative values of RC and RE when appropriate are also chosen for desired gain]
Brigitte VACHON
December 7, 2015
52
CHAPTER 4. TRANSISTORS
Brigitte VACHON
December 7, 2015
53
CHAPTER 4. TRANSISTORS
4.1.3
Intrinsic resistance
One feature of bipolar junction transistor is the intrinsic (typically small) resistance present within
the emitter junction region.
Consider the base-emitter junction as a simple diode. The current as function of voltage across the
diode can be described by the Ebers-Moll equation:
IC = IS eVBE/VT 1 ' IE
Brigitte VACHON
kT 1
e IC
25
IC (mA)
December 7, 2015
54
CHAPTER 4. TRANSISTORS
Now, imagine fixing the emitter voltage and look at how much the base voltage change in response
to a change in base current. This defines the intrinsic resistance rBE to the emitter as seen at the
base, i.e.
25
kT 1
VBE
=
IB VE
e IC
IC(mA)
rBE
rBE = re
The base-emitter junction appears to have a low resistance when viewed from the emitter end, but
a higher resistance (by a factor of ) when viewed from the base.
Note: This effective resistance rBE may have to be taken into account when you choose values of R1
and R2 to bias your transistor.
Brigitte VACHON
December 7, 2015
55
CHAPTER 4. TRANSISTORS
Before introducing different amplifier circuits, let me introduce some quantities that we will want
to be able to calculate for each circuit. In general, what we want to know about an amplifier is:
Voltage gain: Gain = a =
Vout (RL =)
Vin
Iout
Iin
Vin
Iin
Brigitte VACHON
Vout (RL =)
Iout (RL =0)
December 7, 2015
56
CHAPTER 4. TRANSISTORS
4.1.4
This circuit shows what is called a common-collector amplifier (or also often referred to as emitterfollower). The main feature of this circuit is that the output is almost a mirror image of the input
(i.e. the output follows the input). It has no voltage gain! (...and you dare call this an amplifier?)
Ve
Brigitte VACHON
December 7, 2015
57
CHAPTER 4. TRANSISTORS
Brigitte VACHON
December 7, 2015
58
CHAPTER 4. TRANSISTORS
What is the input impedance as seen directly at the base of the transistor (i.e. not including
the voltage divider stage)?
To answer this question, we need to figure out by how much IB changes as a result of a change
in VB .
VB
IB
= (RE + re)
What is the input impedance of this circuit amplifier (i.e. including the voltage divider stage) ?
[Lets ignore the decoupling capacitor for simplicity]
Brigitte VACHON
December 7, 2015
59
CHAPTER 4. TRANSISTORS
What is the output impedance of the amplifier (i.e. with the load removed, looking back into
the emitter output)?
Since, Vout = VB 0.6 then
vout = vB
and also, since VB = Vin IB RB , then for fixed Vin
vB = IB RB
Putting these two equations together you get:
vout = IB RB =
IE
RB
But what is RB ? This is the effective resistance at the base of the amplifier. Ignoring the bypass
capacitor and any output impedance from the AC input signal generator, then
RB = R1 // R2
The output impedance of this emitter-follower amplifier is therefore
v
RB
out
=
Zout =
.
IE
Strictly speaking, Zout should also include RE in parallel but typically RB / dominates.
The effect of resistors attached to the base is reduced by significantly lowering the output
impedance of this amplifier circuit. Thats a good thing!
Brigitte VACHON
December 7, 2015
60
CHAPTER 4. TRANSISTORS
4.1.5
Common-emitter amplifier
The following circuit shows what is called a common-emitter amplifier. It is called a commonemitter because both the base and emitter leads are common (grounded through resistors). The
main feature of this circuit is that it provides a voltage gain.
(source:
T.Floyd, Electronic Devices (Conventional Current Version))
Brigitte VACHON
December 7, 2015
61
CHAPTER 4. TRANSISTORS
vE
RE
Finally, we find that the voltage gain of the circuit is: Gain =
vout
vB
RC
= R
Note the negative sign!! This means the output is inverted. Why? If Vin increases, then IC
increases and, by Ohms law, VC (= Vout) then decreases, voila!
Brigitte VACHON
December 7, 2015
62
CHAPTER 4. TRANSISTORS
But now, wait a minute! Does that mean that if RE = 0 that the gain of the amplifier is infinite?
No.
What we have forgotten here is to take into account the intrinsic resistance within the emitter
junction region of the transistor. Looking into the base of the transistor, this small resistance
re appears in series with the resistance RE .
C
The actual voltage gain of the amplifier is therefore Gain = R R+r
e
E
If RE = 0, then it looks like you get a much higher voltage gain out of your amplifier than if
you actually put a resistance RE in the circuit. Why would you put a non-zero RE ?
Well, the value of the intrinsic resistance re is quite unstable and depends for example on temperature. Therefore, to avoid unstable amplifier gain, a resistor RE is placed at the emitter at
the expense of loss in voltage gain.
But the story is not over yet, theres a trick you should know about. If you put a capacitor in
parallel with RE , then a high-frequency output signal will by-pass RE making it effectively
disappear. And therefore, for high frequency input signals the gain is dictated by re and no
longer by RE , thereby recovering high voltage amplifier gain but with temperature stability.
Neat eh?
Brigitte VACHON
December 7, 2015
63
CHAPTER 4. TRANSISTORS
IC
VB
And re-using the work weve done earlier, we find (taking into account the intrinsic emitter
resistance):
1
gm =
RE + re
Brigitte VACHON
December 7, 2015
64
CHAPTER 4. TRANSISTORS
Brigitte VACHON
December 7, 2015
65
CHAPTER 4. TRANSISTORS
4.1.6
Differential amplifier
Circuit that amplifies the difference between two signals and rejects any common components.
Long-tailed pair differential amplifier circuit
Brigitte VACHON
December 7, 2015
66
CHAPTER 4. TRANSISTORS
Looking at the quiescent state of the long-tailed pair differential amplifier circuit.
Brigitte VACHON
December 7, 2015
67
CHAPTER 4. TRANSISTORS
Brigitte VACHON
December 7, 2015
68
CHAPTER 4. TRANSISTORS
vout = RCi2 =
Brigitte VACHON
(4.1)
(v2 v1)
2RE
Then we find
Gdiff =
RC
(v2 v1)
2RE
vout
v1 v2
RC
= 2R
E
December 7, 2015
69
CHAPTER 4. TRANSISTORS
Brigitte VACHON
December 7, 2015
70
CHAPTER 4. TRANSISTORS
Brigitte VACHON
December 7, 2015
71
CHAPTER 4. TRANSISTORS
v2RE
v2
=
2R1RE + RE2 2R1 + RE
RC
2R1 + RE
Brigitte VACHON
vout
v1,2
RC for R R
C
= 2RR
'
E
1
2R1
1 +RE
December 7, 2015
72
CHAPTER 4. TRANSISTORS
Ideally, you want your differential amplifier to provide a high gain for your desired signals and zero
gain for common-mode signals.
A differential amplifier performance is typically expressed in terms of its Common-mode rejection ratio (CMRR).
Gdiff
GCM
R1 for R R
R
E
1
E
In general you want to make R1 large (i.e. a long tail) to have a high CMRR.
Brigitte VACHON
December 7, 2015
73
CHAPTER 4. TRANSISTORS
4.2
FET is the name of a very large family of transistors that include Junction Field Effects Transistors
(JFET) and Metal Oxide Semi-conductor Field Effect Transistor (MOSFET).
General characteristics:
Unlike BJTs, FETs are voltage-controlled devices (they require essentially no current).
Extremely large input impedance
Generally lower gain than BJTs
Susceptible to dammages from static electricity (especially MOSFETs)
Easy to build in small size for use in integrated circuits (ICs)
Large manufacturing spread in FET parameters (larger than BJTs)
Brigitte VACHON
December 7, 2015
74
CHAPTER 4. TRANSISTORS
4.2.1
JFET
n-channel JFET
D
p-channel JFET
D
G
S
Brigitte VACHON
A negative voltage applied to the gate (compared to the source voltage) creates a depletion region which chokes the flow of electrons thereby limiting the current flowing
through the transistor. Gate-source always reverse-biased.
December 7, 2015
75
CHAPTER 4. TRANSISTORS
4.2.2
MOSFET
There are several different types of MOSFETs and going through the detailed of them all is beyond
the scope of this course. For completeness, here are some of the MOSFET symbols you may come
across.
Enhancement MOSFETs
n-channel
p-channel
D
D
G
Depletions MOSFETs
n-channel
p-channel
D
D
G
S
G
S
G
S
To give you a general idea, here is how a typical (n-channel enhancement) MOSFET works:
(source: http://volga.eng.yale.edu/index.php/main/semiconductors)
Brigitte VACHON
Positive gate voltage repels all holes and actually attracts electrons (from source and/or
drain) into the channel. A current between
the Drain and the Source can now flow.
December 7, 2015
76
CHAPTER 4. TRANSISTORS
4.2.3
Theory
Since both JFETs and MOSFETs have similar general behaviour, for the rest of the course, well
focus more or less exclusively on JFETs, although much of the circuits we will study and build could
also be built using MOSFETs.
We will first start by defining the following voltages and currents
Brigitte VACHON
December 7, 2015
77
CHAPTER 4. TRANSISTORS
The current ID flowing through the transistor varies as function of the voltage difference between
the drain and the source (VDS = VD -VS).
A maximum current will flow through a JFET when the gate voltage is identical to the source voltage
(i.e. VGS = VG - VS = 0 V).
This maximum drain current (IDSS) varies for different FETs and is given in the transistors data
table . Typical values range from about 1 mA to 1 A.
Brigitte VACHON
December 7, 2015
78
CHAPTER 4. TRANSISTORS
The cutoff voltage (VGS(off)) is the value of VGS that makes ID zero. Typical values range from -0.5
to -10 V. This value is normally given in the data table of your transistor.
