Energy Heat and Processor Performance by nikeborome

VIEWS: 2,748 PAGES: 17

									                                                                Lecture 1: Dr. Tom Butler

Energy, Heat and Computer Performance
Table 1 Processor Transistor Counts
                                         Date of
  Processor     Transistor count                        Manufacturer
Intel 4004     2300                1971              Intel
Intel 8008     2500                1972              Intel
Intel 8080     4500                1974              Intel
Intel 8088     29 000              1978              Intel
Intel 80286    134 000             1982              Intel
Intel 80386    275 000             1985              Intel
Intel 80486    1 200 000           1989              Intel
Pentium        3 100 000           1993              Intel
AMD K5         4 300 000           1996              AMD
Pentium II     7 500 000           1997              Intel
AMD K6         8 800 000           1997              AMD
Pentium III    9 500 000           1999              Intel
AMD K6-III     21 300 000          1999              AMD
AMD K7         22 000 000          1999              AMD
Pentium 4      42 000 000          2000              Intel
Itanium        25 000 000          2001              Intel
Barton         54 300 000          2003              AMD
AMD K8         105 900 000         2003              AMD
Itanium 2      220 000 000         2003              Intel
Itanium 2      592 000 000         2004              Intel
Athlon 64 X2   243 000 000         2006              AMD
Core 2 Duo     291 000 000         2006              Intel
Core 2 Quad    582 000 000         2006              Intel
Phenom         463 000 000         2006              AMD
Phenom II      758 000 000         2008              AMD
i7 Core        731 000 000         2008              Intel

                                                                       Lecture 1: Dr. Tom Butler

   The computer industry has obsessed with increasing the performance of processors and
   addressing the associated increase in power consumption and temperatures. They have
   not always been successful as the Intel Pentium 4 Prescott became known as the Presshot
   due to its high operating temperature: worse still, Intel Pentium 4 955 Extreme Edition
   barely worked, and when it did, the temperature readings in the BIOS were in the high
   60s to 70s Fahrenheit, critically close to thermal limits of a pro cessor’s operating
   Heat generation and energy consumption are the downside of more and more powerful
   processors: powerful in terms of their speed of operation, which is related their clock
   speed (e.g. 2 GHz, 3 GHz etc.) and the number of functions they can accomplish
   simultaneously, which is basically a function of the number of transistors they have on
   board. The above table illustrates the typical number of transistors each type of processor
   going back to 1971.
   It is significant, however, that the size of the die, i.e. the container for the central
   processing unit (CPU) chip have not increased in size, if anything they have gotten
   smaller. You can see a typical Pentium 4 CPU in the next figure.

                               L1         Unified
                      Core                            L3        Die                   Instruction
Pentium Model                  Data       L2                             Process
                      Freq.                           Cache     Size                  Sets
                               Cache      Cache

                      < 2.0    8kB 4-                           217                   MMX, SSE,
Willamette                                256 kB      n/a                180 nm
                      GHz      way                              mm2                   SSE2

                      < 3.4    8kB 4-                           146                   MMX, SSE,
Northwood                                 512 kB      n/a                130 nm
                      GHz      way                              mm2                   SSE2

                      3.2 -
ExtremeEdition                 8kB 4-                           237                   MMX, SSE,
                      3.4                 512 kB      2 MB               130 nm
(Gallatin)                     way                              mm2                   SSE2

                      2.8 -                                                           MMX, SSE,
                               16kB       1024                  112
Prescott              3.4                             n/a                90 nm        SSE2,
                               8-way      kB                    mm2
                      GHz                                                             SSE3

   Table 2 Comparing Pentiums for Process and Die Size

                                                                   Lecture 1: Dr. Tom Butler

Figure 1 Intel Pentium 4 Die

Table 3 High-End/Mainstream: Phenom II X4 (Deneb Quad Core)

 Model      Clock     Cores    Socket /    TDP     L2      L3      Release       Hyper
            Speed              Memory             Cache   Cache                Transport

Phenom      3.4 GHz   4       AM3/AM2+     140W   4x      6MB     08/13/2009   2.0 GHz
II X4 965                     DDR3, DDR2          512KB

