Documentos de Académico
Documentos de Profesional
Documentos de Cultura
For
A+ Core Hardware Technology
Examination 220-221
Authors
Tcat Houser
CTT+, A+, Network+, Server+, MCSE,
Helen O'Boyle
CTT+, A+, Network+, MCSE, MCSD
Published by
TotalRecall Publications, Inc.
1103 Middlecreek
Friendswood, TX 77546
281-992-3131
THIS IS BOOK IS GUARANTEED:
See details at www.TotalRecallPress.com
Dedication
ExamInsight
For
CompTIA
A+ Core Hardware Technology
About the Authors
Tcat Houser
Tcat Houser (CTT+, A+, Network+, Server+, I-Net+, MCSE) has garnered other
milestones such as IBM PSE and Microsoft MCSE/MCT. This is the result of almost 40
years of fussing with electronics. When not writing or teaching, he is doing research.
Tcat accomplishes so many tasks because to him, it isnt work, its fun, and he has fun 18
hours a day, 7 days a week. You can reach him by sending mail to
Author@TotalRecallPress.com or Tcat@Tcat.net.
Looking back at the last book where I (Tcat Houser) was the lead author (i-Net+ Exam
Prep 1576105989) the independent reviews were very good, and most readers understood
that I had two intentions. One, ace the test. Two, supply the information to be a reference
manual and/or alert the reader to upcoming technologies so as todays hero, the reader
was not tomorrows zero.
In the two years since that release, I led a number of career changers through CompTIA
and wrote my own courseware. Building on that learning, you have this release.
Helen OBoyle
Helen OBoyle (CTT+, A+, Network+, Server+, MCSE, MCSD) has been working with
computers for 20 years. Her background is quite varied in regards to both platforms and a
complete interest in both the engineering of software and hardware. You may email her at
Author@TotalRecallPress.com or Hoboyle@mindspring.com.
Editor
Chad Bayer has been working in the computer industry for over 6 years and is A+
Certified. His background is quite varied and is interested in both the engineering of
software and hardware. Chad is working towards his computer engineering degree at the
University of Texas in Dallas. You can reach him by sending mail to
Info@TotalRecallPress.com
Table of Contents
About the Authors ..............................................................................................IV
About the Book ...................................................................................................V
Acknowledgments......................................................................................... XVIII
How to Read This Book .................................................................................. XIX
Real-World Lessons ........................................................................................ XX
Icon Alerts ...................................................................................................... XXII
Exam Specifics .............................................................................................XXIV
Forward.........................................................................................................XXVI
CompTIA A+ Core Technologies Exam Blueprint .......................................XXVII
Terms to Know..................................................................................................XL
XII
XIII
VI
VII
VIII
Table of Contents IX
Expansion Buses........................................................................................ 90
ISA.............................................................................................................. 90
MCA............................................................................................................ 92
EISA ........................................................................................................... 92
VESA-VLB.................................................................................................. 93
PCI.............................................................................................................. 94
Plug and Play ............................................................................................. 96
Why Plug n Play is Sometimes Plug and Pray.......................................... 97
PCI Steering ............................................................................................... 97
AGP ............................................................................................................ 98
AMR............................................................................................................ 99
Demanding Attention Digitally ..................................................................100
Interrupting ...............................................................................................101
IRQ ...........................................................................................................102
IRQ Conflict ..............................................................................................102
DMA..........................................................................................................103
DMA, the PC and the AT..........................................................................103
Bus Mastering ..........................................................................................104
Memory Addresses ..................................................................................104
Memory Mapping......................................................................................105
Logical Names..........................................................................................106
Pop Quiz 0001.10 ..............................................................................107
IX Chapter 0001: Summary ..........................................................................109
X Chapter 0001: Test for Success Questions .............................................110
Test for Success Answers........................................................................117
X Table of Contents
Cyrix .........................................................................................................143
M1.............................................................................................................144
AMD..........................................................................................................144
Sockets & Slots ........................................................................................146
Socket 1....................................................................................................147
Slot 1 ........................................................................................................150
Slot 2 ........................................................................................................151
Socket 370 ...............................................................................................151
Slot A ........................................................................................................151
Socket 462 ...............................................................................................152
III Chapter 0010: Summary .........................................................................153
IV Chapter 0010: Test for Success Questions .............................................154
Test for Success Answers........................................................................161
Table of Contents XI
Troubleshooting Memory Challenges ......................................................196
IV Chapter 00II: Summary ............................................................................199
V Chapter 0011: Test for Success Questions .............................................200
Test for Success Answers........................................................................207
IV
V
VI
VII
VIII
IX
X
Table of Contents XV
I
II
Introduction...............................................................................................448
Safety and the Environment.....................................................................449
3 Demons .................................................................................................449
EMI ...........................................................................................................449
UPS ..........................................................................................................453
RFI............................................................................................................454
ESD ..........................................................................................................455
Contamination ..........................................................................................455
Safety .......................................................................................................457
Chemicals.................................................................................................458
Fire Safety ................................................................................................458
III Chapter 1010: Summary ..........................................................................460
IV Chapter 1010: Test for Success Questions .............................................461
Test for Success Answers........................................................................464
XVIII Acknowledgments
Acknowledgments
Whoa.
For the first time, this page (at least for me) isnt the usual yada-yada.
This book is one of a series that is part of the ExamInsight series. Combining all the ways
a human mind takes in information and teaching both sides of the brain has taken quite a
bit of development time and patience from literally hundreds of people. Yes, I sincerely
thank you all. And for those whos patience was particularly well-tested, my apologies.
First, a tip of the hat to Helen OBoyle somehow, we managed made contributions both
to the book and making impossible QuickTime movies as well as late night/early
morning enforcement of our Success
Bruce Moran of TotalRecall Publications along with the rest of his team deserves several
Tips of the hat for having the courage to run with such a far-reaching concepts as Helen
and myself presented them for the ExamInsight series. A huge thank you is due to my
family and friends such as Pam Fanstill, offering moral support as well as technical
assistance. All of you may be geographically distant, and you didnt let that stop you
from support by phone and email. And I cannot forget the students and interns at
TcatU.net for the thousands of hours of testing and offering advice. Now we all have a
better understanding of the saying, Rome wasnt built in a day.
Tcat Houser
Tcat@tcat.net
Humor
All the facts and figures can be boring, by themselves. Wherever possible, we explain the
economic politics from a historical perspective.
At least this way, you know why sometimes our industry can be both brilliant and
FUBARd all at the same time. No that is one acronym we won't spell out for you!
The tone is informal, slightly publicity incorrect, at least more so than other books, less
so than Maximum PC magazine.
Real-World Lessons
If you purchased this book in paper media, you will notice it was designed to fit on a
desk. This is so it takes less space.
If you purchased this book in electronic form, you will have much less scrolling than
with most other publications in electronic form.
Reading by candlelight or the glow of a Liquid Crystal Display (LCD), we know what
we want as students, and what our students told us they would like to see.
Microsoft has a favorite term, heard on their campus. They call it, Eating our own dog
food. They mean they use their own tools to create new work. (This is great until some
test version of an email distribution system running the campus dies.)
Since you the reader are attempting to learn the INs and OUTs of the IT industry, we
have brought you what it looks like to eat your own dog food.. This book doesn't have
the traditional label of introduction. Instead, we used a popular form in IT as your first
clue to look for whenever you are unpacking a new piece of computer hardware or
software. It is called, Readme.1st, READ.ME, or Read.ME, or other clearly labeled
warning that is too often ignored.
The bottom line for ANY PC is, either On or Off. Early in this book, you will learn that
this is a Binary option. Binary means two. Computers work in binary. So, to assist you in
thinking in binary, the chapter numbers are in binary.
We have included high-resolution close up pictures to help Identify what is being
discussed, or used large clear graphics to help convey the message.
Continued Support
We encourage questions! We find that if you are patient, and go through and complete
the exercises this will reinforce your learning and the information will stick with you.
We ask the average student to plan on a minimum of 10 hours of study time after
completing the course. A visual person you say? We have included lots of pictures and
screen shots of the topics under discussion to help you correlate what you are reading
with what you will see live.
As a purchaser of this work, you are automatically entitled to join your fellow readers and
authors on a private email list. Experience tells us that learning is more fun and easier
when done with a group. You are encouraged to join the students from our physical
(brick and mortar) school who are using this book in a classroom, and who keep in touch
via email.
To join the others, send an email to: AsktheAuthor@TotalRecallPress.com
Subject: Subscribe A+
If everything goes right you will get an email telling you to respond. (This is to prevent
someone else from using us for creative revenge, on you.)
If you don't get a reply, check your email to see if you sent Plain Text (that helps).
Got a thought about this book?
AsktheAuthor@TotalRecallPress.com
Praise
us
or
pan
us
by
writing
to:
We always read each email, and unless we are teaching a class, we respond within hours.
Icon Alerts
These are the symbols or keys to information alerts; use them as another study aid.
Historical
Historical Owl! This is material that is not directly testable by CompTIA and the A+
test, but may either help you 'pull on the thread' so you can see why something is.
Geek
Geek Owl! This is material that is too deep for CompTIA and the A+ test. But
knowing it will help you put the topic together.
Future
Future Owl! This indicates a topic that should be on your radar screen. CompTIA
generally only tests for what are 'commonly accepted' items in the industry. Future OWL
tells you about something coming up. You will look good in a job interview and/or
should know about something that will be hot tomorrow, or today great deal being
tomorrows fish wrap.
Real World
RW OWL! (a.k.a real world) describes tips for understanding the difference between
by the book and the real world out there in IT.
Who is?
Success Owl! The OWL is an honest presentation of how the computer industry
looks to us. You will notice in our logo, Success Owl is setting on a branch with leaves.
The gold of the OWLS feet and the green of the leaves signify the money you can make
in this business.
The Branch is you. The branch represents your strength, energy and diligence in
achieving your goals. Like any serious players in the computer game, we have to run very
fast just to attempt to keep up. Each of the OWLs feathers is something to learn. No
matter how fast you run, you cannot get to the top. (You just fly for more gold.) Further,
each Feather learned will become obsolete too quickly for comfort. That is what this
industry is about -- IT is in a constant state or refreshment and therefore a lively, never
boring chosen field. Finally, once the decision to get on the IT knowledge is entered,
there is no getting off. Life long study is mandatory.
Exam Specifics
Taking the Test
Please arrive at the testing center at least 15 minutes before the test is
scheduled to begin. The administrator of the testing center can
demonstrate how to use the computer-based testing system before the
actual test begins. Two forms of identification must be presented to the
test center administrator. One form should be a photo ID, such as a valid
driver's license. The other can be a major credit card, or a passport. Please
be aware that both forms of identification must have a signature. Books,
calculators, laptop computers, or other reference materials are not allowed
during the test. Because the test is computer-based, pens, pencils, or paper
will not be needed. It is CompTIAs policy to make reasonable
accommodations for individuals with disabilities.
After the Test, how it works!
As soon as you finish the test, you receive the final score. You will see the
results immediately on the computer screen. In addition, a hard copy of the
score report is provided at the testing center. The score report shows
whether or not you passed the certification. It will also show all objectives
related to every item not answered correctly. It can be used to verify your
certification until your certificate arrives.
If you pass the examination, a certificate will be mailed to you within 4 to
6 weeks. Should you not receive your certificate and information packet
within 6 weeks of passing your exam, please contact CompTIA at
fulfillment@comptia.org . You can also contact the fulfillment department
for replacement certificates.
About the scoring for the exam;
A+ is adaptive. The number of questions you get will vary between 20 and
30 questions. A minimum score of 596 on a scale of 0 to 1300 for the core
test (220-221) to pass. Time 30 minutes. The OS test is also a maximum of
30 minutes with a minimum score of 600 on the same 0 to 1300 scale,
with 20 to 30 questions offered. Due to the nature of adaptive testing, a
score of 0 or 1300 is highly unlikely. CompTIA has a PDF available
which explains in detail the ins and outs of adaptive testing.
Price of Exam;
There is a discount for CompTIA members, for non-members the US price
is: $139 for each test.
Exam Assessment:
A+ Certification is a CompTIA-sponsored testing program that certifies
the competency of entry-level (6 months experience) computer service
technicians. The A+ test contains situational, traditional, and identification
types of questions. All of the questions are multiple choices, with only one
correct answer for each question. The test covers a broad range of
hardware and software technologies, but is not bound to any vendorspecific products.
Major computer hardware and software vendors, distributors, resellers and
publications back the program. A+ certification signifies that the certified
individual possesses the knowledge and skills essential for a successful
entry-level (6 months experience) computer service technician, as defined
by experts from companies across the industry.
Forward
This appendix is present in two parts. Part I contains the official CompTIA A+ Core
Technologies exam blueprint. Before you take your examination, be sure that you feel
comfortable with every point listed in the blueprint.
Part II maps each objective to the chapters in the book that cover the test point. When
preparing for the test, use this as a tool to quickly find where your weak spots are covered
in the book. Good luck!!!
Domain
% Of Examination
30%
30%
5%
15%
5.0 Printers
10%
10%
Total
100%
Response Limits
The examinee selects, from four (4) or more options and the option(s) that best completes
the statement or answers the question. Wrong answers are response options that
examinees with incomplete knowledge or skill would likely choose, but are generally
plausible responses fitting into the content area. Test item formats listed in this
examination are:
Multiple-choice: the examinee selects one option that best answers the question or
completes a statement. The option can be embedded in a graphic ware the examinee
points and clicks on their selection choice to complete the test item.
Multiple-response: the examinee selects more then one option that best answers the
question or completes a statement.
Sample Direction: Read the statement or question and from the response options, select
only the option(s) that represent the most correct or best answer(s).
System Board
Power Supply
Processor/CPU
Memory
Storage Devices
Monitor
Modem
Firmware
CMOS
LCD
Ports
PDA
1.2 Identify basic procedures for adding removing field replaceable modules for
both desktop and portable systems.
Module Types:
System Board
Storage Devices
Power Supply
Processor/CPU
Memory
Input Devices
Hard Drive
Keyboard
Video Board
Mouse
Network Interface Card
AC adapter
Digital camera
DC Controller
LCD panel
PC Card
Pointing devices
1.3 Identify available IRQs, DMAs, and I/O addresses and procedures for device
installation and configuration.
Cable types
Cable orientation
Serial vs. parallel
Pin connections
Types of connectors:
DB-9
DB-25
RJ-11
RJ-45
BNC
PS2/Mini-DIN
USB
IEEE 1394
1.5 Identify proper procedures for installing and configuring ID/E IDE devices
Master/slave
Devices per channel
Primary/secondary
1.6 Identify proper procedures for installing and configuring SCSI devices.
Address/termination conflicts
Cabling types (example: regular, wide, ultra-wide,) internal vs.
external expansion slots, EISA, ISA, PCI
Jumper block settings (binary equivalents)
1.7 Identify proper procedures for installing and configuring peripheral devices.
Monitor/video card
Modem
USB peripherals and hubs
IEEE 1284
IEEE 1394
External storage
Portables
Docking stations
PC cards
Port replicators
Infrared devices
Memory
Hard drives
CPU
Upgrading BIOS
When to upgrade BIOS
Portable systems
Battery
Hard drive
Types 1, II, III cards
Memory
Processor/memory symptoms
Mouse
Floppy drive
Parallel ports
Hard drives
DVD
Sound
Sound card/audio
Monitor/video
Motherboards
Modems
BIOS
USB
NIC
CMOS
Power supply
Slotcovers
POST audible/visual error codes
Troubleshooting tools e.g. Multimeter
Large LBA, LBA
Cables
Keyboard
Peripherals
2.2 Identify basic troubleshooting procedures and how to elicit problem symptoms
from customers
3.1 Identify the purpose of various types of preventive maintenance products and
procedures and when to use them.
3.2 Identify issues, procedures and devices for protection within the computing
environment, including people, hardware and the surrounding workplace.
High-voltage equipment
Power supply
CRT
Batteries
CRT's
Toner kits/cartridges
Chemical solvents and cans
MSDS (Material Safety Data Sheet)
4.0 Motherboards/processor/memory
This domain requires knowledge of specific terminology, facts, ways and means of
dealing with classifications, categories and principles of motherboards, processors in
memory in microcomputer systems.
4.1 Distinguish between popular CPU chips in terms of their basic characteristics.
4.2 Identify the categories of RAM (random access memory) terminology, their
locations, and physical characteristics.
Terminology:
Memory bank
Memory chips (8-bit, 16-bit, and 32-bit)
SIMMS (Single in-line memory module)
DIMMS (dual in-line memory module)
Parity chips vs. non-parity chips
4.5 Identify the most popular types of motherboards, the components and
architecture (bus structures and power supplies)
Types of motherboards:
Components:
Communications ports
SIMM and DIMM
Processor sockets
External cache memory (level 2)
Bus architecture
ISA
PCI
AGP
USB (Universal serial bus)
VESA local bus (VL-bus)
5.0 Printers
This domain requires knowledge of basic types of printers, basic components, and printer
components, how they work, how they print onto a page, paper path, care and service
techniques, and common problems.
Parallel
Network
USB
Infrared
Serial
5.2 Identify care and service techniques and common problems with primary
printer types.
6.1 Identify basic networking concepts, including how network works and the
ramifications of repairs on the network.
Chapter(s)
0100
0000, 0101
0111, 1001
Chapter(s)
0001, 1100
Chapter(s)
1010
0000, 1010
Chapter(s)
0000, 0011,1000
Chapter(s)
0111
0111, 1010
Chapter(s)
1011
Terms to Know
80686
Baud
80386-DX
BIOS
80386-SX
BNC
80486-DX
Bridge
80486-DX2
Bus
80-Conductor
Bus Mastering
100 Mbps
Cable Select
40-Pin
Capacitive Keyboard
Cat 3
Active LCD
Cat 5
Actuator Arm
CAV
ADC
CD-DA
Address Bus
CD-R
AGP
CD-ROM
AMD
CD-ROM XA
Anti-static pad
CD-RW
Anti-static strap
Celeron
ASIC
Centronics 36
AT
Checksum
ATA
Chipsets
ATAPI
CHS
ATDT
CISC
Athlon
Class A
ATX
Class B
ATZ
Class C
Baby AT
Clusters
Backbone
CLV
Backside Bus
CMOS
Bandwidth
Color Depth
EMI
CRT
EPP
Cylinders
EPROM
DAC
ESD
DAT
Ethernet
Data Bus
DB-9
Fast Ethernet
DB-15
Fast SCSI
DB-25
DDR SDRAM
FCC
DHCP
FDD
DIMM
Firmware
DIN-5
Flash BIOS
DIP
Flashing
DMA
Floating Point
DNS
FM Synthesis
Docking Station
Form Factor
Dot Matrix
Frame Buffer
Dot Pitch
Frontside Bus
Double Density
Fusing Assembly
DRAM
Gateway
Drivers
Gigabit Ethernet
Duplex
Hazardous Waste
Duplexing
HD-50
Duron
HD-68
ECC RAM
Heads
ECP
High Pin 1
EDO RAM
High Voltage
EEPROM
Host
EIDE
Hot Swapping
MDRAM
HVD
Memory Controller
HVPS
Memory Effect
I/O Address
Memory Mapping
ID Jumpers
Mesh
IDE
MIDI
IEEE 1284
MiniDIN-6
IEEE 1394
Mirroring
Infrared
Modem
Interleaving
NetBEUI
Interrupt
Netware
IP Number
NiCD
IPX
NLX
IrDA
NMI
IRQ
Non-Maskable Interrupt
ISA
Northbridge
ISO 9660
Optical Mouse
JBOD
Opto-Mechanical Mouse
Landing Zone
Packet
Latency
Parallel
LCD
Parity
Level 1 Cache
Passive LCD
Level 2 Cache
PcCard
Level 3 Cache
PCI
LUN
PCMCIA
LVD
Pentium
LVD/MSE
Pentium II
M1
Pentium III
MAC Address
Pentium IV
Master
Pentium MMX
RJ-45
Photosensitive Drum
ROM
Piezo-Electric
Router
PIO
RS-232C
Pixel
Sag
Platter
Sample Rate
Plug n Play
Sample Resolution
Port Replication
SCSI
POST
SCSI 2
Power Good
SCSI 3
Primary
SCSI ID
SDRAM
PROM
SE SCSI
PS/2
SEC
QAS
Secondary
QIC
Sectors
RAID 0
Seek Time
RAID 1
Serial
RAID 5
SGRAM
RAM
SIMM
RAMDAC
Simplex
Read/Write Head
Slave
Refresh Rate
Slot 1
Resolution
Slot 2
RFI
Slot A
RG-58
SMP
Ribbon Cable
Socket 7
Ring
Socket 8
RISC
Southbridge
RJ-11
SPGA
Type 2
Spindle
Type 3
Spindle Motor
UART
SRAM
UDMA
S-Register
UL 1449
Star
Ultra IDE
Static
Ultra SCSI
STD
STP
Ultra SCSI 2
SubNet Mask
Super Socket 7
SVGA
UPS
Switch
USB
Switched Keyboard
UTP
V.90
TCP/IP
VBE
TCQ
VESA
Term Power
VGA
Termination
VR
TFT
VRAM
Thermistor
VRE
Thinnet
Wavetable Synthesis
Thunderbird
Wide SCSI
Token Ring
WORM
Toner
WRAM
Tracks
Xeon
XGA
Transport Protocol
Type 1
To your Success!
Patience is the companion of wisdom.
-Augustine
2 Chapter 0000
4.1 Distinguish between popular CPU chips in terms of their basic characteristics.
Popular CPU chips (Intel, AMD, Cyrix)
Characteristics
Physical size
Voltage
Speeds
Sockets
4.2 Identify the categories of RAM (random access memory) terminology, their
locations, and physical characteristics.
Terminology:
DRAM (dynamic random access memory)
4.3 Identify the most popular types of motherboards, the components and
architecture (bus structures and power supplies)
Bus architecture
ISA
PCI
AGP
VESA local bus (VL-bus)
4 Chapter 0000
Introduction
To get a sense of where we are, it helps to see where we came from. In other words, to
know where Information Technology (IT) is going, it helps to know where it has come
from. The first few pages of this chapter are historical in nature, and not required for
the A+ Core test.
II
Start of IT
If you are very familiar with PCs (personal computers) you can safely skip the history
lesson. Fair warning that skipping yesteryear data will deprive you of the data for some
fun bets you are sure to win (unless the opposing party has also read this book .
Data you really need to know begins with the Success Owl regarding 8-bit vs. 16-bit
data paths. If that sounds like Greek to you, then it is important to read the history lesson.
You will learn the different ways computer manufactures offered expansion capabilities
to their hardware.
After examining the sub systems that are the parts of the PC you will learn the
principles of electricity. Building on your newfound understanding of electricity, you
will discover how electronic components behave and how to visually Identify them.
Visual explanations augment the discussion.
Finally, you will garner an understanding of how Base2 (binary) math relates to both
Base10 (everyday math) and Base16 (hex) math, without needing to be a propeller
head
IT STARTS
Many consider the first computer to be an invention by the American
known as Herman Hollerith. This machine read punch cards, to tie
the latest statistics for the 1890 census. (He got the idea from his
brother back in Europe, when reading his letter about the new weaving looms operated.)
Hollerith sold the US Congress on the idea. (Computing first vaporware product) The
1890 Census was completely done in months, not years. And distilling data such as how
many wheat farmers live in Ohio? became feasible. Hollerith was a smash hit. Later,
financial difficulties forced the sale of this machine to a firm known as Computer
Tabulating Recording (CTR). A former salesman of player pianos at the time was
Thomas Watson, who was working with CTR. Eventually, Watson took over the
company, renaming it IBM.
6 Chapter 0000
III
IT Starts here
(http://www.exo.com/~wts/mits0013.htm)
IV
Micro Soft
The issue facing MITS was the computer was programmed with 8 on/off switches. This
made programming a tedious effort at best.
The appearance of a solution showed up in a letter sent to MITS from a firm marketing a
program called Traf-O-Data. This company was a part-time business for a couple of
high school students named Bill Gates and Paul Allen.
They offered a computer language known as BASIC to MITS. Let's not forget the trivial
point that neither of these individuals actually had this language for the Altair.
BASIC was available in the public domain for different CPUs (Central Processing
Unit), which is considered the computers brain.
Their prior experience of using a simulator2 used on a DEC computer for Traf-O-Data
allowed them to work a few frantic weeks through which they were able to make BASIC
work on an Altair without the experience of ever having seen one.
With the success of their efforts, Paul Allen went to work at MITS and Bill Gates
dropped out of Harvard and moved to Albuquerque.
The two individuals granted a non-exclusive license to MITS, which made many
hobbyists, resent them. Their point was that BASIC had come from the public domain
and they had used government-funded computers to pirate BASIC for the Altair.
These young men founded a new firm to market BASIC to other firms making
microcomputers. They called this firm Micro Soft.
8 Chapter 0000
Apple
At this same time, another group of young men, comprised of Steve Wozniak, Ron
Wayne, and Steve Jobs were looking at the Altair with the realization that a home
computer was a really cool idea.
The down side of this idea was that the hobbyist needed specific skills such as soldering.
So, Steve Wozniak sat down in his bedroom and created a circuit board, dropping it into
a wooden box, complete with its own version of BASIC hard-coded (fixed) into the
computer. Ron Wayne wrote the manual, and Steve Jobs marketed the intended game
machine to anyone who would listen.
Those listening included Paul Terell, the owner of the only computer store chain in
existence, The Byte Shop. One year later, the Apple II came out, and this time,
Wozniak put in 8 expansion slots. Expansion slots let you add functionality to a
computer.
VI
C/PM
To round out the early days of the microcomputer revolution, we need to introduce a
Seattle native and student of the University of Washington.
Enter Gary Kindall, computer science graduate of 1972. At this time, the pair who
founded Micro Soft was working part-time in the university district of Seattle.
Gary, like Paul & Bill had a love for the DEC computer. After completing a tour for the
Navy, Gary founded the company Intergalactic Digital Research.
One of his projects was to write an operating system for microcomputers called the
Control Program Microcomputer (CP/M). Gary incorporated features he liked from the
DEC operating system to the Control Program Microcomputer.