Ohmic region: When VDS is small, current flowing through the transistor is proportional to VDS.
The transistor behaves like voltage-controlled resistor.
ID =
Brigitte VACHON
IDSS
VGS(off)2
(4.2)
79
CHAPTER 4. TRANSISTORS
Active region: The current flowing through the transistor (ID) is dictated by the voltage difference
between the gate and the source (i.e. VGS), and is mostly independent of VDS. This occurs because
the channel between the drain and source is so narrow that raising VDS does not increase ID.
ID =
IDSS
VGS(off)2
VGS VGS(off)
2
(4.3)
Brigitte VACHON
December 7, 2015
80
CHAPTER 4. TRANSISTORS
Exercise: If VGS(off) =-4 V and IDSS =12 mA, find the value of ID when VGS is set to -2 V and
0 V. Assume the JFET is in the active region. Indicate approximately where these operating points
would like on a typical I-V characteristic curve of a FET.
Brigitte VACHON
December 7, 2015
81
CHAPTER 4. TRANSISTORS
4.2.4
Transconductance
Since the gate current of a FET is zero, the natural gain parameter for FETs is transconductance.
We can even say that FETs are transconductance devices, i.e. a change in input voltage results in
a change in output current.
For a FET, transconductance is defined as:
ID
gm =
VGS VDS
Lets define the transconductance of a FET with its gate grounded such that VGS =0:
IDSS
gm(0) = 2
VGS(off)
The general transconductance can then be written as:
1
gm(VGS) = gm(0)
VGS
VGS(off)
v
u
u
u
t
December 7, 2015
ID
IDSS
82
CHAPTER 4. TRANSISTORS
The value of transconductance changes with ID. If you change the value of the gate-source
voltage VGS , this in turn changes the current flowing through the transistor ID which in turn affects
the value of transconductance of the transistor.
You can think of transconductance as a measure of the amplification of a transistor.
You can also equivalently think of the transconductance as being related to a small internal resistance within the drain-source channel with value 1/gm, which therefore contributes to the output
impedance of the transistor. [Note: This is analogous to the value re for BJTs]
Brigitte VACHON
December 7, 2015
83
CHAPTER 4. TRANSISTORS
4.2.5
What is the current produced by this current source (assuming the transistor is operated in the
active region)?
December 7, 2015
84
CHAPTER 4. TRANSISTORS
Lets add a resistor RS and see how this improves the design of our
current source:
Again, assuming the transistor is operated in the active region:
VS IDRS = 0
and therefore,
VGS = VG VS = IDRS
What we find is that the value of VGS is inversely proportional to the
current flowing through the transistor. RS provides feedback to
the circuit!
[As ID increases, VGS increases in magnitude so we go to lower spot
on ID vs VDS plane, and ID decreases - an equilibrium is established]
What is the current ID once you add a load resistance ? We have two equations:
ID =
IDSS
VGS(off)2
VGS VGS(off)
ID =
2
VGS
RS
With two equations and two unknowns, you can solve for ID.
Brigitte VACHON
December 7, 2015
85
CHAPTER 4. TRANSISTORS
Brigitte VACHON
December 7, 2015
86
CHAPTER 4. TRANSISTORS
4.2.6
Transistor biasing
Just like for BJT, to keep the operating point of a transistor in the intended region of operation, one
needs to carefully set the DC voltages of the transistors terminals.
To operate a JFET, you need to reverse-bias the gate-source junction, which requires a negative gate
voltage VG as compared to the source voltage VS .
The most common type of JFET bias circuit is called self-bias.
What is the purpose of RG?
1. Pick ID = IDSS /2 to allow maximum swings
in drain current.
2. Calculate the required VGS using Equation 4.3.
3. Calculate the required value of RS (for a given
ID and VGS ).
4. Choose RD such that VD = VDD /2 to allow
maximum swings in output voltage.
5. Arbitrarily choose RG to be large to prevent
loading previous stage of the circuit.
(source: T.Floyd, Electronic Devices (Conventional Current Version))
Brigitte VACHON
December 7, 2015
87
CHAPTER 4. TRANSISTORS
4.2.7
Common-source amplifier
This circuit is equivalent to the BJT common-emitter amplifier, and provides voltage gain.
Brigitte VACHON
December 7, 2015
88
CHAPTER 4. TRANSISTORS
To determine the Q-point of the amplifier circuit you need to find ID and VDS:
In the active region, the characteristic curve can be expressed as:
2
IDSS
ID =
VGS VGS(off)
VGS(off)2
Also, without any input signals (i.e. DC state),
VGS = IDRS
Solve for ID. Then once you have ID you can find VD, VS and VDS.
Brigitte VACHON
December 7, 2015
89
CHAPTER 4. TRANSISTORS
Show that the gain of a common-source amplifier circuit (that does not have a C2 capacitor) is
gmRD
Gain = 1+
g R
m
If you add a capacitor C2 to the circuit, then Gain =gmRD. Do you know why?
What is the input impedance of this amplifier circuit?
Brigitte VACHON
December 7, 2015
90
CHAPTER 4. TRANSISTORS
4.2.8
Source-follower amplifier
This circuit is equivalent to the bipolar emitter-follower amplifier, it provides current but no voltage
gain; the output voltage follows the input voltage (i.e. approximately unity voltage gain).
Brigitte VACHON
December 7, 2015
91
CHAPTER 4. TRANSISTORS
S
Show that the gain of this amplifier is Gain = R R
+1/
gm , and that indeed Vout Vin.
S
Is the output in phase or 180 degrees out-of-phase with the input signal?
Brigitte VACHON
December 7, 2015
92
Chapter 5
Operational amplifiers
Operational amplifiers (op-amps) are high performance differential amplifiers packaged in an integrated device.
large (open-loop) gain (at least at low frequencies)
large input impedance
Op-amps were originally conceived for doing mathematical operations in analog computers. Their
use has been partially superceeded by digital computers but still used extensively in analog circuits.
93
(source: https://www.sparkfun.com/products/9456)
A typical op-amp is made up of three types of amplifier circuits: a differential amplifier, a voltage
amplifier, and a push-pull amplifier.
Differential amplifier: Amplifies the difference between the two input signals.
Voltage amplifier: Provides additional voltage gain.
push-pull amplifier: Provides a low output impedance and a high current driving capability (i.e.
high power capability).
Brigitte VACHON
December 7, 2015
94
An op-amp is an integrated device that contains several transistors, resistors and capacitors. You
dont need to know the internal circuitry of an op-amp but heres what a typical one could look like
Brigitte VACHON
December 7, 2015
95
5.1
The output of an ideal op-amp is the difference between the non-inverting input (V+) and the
inverting input (V), amplified by the open-loop gain Ao.
Vout = Ao(V+ V)
December 7, 2015
96
5.2
5.2. COMPARATOR
Comparator
A comparator circuit compares an input signal to a reference voltage, and indicates by a two-state
output whether the input signal is lower or higher than the reference voltage.
This kind of circuit is useful to know when an input analog signal is larger than some predefined
value (i.e. when a signal passes a threshold).
Brigitte VACHON
December 7, 2015
97
5.3
5.3. FEEDBACK
Feedback
Feedback: Take some of the output and mix it with the input.
Positive feedback: Add some fraction of the output to the input. Usually not desirable (ex. microphone in front of a speaker), unless you want to make an oscillator circuit.
Negative feedback: Subtract a fraction of the output from the input. Obtain more predictable and
stable operation.
Feedback can make an amplifier circuit less dependent on actual features of the amplifier
(Harold S. Black, 1928) [See Appendix E]
Op-amps almost always used with
feedback, and mostly using negative
feedback.
Lowers gain of the circuit, but...
Improves linearity (gain stability)
Increases bandwith of amplifier
Can increase input impedance and
lower output impedance of a circuit
Brigitte VACHON
December 7, 2015
98
5.3.1
5.3. FEEDBACK
A0
Vout
=
Vin
1 + f A0
Negative feedback provides stabilization of the voltage gain of an amplifier against dependence on
parameters such as temperature, frequency, etc.
Heres an illustration of this gain stabilization. Lets assume a feedback network that feeds back a
fraction of f=0.1 of the output voltage back to one of the amplifier input. The the closed-loop gain
of the circuit with feedback corresponding to different values of the open-loop gain is shown in the
table below:
A0 ACL = A0/(1 + f A0)
1,000
90.9
5,000
98.3
10,000
99.0
20,000
99.5
Brigitte VACHON
December 7, 2015
99
5.3.2
5.3. FEEDBACK
A0
Vout
=
Vin
1 + f A0
and therefore,
A0
Vin = Vd + f
Vin
1 + f A0
Now replacing Vd by IinZin (where Zin corresponds to the impedance of the amplifier), we find
Zin(CL) =
Vin
= (1 + f A0)Zin
Iin
Therefore, the use of negative feedback increases the effective input impedance of the circuit (which
is typically a good thing!).
Brigitte VACHON
December 7, 2015
100
5.3.3
5.3. FEEDBACK
A0Vin ZoutIout
=
(1 + A0f )
Version))
Vout
Zout
=
=
Iout Vin (1 + A0f )
Therefore, the use of negative feedback decreases the effective output impedance of the circuit (which
is typically a good thing!).
Brigitte VACHON
December 7, 2015
101
5.4
5.5
Buffer
Looking at the circuit you can see that V = Vout.
By the Golden rule (2), V+ = V and therefore
Vin = Vout.
The use of feedback increases the input impedance and decreases the output impedance of the circuit
(as compared to the open-loop input/output impedance of op-amp).
Brigitte VACHON
December 7, 2015
102
5.6
Inverting amplifier
What is the closed-loop gain of this op-amp
circuit ?
By golden rule (2), V = 0 V (virtual
ground)
Voltage across R2 is -Vout.
Voltage across R1 is Vin.