Phenom      3.2 GHz   4       AM3/AM2+     125W   4x      6MB     04/23/2009   2.0 GHz
II X4 955                     DDR3, DDR2          512KB

Phenom      3.0 GHz   4       AM3/AM2+     125W   4x      6MB     04/23/2009   2.0 GHz
II X4 945                     DDR3, DDR2   95W    512KB           06/12/2009

Phenom      3.0 GHz   4       AM2+ DDR2    125W   4x      6MB     01/08/2009   1.8 GHz
II X4 940                                         512KB

Phenom      2.8 GHz   4       AM2+ DDR2    125W   4x      6MB     01/08/2009   1.8 GHz
II X4 920                                         512KB

Phenom      2.6 GHz   4       AM3/AM2+     95W    4x      6MB     02/09/2009   2.0 GHz
II X4 910                     DDR3, DDR2          512KB

Phenom      2.5 GHz   4       AM3/AM2+     65W    4x      6MB     06/02/2009   2.0 GHz
II X4                         DDR3, DDR2          512KB

Phenom      2.4 GHz   4       AM3/AM2+     65W    4x      6MB     06/02/2009   2.0 GHz
II X4                         DDR3, DDR2          512KB

Phenom      2.8 GHz   4       AM3/AM2+     95W    4x      6MB     02/09/2009   2.0 GHz
II X4                         DDR3, DDR2          512KB

Phenom      2.6 GHz   4       AM3/AM2+     95W    4x      6MB     02/09/2009   2.0 GHz
II X4 810                     DDR3, DDR2          512KB

Phenom      2.5 GHz   4       AM3/AM2+     95W    4x      6MB     02/09/2009   2.0 GHz
II X4 805                     DDR3, DDR2          512KB

                                                                    Lecture 1: Dr. Tom Butler

Table 2 compares the various types of Pentium 4 processors. The process by which these
processors were created had transistor gates of 80 nanometres (a nanometre is
1/1,000,000,000 of a meter—really, really small) to 90 nm in size. The transistor switch
is the key building block of all processors, just as the neuron cell is the key building
block of the human brain. Table 3 illustrates AMD 45nm offerings. The 65 nanometre
(65 nm) process, which results in a the gate width of a transistor being just 65 nm was in
2006 the most advanced approach for volume semiconductor manufacturing; however
was soon replaced when 45 nanometre processes becomes commercially viable—Intel
was the leader in innovating around this. The significance of these transistor sizes is that
the Human Immunodeficiency Virus (HIV) is around 120 nm in diameter, a human red
blood cell typically 6000-8000 nm in diameter, and a human hair typically 80000 nm in
diameter. There are however limits to the minimum size attainable.

                                           Lecture 1: Dr. Tom Butler

Figure 2 Powe r Cons umption comparisons

                                                                     Lecture 1: Dr. Tom Butler

So what are the consequences of transistor size and energy? Each transistor consumes
power and generates heat. Therefore, you would say that the more transistors the more
heat generated: that is true. However, if the operating temperature exceeds a critical value
(and they can get hotter than the temperature require to fry an egg), then the processor
will malfunction and will eventually be damaged. If we take the example that, relatively
speaking, small cars are more energy efficient than big cars, then small transistors (65
nm, and lately 45nm and 32mn processors) will be more energy efficient than large
transistors (180 nm process): then 200 million 65 nm process transistors on a die (e.g.
Core 2 Duo) will run cooler than 200 million on a 90 nm processes (e.g. Itanium): that is,
if we keep the speed of operation, i.e. the clock rate the same for both (e.g. 1.6 Ghz).
However, the higher the clock rate, the higher the power consumption and heat. You can
see from the above figure in that the lower clocked processors consume less power and
energy efficient designs run cooler even at the same clock speed.
Figure 3 attempts to put things into perspective by illustrating that the heat generated in
today’s processors and computer assemblies is very like that of a 100 watt filament light
bulb, which is extremely hot. This is a motherboard assembly for the AMD Athlon X2
EE processor. Imagine the energy and environment (air-conditioning) problems that
medium or large business organisations have with hundreds if not thousands of PCs,
workstations and servers.