Because CPU designs need to be bonded to an operating system, Gary wrote a second
operating system specifically for the Intel Microprocessor, calling it Programming
Language for Microcomputers PL/M3.
10 Chapter 0000
VII
Intel
Intel currently was busy creating a fortune with the 4004 CPU that is used in calculators.
They were not a dominating factor in the supply of CPUs for microcomputers. Other
CPUs such as the Zilog Z-80, as well as offerings from Motorola dominated the field.
With this, CP/M from the renamed Digital Research, Inc. (DRI) became the dominant
operating system, with Microsoft (now without a space) being the dominant computer
language firm.
VIII
Killer App
All that was missing at this time was a task automation tool. It was so widely needed; it
was virtually guaranteed that everyone would buy it. Today that market hole is
known as a Killer App. A man named Donald Bricklin unveiled the first killer app.
Donald Bricklin was a programmer of (what else?) DEC computers. Seeing the
possibilities for microcomputers, he came up with the idea of creating a program to
build financial models and have the computer crunch (process) the numbers. The basic
idea was simple, create a traditional accounting worksheet but hook everything
together so that values you change in one area is then reflected in the rest of the
worksheet.
Think of it as a calculator program that would show results visibly on the screen. And,
that is the where name of the program, VisiCalc, came from. By now, you probably won't
be surprised to hear that Donald didn't own an Apple II, or for that matter, any
microcomputer at all. The VisiCalc development project began using a simulator
(approximate) running on a, you guessed it, DEC.
Businesses large and small could create what-if scenarios and business forecasts in a
matter of hours, as opposed to weeks or months, for only a few thousand dollars.
Naturally this phenomenon caught the attention of the world's largest computer supplier,
IBM.
12 Chapter 0000
IX
IBM
What to do about "toy computers" caused some rather heated discussions within IBM. In
the end, IBM management selected "Dashing" Don Eldridge to create an IBM
Personal Computer. They sent Don down to the Florida coast, where if the project
failed, one swift kick would drop the project into the ocean.
Various rumors and stories about what happened between Gary Kindall and IBM
abounded. The bottom line is that the two companies did not make a serious connection.
IBM had Bill Gates sign a strict Non-Disclosure Agreement and asked him what he
wanted to see in a microcomputer. Always ready with an opinion, Bill told them exactly
what he wanted to see. After the misfire between IBM and DRI, and since Bill and
company had previously sold IBM the computer language BASIC, it was a good
opportunity to see if the Seattle company had an operating system.
Not one to miss an opportunity, Bill Gates answered with an affirmative. While
Microsoft didn't actually have an operating system, a poker buddy named Tim Patterson
did.
DOS
Tim Patterson was working hard at a small firm called Seattle Computer Products. In
reality, Seattle Computer Products wasn't in Seattle at all; it sat in a little Seattle suburb
on Industry Drive in Tukwila, Washington. Their flagship product was a Computer
Aided Design and Drafting (CADD) computer. Tim had a challenge. LP/M4 crashed
when it attempted to work with the then super fast 8086 CPU. So, Tim created Gazelle
DOS.
Bill offered to buy Gazelle DOS, outright, neglecting to mention his relationship with
IBM (remember the NDA?). In the end, Microsoft came to an agreement with Seattle
Computer Products, which allowed Tim to keep Gazelle DOS for his own use as long as
it bundled (sold) with a CPU, and received $50,000.00 to allow Microsoft's use of DOS.
(Tim later used that funding to start Falcon Computer Products, with the first item being a
very fast hard drive controller.)
Reportedly, Microsoft licensed DOS to IBM for one dollar per copy, with the provision
that Microsoft could license DOS to other vendors.
L/PM footnote from C/PM sub-head footnote Intel and DOS appear later
14 Chapter 0000
XI
IBM PC
IBM released the PC Model 5150; it had five expansion slots utilizing an eight-bit data
path. It could accept up to four floppies (5.25 inch disks), or work with an audiocassette
to store programs. The marketing clout of a PC with IBM's name on it surprised even
IBM. However, it was a few third-party vendors that made the IBM PC a run-away
success.
Lotus
A small company in Massachusetts created a rival to VisiCalc to run on the 5150. That
firm was called Lotus Development Corporation. The VisiCalc competitor was known
as 1-2-3. What was missing in the IBM line up was an impressive video display. You
either had to settle for monochrome and text, or low color, low-resolution color
graphics.
Hercules
A firm called Hercules created a monochrome card (video card) that could display
mono graphics. This is just what 1-2-3 needed to convert what-if worksheets into
graphic representations.
The Combination
Combining these elements created such a market force that IBM had to release a new
version, the PC-XT. This unit had a larger power supply, which it needed to support the
built-in 10MB hard drive. Within two years, IBM went from zero to hero in personal
computing. At that time, it appeared that IBM had won the race. The industry wasn't
keeping a close enough eye on the Apple.
Xerox
Back in California, there was more to computing than Apple. In fact, right around the
corner from Apple, the copier company Xerox had a little think tank known as Palo Alto
Research Center (PARC). Since Xerox had invested $1 million in Apple, the firm didn't
have major heart burn letting a few key Apple executives see what they were doing at
PARC.
What Steve Jobs and a few others saw in action was science fiction stuff like: the
computer mouse, graphical icons on the screen to represent programs and computer
management, laser printing, and computer networking (what would become Ethernet),
as well as other Buck Rogers sci-fi stuff.
Xerox management didn't really care about letting Apple see the cool work at the think
tank, however, some of the managers at PARC did! After all, they were a copier company
right? What came out of this was the Lisa. What was the Lisa? At $10,000.00 computer
that was a marketing bomb. Almost no one could afford it.5
Mac
This 52-pound monster may have been on the market for about one year. Despite the illfated future of the Lisa (the last of them became landfill near Logan, Utah), it spawned a
successful offspring called the Macintosh.
http://www.members.nbci.com/applemuseum/computers/iigs.html
16 Chapter 0000
Compaq
IBM was quick to get its PC to market, in less than two years, in part because Don
Eldridge, who was in charge of the project that had designed the 5150, IBM's first
successful PC, using mostly off-the-shelf parts, an idea successfully demonstrated with
the Altair. In fact, only two parts were of IBM design, the BIOS (Basic Input/Output
System) and keyboard decoder. This left many firms attempting to cash in with
computers known as "compatibles". How compatible did compatible mean? Well, it
created untold grief.
The best-known truly compatible computer came from a Texas company known as
Compaq. Besides the fact that it worked, Compaq had one big thing going for them, like
an Osborne 01, and the keyboard could be attached over the front of the floppy drives
and monitor. This enables it to be hauled around, with the same grace as hauling a
sewing machine. This was marketed as a portable PC, and a more accurate term that
soon became popular. These and others became known as "luggable" computers.
18 Chapter 0000
XII
GUI
20 Chapter 0000
Types of Switches
Even with todays Plug and Play, sometimes a device needs to have a
configuration set. There are two typical ways to accomplish this. One
is with switches the second is with jumpers.
22 Chapter 0000
Parking Jumpers
When you want to keep a jumper attached to a device without it
completing a circuit, you hang it (attach it) on one connector. This is
sometimes called parking the jumper.
PS2
Personal computer sales that were generated mostly from companies other than IBM
skyrocketed.
In 1986, Compaq bested IBM by being the first company to release a PC based on the
386. One year later, IBM responded to the clone invasion by releasing a new computer
line-up, known as the Personal System 2 or PS/2 for short.
This new line of computers hosted an array of innovations including a new and
completely incompatible method of accepting expansion cards. This new expansion bus
was known as Micro Channel Architecture (MCA). IBM welcomed its competitors to
use MCA, by paying a 'mere' dollar per slot for every computer made. Nobody was
going to pay that, even if it was a 32-bit data bus.
EISA
Looking at millions of dollars in royalties, nine companies, called the Gang Of Nine6,
formulated an alternative to the MCA expansion slot.
The gang of nine took the royalty free expansion design of the IBM PC, known as
Industry Standard Architecture (ISA), and then, went down below the electrical
contacts of the ISA expansion bus to create a second row of contacts and called this
Extended Industry Standard Architecture (EISA).
The gang consisted of; AST, Compaq, Epson, HP, NEC, Olivetti, Tandy, WISE, Zenith
24 Chapter 0000
With this creation, the battle of the 32-bit expansion bus began.
VLB
The battle of the expansion bus left one group of manufacturers very unhappy. Who
would want to design and make video cards for two different incompatible systems? The
Video Electronics Standards Association (VESA) created the Video Local Bus (VLB),
as an extension of ISA in an attempt to improve video performance and make only one
type of video card.
The VESA-VLB bus was the slot to have when ordering your new 80486 computers.
While this offering greatly improved video performance, VLB died in the bus wars with
the introduction of a new bus from Intel.
PCI
While the bus wars loomed for several years, Intel ended it all by coming out with a
royalty free expansion bus known as the Peripheral Component Interconnect (PCI).
This made Intel's stockholders very happy. Why? Unlike the VESA-VLB, PCI uses a
chip set to control it, and, what does Intel make? Why chips of course.
So, here you have a new and very cool 32-bit expansion bus that will end the bus wars.
Oh, you need a set of chips to control the expansion bus? Well, we at Intel can help you
with that!! (Tcat has never worked for Intel!)
26 Chapter 0000
Sub System
In many ways, a computer system is like an airplane. They are both a collection of
spare parts flying in close formation.
In each case, they are a system that depends on the close timing and performance of a
number of subsystems. In the computer case, there are three major subsystems. They
are the CPU, RAM, and storage.
CPU
The first subsystem is the Central Processing Unit (CPU).
This is the brain that would be considered central command. In a desktop metaphor, the
CPU is you sitting behind the desk, deciding what to do next and controlling the flow of
action.
RAM
The second subsystem is the Random Access Memory (RAM). The computer
functions much like your desktop where you spread your work out.
How much RAM you need, depends on how much work must be spread out. The more
projects you have going at the same time, the more desktop you need. The same could
be said of RAM. A large program or even a series of small programs all consume this
resource.
Storage
The third subsystem is storage, such as a hard drive. Using the desktop analogy, the
hard drive or CD acts as a filing cabinet for your projects. Just as in the physical office,
the more projects you need to store, the more storage space you need.
Further, the more drawers you have to stuff things in, the better the chances are for a
particular file to be lost in the clutter. 'Spring cleaning' a large hard drive can be just as
joyous as cleaning an attic with 30 years of collectibles.
28 Chapter 0000
XIII
Electricity
Frequently misunderstood and often feared, electricity is a basic form of energy found
in nature. To understand the nature of electricity, begin by looking at the small unit
called the atom.
In the center of an atom is a nucleus.
An atom is composed of both positive and negative electrical charges.
Spinning around a nucleus are electrons. Electrons have a negative charge.
Protons have a positive charge, and, as the name implies in neutrons, they are neutral
or have no electrical charge.
Typically an atom has an equal number of electrons and protons.
Many of our elements found on Earth have the possibility to dislodge one or more
electrons spinning around the nucleus. When this happens the atom has more positive
than negative electrical properties, and, is known as a positive ion.
If a dislodged electron joins another atom, the electrical balance becomes negative
and is known as a negative ion.
Electrons Unleashed
Remember that an atom doesn't weigh very much, and the electron is a very small part
of an atom. This means an electron has almost no weight.
Electrons not attached to the orbit of an atom are known as free electrons.
A property fact of free electrons is they move at a very high speed (about 186,000 mi.
per second) traveling through metal, many gases, or a vacuum.
If an electron isn't flying around, it can rest on a surface.
When a few trillion free electrons are sitting around on a surface, this is known as a
negative static electrical charge
Electric Behavior
To have electricity, the free electrons need to have positive ions
waiting for them and a way to get there.
30 Chapter 0000
Conductors
Some materials hold onto their electrons very tightly and resist giving up their
electrons. These materials are known as insulators.
Types of Insulators
Glass, rubber, wood, and plastic
Other materials give up their electrons very easily. These materials are known as
conductors.
Types of Conductors
Copper, silver, gold, aluminum, and iron.
Completing a Circuit
At this point it is clear to you that electricity cannot move without a complete path.
When designing a circuit board, a common return path is created that is called ground.
Ground shows up prominently in all aspects of the PC and in this book.
Magnetic Fields
When electrons travel it is called electrical current. When this happens the electrons
create a magnetic field. This behavior can be useful for creating measurement devices,
or, it can muck up data signals.
32 Chapter 0000
Measuring Electricity
In many ways, electricity behaves like water. When stored up, it has the potential to go
somewhere. When it flows, it wants to take the path of least resistance. To further
understand how electricity is measured, let's compare it to water.
Think of a water tank filled with water mounted on a pole. The water in the tank is the
potential energy. If you hook a hose to the water tank, and open the valve, the water
will flow out the hose. The higher the water tank is on the pole, the faster the water
will come out. This is known as water pressure. When measuring electrical pressure,
the unit of measurement for pressure is called voltage, or, V in its shorthand. Work
performed by electricity is called power or, P in its shorthand. The measure of current
is called amps. The formula is known as E=IR7.
Remember the water valve you turned on to let the water out of the tank? You
controlled the amount of water with the valve by impeding the flow of water. In other
words, you gave resistance. In electricity, resistance is measured in ohms, frequently
expressed with the
symbol, and named after the guy who discovered the math
relationship between voltage, current, resistance, and power. The shorthand for ohms
is either or R.
AC/DC
Electricity can flow in either direction. If it only moves in one direction, it is called
Direct Current (DC). Remember the idea expressed earlier that as electricity flows it
also makes a magnetic field? Well if you spin a bunch of wires between two magnets,
(which you might do with, say a waterwheel) you get electricity that swings back and
forth between positive and negative charges creating Alternating Current (AC).
Multimeters
As stated above, the magnetic field can be used to measure electricity. Typically,
meters have more than one function since the relationship between resistance, voltage,
etc., is so closely bound. Devices with multiple measurement capabilities are called
multimeters.
DANGER!
The enclosed in a triangle on both meters indicate measuring dangerous
electrical levels.
CPS
So far, we know that electricity can flow as AC or DC. Who said the flow must be
continuous? It doesn't have to be. Further, there is no law of nature that demands that
voltage remain constant. It is technically possible to vary the number of oscillations
between positive and negative sweeps in AC, and to vary the amount of DC current or
even turn it on and off to create pulses.
34 Chapter 0000
When you have an AC current the number of sweeps between positive and negative is
measured in Cycles Per Second (CPS). Electricity in North America is delivered at 60
CPS, with the pressure of 120 volts. Sometimes this is expressed as 120VAC. In
Europe, the standard is 50 CPS and 240VAC. In computers, we describe CPS using
Hertz, or Hz. A device that operates at 1000 CPS is running at 1KHz.
Intelligent Life
Vary either voltage or the CPS, a process known as modulation occurs.
This is how radio and TV work. The stream is more or less continuous and
is considered an analog signal.
Taking DC and turning it on and off, or, pulsing the voltage is the basis for a digital
signal. For example, 5VDC could be a one, and 1VDC or even a negative voltage could
signify a zero.
With the basics of electrical properties complete, let's look at a few electronic
components.
Resistor
A resistor can come in a great variety of different packages. And regardless of the form
it comes in, the job of a resistor is to resist or limit current.
If you have ever seen a little tube with wires on both sides and little color rings around
the tube, you have visually identified a resistor.
The colors are values telling you the amount of resistance in ohms.
Capacitor
As with resistors, capacitors can also come in a great range of different packages. The
role of capacitors is to store electrons.
Remember the water tank? Think of a hand pump at a well. The water comes out of the
pump in spurts. If the water is held in a water tank, a valve can be opened, getting a
continuous stream of water. So, a capacitor can behave like a water tank to smooth
out uneven electrical flow.
36 Chapter 0000
Figure 12 Capacitors
Capacitors in PCs
Typical capacitors found on a systemboard or expansion card. Larger
capacitors are found in power supplies.
Transistor
Let's look at the transistor next. If you have water flowing through a pipe, and, you put
a valve in the middle of the pipe, you can control the amount of water that is flowing
through the pipe by controlling the valve. A transistor works the same way.
Electricity flows in one wire and out the second wire. Applying electricity to a third
wire, controls flow that acts as the valve regulating the amount of electricity that
flows.
In the simplest of terms, this is how an amplifier works. A small amount of electricity
is used to control a larger amount of electricity.
Diode
One final electronic piece is the diode. Again, diodes come in many flavors. A diode lets
electricity flow easily in one direction, and, offers resistance in the opposite direction.
Diodes can be used to filter out the negative swings of AC, leaving only bursts of
positive voltage. Take only the positive burps of power and send these to a capacitor,
and, what comes out is DC voltage.
38 Chapter 0000
LED
A specific type of diode is very popular in computer equipment. That is the Light
Emitting Diode (LED).
It is used to indicate something is on, working or busy. A pair of typical LEDs for
computers appears in Figure 15.
Figure 15 LEDs
Some LEDs can have a multi-color output indicating different states of operation.
Fuse
The last piece of the puzzle is still missing. While it is very low-tech in nature, it is,
nonetheless, invaluable. This is the fuse. A fuse can be as simple as a piece of metal held
in either glass or plastic.
The fuse conducts electricity readily in normal operation. If an issue arises in the circuit
causing it to draw more current than normal operation calls for, the excess current heats
the metal to the melting point so the electricity stops flowing.
40 Chapter 0000
With all the pieces of basic electronics defined, think about how much space is taken by
putting all these parts on a board to connect them together so they do something for us.
If we could get rid of some of the packaging and wires for the different parts, we could
have devices that are smaller, lighter, consume less electricity, and, even be more
reliable. This idea is explored next.
IC
The Integrated Circuit (IC) is the same stuff that made the capacitors, resistors, and
transistors possible. The difference is the parts are laid out to perform a specific
function, without all the packaging of the individual parts. The whole range of parts is
then placed in one package. Today, the insulation between the parts is literally only a
few atoms thick!
Due to so much being packed in such a small area, very small amounts of power are used
to keep heat to a minimum, and to prevent insulators within the IC from being overrun.
Today it is not uncommon to see voltage requirements of less than 2 volts for an IC. A
CPU is a good example of a VLSIC8.
At this point, you have seen how electricity works, and, the common components in use
to make electricity do something useful for us. Now it is time to apply some logic so the
computer can do some 'thinking'. How a computer thinks is the next topic.
Pop Quiz 0000.01
In May 2001 a research firm managed to successfully create an insulation layer that is
one atom thick. This is the ultimate goal following the current design approach. A new
method is being researched to keep advances continuing.
8
42 Chapter 0000
2. Dangerous/fatal ranges
3.
4. about 2 volts DC
44 Chapter 0000
The fourth column becomes the value of 8, the third column being a placeholder for
the value of 4 with the second column having the value of 2, and the first column the
value of 1.
If there is a zero in any column, it is off, and it is not counted.
Notice that the zero has to be there, or the values in the column would all be skewed and
the math wouldn't come out right.
If all the columns are on, the maximum value is equal to 128+64+32+16+8+4+2+1 or
255.
=255
VALUE
128
ON
64
ON
32
ON
16
ON
8
ON
4
ON
2
ON
1
ON
46 Chapter 0000
Bit - Byte
Each on or off is called a bit. Package 8 bits together, and, it is called a byte. Because
bytes are easy to work with, you will see packaging such as 8-bit, 16-bit, 32-bit or 64bit. If you think about how the base 10 number doubled with each move out in the
columns, 16-bit is considerably more than 8-bits and 32-bits is a very large base 10
number
Value = to
1 Megabyte
1 Kilobyte
1024 Bytes
Equal (=) to 1 million Bytes
1,024,000,000 Bytes 8192 Bits
Units = to
8,192,000,000
1024 Octets
Bits
32 Bits
8 Bits
4 Bits
4 Bytes
1 Byte
1 Nibble
4 Octets
4 Characters
16 Nibbles
1 Octet
1 Character
2 Nibbles
Byte
Character
Base 10
Base 16
Base 2
0
1
2
3
4
5
6
7
8
9
10
11
12
13
14
15
0
1
2
3
4
5
6
7
8
9
A
B
C
D
E
F
0000
0001
0010
0011
0100
0101
0110
0111
1000
1001
1010
1011
1100
1101
1110
1111
48 Chapter 0000
A Hex Shortcut
Of course we could let you go away from hexadecimal math without
showing you a quicker way of doing base 16 to base 10 conversions. This
method requires a bit more understanding of exponents. As you learned
about binary math, each column that binary number represents from right to left
increases in value by power of 2. Therefore the first column represents 2^0, or 1. The
next column represents 2^1, or 2. The column after that represents 2^2, or 4. The same
kind of "conversion chart" can be made with Base 16. Instead of using powers of 2, we
will instead use powers of 16. Building our chart from right to left our first column
would have a value of 16 ^0, or 1. Our next column would have a value of 16 ^1, or 16.
Our third column would have a value of 16^2, or 256. Refer to the charts to see how this
should be constructed.
163 (4096)
162 (256)
161 (16)
160 (1)
163 (4096)
C
162 (256)
C
161 (16)
3
160 (1)
F
Loose connectors
If you were required to know every connection type ever made, no body would be A+
certified. And for the test that the Real World throws, you may need to know about the
connectors pictured in Figure 0.17. These small connectors fit on pins found as part of a
systemboard.
Figure 19 Reset SW
If you are lucky, leads to a PC case are labeled. Notice that the Power LED skips the 2nd
pin.
50 Chapter 0000
XV
In this chapter, you saw the beginnings of IBM, and the birth of the PC with MITS and
its Altair. It was revealed to you how Apple thought they had a better idea and the
background of other key players and how they affect the PC today. You learned the
history of each major expansion bus and how to Identify them.
The concepts of electricity were examined along with common electronic components.
Finally, you learned how to work with Base 2 and Base 10 math, with the relationship to
Base 10 math, that is part of daily life.
52 Chapter 0000
3. Of the examples given which are conductive materials (choose all that apply)?
A. Aluminum
B. Gold
C. Iron
D. Your finger
E. All of the above
4. Which of the following are insulators?
A. Rubber
B. Glass
C. Plastic
D. All of the above
E. None of the above
54 Chapter 0000
9. Taking DC Current and pulsing it is the basis for what kind of signal?
A. Analog
B. Satellite
C. Digital
D. RADAR
13. In Computer math, including Base 2; 10; and 16, is 0 a valid number?
A. Yes
B. No
56 Chapter 0000
15. In Binary, Base 2, Math what does the number 1101 stand for?
A. 3
B. One Thousand one hundred one
C. 13
D. 32
18. What does the Hexadecimal number 220h stand for in Base 10?
A. 220
B. 512
C. 544
D. 666
19. True or false the jumper sleeve has a piece of metal inside that completes a circuit so
that electricity knows where to go?
A. True
B. False
58 Chapter 0000
3. Of the examples given which are conductive materials (choose all that apply)?
*A. Aluminum
B. Gold
C. Iron
D. Your finger
E. All of the above
Explanation: Answer: E. All metals conduct to some degree. You finger conducts due
to the water, carbon, and iron in your blood.
60 Chapter 0000
9. Taking DC Current and pulsing it is the basis for what kind of signal?
A. Analog
B. Satellite
*C. Digital
D. RADAR
Explanation: Answer: C. Taking DC current, which has a constant voltage, and pulsing
it creates 1s and 0s for binary counting.
13. In Computer math, including Base 2; 10; and 16, is 0 a valid number?
*A. Yes
B. No
Explanation: Answer: A. Zero is always a valid number.
62 Chapter 0000
15. In Binary, Base 2, Math what does the number 1101 stand for?
A. 3
B. One Thousand one hundred one
*C. 13
D. 32
Explanation: Answer: C. Count up the columns. 8+4+1=13.
18. What does the Hexadecimal number 220h stand for in Base 10?
A. 220
B. 512
*C. 544
D. 666
Explanation: Answer: C. Refer to your Base2 and Base16 charts.
19. True or false the jumper sleeve has a piece of metal inside that completes a circuit so
that electricity knows where to go?
*A. True
B. False
Explanation: Answer: A. Jumpers complete an electronic circuit when they are placed
over two metal posts.
"If you have no will to change it, you have no right to criticize
it."
--Anonymous.
Chapter 0001:
Power Supplies - System board
The objective of this chapter is to provide the reader with an understanding
of the following:
1.1 Identify basic terms, concepts, and functions of system modules, including
how each module should work during normal operation and during the boot
process.
System Board
Power Supply
Firmware
CMOS
1.2 Identify basic procedures for adding removing field replaceable modules for
both desktop and portable systems.
System Board
Power Supply
1.3 Identify available IRQs, DMAs, and I/O addresses and procedures for device
installation and configuration.
Standard IRQ settings
Hexadecimal/addresses
66 Chapter 0001
4.3 Identify the most popular types of motherboards, the components and
architecture (bus structures and power supplies)
Types of motherboards:
AT (full and baby)
ATX
Bus architecture
ISA
PCI
AGP
USB (Universal serial bus)
VESA local bus (VL-bus)
4.4 Identify the purpose of CMOS (complementary metal-oxide semiconductor),
what it contains and how to change its basic parameters.
Boot sequence
Date/Time
Plug & play BIOS
68 Chapter 0001
Introduction
The few pages of this chapter are historical in nature, and not required for the A+ Core.
However, the information contained here will create a perspective of understanding both
legacy equipment as well as where we are in current technology. Few sub systems in the
PC have gone through the changes as quickly as the CPU.
II
At first glance, it may seem odd to discuss power supplies and system boards in the
same chapter. In fact, the two are closely related, because power supplies are the energy
source for the computer. You will see the choices between different designs impacting
the system board and the form factor.
It will become clear how the chipset impacts the size of the system board along with
other factors.
The dance between the BIOS interacting with the CMOS to create a POST routine will
be revealed. You will discover how to upgrade BIOS that is not capable of meeting your
needs.
Further details about various expansion buses and the impact of the choices you can
make will become clear.
The sometimes-disastrous interaction with drivers will be demystified so you may
become a driver magician.
Various levels of cache and the interaction between them and how they affect
performance will guide you though an area frequently fraught with inaccuracy, thus
giving a clear understanding to utilize to your benefit.
The ability to improve performance by properly applying techniques such as Bus
Mastering and DMA will be explained.