By golden rule (1), no current flows into
in = I = Vout
the op-amp I1 = V
2
R
R
1
2
The gain is therefore, Gain = R
R1
Brigitte VACHON
December 7, 2015
103
5.7
Non-inverting amplifier
December 7, 2015
104
5.8
Summing amplifier
What is the output Vout of this op-amp circuit ?
Practical examples:
Audio mixer
D/A conversion
Adding DC offset to AC signal
Note that the input voltages (V1, V2, V3) can be
positive or negative, such that this circuit can
also do subtraction.
Brigitte VACHON
December 7, 2015
105
5.9
5.9. INTEGRATOR
Integrator
What is the output Vout of this op-amp circuit ?
By rule (2), V = 0 V , thus IC = C dVdtout
Also, we have IR =
Vin
R
By rule (1), IR = IC
Putting everything together: Vin = RIC = RC dVdtout
[Recall: For a capacitor Q=CV and
dV
I = dQ
dt = C dt ]
Practical examples:
analogue computer
charge amplifier
Brigitte VACHON
1
=
RC
Vin(t)dt + const
106
5.10
Current source
Brigitte VACHON
December 7, 2015
107
5.11
Ao 104 to 106
Zin 106 1012
Zout 10 1000
Input current nA to pA
Ao =
Zin =
Zout = 0
Input current = 0
Brigitte VACHON
December 7, 2015
108
Input offset voltage: For V+ V = 0, Vout6= 0. Amount of voltage that needs to be applied to
one of the inputs to zero the output.
Input offset current: Difference between I+ and I.
Common-mode input range: Range of input voltages for which the op-amp is guaranteed to
work, typically given by range of supplied external voltages or fraction thereof. If range not respected, possible results include saturation, gain sign change, circuit dammage, etc.
Differential input voltage range: Limits on voltage difference between the two inputs. If limit
exceeded (ex. > 0.5 V), can cause dammage.
Slew rate: Maximum rate of change of the output voltage (i.e. cant change output voltage instantaneously). Units: [V/s].
Slew rate becomes critical when dealing with high frequency input signals. (ex. op-amp 741 has 0.5
V/s, op-amp HA2539 has 600 V/s.
Frequency response: The gain of an op-amp depends on frequency of the input signal. Roll-off
as function of frequency deliberately designed into op-amp to avoid oscillation at high frequencies.
Brigitte VACHON
December 7, 2015
109
Decibel definition:
A1
Amplitude: dB = 20 log10 A2
P1
Power: dB = 10 log10 P2 since P A2
Brigitte VACHON
fB(CL)A(CL) = fT 1
Thus, gain-bandwidth product is always equal to
fT = fB(CL)A(CL)
December 7, 2015
110
5.12
5.12. OSCILLATORS
Oscillators
Device that produces a repetitive waveform that can be used to drive other circuits. There is an
oscillator in essentially all electronic devices.
In digital electronics, oscillators producing square waves are called clocks and are used to push
bits of information through the digital circuit.
Heres an example of a simple square wave RC relaxation oscillator. Note that this is the first (and
only time) well use positive feedback with an op-amp!
Brigitte VACHON
December 7, 2015
111
5.12. OSCILLATORS
Brigitte VACHON
December 7, 2015
112
Chapter 6
Digital electronics
Digital electronics used in most modern data processing.
Less sensitive to noise and circuit parameters (ex. temperature dependence of components).
Conceptually easier than analog electronics; you can build circuits like LEGO construction.
113
In digital electronics, unlike analog electronics, there are only two voltage states present at any point
within the circuit: high or low.
These two states can represent one bit of a number, on/off state, true/false, a signal has passed a
given threshold, etc.
Logic levels (high/low) usually given by voltage (but can be current), and usually positive true
(but can also be negative true).
Brigitte VACHON
December 7, 2015
114
There are different conventions of what a high/low signal is, this depends on the logic family you
use. For example,
0 0 to 0.8 V
1 2 to 5.0 V
*TTL
(transistor-transistor logic)
0 0.1 to 1.0 V
1 3.5 to 5.0 V
*CMOS
(Complementary MOSFET)
NIM
(Nuclear instrumentation module logic)
ECL
(emitter coupled logic)
0 0 V
1 -0.8 to -1.0 V (current through 50 )
0 -1.8 V
1 -0.9 V
Brigitte VACHON
December 7, 2015
115
6.1
Binary
There are 10 kinds of people in the world - those that use binary and those that dont
Binary numbers are composed of only two binary digits, 0 and 1, which are called bits (for Binary
digIT).
A binary number of 8 bits is called a byte.
A number in the binary representation is expressed as successive powers of 2. For example,
11012 = 1 23 + 1 22 + 0 21 + 1 20 = 8 + 4 + 1 = 1310
The highest-order bit (leftmost bit) is called the most significant bit (MSB). The lowest-order bit
(rightmost bit) is called the least significant bit (LSB).
Brigitte VACHON
December 7, 2015
116
6.1.2
Hexadecimal
Another useful number representation in digital electronics. Here, there are 16 allowable digits
(0,1,2,3,4,5,6,7,8,9,A,B,C,D,E,F) and numbers are expressed as successive powers of 16. For example,
2D516 = 2 162 + D(= 1310) 161 + 5 160 =
(Decimal)
December 7, 2015
117
Microprocessors systems nowdays typically use 8-bit, 16-bit, 32-bit or 64-bit words. Large binary
words can therefore easily be broken down into a small number of hexadecimal digits of more
manageable size. For example, a 32-bit word can be simply expressed as 8 hex digits.
6.1.3
System where each decimal digit is separately encoded into a 4-bit binary number. For example,
12310 = |0001
0010
0011
{z }
| {z }
| {z }
1
Brigitte VACHON
December 7, 2015
118
This is somewhat of a wasteful number representation in terms of resource usage since it uses 4
bits for each digit even though 4 bits can cover 0-15 values.
Having said that, this is a very convenient representation for passing on information to a digital
(decimal) number display.
December 7, 2015
119
Brigitte VACHON
December 7, 2015
120
6.1.4
Gray code
Decimal
0
1
2
3
4
5
6
7
8
9
BCD
0000
0001
0010
0011
0100
0101
0110
0111
1000
1001
Gray
0000
0001
0011
0010
0110
0111
0101
0100
1100
1101
ASCII
ASCII stands for American Standard Code for Information Interchange. This is a standard alphanumeric system used to transmit information (numbers, letters, symbols, commands) between
computers and peripherals.
The ASCII code consists of 128 unique 7-bit words (an additional bit is typically added to make it
compatible with 8-bit systems). Here are some example of ASCII characters:
Brigitte VACHON
December 7, 2015
121
Brigitte VACHON
December 7, 2015
122
6.2
Clock: Most digital circuits require a clock signal (i.e. series of high and low pulses at a fixed
precise frequency). As we will see later, the clock signal is used to push binary digits (bits) through
the digital circuit.
Serial bit transmission: Bits are transmitted on the same electrical conductor, one after the
other. One bit is sent at each clock cycle (either on the rising of falling edge of the clock, depending
on the design of the digital circuit). Here, you need to know ahead of time how many bits are being
sent or alternatively how to recognize the start/end of a number. [Ex. transmission over USB,
FireWire, ethernet cable, etc.]
Example: Raw data from ATLAS muon chamber electronics
frame cnt
z
}|
header
{z
}|
data
board id
{z
}|
{z
}|
data
{z
}|
data
{z
}|
data
{z
}|
end of data
{ z
}|
December 7, 2015
123
6.3
Logic functions
Combine digital inputs from two or more sources and produce a digital output that depends on the
state of those inputs.
Use devices (typically ICs) called gates that give combinational logic.
This allows you to perform Boolean algebra.
Brigitte VACHON
December 7, 2015
124
Brigitte VACHON
December 7, 2015
125
6.3.1
Here is an example use of one single logic gate. This gate is used to control the passage of an
incoming logic signal (ex. a clock signal).
6.4
Combinational logic consists in combining logic gates together to perform complex functions.
Use rules and symbols of boolean algebra (i.e. algebra defined with variables that can either be
true or false) to design complex combinational logic circuits.
Brigitte VACHON
December 7, 2015
126
Brigitte VACHON
December 7, 2015
127
Brigitte VACHON
December 7, 2015
128
Example: Find Boolean expression for the following circuit and use logic identities to implement the
same function with a smaller number of logic gates.
Brigitte VACHON
December 7, 2015
129
6.4.1
Brigitte VACHON
December 7, 2015
130
6.5
Karnaugh Maps
We have seen how a Boolean expression can be simplified to reduce the number of logic gates that
are used to implement a digital circuit. This can be done quite quickly and easily for logic functions
with up to two inputs.
For more complex logic functions, simplifying by-hand a complex Boolean expression is time consuming and not necessarily an easy task.
For circuits with three to four inputs, one can use Karnaugh maps to convert any truth table
into the simplest possible Boolean expression for your complex digital logic function.
For circuits with more than four inputs, one will typically use specialized computer programs
that take in a truth table or Boolean expression and convert it into a simpler Boolean expression
that can be implemented with logic gates.
To see how to construct and use a Karnaugh map, lets look at a specific example.
Brigitte VACHON
December 7, 2015
131
Example: Design a digital circuit for a car alarm. The car alarm should be turned on only when the
key is in the ignition and either the drivers door is open or his/her seatbelt is not fastened.
A: Door (0=open, 1=closed)
B: Key (0=not in ignition, 1=in ignition)
C: Belt (0=unfastened, 1=fastened)
Q: Alarm (0=off, 1=on)
1. Write truth table describing the logic function you want to implement.
A
0
0
0
0
1
1
1
1
Brigitte VACHON
B
0
0
1
1
0
0
1
1
C
0
1
0
1
0
1
0
1
December 7, 2015
Q
0
0
1
1
0
0
1
0
132
B
0
0
1
1
0
0
1
1
C
0
1
0
1
0
1
0
1
Q
0
0
1
1
0
0
1
0
BC
00 01 11 10
0 0 0 1 1
A
1 0 0 0 1
Brigitte VACHON
December 7, 2015
133
4. For each group, identify the variables that are constant or a common denominator within the
group, i.e. the variables that define the outcome of all cells of the group.