Intel and AMD are in a Catch 22 situation, they must make their processors faster, which
means more functions (i.e. more transistors) and higher speeds of operation (in GHz).
Both companies have introduced power management features to help overcome the
problem: for example, AMD's Cool'n'Quiet™ and PowerNow! technologies are typical
solutions for AMD processor-based desktop and mobile systems. Such technologies can
effectively lower the power consumption and enable a quieter-running system while
delivering performance-on-demand. Intel has similar solutions: for example, the desktop

                                                                     Lecture 1: Dr. Tom Butler

PC version of the Core 2 Duo provide up to a 40 % increase in performance and are more
than 40 %more energy efficient versus Intel's previous best processor.
If we look at OHM’s Law (see below) and the relationship between voltage, current and
power, we might get some idea as to how both companies tried to square the circle.
Older processors ran at higher voltages e.g. 5 volts, now processors operate at between 1
and 1.5 volts.
The amount of power consumed P in watts is a function of operating voltage (V) and
current (I) consumed. I.e. P = V x I. In the light bulb example, V = 220 V ac, then a 100
watt bulb will draw a current of 0.45 amps. In an Athlon 64 4600+ EE How do we get the
power consumption down then? By reducing the voltage then the power comes down.
However, current, I, increases with the number of transistors on a die and the speed at
which they are clocked to operate and thereby carry out processing. AMD and Intel
further reduced the operating voltage and tweaking core frequencies to make their
processors more energy efficient. On the other hand, reducing the size of the transistors
(e.g. with 65, 45 and in the future 32 nm gate widths) reduces the amount of current
drawn. Thus, for Intel and AMD reducing voltage and making smaller and smaller
transistors would seem to provide some of the answers to the question of how to reduce
energy consumption: however, this is only part of the answer.
Intel’s Core (latterly i7 Core) microarchitecture provides the basis for their mobile,
desktop and server processors. The first Core microarchitecture took several of the key
microarchitectural benefits developed for the Pentium M, and integrated them with the
Net-burst Microarchitecture of the Pentium 4 to produce a high performance, low power
processors. Previously, Intel operated on the assumption that all that mattered was the
processor’s operating frequency. This was the way that consumers evaluated a
processor’s performance. AMD was the first to topple this myth by focusing on the
increase in instructions executed per clock cycle (Hz of CPU Clock) with their
processors. While it is true that instructions executed per clock cycle is vital to processor
performance, operating frequency still plays an important role in overall throughput.
Intel’s Core microarchitecture addresses these issues with lowered frequency, but a much
improved throughput in terms of instructions per cycle. However, Intel also addressed the
power problem by focusing on “dynamic capacitance”, which is defined by Intel, as the
switching capacitance of the transistors. In this Intel engineers have concluded that
power, as measured in watts, is equal to the dynamic capacitance value times the square
of the operating voltage times the frequency (P= C x V2 x F). In order to balance the
power consumption in the Core microarchitecture, Intel’s engineers took all of these
factors into account to find deliver the maximize the number of tasks the processor forms
in terms of instructions per cycle at the lowest possible power.

                                                                      Lecture 1: Dr. Tom Butler

Primer on Electricity, Electric Motors and Generators
Electricity consists of charges carried by electrons, protons, and other particles. Electric
charge comes in two forms: positive and negative. Electrons and protons both carry
exactly the same amount of electric charge, but the positive charge of the proton is
exactly opposite the negative charge of the electron. If an object has more protons than
electrons, it is said to be positively charged; if it has more electrons than protons, it is
said to be negatively charged. If an object contains as many protons as electrons, the
charges will cancel each other and the object is said to be uncharged, or electrically
neutral. Electricity occurs in two forms: static electricity and electric current. Static
electricity consists of electric charges that stay in one place. An electric current is a flow
of electric charges between objects or locations.