Confusing issues such as PCI Steering and Memory mapping are explained. So lets get
started.
70 Chapter 0001
III
Power Supplies
The term power supply in the PC is actually a bit of a misnomer. Todays power supply is
a sealed box, with a cooling fan and power leads. Inside the power supply box is a step
down transformer that converts 120 or 240 volts AC to approximately 12 and 5 volts
AC. From this point, the AC voltage is run through four diodes filtering off the negative
voltage. This is known as a bridge rectifier. Newer power supplies use switching
transformers in place of the diodes to accomplish this. They are known as switching
power supplies.
From there, the power is passed to a capacitor, which outputs clean DC power. Inside
the case of the power supply are only two more parts. They are a fuse and a fan. The
fan is used to pull air through the computer case as well as to cool components of the
power supply itself.
72 Chapter 0001
IV
AT
Regardless of the type of PC computer, these components are there. What makes power
supplies different is the actual form factor or design that contains these components.
There are two major types of form factors for power supplies, AT and ATX. The first
one is known as an AT power supply. Two features distinguish the AT power supply
from its relative. An AT power supply has a mechanical switch used to control the
availability of AC power.
The second feature that makes an AT power supply instantly recognizable has the
power connectors for the motherboard, sometimes called a system board. An AT
power supply has two power connectors for the motherboard - labeled P8 and P9.
Note: I is power for a 3.5 drive and III is the visual indication for the 5.25 drive
power. The mechanical switch for the AT is seen next to III. Two connections for
power with the AT P8 indicated with IV, with P9 designated by V.
P8 & P9 Grounds
When connecting the power supply to the motherboard, make sure the black
ground leads are side by side. This insures P8 & P9 are properly arranged.
Computer manufacturers faced one challenge when working with the AT power
supply. Remember the fact that the AT power supply uses a mechanical switch to
supply power. That makes it impossible to turn the computer on remotely. To
accomplish remote power-on, a small redesign on the power supply was needed. That
redesign is found in the ATX power supply.
AT is not ATX
Since the AT and ATX designs have different power connectors and case
configuration they are not interchangeable.
74 Chapter 0001
ATX
The ATX power supply has a small power lead going to the case and the switch is used
as a signaling device to turn the computer power supply on or off. By using signaling,
other signals may be installed in the computer such as on a Network Interface Card
(NIC).
This feature is marketed under the handle Wake On LAN (WOL). This makes it
possible for a network administrator to wake up an entire office full of computers,
download updates, for example a new version of word processor.
This saves considerable labor because the administrator did not have to visit each PC.
Further, productivity is not lost because computers are upgraded when workers were
at home. And users don't have the frustration of incompatible file formats between two
different versions of software.
Another feature of the ATX power supply is there's only one connector for the power,
making it impossible to reverse the leads as can be done with P8 and P9 on an AT
power supply.
Of course differences in how the power switch operates means a different computer
case must be used. With the rationale of the two different power supplies, cases are
known as AT or ATX cases.
That isn't the only difference. You just learned the connectors between the two power
supplies are different. That means the systemboard must be mated to accept the correct
type of power connector.
76 Chapter 0001
Form Factor
Each design layout is known as a form factor. While not all computer systems fit the
categories listed below, most do. Starting with the largest design, it is time to examine
the AT.
Depending on the case design, several different form factors may be used when
creating a system board. In late 1984, IBM released the PC- AT, featuring an 80286
CPU. Because of all the electronics involved, the system board was BIG, over a square
foot of systemboard. The original AT power supply had inverted L- shape to
accommodate space for the system board.
Baby AT
About 1997 electronics had shrunk in both size and heat output so the original AT form
factor was no longer needed. Anytime a manufacturer can use fewer raw materials, they
do so, to reduce costs. The newer smaller form factor was called the Baby AT form
factor. The footprint of the Baby AT is usually about 8.5"13". This is a bit smaller than
the original 12"13" AT form factor.
ATX
Today the ATX design reigns. The change is not so much about square inches. It is
about how the square inches are laid out. The orientation on the systemboard is rotated
90 degrees. Cable connections for drives are closer to the part of the case where the
drives are usually mounted, and the CPU is closer to the power supply, and therefore
the cooling fan.
NLX
In some situations, such as replacing a cash register with a computer to create a Point
Of Sale (POS) system, space is at a premium. In these situations may call for a system
board with no CPU on it. Instead all electronics are placed on cards with the system
board be nothing more than a passive backplane. This also allows for simplified repair
and upgrade ability.
With the physical description of what is inside the case complete for this chapter, it is
time to review the components of the system board.
Chipsets
Recall in Chapter 0000 how IBM quickly got the IBM PC ready for market. The
company used almost entirely, 'off-the-shelf' electronic components. Major parts of the
system board were consumed by purchasing a standard component known as Transistorto-Transistor Logic (TTL). IBM arranged these TTLs on the system board to create a
computer in very little time.
One inventive company looked at all the real estate on the system board being used by
the TTLs. They engineered the exact same circuit design, and put the TTLs,
resistors, capacitors and transistors into five different IC's. With this stroke of
brilliance, hundreds of components were reduced to five. This may be the most popular
form of IC in use today. This type of IC is known as an Application Specific Integrated
Circuit (ASIC).
Figure 24 Chipsets
Note: One of the first chipsets (from Suntac) was created to replace the discrete
components found on the IBM AT.
Prior the to ASIC, ICs were designed to perform a specific function, in a very generic
manner. For example, the 555 IC was a clock chip. But it was not specifically built to
power a jumbo digital clock, with alarm, complete with the electronics to drive a display.
An ASIC would have the 555 design along with the necessary electronics to drive display
and a alarm buzzer.
78 Chapter 0001
As computer requirements became more sophisticated, the ASIC grew more complex,
and the number of chips needs became less. The chipset of 5 chips became one chip.
Today two chips not only control the computer, they can have features that used to
require expansion cards be added into the computer. For example, on-board sound is
now a common feature in a chipset.
As time has marched forward, the chipset has been reduced in count, dropping to either
2 chips or even the minimum count of 1 chip.
Northbridge
The first of the two chips that make up the chipset of a motherboard is called the
Northbridge. The Northbridge handles data buses, cache controllers, system buses and
CPU memory.
Due to the nature of rapid changes in these sub-sections, Northbridge chips see updates
more frequently than it companion, the Southbridge.
Southbridge
The Southbridge handles peripherals such as printer ports and drive controllers.
While there are advances in these areas, they do not happen as quickly as the fluid state
of the Northbridge.
Because it is possible to have changes occur more quickly in the Northbridge the name
flag for the chipset resides with the Northbridge side.
Technically, it is even possible to have two different chip makers between the
Northbridge and Southbridge. Different names you may run into include: Intel, AMD,
SiS and VIA. This sampling is by no means exhaustive however these makers are
popular today.
You and I cannot do too much configuration to a chipset. The most we get to do is turn
the on-board sound on or off, and other small things. In our next section we examine
something really configurable.
2. The AT power supply uses a ___________ switch. ATX uses a ________ ________.
3. The generic names for PC chipsets today are the ___________ and the ________.
5. The ___ power supply can be turned on and off via ___________.
80 Chapter 0001
3. Northbridge, Southbridge
5. ATX, software
VI
CMOS
This mouth full of acronym is pronounced see-moss, and is short for Complementary
Metal Oxide Semiconductor (CMOS). What makes a CMOS different from an ordinary
transistor is that it requires less power. A PC uses CMOS to hold data when the
computer is turned off.
The CMOS retains data by using a small battery on the system board.
Sometimes, because the information in CMOS has retained while the computer is
turned off, it may be referred to as Non-Volatile RAM (NVRAM). This is true only as
long as the battery is in good shape and connected.
The question arises, what sort of information must be retained when the computers
turned off? How about the current date and time? Do you have one floppy or two? Are
your floppy drive(s) 5.25" or 3.5"? So tell me. Or more precisely, your computer wants to
know what kind of hard drive is that anyway? These are all questions that must be
asked, and answered, before you can do anything with your computer besides look at
it sitting there.
The idea of a CMOS has been around since the AT or 80286 computers. Before then,
switches or jumpers had to be set to tell the computer more about what was in it when
it woke up with power. Today's modern computers have a bit more data in the CMOS.
This information is titled: Extended System Configuration Data (ESCD). The ESCD
has a special part of CMOS where certain settings are held. These settings are a bit
more advanced than the date and time. Modern expansion cards use what is known as
plug and play to simplify installation. Plug and play is discussed in detail later on. The
CMOS works in tandem with another critical part of the PC, the BIOS.
82 Chapter 0001
VII
BIOS
If you think it your PC is having a brain, then the BIOS is what gives it a personality.
The Basic Input/Output System (BIOS), which is pronounced bye-ose, contains software
that has been written to Read Only Memory (ROM).
Because ROM is physical, and physical is hardware, but software, is a collection of
commands, the marriage of the two is sometimes referred to as firmware.
The BIOS is typically quite easy to locate, because the manufacturer labels it. The
most popular vendors are: Award, American Megatrends, Inc. (AMI), and Phoenix.
Several life times ago (about 12 standard calendar years), BIOS was literally a ROM
chip. The manufacturer would write the instructions needed to support a type of
video card, floppy drive, etc. This was great, until someone wanted to install a new
widget, for example, a 3.5" 1.44Mb floppy that we all take for granted today. No, 15
years ago, the 3.5" 1.44Mb floppy didn't really quite exist. So installing one, say 14 years
ago, meant you had to order a new ROM BIOS, pull the old one out, and put in the new
one with the code in it to support this pile of data that could fit in your shirt pocket.
EEPROM
After several improvements on this idea, today we have the Electrically Erasable
Programmable Read Only Memory (EEPROM).
Now when you want to add something to a computer that isn't supported in BIOS, we
simply electrically erase the old data, and insert new data to support your new toy.
This process is called Flashing, or Flashing the BIOS.
BIOS Lobotomy
The one small drawback to this process is this. Notice in the paragraph
above, you must erase (completely) the old software code before you can
add new code. If you either install the wrong set of code for your
computer, or if you're really unlucky, lose power in the process, you have BIOS that is
brain-dead.
Due to this 'little challenge', in addition to the fact that some viruses (destructive code)
have been written to attack BIOS, motherboard manufacturers such as Gigabyte are
looking at a second BIOS that backs up the first BIOS.
At this point, your computer has a 'personality' (the BIOS) and CMOS entries to tell it
what it has.
It is time to review what happens when you first turn your computer on.
Pop Quiz 0001.01
84 Chapter 0001
3. BIOS
4. audible, visual
5. Dead BIOS
VIII
POST
When you first turn on the PC, the system memory is empty, and even what the PC is
made of, is not known. The system goes through a system boot sequence. While this
process may vary slightly, overall it goes like this:
The power supply converts the AC power to DC, and starts charging capacitors.
When the power supply is fully charged that sends a Power Good signal to the system
board.
The CPU manufacturer pre-programmed the processor to look for the start of the
BIOS boot sequence. The location of the BIOS the CPU is looking for is specifically set
right at the end of system memory.
This allows for changes to be made in the actual size of the BIOS. Note that this point
is only a reference point, with the location being instructions to jump to an actual
starting point of the particular BIOS.
The first thing BIOS does is queries the CMOS about what is supposed to be in the
PC. Then the Power On Self Test (POST) begins. This is a built-in diagnostic routine
to ensure all parts and pieces are ready and accounted for.
Suppose you just installed a new video card, and in your excitement you neglected to
fully press into place (seat) the video card.
The POST sequence cannot find a video card that isn't there. Without a video card it
would be difficult to display an error message on a video screen. With this in mind, the
BIOS sends a code through the PC speaker. Note that the codes vary from one BIOS
manufacturer to another.
Assuming you have realized the error, powered off the computer, properly seated the
video card, and re-started, a BIOS startup screen is displayed.
This screen displays the BIOS manufacturer and version number, along with the date
of creation, and the one or more keys that must be pressed to get into the BIOS setup
program. The particular key or keys must be depressed at this time to get into the
setup program.
86 Chapter 0001
If the setup program is not entered, the System Configuration Summary is displayed.
This data includes the:
CPU type
Math coprocessor
Clock speed of the processor
Style and capacity of all detected floppy drives
Hard drives
CD-ROM
Base memory
Extended memory
Type of memory
Display type
Serial or communication ports
Parallel or LPT (printer) ports
From this point the PC begins to attempt to find a device that has an operating
system. That could be a floppy or more commonly, a hard drive.
Once that device is found, the Operating System begins to load. If everything is in
order, you are ready to being your work or play.
With your computer ready to assist you with your work or play, the question arises, is it
up to the task at hand? Some of that answer depends on what capabilities your PC has.
Curiously enough, that is the next topic.
Bus
While the term bus has a slightly different meaning in a PC than it does on a roadway, the
concept is identical. A Greyhound bus leaving Seattle and heading to Los Angeles makes
stops in towns along the way to pick up or drop off passengers.
Inside a PC (or even a network of computers) the bus is a transmission path that drops
off or picks up signals (instead of people) from every device (instead of town) attached
to the line.
If you were on the bus from Seattle to Los Angeles, you would still stop at Olympia,
Washington. Since Olympia was not your destination, you ignored it. In the same
fashion, devices only pay attention to signals addressed to them. If the signal is not
addressed to the device, it is ignored the same way you disregarded the stop in Olympia.
Just as Greyhound has more than one route, the PC has more than one bus. One area
that had been receiving a great deal of attention is the Frontside Bus.
I/O Bus
Breaking down In/Out abbreviated is I/O. The I/O bus includes communications
(serial) ports, parallel (printer) port, keyboard port, and today Universal Serial Bus
(USB). The I/O bus is the slowest of the bus system in the PC today. For those of you
wondering why not just make everything fast? The simple answer is, "Speed costs. How
fast do you want to go?" It would be money wasted to create a PC where someone could
type at a million words a minute.
If you are wise to the ways of the PC, you may be wondering what about data bus and
address bus? These two buses are discussed in Chapter 0011.
Just a few paragraphs ago, in Backside bus, a reference was made to a special type of
RAM. It is time to flip to the backside and see what is happening back there.
The original IBM 5150 had one thing in common with the other PCs of the day. They
were all excruciating slow. As the components got quicker, some parts became more
capable then other parts. This observation became one of Tcat's laws of Engineering.
"All computers wait at the same speed".
88 Chapter 0001
This was not good for Intel. Who would purchase a new, faster CPU if it had to wait for
something to do?
By analysis of how a computer works, it was found that, in general, if a CPU needed a
piece of information now, there is a high likelihood it will be needed again, real soon.
So, if the CPU designers created an intelligent process to determine what might be
needed again soon, they could put that data in super high-speed memory. By limiting
the amount of super-high speed memory, they could keep costs low, to only one or two
limbs (say one arm and a leg), yet improve the working time of the CPU. That is the
concept known as cache.
Cache
Todays PC has several layers of cache. The first level has the catchy title of Level 1
cache. This is the fastest hunk of memory in your computer. Its easy to find the
general location of it, because it is inside your CPU. When your CPU makes a request,
it looks at the Level 1 cache. If what the CPU needs is there, it takes it without ever
leaving the CPU. Finding what it is looking for is known as a hit. Naturally, the lack of
a hit is a miss. The 80486 CPUs were the first CPU with cache built-in. This explains
why if you read other technical works, you may find a reference to Level 1 cache being
outside of the CPU. Before the 486, the first level of cache had to be outside the CPU.
If we have a miss in the L1 cache, the next place to look is the Level 2 cache. This is
where the Backside Bus comes into play. In earlier designs, the Level 2 caches had to
fight for time with other buses, like the expansion bus. This became a digital version
of rush hour traffic. And since the CPU was left waiting, well you already read that
outcome. So, the solution became to put a Backside Bus that typically runs at full
processor speed just to the L2 cache.
Before closing on cache. Yes, there can be a Level 3 cache. And typically, hard drives,
CD-ROMs, almost all drives have some cache. Cache works as well as it does
(typically more than 90% of all CPU request are in Level 1 or Level 2 cache) because the
design engineers have developed some really slick routines, known as algorithms to
determine what is really needed to keep the CPU busy.
Almost done with buses. There is one more bus that arrived only a few years ago. As you
will see later in this chapter, the video sub-system has been starved for data. This isn't
good, since this affects the users perceptions. Leave to Intel to address this quandary.
AGP Bus
Originally the video card had to compete with other devices on the same bus, causing
saturation. Intel designed an expansion bus structure as a slot (the slot is seen later in
this chapter) known as the Accelerated Graphics Port (AGP).
AGP is a port and not a bus, because it cannot be expanded, AGP does have its own
'pipe' between the AGP and the CPU and the main system memory. And the bus
speed is faster than the rest of the expansion bus. AGP was a hit with game players
who needed the performance boost.
AGP design calls for a combination of on-board RAM on the video card, as well as
borrowing some of the PCs main memory for some processing.
When the AGP card is not using the main memory, it is released for use by other
applications.
When AGP was released it was available as both a 1X and 2X specification. Two years
later Intel released a 4X APG standard. Currently in late draft format is the 8X AGP
specification.
The physical slot for 8X remains the same, changing only in how signaling occurs.
AGP Specification
1X AGP
2X AGP
4X AGP
8X AGP
Bandwidth
266MBps
533MBps
1.07GBps
2.14GBps *Draft standard
90 Chapter 0001
Expansion Buses
Over the past 20 years, expansion buses have seen a series of "turf wars" in mind
numbing propaganda. For a number of years the outcome was mass confusion. Intel
brought peace to the battle of the expansion bus in the end.
Today's A+ core test reflects this outcome. And it is still possible to find computers with
these earlier buses.
The older bus styles today are considered legacy architecture. And they have been
included here, so if you see one, you're not saying, "what's that?" Our brief tour begins
with the original IBM PC Model 5150.
ISA
The first bus was called Industry Standard Architecture (ISA). This was a simple bus,
and the Model B had five expansion slots. Each one offered an 8-bit data path, and each
card had to 'wait its turn,' if another expansion card was busy with the bus.
It quickly was realized by IBM that it was business, not homeowners who would buy the
5150. That led to the Model Z, which featured 8 expansion buses.
When IBM released the AT, the corporation extended the ISA slot with an extra
connector to make the ISA an 8 or 16-bit slot. By just adding the extra connector, the
slot was backwards compatible with 8-bit cards.
To this day, the 16-bit ISA slot runs at 8Mhz. What 8Mhz means in real delivery of
data on the bus is seen in Chapter 0100. In plain English, and in todays world, the nice
word is legacy. To be more blunt, today it is considered unacceptably slow.
Figure 27 80286 with six 8/16-bit slots and two 8-bit slots.
92 Chapter 0001
MCA
By 1987 IBM was getting a bit annoyed. Here 'the real computer company' was getting
its clock cleaned by companies like Compaq, PC Limited (now Dell), Heathkit/Zenith,
along with hundreds of 'garage shops' that were purchasing ' white box' parts and
assembling computers with better performance at a lower cost. IBMs market share was
dismal, and getting worse. IBM, like Ford, thought they had a better idea.
That better idea was the PS/2 computer series, with all but the lowest end model
featuring a new expansion bus known as Micro Channel Architecture (MCA). As we
saw in Chapter 0000, IBM was going to let everybody have MCA, for a price. Now while
everybody agreed that MCA was a superior alternative to ISA, just about nobody was
going to pay that kind of money ($1. per slot, per system board!). Not even for 32-bit
performance.
EISA
Nine third party 'heavyweight' computer manufacturers got together and created the
competitive answer to MCA. That answer was Extended Industry Standard Architecture
(EISA). This open alternative was backwards compatible to 8-bit ISA as well as 16bit ISA cards. Transfer rates screamed along at 33 MB a second. Chapter 0000 Fig. 0.6
has a side-by-side picture of ISA and EISA.
VESA-VLB
As it was revealed in Chapter 0000, makers of video cards did not want to be making
cards for ISA, EISA and MCA.
What formed were the Video Electronics Standards Association (VESA), and the VESA
Local Bus (VLB). The VLB expansion bus was backwards compatible at the ISA
business end. The extra socket was located at the other end of the card.
Figure 29 VESA-VLB
Note: VESA-VLB featured standard ISA as seen on the right, with a new high-density
connector as seen on the left.
VLB Hell
About the same time VLB was becoming popular, the price point for
computer components was dropping rapidly, in part to inexpensive off
shore manufacturing. One outcome was very inexpensive cases, which
were not shaped with great precision.
After working on a system using VLB, (and successfully testing) then installing a screw
to retain the card, a poorly bent case would slightly pull up the card from the high
density VLB connection, disabling the video! This can be the cause of endless frustration
until the cause is found. The solution is to remove everything from the case and bend the
metal chassis until it is truly square.
94 Chapter 0001
PCI
Once again, looking back at our brief history tour in Chapter 0000, it was revealed that
Intel ended the bus wars for a number of years by releasing the Peripheral Component
Interconnect (PCI) expansion bus.
The original PCI bus was similar in design to VLB. With the advent of PCI Version
2.0, it was no longer a local bus, as it got its own data path, and designed to be
independent of the CPU design. Most PCI buses operate at 33Mhz, offering a 32-bit
bus using 124 pins. The extra pins are used for power and grounding.
Adaptec has been the champion of a 64-bit PCI11 bus, utilizing 188-pins, and being
backwards compatible with regular PCI by extending the slot, just as in 8-bit to 16-bit
ISA. Not many manufacturers of PC equipment have jumped on the 64-bit PCI
bandwagon promoted by Adaptec, although it is seen in data communications equipment.
Because this card is so rare, take a look at one below, compared to a regular PCI card.
(At least this way you can say you have seen one).
While Adaptec has not been entirely successful with creating a market standard for 64-bit
PCI, the point they are making has not gone unnoticed. Advances in RAM, CPUs and
even storage have reached the point where the once waiting for other sub systems are
close to saturating the PCI bus. To that end, look for PCI-X.
11
http: // www.adaptec.com
96 Chapter 0001
PCI-X
An extension of PCI running at speeds up 133MHz - burst transfer
exceeding 1 GB /Sec. Using 64 bits with increased fault tolerance
attained by tighter bonding to the Operating System. This is too new for
the A+ 2001 test, however you can buy this today from system board manufacturers such
as Asus.
Expansion hardware. The device you are adding must be Plug and Play capable.
Being capable means the device must be able to Identify it when requested and be able
to accept resource assignments.
NT4 contains a sub-folder under the Drivelib folder called PnPISA. By standing on
your head, and following the odd instructions in Readme, ISA based Sound Blaster cards
are detected.
12
System BIOS. In Plug and Play, BIOS is the conductor of the quartet. Program
routines collect information about the devices to determine which device should get
what resources. The BIOS communicates this to the Operating System.
Operating System. For an Operating System to use Plug and Play devices it must
install software that interfaces between the device and the main Operating System,
automatically. These program routines are known as Drivers.
Sharing IRQs
Prior to PCI Version 2.1 all expansion cards needed their own IRQ (if a
device was to be used at the same time another device with the same IRQ
would be in use). With the release of PCI V.2.1 that is no longer the case.
Expansion cards can share IRQs.
PCI Steering
Assuming a modern system board, beginning with Windows 95B (not Windows 95
Release 4.00) a new feature called PCI steering was introduced. For example: An ISA
sound card uses IRQ 9. You install a PCI card that would like to use IRQ 9.
PCI Steering will:
98 Chapter 0001
The only challenge with PCI Steering is if one device goes haywire, the whole
scheme falls apart like a house of cards.
IRQ Steering
IRQ Steering only works on PCI, not ISA.
AGP
Earlier in this chapter you saw a view of AGP from a port perspective. While not
everyone agrees that AGP is a bus, it certainly is a type of expansion slot. Note that in
many cases a PCI slot is near the AGP port. The design is such that, if you use the AGP
slot, you cannot use the PCI slot next to it, or vice versa.
PCI or AGP
You can use the PCI slot OR the AGP slot, and NOT both at the same
time.
AMR
The Audio/Modem Riser (AMR) is another Intel innovation. This little slot can be used
to give the end user an inexpensive solution for sound and modem functions.
Physically partitioning the analog portion to the riser card, and putting the digital
functions on a chip set achieves this.
This solution lowers costs, however it consumes CPU time. Sometimes these modems
are referred to as software-based modems. Software-based modems are acceptable for
computers that have CPU time idling, and are not recommended by your authors for
servers or other computers doing resource intensive work.
CNR
(Communication and Networking Riser) CNR is a variation of the AMR from
Intel. However there are differences. They include:
CNR does not have to give up a PCI slot
CNR is not backward compatible to AMR
Many of the parts and pieces of a computer have been laid out here. The question arises,
how do all of them work together in a logical order? It has been said that time is nature's
way of keeping everything from happening at once. This is true in a computer as well.
A computer can get really busy doing what you told it to do. And as we have seen, this
collection of spare parts flying in close formation is a bunch of simple sub-systems tied
together to make a single more complex system.
The way all the sub-systems keep from colliding into a hopeless mess is by
interruption.
Polling
Interrupting
This approach involves the CPU asking each device if it needs the CPU's time.
Typically this is not done on a PC. Just think of how much time would be wasted with
the answer typically being, no.
Interrupting
So plan B is to let devices tell the CPU when they need attention. While this might
seem inefficient, it is more practical than polling. Imagine you don't have a doorbell. If
the doorbell didn't interrupt you, you would have to go check the door every minute or so
to see if anybody was there.
Interruptions are handled by Interrupt Controllers. As it was revealed previously, the
system board shrank in size by combining parts into ASICs called chipsets. The
interrupt controller was one of those devices absorbed into the chipset.
The original IBM PC and XT were designed with a grand total of eight interrupt
channels.
The interrupt count started with zero, leading up to seven. When the IBM AT was
released with the 80286 CPU, the expansion bus jumped to 16-bits. At the same time, a
second interrupt controller was added.
Rather than trash compatibility with the PC and XT, the designers created a cascade so
both interrupt controllers could talk to the CPU on the single control line.
The cascade point the designers selected was interrupt number 2 and sent it over to the
second interrupt controller. They choose to land at interrupt number 9. If on the rare
chance you have an expansion device that needs interrupt number two, it will be pushed
over to number nine. Note that interrupt number 9 is usable, if you have any devices
that need interrupt number two, the 'faking out' process consumes interrupt number nine.