AB
BC
5. Just make an OR of all expression describing all the groups you have identified. Voila!
+ B C
Q = AB
A
Q
B
C
Brigitte VACHON
December 7, 2015
134
Exercise: Design a simple logic circuit that implements the following truth table.
A
0
0
0
0
0
0
0
0
1
1
1
1
1
1
1
1
B
0
0
0
0
1
1
1
1
0
0
0
0
1
1
1
1
C
0
0
1
1
0
0
1
1
0
0
1
1
0
0
1
1
D
0
1
0
1
0
1
0
1
0
1
0
1
0
1
0
1
Q
0
1
0
1
0
1
1
1
0
1
0
1
0
1
0
1
CD
00 01 11 10
00
01
AB
11
10
Answer:
Brigitte VACHON
December 7, 2015
135
Brigitte VACHON
December 7, 2015
136
6.6
Sequential Logic
Up to now, our digital electronics experience has only involved combinational logic gates which have
the property of immediately changing their output as a results of a change in input.
In order to provide memory to a circuit, the circuit needs to be able to latch (i.e. hold-on) to data
(i.e. bits) at a particular time.
Digital electronics circuits which have memory fall into the category of sequential logic. The word
sequential refers to the fact that to store/retrieve data involves a sequence of steps.
A clock signal is often used to push bits of information (i.e. 1s and 0s) through a sequential
circuit, and also provides an absolute time reference for all the sequences of steps involved in the
digital circuit.
Asynchronous circuit element: Output changes as soon as the input values changes.
Synchronous circuit element: Output changes as a result of changes on its inputs, only in synch
with a clock signal.
Brigitte VACHON
December 7, 2015
137
6.6.1
SR flip-flops
The most basic building block of all circuits storing data is a Set-Reset flip-flop, also known as SR
latch.
A SR flip-flop is a two-input one-bit memory bistable device.
Heres a simple flip-flop implemented using two NOR gates.
Brigitte VACHON
December 7, 2015
138
Transition table
condition
S R Q Q
Brigitte VACHON
December 7, 2015
139
condition
S R Q Q
Brigitte VACHON
December 7, 2015
140
Brigitte VACHON
December 7, 2015
141
Q Q condition
S R
Brigitte VACHON
December 7, 2015
142
Logic device where putting set input HIGH gives an output Q=1 is called active high.
Example: SR flip-flop implemented using NOR gates.
Logic device where putting set input LOW gives an output Q=1 is called active low.
Example: SR flip-flop implemented using NAND gates.
flip-flops
Logic symbols for an SR and SR
Brigitte VACHON
December 7, 2015
143
Brigitte VACHON
December 7, 2015
144
6.6.2
This is the first synchronous device we will study. The inputs of the flip-flop are enabled/disabled
by an external control signal (ex. clock). Only when the clock is HIGH are the inputs sampled, and
the resulting output is held unchanged while the clock is LOW.
December 7, 2015
145
6.6.3
Edge-triggered SR flip-flops
An edge-triggered SR flip-flop only samples the inputs during the edge of a clock signal. The circuit
can be designed to respond either to the positive () or the negative () edge of the clock signal.
pulse
transition
detector
Brigitte VACHON
December 7, 2015
146
CLK
pulse
transition
detector
S
0
1
0
1
R Qn
Qn
mode
0 Qn1 Qn1
hold
0
1
0
SET
1
0
1
RESET
1
1
1
indeterminate
Brigitte VACHON
December 7, 2015
147
6.6.4
CLK HIGH: Master produces Y,Y according to inputs S,R. Slave keeps old values of Q,Q.
are updated according to values of Y,Y.
CLK LOW: States Y,Y remain the same but Q,Q
Brigitte VACHON
December 7, 2015
148
6.6.5
Symbolic representation
6.6.6
A D-type flip-flop (D for Data) is a single input device. The circuit of a D-type flip-flop consists
in a SR flip-flop where you tie the S and R inputs together with an inverter such that R = S always.
This configuration avoids ever having the flip-flop be in an indetermined state.
This type of flip-flop also does not have a hold state, it only has a set and reset condition.
Brigitte VACHON
December 7, 2015
149
Brigitte VACHON
December 7, 2015
150
Just as weve seen before for a basic SR flip-flop, D-type flip-flop also comes with the following
variations:
Level-triggered (clocked) D-type flip-flop
December 7, 2015
151
December 7, 2015
152
6.6.7
JK flip-flop
December 7, 2015
153
December 7, 2015
154
Brigitte VACHON
December 7, 2015
155
6.6.8
Since it can easily get confusing, lets now summarize the different types of flip-flops we have seen
so far:
SR flip-flop: Two inputs. Can be in set, reset, hold or not used states.
asynchronous or default
Level-triggered (clocked or gated)
Edge-triggered
Pulsed-triggered (Master-Slave)
D-type flip-flop: Single input. No indeterminate nor hold states, only set and reset
states possible.
Level-triggered (clocked or gated)
Edge-triggered
Pulse-triggered (Master-Slave)
JK flip-flop: Two inputs. Allowed states are set, reset, hold and toggle.
Edge-triggered
Pulse-triggered (Master-Slave)
Brigitte VACHON
December 7, 2015
156
6.6.9
Timing considerations
In the design of digital electronics circuits, one needs to avoid possible synchronization problems
that could results in an undetermined outcome. This is called a race condition.
To avoid race conditions, one often builds in a hold time where inputs have to stay in a state for a
minimum time after the clock to define the output.
Race conditions can arise due to:
propagation time of signals
delays in logic function
Brigitte VACHON
December 7, 2015
157
Brigitte VACHON
December 7, 2015
158
6.6.10
Can link together JK flip-flops to make a circuit that counts how many clock pulses have occured.
Use the data output from the previous element to clock the next element.
Recall that J=K=1 corresponds to toggle mode.
This is called a ripple counter because the effect of the clock signal moves down the chain of
JK flip-flops.
The example circuit on the next slide is called a modulus-16 counter because it returns zeros
after 16 clock pulses. To make a modulus-2n counter, you need n flip-flops.
Brigitte VACHON
December 7, 2015
159
Brigitte VACHON
December 7, 2015
160
If you need a modulus-N counter where N 6= 2n, you can build in some logic that will reset all the
flip-flops to zero when you get to that number.
Brigitte VACHON
December 7, 2015
161
Brigitte VACHON
December 7, 2015
162
6.6.11
Synchronous counters
The last output stages of an asynchronous (ripple) counters take time to react to the original
incoming clock signal due to propagation delays.
Can instead construct a synchronous counter where all flip-flops are driven at the same time
by a common clock signal.
Use AND gates to change flip-flop states from hold to toggle.
Brigitte VACHON
December 7, 2015
163
Brigitte VACHON
December 7, 2015
164
6.6.12
Memory Registers
Brigitte VACHON
December 7, 2015
165
6.6.13
Shift registers
A shift register is made by connecting a series of (D-type) flip-flops such that each Q output
drives the next D input, and all clock inputs are driven simultaneously.
The idea is that at each clock pulse, the pattern of 0s and 1s in the register shifts to the right
(with data D entering from the left).
Shift registers are useful for conversion of parallel data to serial data and vice-versa, and can
also be used as a way to queue information.
CLK Q0 Q1 Q2 Q3
1
1 0 0 0
2
0 1 0 0
3
0 0 1 0
4
1 0 0 1
5
1 1 0 0
0 1 1 0
6
7
1 0 1 1
.....
(source: P.Scherz and S.Monk, Practical Electronics for Inventors)
Brigitte VACHON
December 7, 2015
166
Brigitte VACHON
December 7, 2015
167
This particular use of a shift register requires additional logic to load the parallel data into the
D-type flip-flops before enabling the serial reading out of the shift register.
Brigitte VACHON
December 7, 2015
168
Chapter 7
Analog/Digital interfacing
Almost any electronic device relies on the ability to convert an analog signal into a digital one, and
vice-versa. [Give examples of where ADC and DAC are used in everyday life.]
169
7.1
7.1. SAMPLING
Sampling
Process of taking a number of discrete values at points along the waveform. (the more samples you
take the more accurately you can define a waveform).
Brigitte VACHON
December 7, 2015
170
CHAPTER 7. ANALOG/DIGITAL INTERFACING Electronics for Scientists 7.2. NYQUIST(-SHANNON) SAMPLING THEOREM
7.2
December 7, 2015
171
7.3
7.3. ALIASING
Aliasing
Aliasing: Frequency not present in an original signal but that nevertheless could match the sampled
points. This occurs when a signal is undersampled.
Amplitude (dB)
falias(N ) = |f N fs|
fsignal =
900 Hz
fsamplingl =
1000 Hz
(source: http:
falias1 =
100 Hz
//www.dataq.com/blog/faqs/really-need-know-sample-rate/)
falias2 =
1100 Hz
falias3 =
2100 Hz
falias3 =
3100 Hz
f(Hz)
December 7, 2015
172
7.3. ALIASING
You can experiment with the phenomena of aliasing using an applet at http://www2.egr.uh.
edu/~glover/applets/Sampling/Sampling.html
Left: No aliasing
Right: Aliasing
Note that when you go from digital to analog conversion, the analog signal is reconstructed using
only frequencies less than the Nyquist freqency which is defined as half the sampling rate.
Brigitte VACHON
December 7, 2015
173
7.3. ALIASING
To avoid undesirable effects of aliasing, you can oversample your analog signal.