Static Electricity

Static electricity can be produced by rubbing together two objects made of different
materials. Electrons move from the surface of one object to the surface of the other if the
second material holds onto its electrons more strongly than the first does. The object that
gains electrons becomes negatively charged, since it now has more electrons than
protons. The object that gives up electrons becomes positively charged. For example, if a
nylon comb is run through clean, dry hair, some of the electrons on the hair are
transferred to the comb. The comb becomes negatively charged and the hair becomes
positively charged. The following materials are named in decreasing order of their ability
to hold electrons: rubber, silk, glass, flannel, and fur (or hair). If any two of these
materials are rubbed together, the material earlier in the list becomes negative, and the
material later in the list becomes positive.

Electric Current

Current (electric), flow of electric charge. The electric charge in a current is carried by
minute particles called electrons that orbit the nuclei of atoms. Each electron carries a
small electric charge. When a stream of electrons moves from atom to atom—for
example, inside a copper wire—the flow of the charge they carry is called electric
current. Batteries and generators are devices that produce electric current to power lights
and other appliances. Electric currents also occur in nature—lightning being a dramatic
Electric currents flow because atoms and molecules contain two types of electrical
charge, positive and negative, and these opposite charges attract each other. If there is a
difference in the overall charge of atoms between two points—for example, between two
ends of a wire—the negatively charged electrons will flow toward the positively charged
end of the wire, creating electric current. Direct current (DC) is the flow of electricity in
one direction. Alternating current (AC) intermittently reverses direction because of the
way it is generated.
Electric current flows easily in some substances but not at all in others. Solids, liquids,
and gases that carry electric currents are called conductors. Many metals are good

                                                                     Lecture 1: Dr. Tom Butler

conductors. More than one conductor may be needed to build an electric circuit—a path
for electric current to move from one place to another.
The rate at which electric charge flows in a current is measured in amperes. The unit is
named for French physicist Andre Marie Ampere, who contributed to the study of
electrodynamics in the early 19th century. One ampere of electric current is equal to
about 6 billion billion electrons per second flowing past a point.
The difference in charge between two points creates a force called the electric potential
that drives the current ahead. This force is measured in volts, and is named after Italian
scientist Alessandro Volta. A typical flashlight battery produces 1.5 volts. In the US
household appliances run on 110 volts (60 Hertz or Hz), while but some require 220 volts
@ 50 Hz. Voltage in power lines that deliver electricity around the country is measured
in tens of thousands of volts.
Except for the special circumstance in which a substance becomes a superconductor, all
conductors resist the flow of current to some extent. The measurement of a conductor’s
resistance to electric current is measured in ohms, named after German scientist Georg
Simon Ohm. Ohm’s Law states that the amount of current (I) in a conductor is equal to
the voltage (V) divided by the resistance (R), when these values are measured in amperes,
volts, and ohms.
An electric current’s overall power (measured in Watts) depends on the amount of
current flowing (measured in amperes) and the electric potential driving it (measured in
volts)—i.e. the voltage value multiplied by the measured current. Electric power is
measured in watts, named after Scottish scientist Ja mes Watt. One watt is equal to one
ampere moving at one volt. Multiplying amperes by volts produces the number of watts.
An appliance that uses 10 amperes and runs on 115 volts consumes 1150 watts of power.
Electricity suppliers charge customers a rate based on the amount of watts the customers
use. Since the number of watts is usually large, it is calculated in a more convenient unit,
the kilowatt, which equals 1000 watts. The standard measure for large amounts of
electricity is the kilowatt-hour, which equals 1000 watts per hour.

Conductors and Insulators
Conductors are materials that allow an electric current to flow through them easily. Most
metals are good conductors. Substances that do not allow electric current to flow through
them are called insulators, nonconductors, or dielectrics. Rubber, glass, and air are
common insulators. Electricians wear rubber gloves so that electric current will not pass
from electrical equipment to their bodies. However, if an object contains a sufficient
amount of charge, the charge can arc, or jump, through an insulator to another object. For
example, if you shuffle across a wool rug and then hold your finger very close to, but not
in contact with, a metal doorknob or radiator, current will arc through the air from your
finger to the doorknob or radiator, even though air is an insulator. In the dark, the passage
of the current through the air is visible as a tiny spark.