Interrupt Channels
A few of the interrupt channels are reserved right out of the gate for the
PC control and cannot be borrowed. They are 0, 1, 2, 8, and 13.
IRQ
The use of these interrupt channels to make a request carries the term Interrupt
ReQuest and is abbreviated IRQ. From this you will hear the term, IRQ #.
Depending on the priority of what is happening in your PC, a request can be
temporarily ignored if the PC has something going of a more critical nature. This
state is known as masking the interrupt.
Suppose all heck is breaking out in your PC. Say, your memory has just gone brain dead.
Since this is a (not good) condition that is more serious than anything else (to prevent
data loss), your PC will report a Non-Maskable Interrupt (NMI).
NMI Errors
NMI errors are a typically an indication of bad RAM
IRQ Conflict
Until the PCI slot, sharing an interrupt was a bad idea. The CPU wouldn't know
which device was really talking. This is known as an IRQ conflict. This is not as an
uncommon experience as we would like. As folks added more toys to the expansion bus,
the more IRQs were used. When the IBM PC came out, it had one serial
(communications) port. In theory, it could handle up to four communication ports.
However, the designers allowed both communications port 1 and 3 (COM1 & COM3)
to have the same IRQ! In addition, COM2 and COM4 also share IRQ 3. Now this
wasn't an issue when there was no mouse, and a 300-baud modem was over a
thousand bucks.
Today the answer is assign additional needs for the communication ports to nonstandard IRQs or install a serial board that uses one IRQ and has a dedicated
processor.
DMA
Going back to the CPU and using interrupts to move data from one sub-system to
another sub-system is not very efficient. The solution comes from the use of Direct
Memory Access (DMA) channels.
By creating DMA, designers enabled devices to cut out speaking directly with CPU,
leading to increased performance. DMA channels were created along with the
appropriate circuitry to control them. This is another example of circuitry that was
absorbed into the chipset.
DMA Details
DMA channels 0-4 originally were 8-bit (only) and capable of a maximum
of 64KB per transfer. In the AT and above, 0, 5, 6 and 7 are 16-bit
channels and capable of 128KB per transfer. Only one transfer can occur
at a time.
DMA must reside in conventional memory.
Not many ISA based cards were designed to take advantage of the new DMA channels
available in the AT.
DMA channel 3 presents the greatest opportunity for a DMA conflict, which will result in
very erratic behavior.
Keep in mind that while DMA channels became standardized, PCI does not use
standardized DMA the way other expansion bus slots use them. Some refer to DMA as
'third-party DMA. The first and second to the party are the two sub-systems, with DMA
being the third party.
Bus Mastering
Newer, more sophisticated circuitry unveiled the situation where the peripheral that is
transferring data actually takes control of the system bus to do its duty.
This process is known as bus mastering.
This technique really shines when using a true multitasking operating system, such as
NT or Windows 2000.
If you read the Geek Tip, and are wondering about conventional memory, relax. That will
be covered later in this book. Right now, it is time to introduce one more topic, and now
you will see why, in Chapter 0000, we spent all that time ensuring an understanding of
binary and hexadecimal math.
Memory Addresses
While many users are uninformed about how a computer works, most of them can at least
tell you, "I have XXX megs of RAM." What is not considered is how RAM works. Think
of a city say, New York City. This city has millions of residence. Going to the city
looking for a certain individual, your odds of success are nil without knowing the
person's address.
Computer memory works very similar fashion. To find where something is, you need
its address. Unlike a physical city, memory addresses work in hexadecimal format.
For a device such as a modem to work in concert with the PC, it must be connected to a
communications port. The communications port must have a place to put the data
after the modem converts it from an analog signal to a digital signal that is given to the
communications port.
The Operating System must know where the newly arrived data landed if you expect
your computer to do anything with the data. This gives us the need for specific
memory location so the Operating System knows what to do.
Memory Mapping
Memory mapping is the process of reserving blocks of system memory, using
hexadecimal address, to place bits of information that the CPU and operating system
need to find.
In this particular example, we are memory mapping the I/O. It is important to note that
I/O is not stored in RAM, but in different areas of the system memory, such as ROM
chips on an expansion card.
By assigning a particular memory map to a communications port, when the CPU is
ready, it knows where to find the data (the In part of I/O). When the CPU needs to
send data out, it uses this address again. This process is known as I/O Addressing.
A typical example is communications port 1. The standard memory address for Com1 is
3F8h.
I/O addresses are not exactly like IRQ and DMA. IRQ and DMA are typically
assigned one per device and are of uniform size.
While a sound card needs an I/O address, since it really is several devices on one card,
the I/O address needs can (and do) vary.
The challenge arrives with the fact that when I/O addresses are specified, the only part
specified is the beginning of the I/O address. A device may only need as little as one
byte, or as much as 64 bytes.
It would seem with 32, 64, or even 256 MB of ram, there is plenty of room to find an
address. And there is, however we have one little issue that shoots a hole in that logic.
That would be the need for backward compatibility.
Prior to Windows 95, finding I/O overlap was difficult. The routine known as Device
Manager in the Control Panel->System of Windows 95 made this a much simpler
process to discover I/O overlap.
Of course it isn't the simplest of tasks to say, my modem is on IRQ 3 with an I/O address
of 2F8-2FFh. What to do? Read on for the answer.
Logical Names
The BIOS assigns Logical Names to the devices using an IRQ and memory address.
Now IRQ 3 and I/O address 2F8-2FFh becomes Com 2.
This does not mean that you cannot change the standardized IRQ and I/O address. Be
strongly cautioned however against doing this, "just because you can". And whenever
you are building or working on a new system, be strongly advised to create a map of the
system before you start making any changes.
Remember some programs will not ask the Operating System for details such as IRQ and
I/O. Some programs require this information in their own setup. And there are a few
older programs, such as games, that require the popular default settings for the ISA
Sound Blaster. That is:
IRQ 5, I/O 220h, DMA 1
IRQ
0
1
2
3
4
5
6
7
8
9
10
11
12
13
14
15
Used By
System Clock
Keyboard
Cascade
Com 2/4
Com 1/3
LPT2 / Sound
Floppy
LPT1
Real Time Clock
Open
Open / NIC
Open / VGA
PS/2 Mouse
Math Coprocessor
Primary IDE
Secondary IDE
I/O Address
40h
60h
2F8h / 2E8h
3F8h / 3E8h
278h / 220h
3F0h
378h
70h
A000
170h
1F0h
IRQ TIP
Write the chart until you can do so from memory. Before you begin your
test, re-create the chart.
2. CPU
3. PCI
4. 0, 1, 2 ,8, 13
5. ignored
IX
It was revealed how a power supply behaves which displays how power surges can pass
through to the PC, regardless of the connection type AT or ATX. The similarities and
differences between these two form factors have been explained from power connection
and capabilities.
The function of chipsets and how they affect capabilities is unveiled as well as how BIOS
enhances or limits the PC. You have seen that the BIOS may be flashed if you have an
EEPROM.
The functions of POST have been explained along with the fact this is only performed
during a Cold Boot.
The various bus structures from I/O to ISA, PCI and AGP has been discovered. The roles
of IRQ and DMA have been shown as critical components. You discovered the role of
memory mapping and how logical names make life simpler for us humans.
6. Which of the following is not an I/O Bus (Pick all that apply)
A. Serial Ports
B. USB Ports
C. Parallel Port
D. Keyboard Port
E. None of the above
10. The correct choice for the two data path widths of ISA is
A. 8 and 32 Bit
B. 16 and 24 Bit
C. 8 and 16 Bit
D. 4 and 8 Bit
12. True or False: ISA cards can be put into an EISA slot and still work?
A. True
B. False
14. Which of the Following are required for the Plug and Play capability?
A. Plug and Play BIOS
B. Plug and Play O/S
C. Plug and Play Software
D. Plug and Play compatible cards
E. All of the above
17. Which of the following Bus Architectures Does NOT support bus mastering?
A. PCI
B. AGP
C. ISA
D. EISA
20. On Which IRQ would you typically find the Sound Card?
A. 6
B. 10
C. 12
D. 5
E. 15
6. Which of the following is not an I/O Bus (Pick all that apply)
A. Serial Ports
B. USB Ports
C. Parallel Port
D. Keyboard Port
*E. None of the Above
Explanation: Answer: E. All of the devices listed are part of the I/O bus. This is the
slowest bus in the system, and handles data flow between the chipset and the devices
attached to the bus.
10. The correct choice for the two data path widths of ISA is
A. 8 and 32 Bit
B. 16 and 24 Bit
*C. 8 and 16 Bit
D. 4 and 8 Bit
Explanation: Answer: C. ISA defines both a 8-bit and 16-bit bus, running at 4.77 MHz
and 8 MHz respectively.
12. True or False: ISA cards can be put into an EISA slot and still work.
*A. True
B. False
Explanation: Answer: A. The EISA architecture was designed to accept both 8-bit and
16-bit ISA cards as well as 32-bit EISA cards.
14. Which of the following are required for Plug and Play capability? Choose all that
apply.
A. Plug and Play BIOS
B. Plug and Play OS
C. Plug and Play Drivers
D. Plug and Play compatible cards
*E. All the Above
Explanation: Answer: A, B, C, D. For Plug and Play to work properly, you must have a
Plug and Play system BIOS, a Plug and Play OS, Plug and Play expansion cards, and
Plug and Play drivers for the card.
17. Which of the following bus architectures does not support bus mastering?
A. PCI
B. AGP
*C. ISA
D. EISA
Explanation: Answer: C. The only bus architecture that does not support bus mastering
is ISA.
CPUs 125
4.1 Distinguish between popular CPU chips in terms of their basic characteristics.
Popular CPU chips (Intel, AMD, Cyrix)
Characteristics
Physical size
Voltage
Speeds
On board cache or not
Sockets
(SEC) single edge contact
CPUs 127
Introduction
This very interesting chapter opens by revealing the difference between RISC and CISC
and how they are not that much different today. You will discover the difference between
an 80386-DX and 80386-SX. How the math co processor became standard equipment
as CPUs evolved will be explained
CPUs 129
II
CPUs
Lets begin by covering a few of the insider things to know such as; A judge telling Intel,
No becoming the reason for the Pentium is unveiled, and that is why there are different
sockets and slots.
The details of different acronyms and voltages for the Pentium line and the complications
brought by competition arise to clarity.
As processors evolved, you will see the maneuvering done by the processor
manufacturers that created physically incompatible designs, while competing for mind
share and profits. You will also discover new CPU technologies, and how they compare
between the vendors.
The mystery of what happened to Intel competitor, Cyrix is explained.
And the almost never seen inside of a CPU can been seen through several photographs
using photomicrography13.
RISC
From the minds at University of California, Berkeley and Stanford University sprouted
the idea that maximizing computer performance could be accomplished by putting most
functions in software, except for the teachers who were putting them in hardware would
yield a net performance gain. This is the idea behind the Reduced Instruction Set
Computer (RISC). The basic design concepts include:
Hardwired control. RISC designs eliminate micro code in ROM and put the instruction
set directly in hardware.
Simple instructions with few addressing modes. Conventional CPU design includes a
variety of ways to address memory, which require the processor to calculate effective
addresses on a variety of fronts. Simplifying operations reduces CPU overhead.
13
RISC based CPUs are popular in high-end computers dedicated to resource intensive
tasks such as animation. Perhaps the most well known RISC CPU is the SPARC
(Scalable Processor Architecture)
CISC
RISC CPUs were an idea of the 1980's because hardware was getting less expensive,
while coding was becoming increasingly more costly. CISC was a term created to mean
Complex Instruction Set Computing, giving a name to the existing processor designs
that were not RISC. Today, your authors prefer CISC to mean Conventional Instruction
Set Computing.
80386
In chapter 0000, you saw the details of the 8088 CPU and the 80286 CPU. The 80386
offering was backward compatible to the two previous CPUs, however it represented a
fundamental difference in design philosophy. In simple terms, the 80386 CPU was
designed with multitasking in mind. The idea was to make the 80386 operate as if it
was a bunch of 8088/8086 CPUs, with each able to address its own section of memory.
This way a user could have multiple programs loaded at the same time, and the CPU
could use 'round robin' processing, so the end user would experience several programs
running at the same time.
CPUs 131
80386-SX
The 80386 CPUs were an amazing technical achievement, and luckily for Intel
stockholders, priced as such. Everyone wanted a 386, even if they couldn't afford one.
The marketing folks at Intel found a way to sell more 386's without hurting the
profitability of the original 80386. Reduce the data path external to the CPU, making
for a less expensive system board, add a suffix to the label 80386, oh, SX will do, and
now more people can buy an 80386. This worked so well; Intel will use this idea again
later.
Only one thing was missing in the 80386. That was the ability to do floating-point math.
Math Co-processors
The challenge with the CPU is that the design is a bit of a generalist. A good analogy
may be that of the family doctor. The family doctor is certainly a specialist in medicine,
however probably not somebody you would want to see if you need open heart surgery.
For that type of an operation, you'd see a surgeon specializing with heart surgery. A math
co processor is a specialized processor designed especially to work with numbers where
the decimal point is shifting left and right.
Through the 80386, if you wanted a math co processor, you purchased one. For the
8088, you bought an 8087. For the 80286 the 80287s were just the ticket. And if you had
an 80386, you would need (what else?) an 80387.
And Intel was happy to sell you one. Only one small challenge existed for Intel. A feisty
little company down in Texas by the name of Cyrix was making a math co processor
that plugged right into the math co processor socket, and outperformed the '87 series from
Intel, while costing less. These number crunching competitors from Texas were known
as FastMath chips. How could Intel stop this competition? How about building the
floating-point math co processor right inside the CPU?
80486
The 80486s were not groundbreaking in terms of a radically different design philosophy,
like the 80386. It did have four new features that made the 80486 about twice as fast as
the fastest 80386. The most talked about new features were a built-in cache, and a builtin math co processor. On average, the math co processor built into the 80486 yielded
three times the greater performance than external 80387 Numeric Processing Unit
(NPU).
Figure 2.1 - Notice the Notched corner indicating Pin 1 placement. Incorrect placement
will burn the CPU and melt the pins off, leaving a plugged socket
The speed difference between the 80386 and the 80486 made the Graphical User
Interface (GUI) practical for everyday use. Many consider this simple difference the
primary reason why the 80386, once classified as too powerful to be exported (military
considerations), as now worthless silicon.
CPUs 133
486-DX2
DX2 CPUs were also marketed as OverDrive chips. The actual speed of the CPU was
double the motherboard frequency. The only part of the DX2 chip that didn't run at
double speed is the region of the chip to handle the in and out between the CPU and the
rest of the computer. The OverDrive chip was installed in the second socket on the
system board effectively turning off the 486-SX. In the last days of the 80486, a few DX4
CPUs saw daylight outside of Intels factories.
80486-SX
Remember Intel's marketing of the 80386-SX? Well, in a classic case of taking lemons
and making lemonade, Intel took some of its 80486 CPUs that had defective math coprocessors, burned out the ability to access the defective math-co, and sold them as
80486-SX CPUs. Once again, Intel had a hit.
All too often success comes with a price. In this particular instance, that price came with
competition. Remember Cyrix and FastMath? Well, when Intel looked to bury Cyrix's
FastMath by putting the math-co in the CPU, Cyrix responded by making CPUs. And
Cyrix wasn't alone in looking to take a bite out of Intel. Another firm with a fair amount
of experience in turning sand into dollars called Advanced Micro Devices (AMD)
figures prominently in really ticking Intel off.
Pentium
Cyrix and AMD are out in the marketplace selling their CPUs and math co processors,
calling them 80386, 80387, just like Intel's. As you can guess, Intel is not happy about
this. The firm was so mad; they went to court to stop their competitors. Ha-Ha, you can't
copyright or patent numbers, the judge says. So Intel runs a contest to come up with a
name for the 80586 that isn't a number. Penta means five, right? And on the 19th of
October 1992, the name Pentium was announced.
As usual, the Pentium was backward compatible, while offering new features. The
revolutionary step in this CPU was twin data pipelines. This enabled the CPU to
execute two instructions at the same time. This is known as super scalar technology,
typically found in RISC based CPUs.
The Pentium uses a 32-bit expansion bus, however the data bus is 64-bits which means
the system memory is addressed at 64 bits at a time. This is an important distinction to
remember when working with some types of RAM packaging, which is seen in the next
chapter.
The original Pentium was out (introduced) at 60MHz, and was later upped to 66Mhz.
While the Pentium rocked the digital world, it did have one challenge. Operating at five
volts, the CPU put out tremendous heat, limiting the ability to create even faster
processors. The 66 MHz Pentium consumed a whopping 16 W of power (3.2 Amps).
This made laptop computers fry legs of the laps they sat on. The solution was simple, as
soon as it was technically feasible. Lower the required voltage. And lowering the
voltage to about 3.3 volts exactly what Intel did, starting with the 75Mhz Pentium.
Actually, the 75Mhz Pentium used 3.465 volts, but you get the idea.
Bending Pins
CPUs became more complex by additional pins added to support the
increased load. This increased the possibility of bending the contacts on
the CPU as it was being inserted into the socket. Insertion force
reaching 100 pounds (45 Kilograms) became required. Pushing this
hard requires removing the system board from the case.
Intel worked on getting the requirements down, and labeled each Pentium with a suffix14.
Standard
Voltage Reduced
VR Extended
STD
VR
VRE
3.30V
3.38V
3.50V
Pentium Volts
Majority of Pentium CPUs used 3.3 Volts
Voltage standards begin to get confusing with later Pentium's, all Pentium Overdrive
CPUs, and settles to 3.3Volts and 2.8Volts at the core of the CPU through the PII.
14
CPUs 135
To make sure the newer Pentium using the lower voltage was not inadvertently plugged
into a system board using five volts Intel changed the socket type (Socket 4) to a 296-pin
arrangement and staggered the pinouts to create the Staggered Pin Grid Array (SPGA).
Another feature of the second-generation offering was the Advanced Programmable
Interrupt Controller (APIC) that in concert with a dual-processor interface allowed
system board makers to create systems holding multiple CPUs. This really helps out
some Operating Systems, such as NT handling larger loads. When multiple-processors
are used the process is known as Symmetric Multi-Processing (SMP).
SMP
SMP (Symmetric Multi-Processing) allows the use of multiple processors
sharing a common operating system and memory.
Another issue facing the Intel engineers was the fact that they were able to get CPU
speeds moving along faster than the rest of the system bus could handle. The solution
was to use a clock-multiplier circuit to run the processor at speeds faster than the rest
of the system. Table 2.1 shows a typical multiplier and system board speed.
Processor
Pentium 75
Pentium 90
Pentium 100
Pentium 120
Pentium 133
Pentium 150
Multiplier
1.5 x
1.5 x
1.5 x
2.0 x
2.0 x
2.5 x
Bus Speed
50Mhz
60Mhz
66Mhz
60Mhz
66Mhz
60Mhz
ZIF SOCKET
To avoid damaging CPUs during installation, the ZIF (Zero Insertion
Force) socket was developed.
Pentium MMX
The third generation Pentium, released in 1997 incorporated Multi-Media eXtentions
(MMX). This 3.3/2.8V15 CPU added 57 new instructions to enhance audio and video,
and introduced the Socket 7, utilizing 321 pins. (Socket designs are covered later in this
chapter.)
Pentium Pro
Towards the end of 1995 the Pentium Pro was announced. This Pentium introduced a
new socket (Socket 8), utilizing 387 pins. The Pro series included ability to run multiple
instructions in one cycle, could execute instructions out of order, and had dynamic
branch prediction, as well as speculative execution. Also included was an impressive
cache arrangement. For programmers, the Pro looks like a classic CISC CPU, while
internally the CPU is very RISC oriented in design. This 3.316 Volt CPU was designed
with a 32-bit operating system (OS) such as Windows NT in mind.
While the Pro had Level 1 cache in the CPU, its real forte was the integrated Level 2
cache which allowed upwards of 1MB of cache to reside inside the CPU packaging to
run at processor speed. This really improved performance in SMP based system boards.
The Pro chip was the first chip to be offered in the AT or the ATX format. The ATX
format was preferred, as the Pro consumed more than 25 W of power, which generated a
fair amount of heat.
This CPU introduces the concept of dual voltage. The core of the CPU runs lower than
3.3 Volts, while the I/O voltage remains the same.
16
3.1 Volt for 150Mhz
15
CPUs 137
Pentium II
As anyone who has used a Pentium Pro can tell you, this CPU had remarkable
performance. The challenge Intel faced was the cost of production creating the Pro
chip.
The built-in L2 cache had a high failure rate at Intel fabrication plants. How to get out
of this pickle?
The answer lies in the Pentium II (P2). Intel began by separating the processor, and
cache of the Pentium Pro, mounting them together on the circuit board with a big heat
sink.
Then by dropping the whole assembly onto the system board, using a Single Edge
Contact (SEC) with 242 pins in the slot, and adding the 57 MMX micro-code
instructions, then Intel had the Pentium II.
This way, defective cache modules don't force throwing out of a perfectly good CPU,
because of a bad cache. And to further improve cache yields, the Pentium II ran cache at
half the speed of the CPU.
SLOT 1
Slot 1 supports Celeron, Pentium II, and Pentium III.
Xeon
In 1998 Intel expanded the Pentium II line by introducing the Xeon. This breed of CPU
uses Slot 2 technology, which is a wider connector with 330 contacts. Starting at 400
MHz, the biggest change is the L2 cache operates at full processor speed.
XEON
Xeon uses a Slot 2-form factor.
Pentium III
Designed as a successor to the Pentium II, the P3 offers higher CPU speeds and an
update to MMX, with 70 new micro code instructions for; speech recognition,
streaming video and 3-D. Pentium IIIs were released in both SEC (Slot 2) design and
PGA370 socket versions.
The Pentium III offers SpeedStep technology. This allows a choice between full speed
or saving battery life. New to the Pentium III was Internet Streaming Single-Instruction
Multiple-Data (SIMD), featuring 70 new instructions for advanced imaging, 3D,
streaming audio and video, and speech recognition for an enhanced Internet
experience, quoting Intels web site.
CPUs 139
P3
The Pentium III introduced Internet Streaming SIMD
Pentium IV Oops!
In what can only be considered a monumental lapse in judgment, Intel
went ahead and threw out the many tried and tested ideas implemented in
both the PowerPC and AMD Athlon processor families and literally took a
step back 5 years to the days of the 486 It seems that Intel is taking the approach
similar to that of their upcoming Itanium chip - that the chip should do less optimization
work and that the programmer should be responsible for that work. An idea not
unfamiliar to RISC chip programmers, but Intel really went a little too far. They literally
stripped the processor bare and tried to use brute force clock speed to make up for it! 17
This expert on emulators ran several tests to compare the new chip to some older ones
and here are his conclusions:
CPUs 141
The real Pentium IV is built using .13 micron technology, and has a 55 million transistor
count. Yes, fifty-five million transistors, powered at 1.5 V, down from the 1.75 V in the
Williamette Pentium IV. Sorry, no SMP support. Reports abound about it being rock
stable, and an over clockers dream. The race between Intel and AMD has served the
industry.
In closing, a warning. Some Pentium IV CPU's were created with the older Williamette
technology. To be safe, don't buy a Pentium IV that is less than 2.2 Ghz
Using photomicrography and a little sharp shooting we bring you pictures of an 80486
CPU18.
CPUs 143
Celeron
Market forces drove the price of a PC to the point where it could be used as a premium
gift. And the year 2000 saw the "free PC" tied to 3 years of Internet Service Provider
(ISP). Naturally, this put pressure on all vendors to have a more entry-level product.
Introduced first in the SEC package is the Celeron. The early Celeron was a market
disaster. Because it was released with absolutely zero Level 2 cache. Performance of
this CPU series was laughable. Quickly Intel saw its mistake. With this they put cache
into the Celeron, making it a viable contender in the low-end market, as well as
causing AMD some fits.
@@@Real-World Owl @@@ Celeron The Sequel
Take the Willamette core (.18 Micron) of the Pentium IV, cut the cache in half to 128KB,
switch to the mPGA 478 socket, start at 1.7Ghtz, and you have a 'new' Celeron CPU.
Northwood based Celeron CPU should be available in 2003. Supporting different types
of modern RAM (with a host of new chipsets), Celeron is returning the kick in the head at
the sensitive price point that the Duron delivered with a solid reply.
Which of course begs the question, what happened to Cyrix and AMD? A tale of two
competitors and their wildly different stories is the next topic.
Cyrix
Remember the FastMath folks? They had an answer to the 80486 with the Cyrix 5x86.
While the CPU sat in an 80486 motherboard, it performed rather well by using some of
the architectural features normally found in Pentium class CPUs. The Cyrix 5x86 was
available at 100 and 120MHz. While fitting a 486 motherboard, it offers performance
comparable to roughly a Pentium 100. While the CPU is a 3.45-volt design, firms such
as Evergreen Technologies, mounted this Cyrix offering to a voltage regulator,
allowing it to be dropped into an older 486 motherboard. A quick check at Ebay found
offerings for $10.00 apiece. This chip is a nice way to squeak out a little more life for a
486 PC running Windows 95 and a couple of applications, or converting to use with
Microsoft Terminal Server.
M1
In the Pentium class Cyrix offered the 6x8619 (also known as the M1 CPU). Just as in the
release from AMD, it was pin and voltage compatible to be dropped into a Pentium
motherboard. The 6 were to indicate a sixth-generation CPU (which it is not). Cyrix did
incorporate some architectural features designed to outperform a Pentium of equal clock
speed. This led to testing processors with a series of applications to create a
standardized test set, create a reference point, and label the CPU based on a
Performance rating (called a P Rating) that was close to what an Intel Pentium would do.
For example, the Cyrix 133 MHz (actual speed) was rated at 166 MHz.
When Intel moved to SEC based CPUs, the firm gave up competing directly with Intel.
Today Cyrix CPUs are alive and well, as CPU's today, sold under the VIA name, the
company which bought Cyrix.