Example: Audio CDs have a sampling rate of 44100 Hz. The Nyquist frequency is therefore 22050 Hz.
If recorded sounds have frequency components higher than 22050 Hz, the aliased frequency images
of the sound will overlap partly with the original sound frequency spectrum, thereby distorting the
output signal produced from the sampled version of the original audio signal.
One way to avoid this distortion is by oversampling the original audio signal at much higher sample
rate than the range of human hearing (20 kHz) demands, thereby ensuring that any aliasing effects
occur at supersonic frequencies (frequencies above those of audible sound for humans). For example,
using a 64x oversampling converter, means that you are sampling an audio signal at 44.1 kHz x 64
2.8 MHz (!). The Nyquist frequency here would therefore be 1.4 MHz, a much higher frequency
than what can actually be recorded by a microphone.
Brigitte VACHON
December 7, 2015
174
7.3.1
7.3. ALIASING
Anti-aliasing filter
Note that the Nyquist-Shannon criteria is to have a sampling rate twice the highest frequency actually in the signal, and not the highest frequency of interest.
One might decide that they dont need the higher frequencies in an original analog signal. These
higher frequencies should be removed prior to signal sampling to avoid aliasing. The filters that do
that are called anti-aliasing filters.
Example: Human speech has components about 10 kHz but it is still easy to understand speech
only considering frequencies of the voice less than 3.4 kHz. Therefore, you can filter out frequency
components of the speech above 3.4 kHz and then sample the resulting analog signal at a rate equal
or greater than 3.4 kHz x 2 = 6.8 kHz. This would satisfy the Nyquist sampling theorem and ensure
that your original (filtered) analog input can be reconstructed completely and without any aliasing.
Brigitte VACHON
December 7, 2015
175
7.4
After filtering and sampling, the sampled voltage levels must be kept constant until the next sample
occurs. This step in the signal processing is important to allow the analog-to-digital converter (ADC)
circuit to have time to process the sampled value.
Brigitte VACHON
December 7, 2015
176
7.5
Analogue-to-Digital Conversion
Brigitte VACHON
December 7, 2015
177
(source: Texas Instruments, Understanding Data Converters, Application Report, SLAA013, 1995, www.ti.com/lit/an/slaa013/slaa013.pdf)
Brigitte VACHON
December 7, 2015
178
7.6
7.6. RESOLUTION
Resolution
Resolution: Step size or quantization levels. The graininess of your digital representation.
The previous example shows that a 4-bit converter does not have a sufficient resolution to make the
analog signal appear as a continous (smooth) function.
To improve the resolution of the signal conversion, you use a larger bit-number, ex. 6-bit, 8-bit,
10-bit, etc.
Example: If your converter has a resolution of 8-bit, then this defines 28 = 256 analog steps (quantization levels). Now if your converter uses a maximum voltage scale of 0 to 15 V, then each analog
step is of (15 V / 256) = 0.058 V
Using high resolution converter is more expensive.
The cost of reducing the reference voltage is a loss of input dynamic range.
Brigitte VACHON
December 7, 2015
179
7.7
Staircase ADC
Brigitte VACHON
December 7, 2015
180
Digital
output
Brigitte VACHON
December 7, 2015
181
7.7.2
The value of each bit of a binary output is found one at a time, starting with the MSB first.
1. Start conversion by setting MSB (D7) bit high and all other lines low.
2. DAC returns half of maximum dynamic scale (+5 V here)
3. Comparator compares DAC value with input analog voltage
4. If analog input greater than +5 V, SAR keeps D7 line high, otherwise puts D7 line low.
Brigitte VACHON
December 7, 2015
182
5. At next clock pulse, next bit (D6) is tried (i.e. put high)
6. etc.
Note that this is a faster (10 to 300 s) search algorithm than with the counter (staircase) method
above.
What happens if the analog input changes during A/D conversion? Errors in the ADC output! (i.e.
you would not get a binary number representative of the analog input value at the beginning of the
conversion process)
7.7.3
Brigitte VACHON
December 7, 2015
183
The priority encoder is enabled by a pulse on the EN input. The frequency of the enabled pulse
consists of the sampling frequency.
The (octal-to-binary) priority encoder transforms the value of the input lines interpreted as an octal
input into a 3-bit binary output.
Brigitte VACHON
December 7, 2015
184
7.8
Note that one will usually use a low-pass filter at the output of the DAC in order to smooth off the
square edges due to the quantization of the digital input. This is called a reconstruction filter.
Brigitte VACHON
December 7, 2015
185
(source: Texas Instruments, Understanding Data Converters, Application Report, SLAA013, 1995, www.ti.com/lit/an/slaa013/slaa013.pdf)
Brigitte VACHON
December 7, 2015
186
7.8.1
Brigitte VACHON
December 7, 2015
187
Note that each resistor is different (R, 2R, 4R, ...). For converting a 10-bit word, you need 10 resistors with a range of values from R to 512R. The different-valued resistors have different temperature
coefficients such that a calibration of this DAC is very temperature-dependent.
In practice, this approach only useful for converting binary words with a small number of bits, for
higher resolution use another technique called a R/2R ladder DAC.
7.8.2
Pros:
Only need two resistor values: R and 2R (and RF for the feedback)
Can easily add steps to the resistor ladder to improve resolution
Brigitte VACHON
December 7, 2015
188
Current drawn through any of the switches is always the same, regardless of whether each switch is
up/down. This is because in both cases the current flows directly to ground.
Total current supplied by Vref is therefore also constant.
Brigitte VACHON
December 7, 2015
189
Diagram showing how the R/2R circuit can be reduced to deduce the fraction of the total current
flowing through each switch.
Exercise: Given R=10 k, RF = 20 k and Vref = +5 V, if switches S3, S2, S1 and S0 are put in
position to represent the binary number 0101 (510), calculate what will be the output voltage?
Brigitte VACHON
December 7, 2015
190
Chapter 8
Memory Devices
Memory devices allow you to store binary data on a temporary or permanent basis.
Different storage technologies exist: Semiconductor, magnetic, optical, paper. [Give examples
for each category]
We will only discuss here semiconductor memory devices.
191
8.1
Memory Organization
Memory devices are typically arranged in an array of individual memory cells, each representing
one bit (0 or 1).
Each bit can be uniquely identified by specifying a row and column number.
(source: http://www.zeepedia.com/read.php?memory_organization_capacity_density_signals_and_basic_operations_read_write_address_data_
signals_digital_logic_design&b=9&c=39)
Brigitte VACHON
December 7, 2015
192
December 7, 2015
193
Memory capacity identified by number of units it can store times the unit size.
Example:
A memory array of 8 rows x 8 columns can store 8 different Bytes = 8 Byte memory.
Rows are selected using a 3-bit address.
A memory array of 64 rows x 1 column can store 64 different bits = 64-bit memory.
Rows are selected using a 6-bit address.
A memory array of 1024 rows x 8 columns can store 1024 different Bytes = 1 kilo-Byte memory
(1 KB).
Rows are selected using a 10-bit address.
Brigitte VACHON
December 7, 2015
194
(source: http://www.zeepedia.com/read.php?memory_organization_capacity_density_signals_and_basic_operations_read_write_address_data_
signals_digital_logic_design&b=9&c=39)
Brigitte VACHON
December 7, 2015
195
The are two different families of semiconductor memories: Read-only memory (ROM) and Random
Access memory (RAM)
Semiconductor
Memory devices
8.2
Non-volatile memories
Read-Only Memory
(ROM)
Volatile memories
Random Access Memory
(RAM)
MROM
PROM
EPROM
EEPROM
Flash
Etc...
SRAM
DRAM
... and lots of other flavours..
ROM devices are a type of memory called non-volatile which means that they do not lose stored
data when power is removed from the device. ROMs are used to store data permanently (or
semipermanently). Typically used to store data used repeatedly in a system (e.g. initialization,
look-up tables, conversions).
Typically used for read-only operations and not to write to after the initial programming. There
exists however ROM-like memory devices on which you can erase and write (relatively slowly)
new data to memory (ex. Flash).
Brigitte VACHON
December 7, 2015
196
8.2.1
MROM
Somewhat expensive to produce ROM (initial mask), so only cost effective when producting a large
number of devices.
Example: Commonly used in computer to store system operating instructions and data to decode
keyboard instructions.
Brigitte VACHON
December 7, 2015
197
8.2.2
Main difference with a MROM is that PROM come from the manufacturer unprogrammed and are
custom programmed by (typically) engineers to meet the users needs. Just like MROM, once a
PROM is programmed, the memory values cannot be changed.
Programming a PROM is done by burning or leaving intact a memory link.
Brigitte VACHON
December 7, 2015
198
8.2.3
Memory cells made of specialized MOSFET transistors with an additional floating gate (electrically
insulated from control gate and drain-to-source channel). Floating gate can store electrons, thereby
affecting normal operation of gate.
Programming one transistor: Apply high-voltage
pulse (e.g. 12 V) to the control gate with
drain-source biased. This sets up an electric field.
Electrons are attracted to the surface of channel
and some with enough energy tunnel through
insulating layer to get trapped in floating gate.
Reading one transistor: Apply voltage on the
control gate.
Electrons on floating gate, if
present, will inhibit voltage from turning on
the transistor. When no (little) current flows, this
corresponds to a 0.
(source: P.Scherz and S.Monk, Practical Electronics for Inventors)
December 7, 2015
199
8.2.4
Flash memory
ROM-like device that can nevertheless be erased and re-programmed (although it is a relatively slow
process).
Also uses transistors with floating gate, similar to EPROM.
Instead of using UV light to erase the memory (remove electrons on floating gate), it uses a large
voltage in opposite polarity than what is used to program the memory cell.
Example: Used for mass storage devices, digital camera, music players, cell phones, etc.
Brigitte VACHON
December 7, 2015
200
8.3
These devices are a type of memory called volatile which means that they lose stored data when
power is removed from the device.