Semiconductor, solid or liquid material, able to conduct electricity at room temperature
more readily than an insulator, but less easily than a metal. Electrical conductivity, which

                                                                    Lecture 1: Dr. Tom Butler

is the ability to conduct electrical current under the application of a voltage, has one of
the widest ranges of values of any physical property of matter. Such metals as copper,
silver, and aluminium are excellent conductors, but such insulators as diamond and glass
are very poor conductors (see Conductor, electrical; Insulation; Metals). At low
temperatures, pure semiconductors behave like insulators. Under highe r temperatures or
light or with the addition of impurities, however, the conductivity of semiconductors can
be increased dramatically, reaching levels that may approach those of metals. The
physical properties of semiconductors are studied in solid-state physics.

Integrated Circuit of a Computer An integrated circ uit (IC) consists of many circuit
elements such as transistors and resistors fabricated on a single piece of silicon or other
semiconducting material. The tiny microprocessor shown here is the heart of the personal
computer (PC). Such devices may contain several million transistors and be able to
execute over 100 million instructions per second. The rows of leglike metal pins are used
to connect the microprocessor to a circuit board.
The invention of the transistor in 1948 was a turning point in the histor y of electronics.
Transistors are composed of a semiconducting material—that is, a substance that can act
as either a conductor or an insulator. Transistors quickly replaced vacuum tubes for
amplifying electronic signals in devices ranging from radios to telephone lines to military
targeting devices. Without the bulky heat- generating vacuum tubes, electronic devices
became much more compact and powerful.

                                                                      Lecture 1: Dr. Tom Butler

Measuring Electric Current
Electric current is measured in units called amperes (amp). If 1 coulomb of charge flows
past each point of a wire every second, the wire is carrying a current of 1 amp. If 2
coulombs flow past each point in a second, the current is 2 amp. See also Electric Meters.

When the two terminals of a battery are connected by a conductor, an electric current
flows through the conductor. One terminal continuously sends electrons into the
conductor, while the other continuously receives electrons from it. The current flow is
caused by the voltage, or potential difference, between the terminals. The more willing
the terminals are to give up and receive electrons, the higher the voltage. Voltage is
measured in units called volts. Another name for a voltage produced by a source of
electric current is electromotive force.

A conductor allows an electric current to flow through it, but it does not permit the
current to flow with perfect freedom. Collisions between the electrons and the atoms of
the conductor interfere with the flow of electrons. This phenomenon is known as
resistance. Resistance is measured in units called ohms. The symbol for ohms is the
Greek letter omega, Ω.
A good conductor is one that has low resistance. A good insulator has a very high
resistance. At commonly encountered temperatures, silver is the best conductor and
copper is the second best. Electric wires are usually made of copper, which is less
expensive than silver.
The resistance of a piece of wire depends on its length, and its cross-sectional area, or
thickness. The longer the wire is, the greater its resistance. If one wire is twice as long as
a wire of identical diameter and material, the longer wire offers twice as much resistance
as the shorter one. A thicker wire, however, has less resistance, because a thick wire
offers more room for an electric current to pass through than a thin wire does. A wire
whose cross-sectional area is twice that of another wire of equal length and similar
material has only half the resistance of the thinner wire. Scientists describe this
relationship between resistance, length, and area by saying that resistance is proportional
to length and inversely proportional to cross-sectional area. Usually, the higher the
temperature of a wire, the greater its resistance. The resistance of some materials drops to
zero at very low temperatures. This phenomenon is known as superconductivity.