Technologies. Their latest offering is the VIA Cyrix 3A that while is very lightweight
on floating point capabilities, it is also so light on power consumption, it does not need
a cooling fan to operate.
AMD
If you're wondering who AMD is for CPUs, refer back to the picture with the Suntac
chipset (Figure 24) and notice the manufacturer of the CPU for this 800286 running at
10Mhz. Why, its AMD. Yes they have quietly been in the game for a long time.
For an answer to the Pentium, AMD launched the K5. It was pin for pin compatible
with a socket for the Intel Pentium. The K5 was like the Pentium, in that it does not
contain MMX instructions. When Intel released the MMX version of Pentium, AMD
released the K-620 with 3DNow (3DNow is actually SIMD licensed from Intel), to answer
Intel. AMD offered almost OK performance at a lower price point than Intel.
Intel successfully shut down the pin-for-pin competitive front by dreaming up Slot 1,
and placing a patent on the design. AMD kept the less expensive socket mounting of the
CPU and extended socket 7 with Super Socket 7, taking the K6 to speeds of 500 MHz.
However, after a time, the laws of physics caught up with AMD. To extend beyond
500Mhz, they had to resort to a SEC design as well, and did so, creating Slot A. This
was the connector design for the Athlon.
19
20
These CPUs require special active heat sinks when installed in a Pentium socket.
And AMD introduces dual voltage or split rail design after Intel.
CPUs 145
SLOT A
Athlon uses a Slot A form factor
AMD didn't like the expense of being SEC or "cartridge based" any more than Intel. So
as soon as AMD could they went back to a socket based CPU. Socket 462 carries the
second generation Athlon CPUs, now split into price or performance solutions. They
are known as the Thunderbird and the Duron. Due to the price/performance ratio of this
line up, AMD has been gaining market share.
CPUs 147
Socket 1
Beginning with the 80486 a rectangular socket with 169 pins ultimately became known
as Socket 1.
21
Figure 38 Socket 7
Socket 7 with ZIF lever opening pins. Closing the lever slides the socket closed, locking
the CPU in place. The plastic tabs on the edges are to hold a CPU cooling fan.
CPUs 149
Slot 1
It was revealed previously in this chapter that the Pentium Pro wasnt great for Intel
profitability, and it was a real dog running Windows 9.x. However the Pro CPU is so
good at running NT/Windows 2000 that manufacturer stuck with the Pro chip until the
Xeon was released.
Slot 1 uses a SEC (Single Edge Contact) that looks a bit like a game cartridge with 242
contacts.
Figure 40 SLOT1
NOTE: Two Slot 1 connectors with a Pentium II inverted to display the contacts.
This is an SMP capable system board.
CPUs 151
Slot 2
Slot 2 is wider than Slot 1, with 330 contacts. The biggest difference is Slot 1 limits L2
cache to speed, while Slot 2 allows full processor speed to the L2 cache, as seen in
the former Pentium Pro or the Xeon processors.
Socket 370
The costs to CPU manufacturers for slot technology hurt financially. As quickly as they
could, the makers of CPUs have dashed back to Socket technology. The first was Intel
with Socket 370. The first CPUs to move over was the Celeron series. Intel hopes to
move all its CPUs back to the Socket 370.
Slot A
AMD could move into Slot 1 with Intel garnering legal protection for its Slot 1. So they
changed the pin array and came out with Slot A. This was used for the original Athlon.
From a visual inspection, Slot A and Slot 1 look virtually identical.
Socket 462
Just like Intel, AMD wants out of SEC based processors very badly. The Athlon name
has become split in the move back to the socket based CPU. Both the Thunderbird and
the Duron use Socket 462. Since this socket uses a ZIF design, starting with a Duron
and changing over (upgrading) to a Thunderbird later is a snap.
CPUs 153
III
In this chapter you learned that the RISC verses CISC battle is now a moot point. You
saw the 80486 as the first CPU incarnation with a built-in math co processor, and the
marketing of the SX chip being reused to sell CPUs without built-in NPUs.
The battle lines between Intel, AMD, and Cyrix for the lucrative X87 market stepped
into a battle for the CPUs market with entries such as the K5 and K6 was shown.
Intel's competitive responses such as Pentium II and MMX being countered with
marketing volleys such as 3D-Now and the Athlon family of processors has been
outlined to you.
It became clear that as processor speeds went up, voltage to the processor had to come
down. You saw the new battle shaping up for the SMP computer systems that will
overcome the physical limitations that are being reached in the quest for even more
processing power.
Finally the evolution started with Socket 1 and moved through to the wildly popular
Socket 7 and the Socket 8 was explained. The dilemma of the Slot based CPU drama
was revealed, and you saw the return of the socket with Socket 370 and Socket 462.
IV
2. True or False. The 80386 CPU was the first to include a math co processor.
A. True
B. False
CPUs 155
9. True / False. The Duron and the Thunderbird share the same Socket type.
A. True
B. False
CPUs 157
11. Select the CPU types that support full processor speed with L2 cache.
A. Pentium II
B. Xeon
C. Pentium Pro
D. 80486
12. The first socket to introduce a staggered pin array to prevent miss insertion was the
A. Socket 1
B. Socket 3
C. Socket 4
D. Socket 7
13. The AMD K-6 competed with Intels Pentium MMX with,
A. 3D Now
B. 3D Windows
C. Legal Challenges
D. No choice is correct
14. True / False. The AMD K5 and K6 are socket and pin compatible with the Intel
Pentium.
A. True
B. False
CPUs 159
17. The maximum amount of Cache available within the CPU die of the Pentium Pro
CPU was
A. 16K
B. 256K
C. 512K
D. 1Meg
CPUs 161
2. True or False. The 80386 CPU was the first to include a math coprocessor.
A. True
*B. False
Explanation: The first processor to include the math coprocessor in the CPU was the
80486DX. The 80386 required that a separate chip be installed for the math
coprocessor.
CPUs 163
9. True / False. The Duron and the Thunderbird share the same Socket type.
A. True
B. False
Explanation: The Duron and Thunderbird both use Socket 462 ZIF sockets.
11. Select the CPU types that support full processor speed with L2 cache.
A. Pentium II
*B. Xeon
C. Pentium Pro
D. 80486
Explanation: The Xeon uses Slot 2, which allows L2 cache to operate at full processor
speed.
12. The first socket to introduce a staggered pin array to prevent miss insertion was the
A. Socket 1
B. Socket 3
*C. Socket 4
D. Socket 7
Explanation: The Socket 4 ZIF socket was the first to use a Staggered Pin Grid Array
(SPGA) to prevent incorrect insertion of the processor.
CPUs 165
13. The AMD K-6 competed with Intels Pentium MMX with,
*A. 3D Now
B. 3D Windows
C. Legal Challenges
D. No choice is correct
Explanation: The AMD K6 processor used 3D. Now technology in competition with
Intels MMX technology.
14. True or False. The AMD K5 and K6 is socket and pin compatible with the Intel
Pentium.
*A. True
B. False
Explanation: In an effort to compete better with Intel, AMD used Socket 7 and Super
Socket 7 ZIF sockets with their K5 and K6 processors.
17. The maximum amount of Cache available with the die of the Pentium Pro CPU was
A. 16K
B. 256K
C. 512K
*D. 1MB
Explanation: The Pentium Pro was available with up to 1MB of on-chip cache.
CPUs 167
4.2 Identify the categories of RAM (random access memory) terminology, their
locations, and physical characteristics.
Terminology:
EDO RAM (extended data output RAM)
DRAM (dynamic random access memory)
SRAM (static RAM)
RIMM (Rambus in-line memory module 184 pins)
VRAM (video RAM)
SDRAM (synchronous dynamic RAM)
WRAM (Windows accelerator card RAM)
Locations and physical characteristics:
Memory bank
Memory chips (8-bit, 16-bit, and 32-bit)
SIMMS (Single in-line memory module)
DIMMS (dual in-line memory module)
Parity chips vs. non-parity chips
Introduction
The system memory of a computer is what stores all of the program information and
data that the CPU uses for processing. Recalling back to our office analogy in Chapter
0000, the system memory was our desktop.
II
Memory
We removed files from the filing cabinet (Hard Drive), placed them on the desktop
(System Memory), and the worker (CPU) processed the file. Like all other computer
components we have discussed to this point, system memory comes in all different
flavors and sizes.
III
First, you will learn the difference between Read Only Memory (ROM) and Random
Access Memory (RAM). These two types of system memory are the basis for all the
other types of memory available.
Next, we will discuss how RAM is read from and written to by the processor and
memory controller. You will then learn about how memory is packaged and used in a
PC.
Following this, you will learn about special types of RAM that are used for error
checking and correcting. Finally, we will discuss all of the various flavors of RAM that
have been seen in the last few years, concluding with discussion about what technologies
are on the horizon for system memory.
FIRMWARE
When software is stored on hardware, it is called firmware
The significant difference between ROM and RAM is that the information that is
stored in ROM is maintained even after the power to the system is shut off. We call
this type of memory non-volatile memory. As a matter of fact, one could remove a
ROM chip from a motherboard (very carefully), put it away for a few years, and then
put it back in the motherboard (again very carefully), an the information that it
contains would still be there, ready to use.
Originally, ROM chips had their instructions hard wired into the IC chip, and could
never be altered after they were produced. Because of this inflexibility, this type of
ROM chip is used for programs that are unchangeable (static), and are usually massproduced.
Of course, this type of ROM has its drawbacks, specifically the fact that it can never
be altered. Many manufacturers wanted a ROM chip that could be programmed with
their own software that didn't rely on the circuitry of the chip itself. This type of
ROM is known as Programmable Read Only Memory, or PROM.
A special software program that is often stored on a floppy disk controls the erasing
and reprogramming routine.
This process of erasing and reprogramming an EEPROM chip is referred to as
"flashing." A common implementation of this technology is used to upgrade the
system BIOS program. In most new motherboards, the BIOS program is stored on an
EEPROM, and can be updated by the user of the PC. The process is as follows:
The user gets a special program from the manufacturer of the motherboard, usually
via the companies website.
The user runs this program in their OS environment. The program usually requires a
blank, formatted floppy disk. It then writes a small operating system and the flashing
program on the disk.
The user then reboots the computer with the disk in the floppy drive. The system boots
the operating system from the floppy, bypassing the OS loaded on the hard drive of
the PC.
The OS then executes the flashing program that erases and reprograms the ROM
BIOS with a newer BIOS program.
The user then reboots the system without the floppy disk, and the process is complete.
WARNING!!! -- The flashing process can be very hazardous to the health of your PC
and should only be done when absolutely necessary. Anything that interrupts the flashing
process, such as a power surge or power loss may cause the system to stop functioning in
the middle of the process. If this happens, you will more than likely have a partially
programmed ROM BIOS that will not function properly. That being the case, you will be
lucky if you can replace the chip itself. Often times, you may need to buy a new
motherboard. So be careful!
FLASHING
The process of erasing and reprogramming an EEPROM is known as
flashing.
SHADOWING
The process of coping the ROM BIOS program into RAM is called
shadowing the BIOS.
Unlike ROM, RAM can be read from and written to by the processor at any time.
However, RAM requires power from the system in order to keep the instructions and
data that it contains. If the power to the system is removed, the contents of the RAM
are lost. Therefore, RAM is often referred to as volatile memory.
Much like ROM, RAM comes in many different types, sizes, and speeds. The most
basic delineation between types of RAM is Static RAM (SRAM), versus Dynamic RAM
(DRAM).
SRAM
System cache (L1 & L2) uses SRAM
Data Bus
The portion of the memory bus that carries the actual data to and from the memory is
known as the data bus. Like the other buses that you have learned about thus far, the
performance of the bus, or how fast the data travels, is determined by how fast the
data moves over the bus and the width of the bus.
One can think of a bus much like a highway system. The higher the speed limit, the
more cars can travel over the highway. Additionally, the more lanes that make up the
highway, the more cars can travel over the highway simultaneously. The same holds
true for computer buses. Your typical modern PC has a memory bus speed of 100
MHz or greater. This means that there are 100 million cycles per second that data can
be sent over the bus. Newer PC's have memory bus speeds of 133 MHz. So we have
effectively raised the speed limit on our digital highway 33 million cycles per second.
Taking our analogy further, if we add more lanes to our highway (more wires on the
bus), the more cars (bits) can travel over the bus per cycle. The function of the bus
speed and width describes how much data can flow over the bus and is known as
bandwidth.
Calculating Bandwidth
For those of you who are into math, here is how you can calculate the
bandwidth of a given data bus (for this example, we have chosen a
Pentium III processor with a data bus of 64 bits, running in a 100 MHz
system):
Data Bus width (in bits) / 8Bus Speed (MHz) * 1,000,000 / 1,048,576 = Bandwidth (in
MB / sec)
64 / 81001,000,000 / 1,048,576 = 763 MB / sec (Pentium III, 100 MHz bus)
Okay, lets see what would happen if we upgraded to a 133 MHz system:
64 / 81331,000,000 / 1,048,576 = 1015 MB / sec (Pentium III, 133 MHz bus)
As you might have expected, the extra 33 MHz afforded us with about 252 MB / sec of
additional bandwidth to slightly under 1 GB per second of data transfer.
Address Bus
The second piece of our memory bus that is responsible for addressing the individual
cells (typically a cell is comprised of 4 bits of data) of RAM is known as the address
bus. The width of the address bus defines how much memory the processor could
potentially address. For example, if we had a 24-bit address bus (like in an Intel 80286
processor), the processor would have the ability to address 16 MB of RAM (224 =
16,777,216).
If we had a 32-bit address bus (Pentium Pro, Pentium II) the processor could access a
whopping 4 GB of RAM (232 = 4,294,967,296)! Now before you whip out you Visa card
and dump ten grand on RAM, you must know that most motherboards today allow for a
maximum of 2 GB of RAM. This limit is typically controlled by the chipset and BIOS
of the motherboard.
RAM Packaging
The way in which RAM is organized both physically and logically in a PC is referred to
as RAM Packaging. The best way to understand how RAM is organized is to view it in a
hierarchal structure. At the bottom of the hierarchy is a memory cell. This is an area
of memory made up of bits that can hold a series of 1s or 0s. The number of bits that
make up a cell is expressed by the bit width of the memory. The cell is the smallest unit
of memory that the system can read from or write to.
Moving up the hierarchy, we have memory chips. These chips are typically made of
ceramic or plastic, and have a row of pins one each side. Chips are made up of cells. The
number of cells on a memory chip, multiplied by the bit width of the cells will give
you the storage capacity of the chip. In older PCs, these chips where installed directly
onto the motherboard.
On the next level of the memory hierarchy is the module. A module contains multiple
chips soldered to a small epoxy circuit board. The entire module fits into a socket on
the motherboard.
Finally, at the top of the hierarchy, memory modules are grouped into banks. The
width of the system data bus compared to the width of the modules data bus will
determine the number of modules needed to make up a bank. For example, a Pentium
class system has a 64-bit data bus. If we were to use 32-bit wide modules, we would need
two modules of matching size and speed to make a bank. The system cannot access
portions of a bank.
When they were soldered directly into the motherboard, the user was unable to
replace them should one of the DIPs go bad. When they were placed into the sockets it
was too easy to bend the pins on the DIP when installing them. They were also prone
to what is known as to chip creep. The heating and cooling of the system when it is
powered on and off causes the chips to expanded and contract, which eventually
works the chips out of their sockets. Because of these limitations, the industry began
packaging chips on small circuit boards that could be inserted into a special socket on
the motherboard.
One easy way to Identify a 72-pin SIMM is the notch that separates the two sets of
connectors on the bottom edge of the module (see figure 3.6 below).
SIMMS
SIMMs come in 30-pin and 72-pin modules. In a Pentium-class system,
it takes two 72-pin SIMMs to make a bank of memory.
To install a SIMM module, place the bottom edge of the module into the socket at a
45 angle, and rotate the module until is at a right angle to the motherboard. When
the module is in place, you will hear a small clicking noise as the two metal spring
clips pop into place in front of the module, holding it in place.
DIMMs
DIMMs come in 168-pin modules. In a Pentium-class system, one
DIMM is needed to make a bank.
S: The S specifies the speed of the module nanoseconds. This be defines how fast a
module can read or write to the memory bus. Refer to the table below to see the
different depths, widths, and speeds of various SIMM and DIMM memory modules.
SPEED O RAM
The speed of a RAM modules access time is measured in nanoseconds.
30-pin
30-Pin
SIMM
SIMM
Size in MB Non-Parity Parity
1 MB
1x8
1x9
256kx32*
256kx36*
--
--
2 MB
4 MB
8 MB
16 MB
32 MB
2x8
4x8
8x8
16x8
--
2x9
4x9
8x9
16x9
--
512kx32*
1x32
2x32
4x32
8x32
512kx36*
1x36
2x36
4x36
8x36
--1x64
2x64
4x64
--1x72
2x72
4x72
64 MB
--
--
16x32
16x36
8x64
8x72
128 MB
256 MB
---
---
32x32
64x32
32x36
64x36
16x64
32x64
16x72
32x72
512 MB
--
--
128x32
128x32
64x64
64x72
2. flashing
3. static
4. two (2)
5. one (1)
Parity RAM
The best way to overcome soft memory errors is to use special kind of RAM, called
parity memory. Instead of using traditional 8 bits, parity memory uses 9 bits. The
ninth bit is used to hold a value that is derived from a mathematical calculation called
a checksum. This calculation is found by a adding all of the 1 bits in the first 8 bits. If
the sum of all of the 1 bits is an even number, then the parity bit is a 1. Conversely, if
the sum of all the 1 bits is an odd number, then the parity bit is a 0. The parity circuit
then adds the one or the zero to the original 8 bits of data. This is called odd parity
and is the most typical type the parity check used in PCs. When the processor requests
the data, the parity circuit then checks the data that is being held with the checksum
bit. If there is an error in one of the bits, the parity circuit generates a non-maskable
interrupt error, or NMI.
NMI
Reoccurring non-maskable interrupt (NMI) errors are usually a sign of
faulty RAM.
All that parity RAM does however is check for errors and report them to the user. It
does not do anything about repairing an error. It is also vulnerable to multi-bit
memory errors. For example, say our original 8 bits of data has six ones in it. This
would give it a checksum of one, because it has an even number of bits. Say for
whatever reason, two of those bits was changed to zeroes. The parity bit would still
report a one, as we still have an even number of one bits. In this case, the processor
would be delivered bad data. Most of the time, memory errors only involve one bit.
PARITY RAM
Parity RAM uses checksums to check for errors in the data.
ECC
The more advanced and reliable form of error detection and correction protocol in the
memory system is Error Checking Correcting (or Error Checking Code, or Error
Checking Circuit), also known as ECC. ECC became available in Pentium class
systems, and is only available in systems that support ECC mode. The ECC protocol
will detect single and multi-bit errors, and correct single bit errors automatically. It
does this by a complex algorithm that requires 7 bits to protect 32 bits, or 8 bits to
protect 64 bits.
ECC requires special circuitry that is encoded in the chipset of the motherboard. Most
chipsets that support ECC also include a way to report corrected errors to the
operating system. Therefore it is up to the operating system to support these reports.
Windows NT and Linux both support these messages.
Is also important to note that ECC mode will cause a 2% to 3% slowdown in system
performance as the memory controller will need to generate and work with this error
detection and correction algorithm.
ECC RAM
ECC RAM is capable of correcting single-bit memory errors.
When detecting multi-bit errors, ECC mode will perform in a similar manner to that
of parity RAM, in which the parity circuit will report a not-maskable interrupt (NMI)
error and halt the processor. It does not however, do any correcting of multi-bit
errors. It is important to note however, that multi-bit errors are about as common as
being struck by lightning twice in the same day. The upside is that ECC's reporting
features make troubleshooting hard memory errors and software bugs much easier.
EDO RAM
EDO DRAM is about 10% faster than asynchronous DRAM.
EDO memory was quickly replaced because EDO memory is really not suitable for
system buses above 75MHz. Be mindful when buying EDO RAM for an older PC.
Many motherboards, old and new, will not support EDO RAM. Some motherboards
are "EDO tolerant" in that they will treat EDO as FPM memory and you will lose any
performance gains from the EDO memory module.
RDRAM
RDRAM uses a 184-pin RIMM module.
However, the partnership has apparently not learned a lesson from the MCA fiasco,
and is planning to charge licensing fees to motherboard manufacturers who wish to
use DRDRAM technology. It is likely that DRDRAM will fail to become dominant in
the marketplace and may be replaced by other technologies, such as DDR SDRAM.
DDR/PCxxxx
Just as with the processor, memory technologies are marching forward.
Double Data Rate RAM currently comes in three speeds, with a fourth coming sometime
in 2003. Note that the 2003 entry, DDR400 requires another physical change in
packaging. This means new system board to go with the new RAM.
The more common mistake is to try and install two different types of RAM on the
motherboard. For example, when installing SIMMs, it is easy to install a standard
DRAM in one socket, and EDO RAM in the other socket of the same bank. Another
easy mistake to make is to install two different sizes modules of SIMMs in the same
bank. You may also have challenges mixing ECC and non-parity memory units in a
system. A possible workaround is to disable ECC mode in the BIOS, which may then
allow you to use ECC modules as if they were standard DRAM.
Another common mistake is to buy RAM modules of similar size and type, yet
mismatch speed of the RAM modules. Again sometimes with SIMM modules you can
use RAM of one speed in one bank, and of a different speed in a separate bank. With
the advent of higher performance SDRAM, an even more obscure challenge is that of
mismatching the manufacturers of RAM. On more than one occasion I have seen two
modules of RAM of the same type, size, and speed, not work together, and their only
difference being who made them. As much as possible, you should always try to buy
RAM of the same type, speed, and manufacturer. Following this simple guideline will
prevent you from endlessly chasing your tail.
IV
In this chapter you learned about the various components that make up the system
memory architecture in a PC. You learned that system memory is made up of two types
of memory, Read-Only Memory (ROM) and Random Access Memory (RAM).
Each of these types of memory can be broken down into various sub-types. You learned
that ROM evolved into Programmable ROM (PROM), then into Erasable Programmable
ROM (EPROM), and finally into Electronically Erasable Programmable ROM
(EEPROM). You also learned that RAM could be found in either Static RAM (SRAM) or
Dynamic RAM (DRAM).
You were then shown how memory access is controlled by the system memory
subsystem. This subsystem is made up of the memory controller, address bus, and data
bus. The size, speed, and capabilities of these three components dictate the performance
of the system memory.
Next, we discussed how memory was organized into cells, then chips, then modules, and
finally banks. You then learned how memory is packaged and rated for size and speed.
You learned to Identify Dual Inline Packaging (DIP), Single Inline Memory Modules
(SIMM), and Dual Inline Memory Modules (DIMM).
You were then shown how Parity and ECC memory is used to check and correct any
errors in the data. These technologies use a checksum to validate the data that is
contained within the memory. If the error cannot be corrected, the memory controller
generates a Non-Maskable Interrupt (NMI).
Next, we discussed the various types of DRAM memory that is commonly used in PCs.
You were shown the evolution of DRAM from Fast Page Memory (FPM), to Extended
Data Out (EDO), to Burst Extended Data Out (BEDO), to Synchronous Dynamic RAM
(SDRAM).
We discussed RAM technology with Rambus DRAM (RDRAM), Double Data Rate
SDRAM (DDR SDRAM), and its different speed offerings. We concluded with a
discussion about troubleshooting common memory challenges. You learned how
mismatching RAM types, speeds, and manufacturers can wreak havoc on your system.
You were also shown the proper way to install both SIMM and DIMM memory modules
to ensure their proper seating.
3. What is the term for memory that looses any of its stored information when the power
to the system is shut off?
A. SRAM
B. ROM
C. Volatile
D. Non-volatile
8. What are DIP modules prone to when they are installed on the motherboard using
sockets? (Choose 2)
A. Chip Creep
B. ESD
C. Mismatched pairs
D. Bent pins
E. Mismatched speeds
9. What is the minimum number of 72-pin SIMMs that would need to be installed in a
486DX2 system to have 64MB of RAM?
A. 2 - 32MB modules
B. 4 - 8MB modules
C. You cant use 72-pin SIMMs in a 486DX2 system
D. 1 - 64MB SIMM
10. What SDRAM technology uses a 184-pin module and utilizes clock doubling?
A. Direct Rambus RAM
B. DDR SDRAM
C. SLDRAM
D. BEDO DRAM
11. You wish to add 64 MB of PC133 SDRAM to your system that has 128MB of RAM
already. What two things would you ideally ensure are identical between the
modules?
A. Size Buy two 64MB modules
B. The speed (in nanoseconds)
C. The manufacturer
D. The DxW-S numbers
12. How many megabytes of RAM could theoretically be addressed with a 32-bit address
bus?
A. 1 GB
B. 2 GB
C. 3 GB
D. 4 GB
13. What component of the system memory subsystem is responsible for synchronizing
read/write cycles between RAM and other devices?
A. Chipset
B. Memory controller
C. Data bus
D. CPU
15. When you power on your system, the speaker sounds three short beeps, pauses, and
then continues repeating the beep code. What is the first thing you should do?
A. Replace the RAM module
B. Replace the CPU
C. Reseat the video card
D. Reseat the memory module(s)
16. Which of the following are common RAM installation mistakes? [Choose two]
A. Not having enough hard disk space
B. Mismatched types of RAM
C. Mismatch speeds of RAM it does
D. Using a wrist strap while installing
17. Which of the following best describes the purpose, or design, of false parity memory?
A. It replaces parity memory where parity mode was unable to be turned off in the
BIOS.
B. It replaces hard disk space when you run out of virtual memory.
C. It effectively reduces your total amount of RAM by 10 percent.
D. It can only be used in ATX- type motherboards.
18. What does ECC stand for, and what is its function? [Choose 2]
A. Error-Checking Conduit, or Error-Checking Control
B. Error-Checking Correcting, or Error-Checking Code
C. A protocol that detects single and multi-bit errors, and corrects single bit errors
automatically.
D. A device that detects single and multi-bit errors, and corrects single bit errors
automatically
3. What is the term for memory that looses any of its stored information when the power
to the system is shut off?