Typically used for short term data storage.
There are a large number of different types of RAM, but let me just mention two of the basic types:
Static RAM (SRAM) and Dynamic RAM (DRAM).
DRAM
SRAM
Brigitte VACHON
December 7, 2015
201
SRAM: Typically found in Application Specific IC (ASIC), Pocket calculators, cache memory for
computer CPU.
DRAM: Computer memory modules (ex. DIMMs)
Brigitte VACHON
December 7, 2015
202
Chapter 9
(If you are interested) IC, ASIC, FPGA,
microprocessor...
[Note on the final exam]
An integrated circuit (IC) consists in electronic circuits built on a small silicon plate (chip).
An Application Specific Integrated Circuit (ASIC) is an integrated circuit customized for a particular use as opposed to a general purpose.
9.1
ASIC
The following slides explain the different ASIC production steps as described on the Nobel website.
(source: http://www.nobelprize.org/educational/physics/integrated_circuit/history/)
203
(source: http://www.icsense.com/icdesign-chipdesign)
Brigitte VACHON
9.1. ASIC
(source: http://www.scienceasart.org/layout-integrated-circuit)
December 7, 2015
204
(source: Wikipedia)
Brigitte VACHON
9.1. ASIC
(source: http://www.tf.uni-kiel.de/matwis/amat/elmat_en/kap_6/illustr/i6_1_2.html)
December 7, 2015
205
(source: http://apcmag.com/picture-gallery-how-a-chip-is-made.htm)
9.1. ASIC
(source:
http://www.newtechpress.net/2014/04/01/economics-and-vengeance-wreak-havoc-with-wafer-market/)
Brigitte VACHON
December 7, 2015
206
Brigitte VACHON
December 7, 2015
9.1. ASIC
207
Brigitte VACHON
December 7, 2015
9.1. ASIC
208
Brigitte VACHON
December 7, 2015
9.1. ASIC
209
(source: Wikipedia)
Brigitte VACHON
9.1. ASIC
(source: http://www.rulestoolsfools.com/?p=803)
December 7, 2015
210
9.1. ASIC
(source: http://www.radio-electronics.com/info/data/smt/smd-bga-ball-grid-array-package.php)
(source: https://learn.sparkfun.com/tutorials/integrated-circuits/ic-packages)
December 7, 2015
211
9.2
9.2. FPGA
FPGA
A Field Programmable Gate Array (FPGA) is a complex integrated circuit containing a very large
number of logic blocks that can be configured after manufacturing.
(source: Wikipedia)
(source: Wikipedia)
Brigitte VACHON
December 7, 2015
212
9.2. FPGA
(source: http://www.ni.com/white-paper/6983/en/)
Brigitte VACHON
December 7, 2015
213
9.2. FPGA
Place-and-route on an FPGA.
You are in control of designing the circuit by configuring the FPGA connections as you wish.
214
9.2. FPGA
FPGAs are configured using a Hardware Description Language (HDL). The FPGA configuration is
called firmware (as opposed to hardware or software).
December 7, 2015
215
9.3
9.3. MICROPROCESSOR
Microprocessor
An Arduino Uno board with a microprocessor mounted on it (long black rectangular IC)
(source: http://arduino.cc/en/Main/arduinoBoardUno)
Brigitte VACHON
December 7, 2015
216
9.3. MICROPROCESSOR
Brigitte VACHON
December 7, 2015
217
Appendix A
Grounding issues
Proper grounding is a major issue in the design of complex and large-scale electronic systems.
The word ground is typically used to define a reference voltage level of zero potential.
Earth ground: Connection terminated at a rod (or pipe) driven into the earth.
Frame or Chassis
218
Simple example of a common floating return (current return line not necessarily tied to earth
ground).
Brigitte VACHON
December 7, 2015
219
A.1
Ground loops
Noise in a circuit is often due to poor grounding design which can lead to ground loops. Ground
loops occur when there is more than one return path for the ground current.
Unwanted current flow due to potential difference between ground points caused by non-zero
impedance in ground lines.
Second return path can be thought of as acting like a big antenna which can pick up EM fields
inducing a loop current.
(source: http://www.prosoundweb.com/article/preventing_hum_and_rfi_in_your_studio/P2/)
A typical sign that a ground loop exists in your circuit, or that a ground is missing, is the presence
of induced power line (60 Hz) noise in the circuit.
To avoid ground loops, use a single-point ground.
Brigitte VACHON
December 7, 2015
220
A.2. AC GROUND
Electronic systems containing both analog and digital circuits should have the analog and digital
grounds kept separate, and eventually connected together. This will minimize noise from the digital
circuit sometimes generating current spikes as the circuit changes state. Analog circuit can also
generate current spikes which could then affect the digital circuitry if the grounds are not kept
separate.
A.2
AC ground
A constant DC point is an AC ground and is equivalent to an ordinary ground for a signal voltage.
Brigitte VACHON
December 7, 2015
221
Appendix B
Semiconductor technology
The following content is an extract from the book: P.Scherz and S.Monk, Practical Electronics for
Inventors, McGraw-Hill, 3rd Edition.
222
Appendix C
Current in semiconductors
The following content is an extract from the book: Thomas L. Floyd, Electronic Devices (Conventional Current Version) Floyd, Basic Electronics for scientists and engineers, Prentice Hall,
2012.
230
C URRENT
Si
Si
Si
Si
Si
Si
Si
Si
Si
Si
C URRENT
Si
Si
Si
Si
1.
2.
3.
4.
5.
6.
7.
8.
9.
10.
11
FIGURE 1 1 1
Si
Si
S EMICONDUCTORS
IN
SECTION 12
CHECKUP
13
Si
Si
IN
Si
Si
S EMICONDUCTORS
The way a material conducts electrical current is important in understanding how
electronic devices operate. You cant really understand the operation of a device such
as a diode or transistor without knowing something about current in semiconductors.
After completing this section, you should be able to
As you have learned, the electrons of an atom can exist only within prescribed energy
bands. Each shell around the nucleus corresponds to a certain energy band and is separated
from adjacent shells by band gaps, in which no electrons can exist. Figure 112 shows the
energy band diagram for an unexcited (no external energy such as heat) atom in a pure silicon crystal. This condition occurs only at a temperature of absolute 0 Kelvin.
12
"
I NTRODUCTION
TO
E LECTRONICS
F IG U RE 112
Energy
Conduction band
Band gap
Valence band (shell 3)
Nucleus
F IG U RE 113
Energy
Heat
energy
Band gap
Valence
band
+4
Free
electron
Conduction
band
Free
electron
Hole
Heat
energy
Hole
+4
Electron-hole pair
When an electron jumps to the conduction band, a vacancy is left in the valence band
within the crystal. This vacancy is called a hole. For every electron raised to the conduction band by external energy, there is one hole left in the valence band, creating what is
called an electron-hole pair. Recombination occurs when a conduction-band electron
loses energy and falls back into a hole in the valence band.
To summarize, a piece of intrinsic silicon at room temperature has, at any instant, a
number of conduction-band (free) electrons that are unattached to any atom and are essentially drifting randomly throughout the material. There is also an equal number of holes in
the valence band created when these electrons jump into the conduction band. This is illustrated in Figure 114.
C URRENT
Si
Si
Si
Si
Si
Si
Si
Si
Si
Si
Si
Si
Si
Si
Si
Generation of an
electron-hole pair
Si
Si
Si
Si
IN
S EMICONDUCTORS
13
FIGUR E 1 14
Recombination of
an electron with
a hole
Si
Heat energy
Si Si Si Si Si
Si Si Si Si Si
Si Si Si Si Si
Another type of current occurs in the valence band, where the holes created by the free
electrons exist. Electrons remaining in the valence band are still attached to their atoms
and are not free to move randomly in the crystal structure as are the free electrons.
However, a valence electron can move into a nearby hole with little change in its energy
level, thus leaving another hole where it came from. Effectively the hole has moved from
one place to another in the crystal structure, as illustrated in Figure 116. Although current
in the valence band is produced by valence electrons, it is called hole current to distinguish
it from electron current in the conduction band.
As you have seen, conduction in semiconductors is considered to be either the movement of free electrons in the conduction band or the movement of holes in the valence
band, which is actually the movement of valence electrons to nearby atoms, creating hole
current in the opposite direction.
It is interesting to contrast the two types of charge movement in a semiconductor with
the charge movement in a metallic conductor, such as copper. Copper atoms form a different type of crystal in which the atoms are not covalently bonded to each other but consist
of a sea of positive ion cores, which are atoms stripped of their valence electrons. The
valence electrons are attracted to the positive ions, keeping the positive ions together and
forming the metallic bond. The valence electrons do not belong to a given atom, but to the
crystal as a whole. Since the valence electrons in copper are free to move, the application
of a voltage results in current. There is only one type of currentthe movement of free
electronsbecause there are no holes in the metallic crystal structure.
FIGUR E 1 15
14
"
I NTRODUCTION
TO
E LECTRONICS
F IG U RE 116
Si
1 A free electron
leaves hole in
valence shell.
2 A valence electron moves
into 1st hole and leaves
a 2nd hole.
Si
Si
When a valence electron moves left to right to fill a hole while leaving another hole behind, the hole
has effectively moved from right to left. Gray arrows indicate effective movement of a hole.
SECTION 13
CHECKUP
14
N-T Y P E
1.
2.
3.
4.
AND
P-T Y PE S EMICONDUCTORS
Semiconductive materials do not conduct current well and are of limited value in their
intrinsic state. This is because of the limited number of free electrons in the conduction
band and holes in the valence band. Intrinsic silicon (or germanium) must be modified by
increasing the number of free electrons or holes to increase its conductivity and make it
useful in electronic devices. This is done by adding impurities to the intrinsic material.
Two types of extrinsic (impure) semiconductive materials, n-type and p-type, are the key
building blocks for most types of electronic devices.