Ohm’s Law
The relationship between current, voltage, and resistance is given by Ohm’s law. This
law states that the amount of current passing through a conductor is directly proportional
to the voltage across the conductor and inversely proportional to the resistance of the
conductor. Ohm’s law can be expressed as an equation, V = IR, where V is the difference
in volts between two locations (called the potential difference), I is the amount of current
in amperes that is flowing between these two points, and R is the resistance in ohms of
the conductor between the two locations of interest. V = IR can also be written R = V/I

                                                                      Lecture 1: Dr. Tom Butler

and I = V/R. If any two of the quantities are known, the third can be calculated. For
example, if a potential difference of 110 volts sends a 10-amp current through a
conductor, then the resistance of the conductor is R = V/I = 110/10 = 11 ohms. If V = 110
and R = 11, then I = V/R = 110/11 = 10 amp. Under normal conditions, resistance is
constant in conductors made of metal. If the voltage is raised to 220 V in the example
above, then R is still 11. The current I will be doubled, however, since I = V/R = 220/11
= 20 amp.

Heat and Power
A conductor’s resistance to electric current produces heat. The greater the current passing
through the conductor, the greater the heat. Also, the greater the resistance, the greater the
heat. A current of I amp passing through a resistance of R ohms for t seconds generates
an amount of heat equal to I2 Rt joules (a joule is a unit of energy equal to 0.239 calorie).
Energy is required to drive an electric current through a resistance. This energy is
supplied by the source of the current, such as a battery or an electric generator. The rate
at which energy is supplied to a device is called power, and it is often measured in units
called watts. The power P supplied by a current of I amp passing through a resistance of
R ohms is given by P = I2 R or VI. E.g., a 100 watt light bulb will draw a current of 0.45
amps with a 220 volt AC supply.

Frequency (Hertz)
Frequency is expressed in hertz (Hz); a frequency of 1 Hz means that there is 1 cycle or
oscillation per second. The unit is named in honor of the German physicist Heinrich
Rudolf Hertz, who first demonstrated the nature of electromagnetic wave propagation.
Kilohertz (kHz), or thousands of cycles per second, megahertz (MHz), or millions of
cycles per second, and gigahertz (GHz), or billions of cycles per second, are employed in
describing certain high- frequency phenomena, such as radio waves. Radio waves and
other types of electromagnetic radiation may be characterized either by their
wavelengths, or by their frequencies.

Electric Generators and Motors
Electric Motors and Generators are a group of devices used to convert mechanical energy
into electrical energy, or electrical energy into mechanical energy, by electromagnetic
means (see Energy). A machine that converts mechanical energy into electrical energy is
called a generator, alternator, or dynamo, and a machine that converts electrical energy
into mechanical energy is called a motor. The examples described in class were much
simplified for the purpose of explanation. This short primer provides a more accurate
treatment of the topic; if students have grasped the basics, then this should be enough, as
the study of generators and motors is highly a complex area of study and best left to
electrical engineers.
Two related physical principles underlie the operation of generators and motors. The first
is the principle of electromagnetic induction discovered by the British scientist Michael
Faraday in 1831. If a conductor is moved through a magnetic field, or if the strength of a
stationary conducting loop is made to vary, a current is set up or induced in the conductor
(see Induction). The converse of this principle is that of electromagnetic reaction, first

                                                                    Lecture 1: Dr. Tom Butler

observed by the French physicist André Marie Ampère in 1820. If a current is passed
through a conductor located in a magnetic field, the field exerts a mechanical force on it.
See Magnetism.
The magnetic field of a permanent magnet is strong enough to operate only a small
practical dynamo or motor. As a result, for large machines, electromagnets are employed.
Both motors and generators consist of two basic units, the field, which is the
electromagnet with its coils, and the armature, the structure that supports the conductors
which cut the magnetic field and carry the induced current in a generator or the exciting
current in a motor. The armature is usually a laminated soft- iron core around which
conducting wires are wound in coils.