A. SRAM
B. ROM
*C. Volatile
D. Non-volatile
Explanation: Answer: C. Memory that looses its data when it has no power is known as
volatile memory. Non-volatile memory keeps its data with or without power.
10. What SDRAM technology uses a 184-pin module and utilizes clock doubling?
A. Direct Rambus DRAM
*B. DDR SDRAM
C. SLDRAM
D. BEDO DRAM
Explanation: Answer: B. DDR SDRAM uses a 184-pin module and utilizes clock
doubling. DRDRAM also employs clock doubling and a 184-pin module, but is not
an SDRAM technology. Remember to read the question carefully.
11. You wish to add 64 MB of PC133 SDRAM to your system that has 128MB of RAM
already. What two things would you ideally ensure are identical between the
modules?
A. Size Buy two 64MB modules
*B. The speed (in nanoseconds)
*C. The manufacturer
D. The DxW-S numbers
Explanation: Answer: B and C. When installing PC133 DIMM modules, the access
time of the RAM (in nanoseconds) and the manufacturer should be matched. You
can get away with different manufacturers, but it may not work for very long.
12. How many megabytes of RAM could theoretically be addressed with a 32-bit address
bus?
A. 1 GB
B. 2 GB
C. 3 GB
*D. 4 GB
Explanation: Answer: D. 232 is equal to 4,294,967,296. Divide that by 1024 three times
(for KB, MB, and GB). This equals 4GB.
13. What component of the system memory subsystem is responsible for synchronizing
read/write cycles between RAM and other devices?
A. Chipset
*B. Memory controller
C. Data bus
D. CPU
Explanation: Answer: B. The memory controller, which is part of the chipset, is
responsible for synchronizing read/write cycles between the RAM and other devices.
15. When you power on your system, the speaker sounds three short beeps, pauses, and
then continues repeating the beep code. What is the first thing you should do?
A. Replace the RAM module
B. Replace the CPU
C. Reseat the video card
*D. Reseat the memory module(s)
Explanation: Answer: D. Three short beeps is a typically POST beep code for a
memory problem. The first step is to ensure the RAM is seated fully. If that does not
solve the problem, replace the module with a known working module.
16. Which of the following are common RAM installation mistakes? [Choose two]
A. Not having enough hard disk space
*B. Mismatched types of RAM
*C. Mismatch speeds of RAM
D. Using a wrist strap while installing
Explanation: Answer: B and C. It is easy to mismatch the type of RAM (parity or nonparity) as well as mismatching the access time of the modules.
17. Which of the following best describes the purpose, or design, of false parity memory?
*A. It replaces parity memory where parity mode was unable to be turned off in the
BIOS.
B. It replaces hard disk space when you run out of virtual memory.
C. It effectively reduces your total amount of RAM by 10 percent.
D. It can only be used in ATX- type motherboards.
Explanation: Answer: A. False parity memory was used to fake out the BIOS of
systems that required parity memory.
18. What does ECC stand for, and what is its function? [Choose 2]
A. Error-Checking Conduit, or Error-Checking Control
*B. Error-Checking Correcting, or Error-Checking Code
C. A protocol that detects single and multi-bit errors, and corrects single bit errors
automatically.
D. A device that detects single and multi-bit errors, and corrects single bit errors
automatically.
Explanation: Answer: B. ECC stands for Error Checking Correcting. It can also be
described and Error Checking Code.
2.1 Identify common symptoms and problems associated with each module and
have to troubleshoot and isolate the problems.
Hard drives
4.3 Identify the most popular types of motherboards, the components and
architecture (bus structures and power supplies)
Basic compatibility guidelines
IDE (ATA, ATAPI, ULTRA- DMA, EIDE)
Introduction
This chapter on hard drives begins with one of the most detailed history review sections
that have ever been brought together in one book .The critical players are revealed, along
with the discoveries they made, and the technologies that came with the historic
breakthroughs. You will discover how some of the decisions made have impacted the
tape and floppy storage options.
II
Hard Drives
First, you will It will be revealed to you the various types of hard drive interfaces that
are still operational in older computers. After the historical perspective you will learn
about the start of the most popular drive interface in the world today and its rapid
development along with its battle with its competing interface.
Additionally, you will learn a series of gotcha pitfalls lurking in the dark just waiting to
be a showstopper at the most inopportune times, and what to do about them.
Finally you will discover the different approaches to redundant data storage and which
is the best choice for a particular need.
III
As you have come accustomed to, the evolutionary steps of PC devices always reveals
details about the current state of affairs. Hard drives are of course not immune to this fact.
The following two history lessons will no doubt reveal some interesting facts about how
we came to use this media for data storage.
Hollerith's machine allowed the 1890 census to be completed in six weeks. The
company's success was assured. And it grew by selling machines all over the world. And
as time went on
Hollerith's firm ran into financial challenges. Remember the traveling piano salesman?
His name was Thomas Watson. He bought Hollerith's company, and after a little time,
renamed the firm to. International Business Machine's (IBM).
For the next 100 years the punch card (sometimes called the 80 card) lived in utility
bills, payroll processes and so on. What is less well known is the player piano paper tape
also played an extensive role in computing. Rather than put the programs on a card
(where one card out of place would crash the program) codes were placed on rolls of
paper, either 4 or even 8 bits wide, and then run.
Obviously reducing years of census work to six weeks presented a great improvement
over pen and ink, the downside was the slow method and if you made an error you might
have to re-punch the card. Needless to say by today's standards these were very
cumbersome and difficult to use. Not only that, there was no easy way to work with the
same data over and over again. The industry realized pretty quickly that something else
had to be done.
The next generation of storage was "magnetic tape". You have probably seen pictures of
computers with large reels of tape. These tapes recorded information in a similar way to
an audiotape. These tapes were flexible, more reliable and faster then the old punch
cards. In fact, some government agencies still transfer data via these old 9-track tapes. (If
you cannot picture one of these devices, there a popular 'high-tech prop in the original
James Bond action/adventure movies).
Tape is still used today on computers, though mostly for backup purposes. These were
the main source of storage for computers before hard drives came along. The main
challenge with magnetic tape is that they must be read linearly, from beginning to end,
and it can take several minutes to access the data you are looking for making random
access a Herculean effort on this type of media.
With all this innovation coming into the mainframe (sometimes nicknamed Big Iron) it
wasn't long before this trickled down to the mini computer (review the role of the mini
computer from Digital in Chapter 0001), and in to the PC arena.
The first PCs were, by today's standards very slow and could not accomplish much.
With the introduction of the PC, technology began to advance at a very fast pace. The
Altair was programmed using eight toggle switches. Later, the idea of using an
audiocassette tape to retain the programming was devised.
Then in 1971 a new device was introduced, a single-sided floppy disk drive. This
device could load your programs and save your data. This was an incredible advance in
personal computing. This revolutionary device was called a floppy disk, with the
nickname referring to its flexibility. IBM engineers led by Alan Shugart created this
remarkable device. These floppy drives had the capability of randomly accessing your
data and were portable. It used an 8-inch flexible magnetic donut housed in a square
flexible plastic enclosure with a rectangular hole punched to allow access to the
magnetic head for reading and writing. If you needed to take your program or data to
another computer to work on it, it was no problem. By today's standards they were very
expensive, around $700, but they were much more reliable than punch cards or tape.
Five years later (1976), Shugart Associates went to work for Dr. An Wang to create the
5 1/4" floppy for Wang Laboratories, which was released in 1978. The 5.25" floppy
reduced the 50 pins needed for the 8" floppy to a 34-pin cable, still in use today. Five
years (1981) after the 5 1/4" floppy story began, Sony released the 3.5" floppy.
22
IV
The first hard drives were the result of experiments conducted mostly by IBM. These
researchers were trying to find a way to develop a disk drive that would be viable for
commercial development. The very first drives were actually not disk drives at all; they
were rotating cylindrical drums that the data was stored on using magnetic patterns.
The drums were very hard to work with and very large, so they were not practical for
commercial use.
The first true hard drives that were developed were designed with the heads touching
the surface of the hard drive. This allowed the low-sensitivity electronics to better read
the surface of the disk, which was magnetic. The techniques used with these drives were
not very sophisticated and did not allow for the surface of the disk to be smooth enough
to allow the head to glide over the disk at high speed. The heads would wear out fairly
quickly or the magnetic coating on the surface of the disk would wear out.
23
The second issue was one of disposal. Military regulations required that obsolete data be
burned. Magnesium is the stuff that emergency flares are made of. Can you imagine what
happened at an US Army burning site when they threw that much magnesium into a fire?
But it wasn't the storage capacity of the new disks that was impressive. It was the fact
that any record -- the equivalent of one punch card -- could be accessed at random in
less than one second. Of course, that five megabyte convenience was about the size of a
very large refrigerator and cost about ten thousand dollars per megabyte.
The story from this IBM researcher, doesn't mention that this was a rogue project. Hiding
in a small town called San Jose, California twelve hours away from IBM headquarters.
When the head honchos wanted the project killed, due to budgetary constraints, they hid
the project and kept going.
The breakthrough in hard drives came in the 1950's when IBM engineers figured out
how to make the heads ride above the surface of the disk without touching it and be
able to read the data stored on the drive as it passed underneath it. This revolution
formed the basis for the modern hard drive.
This drive was the first one with a sealed internal environment and an improved "air
bearing" system that reduced the height of the heads to 17 micro-inches over the disk.
Al Shugart returns to the story, in 1979, forming a company called Seagate, which
distributed the first 5.25" form factor hard drive used in PC's.
This drive, the Seagate ST-506, featured a four head design and had a capacity of 5MB.
IBM chose the next generation of this drive, the ST-412, a 10MB Disk, for it's IBM
PC/XT. It became the first hard drive widely sold in a PC.
We will discuss all these architectures, except SCSI, in this chapter. SCSI is somewhat
unique, and has its own long history therefore it is covered in Chapter 0101 the next
chapter.
MFM
Frequency Modulation (FM) was used as an encoding scheme for the 8-inch floppy on
the earliest hard drives and floppy drives. The challenge with FM was it needed to use
space for a timing signal. Enter Modified Frequency Modulation (MFM) This
encoding doubled the capacity for floppy drives creating what is commonly referred to
as Double Density (DD) disks. MFM is still the standard used for floppy drives today.
RLL
In the ongoing quest for more storage MFM was getting competition from Run Length
Limited (RLL). This new encoding scheme reduces the amount of the data checking
information required, allowing for more data to be stored. By far the most popular
version of RLL was RLL 2,7. It was called this because the sequence of zeros in the
data-checking codes always ranged from 2 to7, yielding a 50% increase in storage over
MFM.
ESDI
A few manufacturers had Maxtor leading the charge for more speed and capacity over
MFM. Using the same set of cables as MFM (the set was a single 34-pin cable that
could support 2 drives, and a 20 pin cable for each drive) it moved the error checking
along with the encoding and decoding to the drive, off the controller, making it faster
than MFM. This was the rationale behind the Enhanced Small Device Interface
(ESDI).
IDE
Integrated Drive Electronics (IDE) surfaced in 1988 when a number of vendors formed
the Common Access Method Committee, originally to standardize a SCSI software
interface. As part of their efforts the outcome was the AT Attachment (ATA), which
became associated with IDE. IDE drives have most of the controller functions as part
of the circuit board attached to the hard drive. The 40-pin connector is now
standardized for most system boards and the IDE hard drive. Because of the logic, or
intelligence being moved to the hard drive, it is sometimes called Intelligent Drive
Electronics.
EIDE
In it's day, IDE was certainly cool stuff. By now, it should be evident that humans want
more, faster, and for (ultimately) less money. With that, Western Digital created new
stuff, and called it Enhanced IDE (EIDE). One of the new features was a faster
transfer rate. Eventually most of what was offered in EIDE became an official standard.
ATA-2
Making EIDE official, was the reason for ATA-2. In the original IDE, the specification
was just for hard drives. By adding AT Attachment Packet Interface (ATAPI), other
devices such as a tape drive or Compact Disc-Read Only Media (CD-ROM) were now
supported in the ATA specification.
Seagate certainly wasn't going to stand around and watch a competitor take the limelight.
They added some functionality, and called it Fast ATA. Quantum joined the fray, and
created Fast ATA-2. Notice these terms are not bolded? That is because these were (are?)
marketing terms to reflect some added functionality.
Ultra DMA
By using Direct Memory Access (DMA), the processor is not tied up. The original
DMA specification moved data at 16.6 MB per second. Ultra DMA (the 2 is implied)
doubled this to 33 MB/sec. Mode 3 didn't really live long, however Ultra DMA 4 or
UDMA 4 is frequently called Ultra ATA/66, named from the transfer rate of 66.7
MB/sec. As of this writing, Ultra ATA/133 is shipping. The fact that it may or may not
become 'official' with ATA/ATAPI-6 won't matter to most of us.
40 Pin/80 wires
As ID transfer rates kept increasing, the limitations of the 40-pin cable became an
issue. The solution was to add a ground wire between each signal carrying line on the
40-pin cable. This extra ground wire absorbs crosstalk. Crosstalk is any stray signal
that 'leaks' from the original wire. Pin 34 on the motherboard side can determine if the
80-conductor cable is present.
Except for some 'value priced' PCs, cable select never caught on, and the regular 40
conductor IDE did not support CS, requiring a CS cable. The drive with the controller
turned on in a CS configuration requires that drive be in the middle of the IDE cable
(perhaps why CS never caught on).
Finally, as mentioned above, CS is automatically supported in the 80 wire IDE cable,
and reverses (to the way it should be) to where the controlling drive is at the end of the
cable. Be careful when working with CS systems.
Pin 1
In almost all cases of designing a circuit board, Pin 1 is ground. It used to be that there
was some sort of keying scheme to avoid putting a cable on backwards. In the case of
IDE drives, Pin 20 was not used, and some cable makers put a plug in, with the idea it
would match up to a missing pin connector on the drive. In other cases, the plastic
housing of the connector had a plastic key.
The challenge was, nobody got all the manufacturers together, and got them to agree on a
standard. (Why do we need a standard? We have a bunch of them already!) So, some
manufacturers did or didn't do one or both, or neither. (It's real 'fun' to stand there with a
pocketknife attempting to either remove a plug, or shave off the plastic tab from a cable
so it fits either a drive or motherboard. I have the scar of five stitches in my hand to prove
what 'fun' it was.)
So today, we have to, 1) be sure not to reverse cables when installing, and 2) carry a very
sharp knife with a small pliers to hold the connector. As an alternative to #2, carry extra
cables without the plug or plastic tab.
Cable Reverse
Reversing the cable will result in either a floppy drive or an IDE drive
with a busy light that will not turn off
In the continued struggle toward more speed, ATA is expected to hit the ceiling at 133
MB/sec. Some analysts feel that 133 MB/sec with ATA/133 is the limit. In either case,
there's no argument there are several issues involving (parallel) ATA. Factors including:
Expense: An 80 conductor cable is more expensive than Serial ATA.
Cooling: All those wires impede airflow, making cooling an issue.
Power: In the move to lower power, (parallel) ATA is stuck at 5 Volts.
Logical Limitation: (parallel) ATA is limited to 2 devices per channel.
All this and more, is simply done away with Serial ATA.
Serial ATA
The first round in Serial ATA may be out around 2003. The first offering
introduces speeds of 1.5Gb/second, or 150MB/second. Higher throughput
simply means increasing the clock speed. Second generation is expected to
support 300 MB/second, and the third generation realizing 600MB/second. Further,
since Serial ATA is a point-to-point protocol, the concept of master/slave and drives
competing for bandwidth goes away.
If your next system board comes with Serial ATA, and you don't have a new Serial ATA
drive, adapters are expected for both the drives and system board/host adapter. Down
the road it may be possible to include Hot Swapping and even power for the drive in
one cable. Other benefits include:
Longer cable length -- 1 Meter (more than doubling (parallel) ATA's 18".
Lower power requirements .5V (500 millivolts)
Improved Error Checking
Lower manufacturing costs
Except for our discussion on SCSI, which is seen in the next chapter (Alan Shugart will
pop up there again) the examination of interfaces is complete. We close with one more tip
before moving into the physical world of hard drives.
Physical
The physical characteristics of a hard drive consist of Platters, Spindle, Read/Write
Heads and the Actuator Arm. These are enclosed in a sealed case to protect them from
any contamination.
It is very important to never open a hard drive because even one spec of dust inside the
unit would most likely cause the drive to become unusable. With this overview, let's
examine each of these in more detail.
Platters
The platter is a flat round disk used to hold the data. The disk is made of a material
that forms a rigid base.
This base gets a magnetic coating that holds the zeros and ones, making data.
It is the platter that determines the drive's form factor.
The form factor is called out in standardized sizes, such as 3.5", which conveniently
fits into a drive bay in a case.
How Small?
IBM unveiled the Microdrive, a hard drive with a 1" form factor. Late in
2000, the company released a 1 GB drive. The Microdrive is being featured in
digital cameras, from manufacturers such as Kodak.24
Figure 53 Microdrive
Microdrive
Microdrive is found in PC Cards & CF2 memory for Digital Cameras. The
trend is towards ever-smaller platters, for a number of reasons including:
Simpler manufacturing
Power conservation
Noise and heat reduction
Improve performance
Each platter has two surfaces, each which can be used for storing data. Typically,
several platters are stacked to form a single assembly that spins as a single unit.
Because at the platters spin at a higher rate of speed, they must be well balanced.
Furthermore, each platter must be very smooth and flat, as you will see shortly.
Spindle
Each of these platters mounts to a direct-drive motor, so they can be spun. This is
known as the spindle. The spindle speed is measured in Revolutions Per Minute
(RPM). The higher the RPM, the better the drive performance.
Read/Write Head
Conceptually, the read and write head is very simple. The principal is based on
electrical theory seen in Chapter 0001. Specifically, applying electrical current to a coil
produces a magnetic field, which alters the magnetic polarity on the magnetic
substrate coated on a platter. To read the magnetic polarity, electrical current in the
coil is modified when it passes over the magnetic substrate.
Giant MagnetoResistive
IBM has pioneered new techniques in hard drives. The latest is Giant
MagnetoResistive (GMR) heads, named from the giant magnetorestive
effect. The effect comes from the fact that by working with different
magnetic materials in thin layers, there is a large resistance to change when in a
magnetic field. How GMR drives are manufactured is too complex a topic for this book.
However if you have wondered how IBM could get a 1 Gig drive about the size of
quarter, now you know.
Flying Height
It was revealed to you that floating the read/write head on an air bearing improves the
life of the hard drive. The distance between the head and the platter is called the
flying height (sometimes called the Float Height). Without getting into a course on
physics, the more data squeezed within a given size, the closer the head must be to the
platter. If the head hits contamination or the platter, it is called a head crash.
Landing Zone
Regardless of the life of a hard drive, the heads cannot be 'flying' forever. Anytime a
drive is not powered, the heads are in contact with the platter. It makes sense, when
the heads are resting on the platter, to ensure that data isn't under it. This is where the
term landing zone comes from. Putting the heads there is known as head parking.
Typically, heads are parked in a landing zone that is one step beyond the last area of
data storage. In recent years, parking the heads is an automatic process that happens
on power down. Appropriately enough this is called auto-parking.
A head crash can occur from more than contamination. Physical shock can cause a
crash. Something notebook makers must concern themselves with. So, with a read/write
head made, it has to go somewhere. If it doesn't move, you basically have a magnetic
head stuck in one position, like a cassette deck. How it moves is the next topic.
Actuator Arm
The magnetic heads are combined into an assembly, and the assembly is attached to an
actuator arm. The actuator arm holds all the magnetic heads and is moved across the
platter(s), as a synchronous assembly. Originally, hard drives moved this assembly by
borrowing concepts from audio speakers, called a voice coil. The voice coil method has
the ability to use feedback for accurate placement of the heads. The definition for this
feedback system is known as a servomotor. This was an expensive process.
Alan Shugart came up with the idea of a lower cost method, using a stepper motor. This
is a motor with predefined steps. This caused issues with expansion and contraction of
the materials used when temperatures changed! (Suddenly the data wasn't where is was.)
Advances led to the demise of the stepper motor, and back to the voice coil. As drive
density increased, even servo based hard drives had challenges. The drive makers
resorted to a process known as thermal re-calibration. This did not fare well for the new
multimedia, Audio-Visual computing environment. Re-calibrating in the middle of a CD
burn that took about an hour on blank media that was $20 a shot made for upset users.
Some drive makers released AV rated drives. The issue of thermal re-calibration has
largely been eliminated as an issue.
Now that the physical attributes of a hard drive have been revealed, its time to demystify how the drive operates, and with that, some new terms you may not have seen
before.
Architecture
Each hard drive manufacturer weighs a great variety of different factors before going into
production. The end result is each manufacturer has a slightly different architecture,
known as the Geometry.
As a premise for understanding the following topics all that is needed at this point is there
has to be a logical structure both for the mechanical and electromagnetic aspects of the
drive. At least if it is ever going to store and retrieve information, it needs logic. Used in
the context of hard drives, geometry carries with it a variety of sub-parts, which are
reviewed in the upcoming section.
Tracks
Information is stored on the hard drive platter in concentric circles, much like a
racetrack with different lanes. The outermost ring is numbered zero (0), and the
numbers increase toward the center.
Cylinders
Recall the discussion that the read/write heads were stacked into a monolithic block on
the actuator arm? That means when one head needs to get some data on a track
numbered bigger than zero, the rest of the heads have to go along for the ride. For this
reason, tracks are not referred to as tracks but as cylinders.
A way to think of cylinders would be to log a tree. If you had a one-foot section of tree,
and cut it into slices each two inches thick, you would have six slices, and each growth
ring, on each side of each piece would be a track. A vertical view would be the
cylinders.
Sector
A track can hold a fair amount of data. It would be a real waste of space to use an entire
track for a small file. The track is broken into smaller parts called sectors. Going back
to the thought of the racetrack. Think of the (human) track race called the relay race. A
runner starts in his/her lane (like the track in a hard drive) and runs with the stick until
they reach a point where they pass it off to the next runner. Each runners area could be
thought of as the sector in a track.
CHS
With the three pieces of Cylinders, Heads and Sectors per track; you have what is
referred to as the geometry of the drive, which is the CHS. Today, with auto-parking,
that is really all you need to know about the 'geometry' of a hard drive. If you have a
drive without auto-park, you should have the landing zone cylinder number.
RAID
Besides being something you might see on a TV show like COPS, it is an acronym called
RAID (Redundant Array of Inexpensive Disks) or (Redundant Array of Inflexible
Disks).
While different flavors of managing hard drives under RAID exist, three flavors are
extremely popular, and seen next.
RAID 0
The easy way to remember RAID 0 is to zero out the capital R that leaves AID. And
that is exactly what this standard is, an AID.
Raid 0
RAID 0 is also called Disk Striping with No Parity. The idea is to take a file, and
distribute the data so different hunks o data are written to two or more drives.
RAID 1
Since we are working with 9 Gig drives, lets stay with that and see what happens with
RAID 1. The R letter kicks in with this standard and Real protection from drive failure
happens. In RAID 1 you get a byte for byte copy of the other drive. One drive goes poof,
and is D.O.A. the other drive is still singing.
The bad news is the cost in both money and using up the inside of the computer. Two 9
Gig drives give you exactly 9 Gigs of storage.
This is because each drive is a mirror image of the other. In fact this is popularly known
as mirroring.
Beyond Mirroring
Ok. So you think life is good because you have a full time backup
happening with a second hard drive.
What happens if the controller working the two drives gets served a digital death
certificate? Well, the controller might just slip quietly into no-more-work land.
And it might get real annoyed at the fact it is being killed. When a controller does give a
last act of defiance, it could scramble the data, perfectly mirrored of course, on both
drives. Now you have a redundant pile of garbage for data.
The solution is known as disk duplexing. By having a separate controller for each
drive, if one gets taken out, it cant force the other controller to misbehave, leaving one
drive with the data.
Disk Duplexing
Disk Duplexing requires 2 controllers.
RAID 5
This scheme follows the idea of RAID 0 getting different drives to hold different parts
of a single file, however a third drive holds what may be best described as a compressed
combination of 2 clumps of data from 2 other drives. Obviously this requires a
minimum of 3 drives.
RAID 5
RAID 5 needs a minimum of 3 drives.
Using the 9 Gig drives, a grand total of 18 Gigs of data is available just as in RAID 1,
and you get to buy a third drive. Such a deal. Well, really it can be because you get two
drives that get to send the heads in each drive their own direction. That means faster
reads and writes than with the mirror which has to well, mirror.
RAID Tricks
You dont have to devote entire drives to RAID. It is possible to grab only
sections of drives. And you cannot create RAID anything unless the drive
space devoted is equal across the board. If you have 2 Gigs, 3 Gigs, and 4
Gigs, you have to use 2 Gigs from each drive.
Striping
Regardless of RAID 0 or RAID 5, dividing the data over different
physical drives is called striping. The former is without parity the
latter is with parity.
JBOD
The number of ways to gang up drives in arrays beyond what has been described (so we
will buy more drives) is amazing. All sorts of RAIDnumbers have appeared.
And the most honest term that has come out for some controllers is known as JBOD.
This is short for Just a Bunch Of Disks.
Interleaving
As hard drives spun (compared to floppies) very quickly, the sectors
of the track would 'fly past' the read/ write head very quickly. The
head could read say, sector 1, and pass the data off to the controller.
Well, even at 3,600 RPM, sector 2 came by really fast. Too fast for the controller to do
any more processing. That would mean wait for an entire rotation to complete. Now the
controller is waiting around (too long). Solutions require more solutions it would appear.
The solution was called interleaving. By storing the data in sectors such as 3, 6, 9 etc. it
gave the controller a chance to catch up. The interleaving could just as well have been in
the example, 6, 12, 18 etc. The ratio of these number is known as the interleave ratio. As
you may have guessed, the smaller the ratio, the faster the controller has to be, to be
able to process the data, and the faster the drive can send and receive data.
Before closing this chapter, a few loose ends should be covered. This is not so much for
the A+ certification test, but rather the ultimate test of successfully performing in the real
world.