After completing this section, you should be able to
Since semiconductors are generally poor conductors, their conductivity can be drastically increased by the controlled addition of impurities to the intrinsic (pure) semiconductive
material. This process, called doping, increases the number of current carriers (electrons
or holes). The two categories of impurities are n-type and p-type.
N-Type Semiconductor
To increase the number of conduction-band electrons in intrinsic silicon, pentavalent impurity atoms are added. These are atoms with five valence electrons such as arsenic (As),
phosphorus (P), bismuth (Bi), and antimony (Sb).
N-T YPE
AND
15
As illustrated in Figure 117, each pentavalent atom (antimony, in this case) forms covalent bonds with four adjacent silicon atoms. Four of the antimony atoms valence electrons are used to form the covalent bonds with silicon atoms, leaving one extra electron.
This extra electron becomes a conduction electron because it is not involved in bonding.
Because the pentavalent atom gives up an electron, it is often called a donor atom. The
number of conduction electrons can be carefully controlled by the number of impurity
atoms added to the silicon. A conduction electron created by this doping process does not
leave a hole in the valence band because it is in excess of the number required to fill the
valence band.
Si
Si
Sb
Si
Si
Majority and Minority Carriers Since most of the current carriers are electrons, silicon
(or germanium) doped with pentavalent atoms is an n-type semiconductor (the n stands for
the negative charge on an electron). The electrons are called the majority carriers in
n-type material. Although the majority of current carriers in n-type material are electrons,
there are also a few holes that are created when electron-hole pairs are thermally generated. These holes are not produced by the addition of the pentavalent impurity atoms.
Holes in an n-type material are called minority carriers.
P-Type Semiconductor
To increase the number of holes in intrinsic silicon, trivalent impurity atoms are added.
These are atoms with three valence electrons such as boron (B), indium (In), and gallium
(Ga). As illustrated in Figure 118, each trivalent atom (boron, in this case) forms covalent
bonds with four adjacent silicon atoms. All three of the boron atoms valence electrons are
used in the covalent bonds; and, since four electrons are required, a hole results when each
trivalent atom is added. Because the trivalent atom can take an electron, it is often referred
to as an acceptor atom. The number of holes can be carefully controlled by the number of
trivalent impurity atoms added to the silicon. A hole created by this doping process is not
accompanied by a conduction (free) electron.
Majority and Minority Carriers Since most of the current carriers are holes, silicon (or
germanium) doped with trivalent atoms is called a p-type semiconductor. The holes are the
majority carriers in p-type material. Although the majority of current carriers in p-type
material are holes, there are also a few conduction-band electrons that are created when
electron-hole pairs are thermally generated. These conduction-band electrons are not produced by the addition of the trivalent impurity atoms. Conduction-band electrons in p-type
material are the minority carriers.
FIGUR E 1 17
16
I NTRODUCTION
TO
E LECTRONICS
"
FIG UR E 1 1 8
Si
Hole from B atom
Si
Si
Si
SECTION 14
CHECKUP
15
1.
2.
3.
4.
5.
6.
7.
8.
9.
10.
Define doping.
What is the difference between a pentavalent atom and a trivalent atom?
What are other names for the pentavalent and trivalent atoms?
How is an n-type semiconductor formed?
How is a p-type semiconductor formed?
What is the majority carrier in an n-type semiconductor?
What is the majority carrier in a p-type semiconductor?
By what process are the majority carriers produced?
By what process are the minority carriers produced?
What is the difference between intrinsic and extrinsic semiconductors?
T HE PN J UNCTION
When you take a block of silicon and dope part of it with a trivalent impurity and the other
part with a pentavalent impurity, a boundary called the pn junction is formed between the
resulting p-type and n-type portions. The pn junction is the basis for diodes, certain transistors, solar cells, and other devices, as you will learn later.
After completing this section, you should be able to
A p-type material consists of silicon atoms and trivalent impurity atoms such as boron.
The boron atom adds a hole when it bonds with the silicon atoms. However, since the number of protons and the number of electrons are equal throughout the material, there is no
net charge in the material and so it is neutral.
T HE PN J UN C T I O N
17
An n-type silicon material consists of silicon atoms and pentavalent impurity atoms such as
antimony. As you have seen, an impurity atom releases an electron when it bonds with four
silicon atoms. Since there is still an equal number of protons and electrons (including the free
electrons) throughout the material, there is no net charge in the material and so it is neutral.
If a piece of intrinsic silicon is doped so that part is n-type and the other part is p-type,
a pn junction forms at the boundary between the two regions and a diode is created, as
indicated in Figure 119(a). The p region has many holes (majority carriers) from the
impurity atoms and only a few thermally generated free electrons (minority carriers). The
n region has many free electrons (majority carriers) from the impurity atoms and only a
few thermally generated holes (minority carriers).
pn junction
p region
n region
Depletion region
p region
n region
Barrier
potential
(a) The basic silicon structure at the instant of junction formation
showing only the majority and minority carriers. Free electrons
in the n region near the pn junction begin to diffuse across the
junction and fall into holes near the junction in the p region.
(b) For every electron that diffuses across the junction and
combines with a hole, a positive charge is left in the n region
and a negative charge is created in the p region, forming a
barrier potential. This action continues until the voltage of
the barrier repels further diffusion. The blue arrows between
the positive and negative charges in the depletion region
represent the electric field.
F IGU RE 119
Formation of the depletion region. The width of the depletion region is exaggerated for illustration
purposes.
HISTORY NOTE
After the invention of the light bulb,
Edison continued to experiment and
in 1883 found that he could detect
electrons flowing through the
vacuum from the lighted filament to
a metal plate mounted inside the
bulb. This discovery became known
as the Edison effect.
An English physicist, John
Fleming, took up where Edison left
off and found that the Edison effect
could also be used to detect radio
waves and convert them to electrical
signals. He went on to develop a
two-element vacuum tube called the
Fleming valve, later known as the
diode. Modern pn junction devices
are an outgrowth of this.
18
I NTRODUCTION
TO
E LECTRONICS
HISTORY NOTE
Russell Ohl, working at Bell Labs
in 1940, stumbled on the
semiconductor pn junction. Ohl
was working with a silicon sample
that had an accidental crack down
its middle. He was using an
ohmmeter to test the electrical
resistance of the sample when he
noted that when the sample was
exposed to light, the current that
flowed between the two sides of
the crack made a significant jump.
This discovery was fundamental to
the work of the team that invented
the transistor in 1947.
electrons across the junction. This occurs as follows. As electrons continue to diffuse
across the junction, more and more positive and negative charges are created near the junction as the depletion region is formed. A point is reached where the total negative charge in
the depletion region repels any further diffusion of electrons (negatively charged particles)
into the p region (like charges repel) and the diffusion stops. In other words, the depletion
region acts as a barrier to the further movement of electrons across the junction.
Barrier Potential Any time there is a positive charge and a negative charge near each
other, there is a force acting on the charges as described by Coulombs law. In the depletion region there are many positive charges and many negative charges on opposite sides of the pn
junction. The forces between the opposite charges form an electric field, as illustrated in
Figure 119(b) by the blue arrows between the positive charges and the negative charges. This
electric field is a barrier to the free electrons in the n region, and energy must be expended to
move an electron through the electric field. That is, external energy must be applied to get the
electrons to move across the barrier of the electric field in the depletion region.
The potential difference of the electric field across the depletion region is the amount of
voltage required to move electrons through the electric field. This potential difference is
called the barrier potential and is expressed in volts. Stated another way, a certain
amount of voltage equal to the barrier potential and with the proper polarity must be applied across a pn junction before electrons will begin to flow across the junction. You will
learn more about this when we discuss biasing in Chapter 2.
The barrier potential of a pn junction depends on several factors, including the type of
semiconductive material, the amount of doping, and the temperature. The typical barrier
potential is approximately 0.7 V for silicon and 0.3 V for germanium at 25C. Because germanium devices are not widely used, silicon will be used throughout the rest of the book.
S UMMARY
Energy
Minority carriers
Majority carriers
Conduction
band
Valence
band
Valence
band
Majority carriers
p region
Minority carriers
pn junction
n region
19
Energy
Conduction
band
p region
pn junction
and depletion
region
n region
(b) At equilibrium
F IGU RE 120
Energy diagrams illustrating the formation of the pn junction and depletion region.
SECTION 15
CHECKUP
1.
2.
3.
4.
5.
6.
What is a pn junction?
Explain diffusion.
Describe the depletion region.
Explain what the barrier potential is and how it is created.
What is the typical value of the barrier potential for a silicon diode?
What is the typical value of the barrier potential for a germanium diode?
SUMMARY
Section 11
According to the classical Bohr model, the atom is viewed as having a planetary-type structure
in the Bohr model. The electrons can be waves or particles and precise location at any time is
uncertain.
The nucleus of an atom consists of protons and neutrons. The protons have a positive charge and
the neutrons are uncharged. The number of protons is the atomic number of the atom.
Electrons have a negative charge and orbit around the nucleus at distances that depend on their
energy level. An atom has discrete bands of energy called shells in which the electrons orbit.
Atomic structure allows a certain maximum number of electrons in each shell. In their natural
state, all atoms are neutral because they have an equal number of protons and electrons.
The outermost shell or band of an atom is called the valence band, and electrons that orbit in
this band are called valence electrons. These electrons have the highest energy of all those in the
atom. If a valence electron acquires enough energy from an outside source such as heat, it can
jump out of the valence band and break away from its atom.
Section 12
Insulating materials have very few free electrons and do not conduct current at all under normal
circumstances.
Materials that are conductors have a large number of free electrons and conduct current very well.
Semiconductive materials fall in between conductors and insulators in their ability to conduct
current.
Semiconductor atoms have four valence electrons. Silicon is the most widely used semiconduc-
tive material.