Direct-current (dc) Generators
If an armature revolves between two stationary field poles, the current in the armature
moves in one direction during half of each revolution and in the other direction during the
other half. To produce a steady flow of unidirectional, or direct, current from such a
device, it is necessary to provide a means of reversing the current flow outside the
generator once during each revolution. In older machines this reversal is accomplished by
means of a commutator, a split metal ring mounted on the shaft of the armature. The two
halves of the ring are insulated from each other and serve as the terminals of the armature
coil. Fixed brushes of metal or carbon are held against the commutator as it revolves,
connecting the coil electrically to external wires. As the armature turns, each brush is in
contact alternately with the halves of the commutator, changing position at the moment
when the current in the armature coil reverses its direction. Thus there is a flow of
unidirectional current in the outside circuit to which the generator is connected. DC
generators are usually operated at fairly low voltages to avoid the sparking between
brushes and commutator that occurs at high voltage. The highest potential commonly
developed by such generators is 1500 V. In some newer machines this reversal is
accomplished using power electronic devices, for example, diode rectifiers.
Modern DC generators use drum armatures that usually consist of a large number of
windings set in longitudinal slits in the armature core and connected to appropriate
segments of a multiple commutator. In an armature having only one loop of wire, the
current produced will rise and fall depending on the part of the magnetic field through
which the loop is moving. A commutator of many segments used with a drum armature
always connects the external circuit to one loop of wire moving through the high-
intensity area of the field, and as a result the current delivered by the armature windings
is virtually constant. Fields of modern generators are usually equipped with four or more
electromagnetic poles to increase the size and strength of the magnetic field. Sometimes
smaller interpoles are added to compensate for distortions in the magnetic flux of the
field caused by the magnetic effect of the armature.
DC generators are commonly classified according to the method used to provide field
current for energizing the field magnets. A series-wound generator has its field in series
with the armature, and a shunt-wound generator has the field connected in parallel with
the armature. Compound-wound generators have part of their fields in series and part in
parallel. Both shunt-wound and compound-wound generators have the advantage of
delivering comparatively constant voltage under varying electrical loads. The series-

                                                                     Lecture 1: Dr. Tom Butler

wound generator is used principally to supply a constant current at variable voltage. A
magneto is a small DC generator with a permanent- magnet field.


In general, DC motors are similar to DC generato rs in construction. They may, in fact, be
described as generators “run backwards.” When current is passed through the armature of
a DC motor, a torque is generated by magnetic reaction, and the armature revolves. The
action of the commutator and the connections of the field coils of motors are precisely the
same as those used for generators. The revolution of the armature induces a voltage in the
armature windings. This induced voltage is opposite in direction to the outside voltage
applied to the armature, and hence is called back voltage or counter electromotive force
(emf). As the motor rotates more rapidly, the back voltage rises until it is almost equal to
the applied voltage. The current is then small, and the speed of the motor will remain
constant as long as the motor is not under load and is performing no mechanical work
except that required to turn the armature. Under load the armature turns more slowly,
reducing the back voltage and permitting a larger current to flow in the armature. The
motor is thus able to receive more electric power from the source supplying it and to do
more mechanical work.
Because the speed of rotation controls the flow of current in the armature, special devices
must be used for starting DC motors. When the armature is at rest, it has virtually no
resistance, and if the normal working voltage is applied, a large current will flow, which
may damage the commutator or the armature windings. The usual means of preventing
such damage is the use of a starting resistance in series with the armature to lower the
current until the motor begins to develop an adequate back voltage. As the motor picks up
speed, the resistance is gradually reduced, either manually or automatically.
The speed at which a DC motor operates depends on the strength of the magnetic field
acting on the armature, as well as on the armature current. The stronger the field, the
slower is the rate of rotation needed to generate a back voltage large enough to counteract
the applied voltage. For this reason the speed o f DC motors can be controlled by varying
the field current.

                                                                   Lecture 1: Dr. Tom Butler

Alternating-Current (ac) Generators (alternators)

As stated above, a simple generator without a commutator will produce an electric
current that alternates in direction as the armature revolves. Such alternating current is
advantageous for electric power transmission , and hence most large electric generators
are of the AC type. In its simplest form, an AC generator differs from a DC generator in
only two particulars: the ends of its armature winding are brought out to solid
unsegmented slip rings on the generator shaft instead of to commutators, and the field
coils are energized by an external DC source rather than by the generator itself. Low-
speed AC generators are built with as many as 100 poles, both to improve their efficiency
and to attain more easily the frequency desired. Alternators driven by high-speed
turbines, however, are often two-pole machines. The frequency of the current delivered
by an AC generator is equal to half the product of the number of poles and the number of
revolutions per second of the armature.