Overlay Software
As you saw before, the BIOS needed to know about the system. One of the factors
written in the BIOS was the CHS information. The challenge arises with the fact that
unless a BIOS is upgradeable, it only knows what it knows. As hard drives moved up in
size, the CHS grew beyond what the BIOS knew. Either this left the user with a drive
that could not be used to capacity, or it wouldn't install. The solution is to use overlay
software. In simple terms, the overlay software is written to track 0, and puts on a 'game
face' that the BIOS can understand. The software than does CHS translation that the
drive actually understands.
The challenge with overlay software is the fact that if it gets corrupted, the drive
cannot be accessed. Given this, the preferable choice is to upgrade the BIOS. If you must
use overlay software, be sure to make a floppy that has the Operating System and the
overlay software. At least this is a good idea if you want to see the data on that drive
after the corruption of the overlay software.
In this chapter you learned about the most commonly used data storage, the hard disk
drive (HDD). You were shown the extensive history of the hard drive and how it has
evolved to what it is today.
From there, you were shown how hard drives are physically constructed. You also
learned about IDE and SCSI implementations of hard drives.
Next, you learned how hard drives are logically divided into tracks, cylinders, and
sectors.
Finally, you learned how hard drives could be grouped together in RAID arrays that
provide fault tolerant redundancy.
VI
7. Choose the appropriate characteristics for Ultra ATA/66 (Choose all that apply)
A. Bypasses CPU
B. Requires a minimum of 32 MB of RAM
C. A maximum transfer rate of 66.7 MB per second
D. Requires 80 conductor cabling
8. A color-coded 80-conductor IDE cable uses a _______ color for the system board;
__________ color for the first drive and a ________ color for the second drive, if it
exists.
A. Black, gray, blue
B. Blue, gray, black
C. Gray, black, blue
D. Blue, black, gray
11. It is possible to have both a C/S and a non-C/S drive in a computer, as long as they
are on different channels.
A. True
B. False
14. Ultra IDE/66 and Ultra IDE/100-133 use 80 conductors with 40 pins to:
A. Make everyone wonder what happened to the other 40 wires
B. Carry the required DC voltage
C. Reduce Cross talk
D. Interface with 80 pin SCSI drives
7. Choose the appropriate characteristics for Ultra ATA/66 (Choose all that apply)
*A. Bypasses CPU
B. Requires a minimum of 32 MB of RAM
*C. A maximum transfer rate of 66.7 MB per second
*D. Requires 80 conductor cabling
Explanation: Ultra ATA66 allows the drive to write data directly to RAM, bypassing the
CPU. It allows transfer rates of 66.7 MB/s, and uses an 80-wire, 40-pin cable.
8. A color-coded 80-conductor IDE cable uses a _______ color for the system board;
__________ color for the first drive and a ________ color for the second drive, if it
exists.
A. Black, gray, blue
B. Blue, gray, black
C. Gray, black, blue
*D. Blue, black, gray
Explanation: On a UDMA 66 cable, the blue connector goes to the motherboard, the
black connector goes to the first drive, and the gray connector goes to the second
drive.
11. True or false. It is possible to have both a C/S and a non-C/S drive in a computer, as
long as they are on different channels.
*A. True
B. False
Explanation: The master/slave and CS configurations are ways of designating which
drive in the chain is being used as the controller. As long as CS and Master/Slave
settings are not mixed in a channel, both can be used in a system.
13. True or False. Serial ATA is incompatible with current IDE standards.
A. *True
False
Explanation: Because it is a completely new interface standard, Serial ATA drives will
be incompatible with IDE cables and controllers.
14. Ultra IDE/66 and Ultra IDE/100-133 use 80 conductors with 40 pins to:
A. Make everyone wonder what happened to the other 40 wires
B. Carry the required DC voltage
*C. Reduce Cross talk
D. Interface with 80 pin SCSI drives
Explanation: The extra 40 wires are for ground, which reduces crosstalk.
You cant expect to this the jackpot if you dont put a few
nickels in the machine.
-Flip Wilson
SCSI 263
Introduction
The possibly most misunderstood interface is the Small Computer System Interface
(SCSI) The proper pronunciation is "scuzzy". In this chapter you will become a SCSI
expert in a very brief time.
SCSI 265
II
SCSI
You will discover the history of SCSI from both an official and real world standpoint.
ADVANCES IN SCSI technology are revealed to you. Challenges in setting a SCSI ID
will become a non-issue. You will discover the mystery surrounding what a LUN is and
why it was created.
The topic of termination will be demystified, eliminating one of the greatest challenges
with creating or maintaining a SCSI environment.
This chapter will show you the planned future of SCSI along with details of
implementation.
III
What is SCSI?
What is SCSI? The first thing to keep in mind about SCSI is it does not have to be a
drive. SCSI is a device. In theory, your household toaster could be a SCSI device. A
popular myth about SCSI is thinking of it as an interface. A more accurate description
would be to call it an I/O bus, because you can attach more than one device to the bus.
Lets take a brief tour of history to see what makes SCSI unique.
Shugart Returns
In 1979 Alan Shugart's company came up with the Shugart Associates
System Interface (SASI). At the time, many drive interfaces worked
at the device level. The SASI interface operated at a logical level. The
next year Shugart attempted to replace the common mainframe/ minicomputer interface
Intelligent Peripheral Interface (IPI), widely used by Control Data Corporation
(CDC). Seagate later absorbed CDC.
The idea allowed changes in drives to occur, leaving the interface intact. In 1980, Shugart
approached the American National Standards Institute (ANSI), which wasn't too
interested. Another firm, National Cash Register (NCR) added functionality to the
SASI standard, and in 1982 approached ANSI to formalize a standard. About this time,
the Write Once Read Many (WORM) optical drives appeared, expanding the standard
to more than magnetic media.
In 1986 ANSI adopted SCSI, which used a 50-pin interface.
Although SCSI follows standards set by ANSI, a clue that life is not a bed of roses with
SCSI is the little 's' added to the end of the word standard. Yes, there are multiple
standards in the SCSI world, which can lead to incompatibility within the SCSI
standards. Currently, there are 7 generations of SCSI, and some variations within each
generation.
How Fast?
Depending on the standard you are working with, SCSI can be very fast. The
latest SCSI standard at the time of this writing is the Ultra SCSI 320. As the
name implies, this standard is theoretically capable of transferring at 320MB of data per
second. Yes, the big 'B' is not a typo, and the big B means bytes, not bits. The SCSI
Trade Association has adopted the roadmap outlining Ultra SCSI 640. (www.scsita.org).
SCSI 267
SCSI has been available for many types of computers, and was a standard interface for
many Macintosh systems. The standard is quite popular in PC servers, and high-end PC
workstations. You'll see why this is a little later in this chapter.
Why SCSI?
Before beginning a thorough discussion of SCSI, the question of why
SCSI over another format say, IDE is in order. There are certain facts that
are unavoidable. Fact: SCSI costs more. Fact: IDE is faster in most cases,
for a lot less money. So what is all the fuss about with SCSI? The answer is: it depends
on the challenge.
Before tackling SCSI, lets look at an IDE drive. Let's use an analogy. Suppose I ask you
to go the store and get a quart of milk. For the purpose of this analogy this is like getting
a file. While I am waiting for you to come back with milk, that is exactly what I am
doing, waiting. While I ask you to get a quart of milk, Matt asks if you would get him a
6-pack of iced tea. With IDE, you have to come back from the store (bringing me my
file) before you can go back out and get Matts iced tea (his file).
Contrast this to SCSI. In this case, it is possible to combine the errands of getting the
milk and the iced tea (both file requests). This is done using Tagged Command
Queuing (TCQ). The commands to get the different files are sent to the drive for
processing, and kept in the command buffer, and sorted into an optimal sequence.
Because the drive knows where both files are, it executes the commands in a sequence
using the least amount of thrashing around the drive surface. Additionally, TCQ cuts
down on overhead by not having to complete the first request before working on the
second (or third) request. TCQ was introduced in SCSI-2 (1995).
Given in most cases, when a single user is not doing many different file accesses at the
same time, IDE makes the most sense. A PC server that needs to access different files
for different users at the same time benefits greatly by having a modern SCSI subsystem.
Command Queuing
Command Queuing was introduced with SCSI-2
SCSI IDs
As it was revealed in the words above, multiple SCSI devices may reside on one bus.
This potentially presents a challenge for who's who in the digital zoo. This issue was
eliminated by the use of SCSI ID. The ID is a simple numbering system, starting with
zero.
SCSI ID 0
By default, SCSI ID 0 is used for the drive containing an Operating
System.
A person installing a SCSI device can make a choice of what SCSI ID the drive will use.
The most common method of setting the SCSI ID on a drive is by the use of jumpers.
SCSI 269
If we refer back to the discussion of binary math in Chapter 0000, we note that this is
how most SCSI drives have their ID set.
All jumpers off on the ID block give the drive the ID of zero.
Putting a jumper on the first pin set creates a SCSI ID of 1.
Putting a second jumper on, right next to the first jumper gives a SCSI ID of 3 (2+1).
Jumpering three pin sets yields 7 (4+2+1).
Review the SCSI table for an expression of this idea. Note that depending on the drive
manufacturer, this table may be reversed.
SCSI ID
0
1
2
3
4
5
6
7
Jumper Settings
off / off / off
off / off / on
off / on / off
off / on / on
on/ off / off
on / off / on
on / on / off
on/ on / on
SCSI ID 7
By default, SCSI ID 7 is used by the SCSI controller card
Unique ID
Each SCSI device on a single bus must have a unique ID.
LUN
Early SCSI devices were limited to 8 unique IDs (0-7). SCSI expanded the capability
on a single bus with the use of the Logic Unit Number (expander). LUNs expand the use
of a single SCSI ID.
A good example is a tape backup system, with an auto-loader. Lets say the tape drive has
the ability to hold 7 tapes in a holder.
By giving the holder a number 1 for Sunday, 2 for Monday, etc., the tape drive would
only have a single SCSI ID, yet a way for the auto-loader to know which tape to insert
into the tape drive.
Currently, 960 LUN numbers may exist for a single SCSI ID.
Termination
Understanding why termination is needed is critical to understanding SCSI. Let's take a
moment to discuss the physics involved. As you saw in Chapter 0000, an electron is the
basis for electricity. It was revealed that the electron is only a piece of an atom, and a
single atom certainly does not have much weight. Further, you saw that an electron
moves at nearly the speed of light (about 186,000 miles a second).
Finally, it became clear that an electron could move through some materials such as
copper, very easily. Unless an electron is being pushed by very high voltage, it does not
move through air. So to end or complete the signal it is necessary to terminate the SCSI
bus. This is done either one of three ways. The first option is resistor packs that are
installed on the SCSI device. The second option is auto termination, available on more
intelligent devices. The third is by jumpers.
SCSI 271
A resistor pack will completely impede the flow of electricity (remember our discussion
in Chapter 0000 on electricity?) Simply put the signal is absorbed by the terminator
preventing signal bounce.
If I were an electron
The following paragraph is somewhat simplified, which leads to some
slight technical inaccuracy. In other words, don't use the following
paragraph to design a SCSI controller ;-) While being a little over
simplified it certainly gets the point across.
So imagine being an electron, or to be more precise, a group of electrons carrying a
command down a copper wire. So, here you are tooling along at a speed that would take
you (literally) to the moon in only two seconds. Suddenly, you reach the end of the
copper, and hit air.
Without much weight, and certainly a lot of velocity, each electron would bounce off the
air much like a super ball would bounce off a wall. (Since the electron is already subatomic, it would be difficult to make it crash into smaller pieces.) And the group of
electrons would come flying back down the copper wire, repeating the same command
sequence, again, and again, and ... (you get the idea).
Can you see how the device on the SCSI bus would get quite confused, getting the same
command repeated to it over and over and over? Actually, in real life, it is even a worse
situation. The electron reflects and subtracts in weird combinations. The circuitry
attached to the SCSI bus is simply responding to electrons. As you can see, complete
havoc and data loss would be the result.
The solution lies in Termination. The idea is to absorb the signaling electrons at both
ends of the transmission line (cable), making signal bounce a non-issue.
Terminate
All SCSI cables Terminate at the end of the cable, on both sides.
Types of Terminators
Terminators come in both active and passive forms. Passive terminators
are only used on older drives and controllers. Passive termination is
powered from the SCSI cable using the term power line on the SCSI bus.
Active termination works by using a small voltage regulator, eliminating possible voltage
fluctuation from Term Power, allowing termination resistance to be lowered to 110 ohms.
Active termination is always recommended. Active termination is required for single
ended Ultra SCSI, and above. Active termination does a better job of matching the
impedance of the SCSI bus.
SCSI 273
If you ever see the term Forced Perfect Termination, it was a great idea. It wasn't
widely implemented, and died some time ago.
QAS
TCQ was not the Holy Grail for SCSI. With Ultra320 (SPI-4) comes the draft
proposal for Quick Arbitrate and Select (QAS), which is expected to appear
in drives and controllers in the second half of 2001
Connectors
You saw before that connectors changed as SCSI improved in performance. A curious
note the SCSI trade association web site does not recognize the DB-25 as the first SCSI
connector. Yet, the real world may well expect you to know that the DB-25 is available
as a SCSI connector. Specifically the DB-25 was seen in SCSI-1 devices.
DB-25
Officially, the DB-25 doesnt exist for SCSI. AND, it EXISTS in the
REAL WORLD.
SCSI 275
For your visual edification, most of the popular SCSI connectors are drawn out.
Name
Visual Identifier
DB-25M SCSI-1
(Cable View Point)
HD-50 SCSI 2
SE/HVD
The original SCSI used positive voltage as a one and zero voltage as a zero. This
became known as Single-Ended signaling (SE). The down side to SE was limits to cable
length and speed limitations. The solution came by using two wires. By having zero
voltage on both wires, the signal was a zero. By having two positive voltages of
different values on the wire pair it gives a logical one (value) for signaling. This was
called differential signaling. Later the term was redefined to High Voltage Differential
(HVD). This allowed for longer cable lengths.
SCSI-1
Fast
Fast-Wide
Signaling Bus
SE Maximum Length
5 MHz
10 MHz
20 MHz
6 Meters
3 Meters
1.5 Meters
25
SCSI 277
LVD/MSE
As SCSI moved up in speed, the cable lengths would have gotten ever shorter, as a quick
review of the table above reveals. At 40MB, SE (single ended) would have been limited
to just over 2 feet (.75 meters).
Something had to be done. A new option was created. Low Voltage Differential (LVD)
combined the best of SE and HVD.
This eliminated the possibility of blowing up a SCSI device with high voltage, and
created the possibility of cable lengths of up to 12 Meters.
Another type of LVD is known as MSE, or to be technically accurate, LVD/MSE.
This is a multimode device allowing connection of both types of drives.
It will automatically switch between an SE device and a LVD device; albeit one device
at a time.
Ultra SCSI
All Ultra SCSI specifications for differential SCSI must be used.
SCSI 279
IV
Summary
You have discovered that SCSI cable quality is almost as critical as SCSI termination
along with termination types such as Active and Passive.
It became clear that each device in a SCSI chain needs its own identification and that
any ID may have many LUN's associated with it.
The various cable distance limitations have been explained.
It became clear that while the DB-25 is not an official connector, it has been widely used
and that gives an opening for destroying a parallel-based printer.
The differences between fast and wide SCSI, as well as the various flavors of Ultra
SCSI have been explained. Finally, you saw the dangers of mixing SCSI and differential
SCSI and the symbols that warn of the two types.
1. The maximum number of devices, not including the SCSI controller for SCSI 1 would
be
A. 2
B. 5
C. 6
D. 7
2. A LUN is used to
A. Provide communication between two SCSI controllers
B. Sub-divide a single SCSI ID
C. Is a guard-band frequency allocation between SCSI devices?
D. To Identify if termination is functioning
SCSI 281
4. True or False. The DB-25 connector is widely used, however is not official.
A. True
B. False
9. You have been assigned the task of adding external SCSI drive to an existing server
that has functioning SCSI drives operating internally. After installing the external
drives, you cannot see them when you power up the server. You know they have
power because you can hear them spin up. What is the most likely source of the
challenge?
A. External Termination is incorrect
B. Internal Termination exists on the SCSI controller
C. The internal drives must be disabled
D. All choices are correct
SCSI 283
11. The default SCSI ID for a SCSI drive containing a bootable Operating System is
A. 0
B. 1
C. 6
D. 7
13. SCSI can be found as both 8-bit and 16 bit in what ranges of throughput
A. 10MB to 160MB
B. 20MB to 160MB
C. 20MB to 320MB
D. 5MB to 80MB
SCSI 285
16. LVD/MSE is a
A. Proposed Standard
B. Multimode
C. Both choices are correct
D. Neither choice is correct
17. True/False. Cable for SE and LVD are keyed differently to avoid damage
A. True
B. False
19. If when reading a SCSI specification, the word Wide is not mentioned, the
assumption is
A. Ultra
B. Narrow
C. Single Ended
D. Ultra 2
SCSI 287
2. A LUN is used to
A. Provide communication between two SCSI controllers
*B. Sub-divide a single SCSI ID
C. Is a guard-band frequency allocation between SCSI devices?
D. To Identify if termination is functioning
Explanation: Logical Unit Numbers (LUN) are used to sub-divide a single ID into
separate channels.
4. True or False. The DB-25 connector is widely used, however is not official.
*A. True
B. False
Explanation: The DB-25 connector was common in SCSI 1 implementations. Many
users commonly mistook it for a parallel port connector.
SCSI 289
9. You have been assigned the task of adding external SCSI drive to an existing server
that has functioning SCSI drives operating internally. After installing the external
drives, you cannot see them when you power up the server. You know they have
power because you can hear them spin up. What is the most likely source of the
challenge?
A. External Termination is incorrect
*B. Internal Termination exists on the SCSI controller
C. The internal drives must be disabled
D. All choices are correct
Explanation: When external devices are added to a SCSI chain, termination needs to be
moved from the SCSI controller to the last external device in the chain.
11. The default SCSI ID for a SCSI drive containing a bootable Operating System is
*A. 0
B. 1
C. 6
D. 7
Explanation: The boot drive of a system uses ID 0 by default. It is recommended that
this default be used.
SCSI 291
13. SCSI can be found as both 8-bit and 16 bit in what ranges of throughput
A. 10MB to 160MB
B. 20MB to 160MB
C. 20MB to 320MB
*D. 5MB to 80MB
Explanation: Current SCSI can range from 5MB/s to 80MB/s. Newer SCSI 160 and
SCSI 320 will support 160MB/s and 320MB/s.
15. True or False. All Ultra SCSI specification use differential SCSI
*A. True
B. False
Explanation: Ultra SCSI always uses differential SCSI
A. 16. LVD/MSE is a
B. Proposed Standard
*C. Multimode
D. Both choices are correct
E. Neither choice is correct
Explanation: LVD/MSE is a multimode version of SCSI.
17. True or False. Cable for SE and LVD are keyed differently to avoid damage
A. True
*B. False
Explanation: Unfortunately, the only way to tell SE and LVD apart is to look at the
symbols or documentation.
18. True or False. Using LVD/MSE overcomes the distance limitations of an SE drive
A. True
*B. False
Explanation: SE has a larger distance than LVD/MSE.
19. If when reading a SCSI specification, the word Wide is not mentioned, the
assumption is
A. Ultra
*B. Narrow
C. Single Ended
D. Ultra 2
Explanation: Unless specified as Wide, a SCSI chain will support 8 devices.
SCSI 293
Introduction
II
Removable Media
We will cover CD-ROM technology and learn how information is read from and written
to a CD-ROM, including the various formats that are used. You will then be shown how
to evaluate a CD-ROMs performance and how to decipher the "X-factor."
Following that, we will discuss a more proprietary, yet commonly used form of
removable media, the ZIP drive. You will be shown how this technology improves many
of the concepts behind the floppy disk, allowing it to store over 200 times more data.
We will then discuss removable tape media. You will learn the advantages and
shortcomings of both QIC and DAT tape storage methods. Finally, you will learn how to
overcome the many challenges that can arise from using tape media.
III
Floppy Drives
Before the advent of hard disk drives, floppy drives were the sole means of saving
data created using a PC. It's scary to think about, but floppy drives were considered
advanced technology and cost significant amounts of money. With modern hard drives
storing huge amounts of data for very little cost per byte, floppy drives have been
relegated to the lowest needs of data storage.
It's frightening to think about, but many PC users still rely on floppy disks for data
transfer from one PC to another and backing up small files. The most common and
understandable use of floppy disks today is for the storage of driver files that come with
hardware components. Because it has been surpassed by so many other faster and
higher capacity media technologies, floppy disk technology has remained unchanged
for well over a decade.
Floppy Disks
The original floppy disk was developed for use as a read-only media
for IBM mainframes in the early 1970s. The used very thin 8 disks
that held less than 100 KB. Because of their flimsy media and large
size, the term floppy caught on.
At the component level, floppy drives are very similar in construction of hard drives
except on a much more primitive scale. Like hard drives, and they use read-write heads
and convert binary data into electromagnetic pulses and vice versa. Instead of using
floating head technology like hard drives, the read-write heads on a floppy drive
actually come in contact with the storage media, known as a cookie.
These read-write heads are also much larger than the ones used in hard drives,
therefore floppy disks have a much lower track density than hard drives. Where hard
drives have a track density of many thousands of tracks per inch, floppy drives have
track density of about 135 tracks per inch.
Another factor in the very low performance of floppy disks is the fact that the spindle
motor that spins the platter rotates at about 360 rpm, where as in hard drives the spindle
motor achieves speeds of 5400 rpm or better. Because of this very low speed, there's no
need for concern over the fact that the read-write heads come in contact with the
platter. Over time wear can occur because of the removable nature of the floppy disks.
Form Factors
There are two form factors used for floppy disks in a PC, 5.25" and 3.5". The sizes
refer to the floppy disks themselves. The original floppy disk drive for PCs used a full
height (3.25" tall) 5.25" drive bay. The disks themselves had a total capacity of 360KB,
known as a Double Density disks. Later, High Density disks were introduced at 1.2MB
capacity.
5
3
Double Density
360 KB
720 KB
High Density
1.2 MB
1.44 MB
Pin 1
A ribbon cable that is attached the floppy disk controller to the floppy disk drive is a 34wire ribbon cable and looks quite similar to an IDE ribbon cable. Often times you may
find a floppy cable with five connectors on it, although typically it will only have three.
On one end you will find a 34-pin connector that attaches to the floppy disk controller
or motherboard.
Pin 1 on the cable is designated with a colored stripe, but it is often difficult to decipher
where it lies on the controller or motherboard. Most modern motherboards and floppy
cables will have a small key on the connector that prevents it being plugged in
backwards.
On older floppy cables this twist did not exist and in order to designate between Drive
A and Drive B, it was necessary to set a jumper on the floppy drive.
A rule of thumb to use is "after the twist, Drive A; before the twist, Drive B."
Another rule that comes in handy is that Pin 1 always faces the power connector on the
drive.
IV
CD-ROM
Due to their small capacity, poor performance, and generally being a pain in the butt, the
industry needed a removable storage option that could eventually replace floppy disks. It
seemed the best option available was to take the existing technology that had been
developed by Philips and Sony, called the Compact Disk (CD), and enhance it by
allowing it to store computer data and be accessed by a PC. Few changes were needed,
like some additional error correction code technology, and the Compact Disk-Read
Only Media or CD-ROM was born.
The physical makeup of the CD-ROM is similar to that of a hard disk drive. There's a
spindle motor that spins the CD, and a 'read head' that reads the data off of the disk.
The reason 'read head' is in quotes, is that it is quite different technology than what is
used in a hard disk drive. Instead of using an electromagnetic method of reading and
writing data to the device, a CD-ROM drive uses an optical read head. This optical
head assembly is made up of an infrared laser, a mirror, and a focusing lens. A typical
read from a CD-ROM drive happens like this:
The infrared laser fires of beam of light on to the reflecting mirror. The mirror is part
of the head assembly and moves with the laser and focusing lens from the inside to the
outside of the CD.
The mirror reflects light through the focusing lens and on to a point on the CD.
Some portion of the light is reflected back from the disk. Depending on the amount of
light reflected back defines whether it is read as a 1 or 0. Data is encoded onto the CD
using a series of 'pits' and 'lands.' A pit is read as a 1 and a land is read as a 0.
A series of mirrors and lenses focuses the reflected light into a photo-detector.
The photo-detector converts the light into electrical energy and sends it to the controller
as binary data.
Clean IT
To clean a CD-ROM use a center to edge motion with a soft lint free
cloth.
Access time is a composite metric composed of the speed change time (for CLV
drives), the seek time, and latency of the drive.
Unfortunately, there's no easy mathematical formula that you can plop these three
values into and determine a drive's access time.
This is due in part to the fact that these three factors can be done in parallel.
CD-ROM Formats
The most critical and confusing (of course) aspects of using a CD-ROM is the various
different formats used to write data. Unlike a hard drive or floppy disk, there's really
no way to 'format' a CD-ROM disk. There's also really no way of partitioning the disk.
The differences lie in how the data is organized on the disk.
For example, a CD audio disc uses bits and bytes to store data. However, this method is
completely different from the way information is written to a CD-ROM disk that stores
PC data. They are encoded in two separate formats.
The first major difference in formats is whether a disk is recorded in single-session or
multi-session. The reason that is so troublesome for many users is that many older CDROMs only supported single-session disks.
With a single-session disc, all of the data that is ever going to be stored on that disk is
placed there when it is manufactured.
In a multi-session disc, one session of data can be recorded and then added to later. A
simple analogy is that a single-session disk is like one book. The book begins with the
table of contents and is followed by the body text. A multi-session disc would be like a
library of books where there are multiple sets of text preceded by a table of contents
for each body of text.
Some older CD-ROM drives only supported single-session disks. To further complicate
matters, some CD-R drives will only support single-session disks.
CD-R / CD-RW
One drawback of a standard CD-ROM drive is that it is unable to write data to the
disk. It is this one sticking point that has kept the CD-ROM from all but replacing
floppy disk drives as a standard for removable media a PC. In response to this, the
Philips Corporation defined a standard for a CD Recorder, known as a CD-R, and a
CD Re-Writable known as a CD-RW. These standards were published in Part II and
Part III of Phillips so-called "orange book" in 1990.