Appendix D
Band gap theory
The following content is an extract from the book: D.Eggleston, Basic Electronics for scientists and
engineers, Cambridge University Press, 2011.
240
Appendix E
Stability of amplifiers using feedback
The following is a re-print of H.S. Blacks original paper describing the invention of negative feedback
as a mean to stabilize amplifiers and also provide larger operational bandwidth.
H.S. Black, Stabilized feed-back amplifiers, Proc. IEEE, vol. 72, no. 6, pp. 716-722, June 1984.
253
BY
H. S.BLACK
MEMBER ALEE
716
Stabilized feed-back possesses other advantages indreduced delay and delay distortion,reduced noise disturbance
from the power supply circuits and various other feature best
appreciated by practical designersof amplifiers.
It is f a r from a simple proposition to employ feed-back in
this way because of the very special control requiredof phase
shifts in theamplifier
and feed-backcircuits, not only
throughout the useful frequency band but for a wide range of
frequencies above and below this band. Unless these relations
are maintained, singing will o c c u , usually at frequencies
outside the useful range. Once having achieved adesign,
however, in which proper phase relations are secured, experience has demonstrated that the performance obtained is perfectly reliable.
The carrier-incable system dealt with in a recent Institute
paper (Carrier in Cables by A. B. Clark and B. W.Kendall.
A.I.EE 'TUNS., k c . 1933, p. 10%) involves many amplifiers
in tandem with many telephone channels passing through each
amplifier and constitutes, therefore, an ideal field for application of this feed-back principle. A field trial of this system
was made at Morristown, New Jersey, in which 70 ofthese
amplifiers were operated in tandem. The results of this trial
werehighly satisfactoryanddemonstrated conclusively the
correctness of the theory and the practicabilityof its commercial application.
CIRCUIT
ARRANGEMENT
In theamplifier of Fig. 1, aportion of the output is
returned to the input to produce feed-back action. The upper
branch, called the p circuit, is represented as containing active
elements such as an amplifier while the lower branch, called
the 8-circuit, is shownas a passive network, The
way a voltage
is modified after once traversing each circuitis denoted p and
/3, respectively, and the product, pp, represents how a voltage
ismodifiedaftermakingasingle
journey aroundamplifier
and feed-back circuits. Both p and 8 are complex quantities,
functions of frequency, and in the generalized concept either
or both may be greater or lessin absolute value than unity: (p
is not used in the sense that it is used sometimes, namely, to
denote the amplification constant of a particular tube, but as
the complex ratio of the output to the input voltage of the
amplifier circuit).
Fig. 2 shows an arrangement convenient forsome purposes
where,byusingbalancedbridges
in the input and output
circuits, interaction between the circuits that connect to the
input and output is avoided. Thereby feed-back action and
amplifier impedances are made independentof the properties
of circuits connectedto the amplifier.
GENERAL
EQUATION
In Fig. 1, 8 is zero without feed-back and a signal voltage,
e,, applied to the input of the p-circuit produces an output
voltage. This is made up of what is wanted,the amplified
CHANGE IN
GAINDUE TO FEED-BACK
From eq 2,theamplificationwithfeed-back
equals the
amplificationwithoutfeed-backdivided
by (1 - a/?). The
is tochange
effect of addmgfeed-back,therefore,usually
the gain of the amplifier and this change will be expressed as
From pB
1 - PB
= Ip&@
10-GCFo= 1- 2lpplcos@
(4)
ljbB12
which is theequationfora
familyof concentric circles of
radius 10-GcF/lo aboutthepoint
1, 0. Fig. 3 is apolar
+ No + Do = peo + n + d ( E , )
(1)
p/3
e.
71 7
Fib3
C.
D. IGPI = I
E CP 90". Improvement in gain stability corrrsponds to twice db re
F.
G.
H.
I.
J.
STABILITY
From eq 2, -is the amplified signal with feed-back
1 - PB
and - therefore, is an index of the amplification. It is
1 - PB'
718
IfpB~l,itisseenthat~orthe~-circuitisstab~by
an amount corresponding to the reduction in amplification
and the effect of introducing a gain or loss in the p-circuit is
to produce no material change in the over-all amplificationof
the system; the stability of amplification as affected by B or
theB-circuit is neitherappreciablyimprovednordegraded
since increasing the loss in the &circuit raises the gain of the
amplifier by anamount almost corresponding to theloss
introduced and viceversa. If both p and B are varied and the
variations sufficiently small, the effect is the same as if each
were changed separately and the two results then combined.
In certain practical applicationsof amplifiers it is the change
in gain or ammeter or voltmeter reading at the output that is a
measure of the stability rather than the complex ratio previously treated. The conditions surroundmg gain stability may
be examined by considering the absolute value of Ap. This is
shorn as fonows:
Let (db)represent the gain in decibels
to A,.
Then
o
c
Acuriousfacttobenotedfromeqllisthatitispossibleto
choose a value of pB (namely, lpBl= set@) so thatthe
numerator of the nght hand term vanishes. This meam that
the gain stabilityis perfect, assuming differential variations in
Ipl. Referring to Figs. 3 and 4, contour C is thelocus of
IpBI = sec @ and it includes all amplifierswhose gain is
the output. Denoting this amplification by u and uo, respectively, it can be shown that the relation between the 2 noise
ratios is (uo/u)(l - p8). This is called the noise index.
If noise is introduced in the power supply circuitsof the last
tube, ao/a = 1 and the noise index is (1 - p8). As a result of
this relation less expensive power supply filters are possiblein
the last stage.
MODULATION
8-Circuit Equalization. Referring to eq 2, the output voltage E approaches - eo/8 as 1 - p 8 = -a8 and equals it in
absolute value if cos@ = -where p 8 = IpBp3. Under
21PBI
these circumstances increasing the loss in the 8-circuit 1 db
raisesthegain
of theamplifier 1 db, and vice-versa, thus
giving any gain-frequency characteristic for which a like lossfrequency characteristic can be inserted in the 8-circuit. This
procedure has been termed 8-circuit equalization. It possesses
other advantages and properties which are beyond the xope
of this paper.
"
Do
it is to be concluded
1 - PB'
that modulationwithfeed-back will bereduceddecibelfor
decibel as the effect of feed-back action causes an arbitrary db
reduction in the gain of the amplifier; i.e., when the feed-back
is negative. With positive feed-back the opposite
is true, the
modulation being increased by an amount corresponding to
the increase in amplification.
If modulation in the 8-circuit is a factor, it can be shown
that usually in its effect on the output the modulation levelat
theoutputduetononlinearity
of the/3-circuitisapproxip8
multiplied by the modulation generated
in the
mately 1 - P8
/?-circuit acting alone and without feed-back.
ADDITIONAL,
EFFECTS
BLACK:
In the
A , 0 is theover-allphaseshift
expression A , = 1
-Pal
withfeed-back,and
it can beshown that the phase shift
through the amplifer with feed-back may be made to apprwch
the phase shift through the 8-circuit p h 180 deg. The effect of
phase shift in the 8-circuit is not reduced correspondingly. It
will be recalled that in reducing the change in phase shift with
is the slopeof the phase shift
frequency, envelope delay, which
with respect to the angular velocity, o = 2 vf, also is reduced.
The delay distortion likewise is reduced because a measure of
delay distortion at a particular frequency is thedifference
betweentheenvelopedelay at that frequency and the least
envelope delay in the band.
Phase Shift, Envelope Delay,DelayDistortion.
AVOIDSINGING
Having considered the theory up to this point, experimental evidence was readilyacquired to demonstrate that p8
might assume large values, 10 to lO,OOO, provided CP was not
at the same time zero. However, one noticeable feature about
thefield of p 8 (Figs. 3 and 4) is that it implies that even
though the phase shift is zero and the absolute value of p 8
exceeds unity, self-oscillations or singing will not result. This
may or may not be true. Whenfirst thinking about this matter
it was suspected that owing to practical nonlinearity, singing
would result whenever the gain around the closed loop equaled
or exceeded the loss and simultaneously the phase shift was
zero; i.e., p 8 = Ip8I + J o 2 1. Results of experiments, however,seemed to indicate somethingmorewasinvolvedand
these matters were described to H. Nyquist who developed a
more general criterion for freedom from instability applicable
to an amplifier having linear positive constants. (For a complete description of the criterion for stability and instability
andexactlywhatismeantbyenclosingthepoint
(l,O),
referenceshouldbemade
to RegenerationTheory, by H.
Nyquist, Bell System Technical J m l , v. XI, July 1932, p.
126-47.)
719
720
FRCWENCV
OB C M f f i E
-.2
IN
CAIN
RILOCYCLCS
WITHOUT CCEDBACK
*P
r4
Fig. 11. Phase shift delay, and delay distortion with and without
feed-back for
a single tube voice frequency amplifier
Fig. 9. Repesentative gain stability of a single amplifier as determined by measuring @ feed-amplifiers in tandem at Morristom, New Jersey
The upper figure shows the absolute value of the stability index. It can be
seen that between 20 and 25 kc the improvement in stability is more than
1,OOO to 1 yet the reduction in gain was less than 35 db
721
mmcvcmmaaum
equnlizerintheB-cimlit
CONCLUSION
The feed-back amplifier dealt within this paper was developed primarily with requirements in mind for a cable carrier
telephone system, involving many amplifiers in tandem with
many telephone channels passing
through each amplifier. Most
of the examples of feed-back amplifier performance naturally
have been drawnfromamplifiersdesignedfor
this field of
operation. In this field,vacuum tube amplifiersnormally
722
possessinggoodcharacteristicswithrespect
to stability,and
freedomfrom distortion are made to possess superlatively
good characteristics by application of the feed-back principle.
However, certain types of amplifiers, in which economy has
been secured by sacrificing performance characteristics, particularly as regards distortion, can be made to possess improved characteristics by the application of feed-back. Discussion of these amplifiers is beyond the scope of this paper.