                                                                     Lecture 1: Dr. Tom Butler

It is often desirable to generate as high a voltage as possible, and rotating armatures are
not practical in such applications because of the possibility of sparking between brushes
and slip rings and the danger of mechanical failures that might cause short circuits.
Alternators are therefore constructed with a stationary armature within which revolves a
rotor composed of a number of field magnets. The principle of operation is exactly the
same as that of the AC generator described, except that the magnetic field (rather than the
conductors of the armature) is in motion.
The current generated by the alternators described above rises to a peak, sinks to zero,
drops to a negative peak, and rises again to zero a number of times each second,
depending on the frequency for which the machine is designed. Such current is known as
single-phase alternating current. If, however, the armature is composed of two windings,
mounted at right angles to each other, and provided with separate external connections,
two current waves will be produced, each of which will be at its maximum when the
other is at zero. Such current is called two-phase alternating current. If three armature
windings are set at 120° to each other, current will be produced in the form of a triple
wave, known as three-phase alternating current. A larger number of phases may be
obtained by increasing the number of windings in the armature, but in modern electrical-
engineering practice three-phase alternating current is most commonly used, and the
three-phase alternator is the dynamoelectric machine typically employed for the
generation of electric power. Voltages as high as 13,200 are common in alternators.

AC Motors
Two basic types of motors are designed to operate on alternating current, synchronous
motors and induction motors. The synchronous motor is essentially a three-phase
alternator operated in reverse. The field magnets are mounted on the rotor and are excited
by direct current, and the armature winding is divided into three parts and fed with three-
phase alternating current. The variation of the three waves of current in the armature
causes a varying magnetic reaction with the poles of the field magnets, and makes the
field rotate at a constant speed that is determined by the frequency of the current in the
AC power line. The constant speed of a synchronous motor is advantageous in certain
devices; however, in applications where the mechanical load on the motor becomes very
great, synchronous motors cannot be used, because if the motor slows down under load it
will “fall out of step” with the frequency of the current and come to a stop. Synchronous
motors can be made to operate from a single-phase power source by the inclusion of
suitable circuit elements that cause a rotating magnetic field.
The simplest of all electric motors is the squirrel-cage type of induction motor used with
a three-phase supply. The armature of the squirrel-cage motor consists of three fixed coils
similar to the armature of the synchronous motor. The rotating member consists of a core
in which are imbedded a series of heavy conductors arranged in a circle around the shaft
and parallel to it. With the core removed, the rotor conductors resemble in form the
cylindrical cages once used to exercise pet squirrels. The three-phase current flowing in
the stationary armature windings generates a rotating magnetic field, and this field
induces a current in the conductors of the cage. The magnetic reaction between the
rotating field and the current-carrying conductors of the rotor makes the rotor turn. If the

                                                                     Lecture 1: Dr. Tom Butler

rotor is revolving at exactly the same speed as the magnetic field, no currents will be
induced in it, and hence the rotor should not turn at a sync hronous speed. In operation the
speeds of rotation of the rotor and the field differ by about 2 to 5 percent. This speed
difference is known as slip. Motors with squirrel-cage rotors can be used on single-phase
alternating current by means of various arrangements of inductance and capacitance that
alter the characteristics of the single-phase voltage and make it resemble a two-phase
voltage. Such motors are called split-phase motors or condenser motors (or capacitor
motors), depending on the arrangement used. Single-phase squirrel-cage motors do not
have a large starting torque, and for applications where such torque is required, repulsion-
induction motors are used. A repulsion- induction motor may be of the split-phase or
condenser type, but has a manual or automatic switch that allows current to flow between
brushes on the commutator when the motor is starting, and short-circuits all commutator
segments after the motor reaches a critical speed. Repulsion- induction motors are so
named because their starting torque depends on the repulsion between the rotor and the
stator, and their torque while running depends on induction. Series-wound motors with
commutators, which will operate on direct or alternating current, are called universal
motors. They are usually made only in small sizes and are commonly used in household


To top