One of the biggest differences between a standard CD-ROM and a CD-R is that of the
media used to record data.
In a standard CD-ROM disk, a glass master CD is made by the manufacturer with
the pits and lands etched into it. This master is then used to stamp copies into blank
plastic CDs.
A thin layer of aluminum is placed over the stamped side of the disk. This is what
makes the bottom of the CD reflective. Then another thin layer of plastic is placed
over the aluminum layer. Finally, any markings or labels are stamped on top of this
final layer of plastic.
The CD-R is constructed in a similar way, but instead of stamping predefined pits and
lands, a single blank track is stamped into a special layer of polycarbonate plastic.
The grooved side of the plastic is then coated with a special dye that has a green or blue
tint. On top of that, a thin layer of gold is used for reflecting the laser beam.
There is then a layer of lacquer and a final layer of polymer plastic to protect it. A
special dye layer is designed to absorb light at a specific frequency.
This absorption of light energy creates a mark, usually by distorting the
polycarbonate underneath it, which is read by any read head in any CD-ROM as a 1.
This technology is also known as Write Once Read Many, or WORM.
SCSI
A second interface used for connecting a CD-ROM to PC is the SCSI interface. SCSI
CD-ROM drives have much higher performance than their ATAPI competitors, but are
less common due to the substantially higher cost of SCSI technology.
As you learned in Chapter 0101, devices on the SCSI chain are controlled completely by
the SCSI adapter.
Therefore, when performing operations that require a read from one device with a
simultaneous write to another device, SCSI outshines IDE.
This becomes even more critical when using a CD-RW device that requires the
uninterrupted data flow from the source of the data (a hard disk or another CD-ROM
drive) to the CD-RW.
What this means is that if you're using an IDE/ATAPI CD-RW, you are forced to not
perform any other operations or run any programs during the burn operation.
However if you're using a SCSI CD-RW, the data transfer is completely controlled by
the SCSI adapter and in turn frees up the rest of the system to do other things.
DVD
The Digital Versatile (Video) Disk, or DVD, is yet another example of the IT industry's
ability to make one technology operate on an ever increasing scale.
A DVD drive is written, constructed and read in an almost identical fashion to that of
a standard CD-ROM. The main difference is that the DVD drive uses a single frequency
red laser to read and write data.
Because a laser light has a shorter wavelength than red light, data bits can be packed
tighter on the same media.
This subtle change in technology allows a DVD to store an amazing 2.6GB of data per
side. I say per side because DVDs are also capable of holding data on both sides of the
disk.
DVD technology also employs a specific form of data compression that requires them
to be read in a DVD drive only. CD-ROM drives are not capable of reading DVDs,
but any DVD drive is capable of reading CD-ROM media.
2. 34
VI
ZIP Drives
A technology that was thought to replace the hapless floppy drive was the ZIP Drive,
developed by the Iomega Corporation. This intriguing piece of technology uses disks
that are roughly the same size as a standard 3.5" floppy disk (although slightly thicker).
Original versions of the ZIP Drive could store over 100MB of data per disk, with
newer versions storing roughly 250MB of data per disk. This was achieved by
addressing some of the more obvious shortcomings of floppy drives, such as:
Head Size: The read-write head of the ZIP Drive is about 1/10th the size of a 3.5" floppy
drive read-write head. This allows bits of data to be written closer together and in a
smaller size, raising the overall capacity of the disk. The smaller heads allow a ZIP Drive
to write data a using 2,118 tracks per inch, compared to 135 tracks per inch on a floppy
drive.
Higher Quality Media: The storage medium itself, called a cookie, is made of a higher
quality material that is coated with the same particles that are used in S-VHS videotape.
These particles have a higher energy level and are not as easy to magnetize. Therefore,
magnetic fields from the read-write heads affect a smaller area on the disk, increasing the
storage capacity of the disk again.
Efficient Storage Format: A conventional floppy disk is divided into radial sectors.
This means that sectors on the outside of the disk take up more surface area than
sectors on the inside of the disk. Due to the fact that a sector is the smallest unit that can
be written to on the disk, the overall surface of the disk loses its efficiency for storage as
data is written towards the outside of the media. This is overcome in the ZIP Drive by
using a method called the zone bit recording. This method formats the disk in such a
way that every sector takes up the same area anywhere on the disk. This translates into
more sectors per track as you move from the inside to be outside the disk.
VII
Tape Drives
The bane of many PC technician and network administrators is the tape drive. Despite its
poor performance, unreliability, and high difficulty of use, tape drives have become a
necessary evil in the IT industry (let me tell you what I really think).
Tape drives provide a high capacity, inexpensive, and widely used media for backing
up large amounts of data to a small, removable cartridge. They typically interface
with a PC via a SCSI port, although lower end models can be used with a parallel port
or floppy disk controller. The two most commonly used technologies of tape are:
Quarter Inch Cartridge (QIC)
Digital Audio Tape (DAT)
These tapes look like a scaled down version of the old reel-to-reel tapes that were used
in large mainframe computers. The basic principle is that of a thin magnetic tape on
which data is written from one end of the tape to the other. Data is written onto the
tape in parallel tracks, usually 20 to 32 tracks that run the length of the tape.
A number of tape manufactures have come together to create the QIC committee. While
consistently advancing the technology, the group has worked to keep a format common to
all members to increase acceptance. The first QIC format was QIC-40. It was called this
because it could back up 40 Megabytes of data. Later the QIC-80 was released.
As time evolved, 3M came out with the Travan series, which extended the QIC lifespan.
Most Travan models can read QIC tapes.
Travan
QIC tapes can be read by most Travan tape drives.
VIII
These tapes use an ingenious method of writing large amounts of data to a very small
amount of magnetic tape.
The read-write head of the drive is actually comprised of two separate read heads and
two separate write heads placed at right angles to each other on a small circular piece
of hardware.
The read write head is then tilted at an angle so that bits of data are recorded
diagonally on the tape. This efficient use of the surface area on the tape allows for more
data to be stored.
Additionally, one write head can write data bits at a diagonal that is perpendicular to
its opposing write-head, effectively doubling the capacity of the tape.
Even more disturbing is a tape drive that has been repaired, due to some malfunction
that cannot read tapes it had written before the malfunction.
Despite all of this, current technology improvements have made it possible to store over
100GB of data on a single tape that is slightly bigger than box of matches.
IX
In this chapter, you learned how data is stored on removable media. We began by
discussing floppy drives; the most commonly used removable media solution.
You learned how the physical mechanics of a floppy drive affect its overall performance.
You were then shown how a floppy drive interfaces with a PC via a floppy controller and
ribbon cable.
You learned how to configure a floppy drive to be Drive A or Drive B by connecting it
in the appropriate place on the cable.
Next, we discussed CD-ROM technology. You learned how information is read from and
written to a CD-ROM.
We then discussed how to evaluate a CD-ROMs performance.
You were shown the differences between CLV and CAV drive, and how access time,
latency, and seek time affect a CD-ROMs performance.
We then discussed how a CD-RW drive writes data to special CD-RW media.
You learned about the three formats that are used to write information, CD-DA, ISO
9660, and CD-ROM XA.
You were then shown how CD-ROMs interface with a PC via ATAPI, SCSI, USB, and
IEEE 1394 interfaces.
Next, you learned about ZIP drives.
You were shown how by improving a few components of floppy disk technology, ZIP
drives are able to hold over 200 times more information that a floppy disk.
We concluded this chapter with a discussion on tape drives.
You learned about the two most common technologies used to store data on tape, QIC
and DAT.
Finally, you learned about some of the more typical challenges that occur when using
tape drives.
3. True/False. It is possible to have both a 5.25-inch and a 3.5-inch floppy on one floppy
cable.
A. True
B. False
4. You completed an upgrade on a users machine. Several days later the user calls you
and tells you that when they change floppies the directory for the old floppy is still
displayed. From the choices below, select the most likely cause of this.
A. The drive cable came loose during the upgrade, pulling the cable off of pin 34.
B. The user has replaced the same floppy diskette and not really changed it.
C. The floppy drive is defective.
D. The floppy diskette is defective.
6. CD-ROM drives can be found in the following formats (Choose all that apply)
A. CLV
B. CRR
C. CAV
D. ECC
12. True or False. Travan drives generally are capable of reading QIC tapes.
A. True
B. False
3. True/False. It is possible to have both a 5.25-inch and a 3.5-inch floppy on one floppy
cable.
*A. True
B. False
Explanation: Provided the cable has an edge connector for the older 5.25 inch style
floppy connector along with the newer pin type connector for the 3.5 inch floppy,
both may be connected, making the answer a true.
4. You completed an upgrade on a users machine. Several days later the user calls you
and tells you that when they change floppies the directory for the old floppy is still
displayed. From the choices below, select the most likely cause of this.
*A. The drive cable came loose during the upgrade, pulling the cable off of pin 34.
B. The user has replaced the same floppy diskette and not really changed it.
C. The floppy drive is defective.
D. The floppy diskette is defective.
Explanation: Pin 34 is used to indicate a diskette change. If the cable was pulled loose
from pin 34 during the upgrade, it could not indicate to the system the fact that the
diskette has changed. This make choice A the most likely cause. A defective floppy
drive would not give the same directory. Instead an error would be created, making
choice c incorrect. The same issue arises with choice d making it incorrect as well.
Since the user has not indicated issues before this, it is unlikely they are
inexperienced enough to mix up diskettes making choice b a low probability, and an
incorrect option.
6. CD-ROM drives can be found in the following formats (Choose all that apply)
*A. CLV
B. CRR
*C. CAV
D. ECC
Explanation: CLV (Constant Linear Velocity) the disk spins slower as you move toward
the outer edge. This makes choice a correct. The other option is to use CAV
(Constant Angular Velocity), which keeps the spin speed the same, making for a
quieter drive. This makes choice c correct. Choice b is bogus while choice d refers to
RAM, making this choice incorrect.
12. True or False. Travan drives generally are capable of reading QIC tapes.
*A. True
B. False
Explanation: 3M released Travan in 1994 as an extension of QIC technology and all but
the largest Travan drives can read QIC tapes, making this statement true
Introduction
The famous saying Garbage in, garbage out has become the battle cry of many PC
users that are frustrated with the results they achieve using their PCs. In this chapter we
will discuss how a PC interacts with devices that are outside of the box.
II
Peripheral Devices
We will begin this topic by discussing the various methods of transferring data from the
PC to the peripheral devices. You will learn about serial communications, parallel
communications, Universal Serial Bus (USB), IEEE 1394 (FireWire), infrared, and radio
communications.
Continuing from there, you will be shown all of the various ports and connectors that are
found on the back of PC. These connector types are defined by the data transfer method
they use. You will also learn what devices commonly use these ports and connectors.
Once we have covered the methods of data transfer and the connectors that are used, we
discuss the most common types of peripheral devices used with PCs. You will learn
about the different types of input devices, such as keyboards and mice.
We will conclude this chapter with a discussion of the most frequently used output
devices for PCs, printers. You will learn about the three main technologies used for
generating printed documents; impact printers, inkjet printers, and laser printers.
III
Serial
Serial communication is the simplest form of data transfer in modern PC. In serial
data transfers, data is sent one bit at a time, single file. As you learned in previous
chapters, data is transferred between two devices inside of the computer using a bus.
The analogy we used to describe this transfer was that of a highway. This highway
could have multiple lanes and travel at various speeds. You can think of serial
communications as being a one-lane road. Only one car is allowed to travel down the
road at a time.
In modern PCs, serial communications are full duplex, meaning that both of the
communicating devices can send data at the same time. This is opposed half duplex,
in which only one other device can send data at a time. Full duplex communications
are like that of a telephone call. Both parties on the line are able to talk same time.
Half duplex communications are like that of the CB radio. In order for one party to
transmit they must hold down a button and talk. The other party must listen to what
they say and are unable to talk at the same time.
Both the receiving and transmitting devices must, however, agree on the number of
data bits and the speed data will be sent and received. Most PC devices use either
seven or eight data bits.
Most modern PC's use asynchronous mode for serial communications. A special chip
called a Universal Asynchronous Receiver Transmitter, or UART chip controls these
communications. Most modern PCs come equipped with a 16550 or 16450 UART chip.
Of the two, the 16550 is faster, allowing transmission speeds up to 153,000 bps.
RS-232C
The RS-232C (Recommended Standard-232C) standard is a standard interface
approved by the Electronic Industries Association (EIA) for connecting serial devices.
The standard defines two types of serial devices. The first of which is the DTE, or Data
Terminal Equipment. DTE devices are responsible for processing data to be received
or sent via a serial connection. Simply put, a DTE device is a PC with a UART chip.
The second type of serial device is the DCE, or Data Communications Equipment. DCE
devices communicate with DTE devices. A good example of the DCE device is a
modem.
The RS-232C standard also defines how DTE and DCE devices are connected. It says
that a DTE device will use a 25-pin D-type male connector, also known as a DB-25
male connector. DCE devices must use a DB-25 female connector. However modern
PCs use a DB-9 male connector on the PC, and a DB-9 female connector on the DCE
device.
RS-232C also defines that a DTE device (PC) is then connected to the DCE device
using a straight through cable. This means that the pins used for transmitting data in a
DTE device are connected to the pins responsible for receiving data in the DCE
device. The standard also limits the length of this cable to 50 ft. Of course this limit can
be exceeded depending on the quality of the cable that is used and the speed of
transmission. If you use a high quality, well-shielded cable, lengths of well over a
1,000 ft. can be used. Of course, the higher the rate of transmission, the shorter your
cable can be.
If you wish to connect two DTE devices, say two PCs, you must use a special type of
cable called a null modem cable. A null modem cable is nothing more than a standard
serial cable that has its receiving and transmit wires reversed. They can be purchased
in any office supply or computer supply store for less than $10.00. By connecting two
PCs with the null modem cable, you can perform some basic networking operations,
such as file transfers.
Parallel
If serial communications are like a one lane road, then parallel communications are
like a freeway. A standard parallel connection on a PC transmits data eight bits at a
time, with the total bandwidth of 12 Mb per second. Originally designed for use by
printers, the parallel port has seen many devices that take advantage of this almost
universal interface on PCs. Common uses on the parallel port range from external
hard drives and CD-ROMs and two special security blocks used by a high-end
software packages called dongles. Before the advent of inexpensive networking
hardware, there were even some software packages that could link to PCs with a special
cable that connected to the parallel ports of the computers.
A parallel port on a PC uses a DB-25 female connector. When connecting a printer
through a parallel port, a special cable called an IEEE 1284 cable is used. This cable
has a DB-25 male connector on one end and a Centronics-36 connector on the other
end. The length of this cable should not exceed 15 ft. Parallel ports can be configured
to use three separate modes that allow for varying degrees of functionality with the
port and any device attached to it. The first mode is Standard Parallel Port mode that
we have discussed thus far.
The second mode is EPP, or Enhanced Parallel Port mode. This mode allows for
synchronous bi-directional communication between the port and any device attached
to it. However, in order to make the most use of this mode, the device and the operating
system must support EPP mode.
The third parallel port mode is known as the ECP, or Enhanced Capabilities Port
mode. ECP mode allows for high data transfer rates that are two to fifteen times that of
standard parallel port mode. Like EPP mode, ECP allows for synchronous bidirectional communication between the port and any device attached to it. The
performance gains come from fact that ECP mode uses a DMA channel that is set in the
BIOS, typically DMA 1 or DMA 3. Of course, you would probably want to choose
DMA 3, as sound cards typically use DMA 1. As is true with EPP mode, the device and
the operating system must support ECP mode.
USB
The Universal Serial Bus, or USB, is a peripheral interface that is rapidly replacing
serial and parallel communications. USB was developed by a group of industry giants
such as Compaq, IBM, Intel, Microsoft, NEC and Nortel Networks. Originally intended
as an interface for low bandwidth devices such as modems, keyboards, mice, and
printers, many hardware manufacturers now use the USB interface for other devices
such as CD-RWs and external hard drives.
The most exciting thing about USB is that it's truly "plug and play." The USB interface
allows you to plug in and unplug devices while the computer is running. This is
known as "hot swapping" or "hot plugging." When devices are plugged into the
interface, the operating system will prompt the user for the proper device drivers.
After the drivers are loaded and configured, the user can unplug the device and plug it
in again any time without reinstalling the drivers.
The current specification found on nearly all new PCs is USB 1.1. It allows for 127
devices to be attached to one USB interface, with an overall bandwidth of 12 Mb per
second that is shared by all the devices in the chain. The interface that is built onto the
motherboard or USB controller card is referred to as a root hub. Most new
motherboards come with two integrated USB root hubs. Devices can be attached
directly to the root hub, or you can attach a USB hub with multiple USB ports to the
root hub. The specification also puts a 10m limit on the length of any USB chain.
USB 2.0
The new USB 2.0 specification hit the market in the 2nd quarter of
2001. This new specification increases the bandwidth from 12 Mb per
second to 480 Mb per second. USB 2.0 (HS-USB) devices are
backwards compatible with USB 1.1 devices.
What is truly exciting about this new specification is that its high-speed bandwidth is
greater than current IDE standards. Theoretically, we could see an external hard drive
with greater throughput than an internal hard drive with hot swapping capabilities. HSUSB is found on the newest system boards, and PCI adapters are available for older
systems. The new logo, courtesy of USB.org, is below
IEEE 1394
Originally developed by Apple Computer under the trademarked name FireWire,
IEEE 1394 may well become an industry standard for high bandwidth peripheral
devices. Sony has also implemented IEEE 1394 under the trade name iLink. IEEE 1394
is described in the SCSI-3 standards and has sometimes been referred to as 'serial
SCSI.'
IEEE 1394 will support 63 devices in a chain, with up to 400 Mb per second of shared
bandwidth for the devices. Like USB, IEEE 1394 allows for hot swapping of devices.
Current uses of IEEE 1394 are commonly seen in the digital video cameras, where
high bandwidth data transfers are necessary. However many developers are using this
technology for other devices such as external hard drives, CD-RWs, and other
bandwidth intensive devices. Windows Millennium and XP operating systems
supports IEEE 1394. Its growth in the PC market remains to be unseen with the
introduction of USB 2.0 with its 480 Mb per second bandwidth capabilities. Currently,
only Apple computers and some PC based laptops come with integrated IEEE 1394
technology.
IEEE 1394
IEEE 1394 is used for high bandwidth devices such as digital video
cameras and external hard drives
Wireless
One of the more exciting advancements in PC data transfer is that of wireless
communications.
Although its uses in peripheral devices have been limited, wireless technologies are
growing rapidly in networking markets.
The two most common methods of wireless communications are infrared and radio
(using either 802.11x or Bluetooth).
Infrared
The Infrared Data Association, or IrDA, is a consortium of device manufacturers had
developed a standard for transferring data via infrared light waves.
Many portable devices, such as laptops and Personal Device Assistants (PDAs), are
increasingly using IrDA ports.
Two IrDA enabled devices can easily transmit data between each other as long as
there's a clear line of sight between the devices, and they are in close proximity to each
other.
Current IrDA technology can achieve transmission speeds close to that of the standard
parallel port (12 Mb per second).
As for peripheral devices, they're commonly limited to the low bandwidth devices such
as keyboards, mice, and printers.
Radio
One of the more popular and effective ways of communicating with peripheral
devices from a PC using wireless communications is with radio waves using Frequency
Modulation (FM).
Wireless radio devices have a significant advantage over infrared devices in that
there's no need for a clear line of sight between the devices. However, they are
susceptible to interference from other devices that use radio waves.
One of the more popular uses of radio technology is for wireless mice and keyboards.
Common Use(s)
Centronics 36
Printer
DB-15
VGA
DB-15
DB-25
Parallel
DB-25
Serial
DB-9
Serial
DIN-5
AT Keyboard
Device Interface
Cable Connector
FireWire (IEEE
1394)
MiniDIN-6
(PS/2)
Keyboard, Mouse
MiniRCA
Speakers, Microphone
USB A
USB B
Keyboards
Keyboards are the most commonly used input devices for PCs. Given this, it makes
perfect sense that keyboards have not seen any major technological developments in a
long time .
Keyboards typically interface with the PC by means of a DIN-5 connector (AT
motherboards) or a PS/2 connector (ATX motherboards). Some newer keyboards are
using wireless (or USB) transmission methods as well. However the underlying
technology that makes a keyboard work has remained unchanged for many years. There
are basically two ways in which a keystroke is converted into binary data to be sent to
the processor.
Switched
Capacitive
The mechanical switch keyboard was the original design for keyboards in AT and XT
based PCs. Each key on the keyboard had its own individual switch, which is made up
of two small metal plates that would come into contact with each other when a key
was pressed. When the plates touched a signal was sent to the processor telling of
which key was pressed.
When the key was released, a small spring would push the key back up into place. This
action gives switch keyboards their familiar clicking noise that could become very
annoying for the user or any of their neighbors. The original models of switched
keyboards also had a design flaw caused by the springing action of the key. This
caused a single keystroke to send multiple signals to the processor.
This was solved by an electronic technique called debouncing. With this technique the
keyboard controller would constantly scan the keyboard for keystrokes. Keystrokes
would only be processed if a key were depressed for two or more scans. This anomaly
was ultimately eliminated with the advent of a newer keyboard technology known as
capacitive keyboards.
Capacitive
Capacitive keyboards first arrived with the introduction of laptop computers. Because
mechanically switch keyboards were too bulky and required too much power, a new
technology had been developed for use in laptops.
Capacitive keyboards eliminate the expensive, power hungry key switch. Instead they
use two sheets of semi conductive Mylar material. A thin layer of air separates the
sheets of material. Each key in the keyboard has a small plunger on the bottom of the
key. When a key is pressed the plunger pushes into the sheets of Mylar, forcing them
into contact. When the sheets come into contact, the small voltages running through
each sheet change. The keyboard controller can tell which key was pressed by the
amount of change detected in the voltages.
These keyboards have the advantage of been inexpensive to manufacture and lower
power consumption. However they are impossible to repair should they be damaged.
Because of their low cost to manufacture, they can easily be replaced for less than $50.
Mice
With the introduction of the Graphical User Interface (GUI) operating systems, such
as Windows, the mouse quickly became the second most common input device found
on a PC. Mice can connect to the PC using a serial port, a PS/2 port, USB, or wireless
technology. Regardless of the way they connect to the PC, there are three basic
technologies used to track the movement of a mouse in order to send input signals to
the processor.
Opto-Mechanical Mouse
Optical Mouse
Digital Tablet
Opto-Mechanical Mouse
The most common type of mouse found in most PCs is the opto-mechanical mouse.
This mouse works by using a small rubber ball that rotates as the mouse is moved
around the mouse pad.
The movements of the ball engage two small rollers that track movements on the xaxis and y-axis. At the end of each roller there's a small wheel with holes in it. On one
side of the wheel is a small LED and on the other side a photo sensor.
As the wheel spins the holes allow light through them that is picked up by the photo
sensor.
This generates the necessary signals to tell the processor how far and how fast to move
the pointer on the screen on both the x-axis and y-axis.
Optical Mouse
The optical mouse works by the same principles as a mechanical mouse, except that it
uses an LED and a special mouse pad to track the movements of the mouse.
The LED shines from the bottom of the mouse and is reflected back by the pad into a
photo sensor.
The mouse pad itself has a grid of horizontal and vertical lines that absorb the light
when it hits them.
The alternating light and dark signals are used to track X-Y movements on the pad.
MS Intellimouse
The new IntelliMouse from Microsoft eliminates the need for special
mouse pad while still being a true optical mouse. The LED and photo
sensor instead use the natural variations of any surface to track the X-Y
movements.
Modems
The most commonly used serial devices in PCs are modems. The word modem is
actually an acronym for Modulate-Demodulate.
Modems allow PCs to send data over a phone line to another computer. Digital
signals from the processor are modulated into analog tones that can be carried over a
phone line.
On the receiving end, the analog tones are demodulated back into digital signals that
can be processed by the receiving PC.
The invention of the PC modem can be directly attributed to a gentleman by the name
of Denny Hayes.
Before the introduction of the Hayes SmartModem, modems worked on a
synchronous basis. Each modem has an internal clock. The clock in each modem was
used to synchronize data transfer. This type of modem is known as a synchronous
modem.
The SmartModem created a relatively low-cost modem by eliminating the clock. Hayes
was able to eliminate the clock by specifying when a data transmission begins and
ends by using start and stop bits.
The one challenge Hayes faced was how to control the modem. To accomplish this task
he created the AT command set.
AT Command Set
The AT command set is a set of text-based instructions that can be used to control a
modem manually. With newer modems, these commands are automatically
performed in the background by the modem's software. You can use any of these
commands on any modem through a program called Hyper Terminal.
This program comes with any of the Windows operating systems and is a great tool for
troubleshooting modem challenges.
The first command is 'AT,' which stands for attention. This command wakes up the
modem and readies it for use. AT commands can be sent in a compound fashion.
Another popular command is 'D' for dial. The command structure uses a T for tone
dialing and a 'P' for dialing with pulse or rotary dialing.
Given this lets say you wanted to dial the number 555-1212. Using Hyper Terminal or
other modem control program, you would enter:
ATDT 555-1212
If the modem were behind a Private Branch Exchange (PBX) it would not be able to
dial without getting an outside line and its dial tone.
For most PBX systems this requires dialing 9 and a brief pause to wait for an outside
line. The AT command set offers a pause for two seconds by using a comma. So now
to dial the example above you would enter:
ATDT 9,555-1212
To get the modem to pick up the phone you would enter:
ATH1
To hang up the phone you would enter:
ATH0
Sometimes a modem will go through a considerable amount of reconfiguration, either
by the user or a software application in the PC. To reset a modem back to its original
configuration, the command ATZ resets it back to its original state.
ATZ
The reconfiguring of the modem for a particular call allows each modem to find the
maximum transfer rate that can be attained between the two modems. This is known
as negotiation or training. Theseries of commands can add more intelligence to the
modem. For example, using X4 adds the ability to detect both dial tone and busy
signals.
Basic AT Modem Commands
AT Command
Description
AT
Attention. Wakes up the modem
Dials the modem in either tone (T) or pulse (P), followed by the
ATD(T|P) xxx-xxx-xxxx
number
ATA
Automatic answer
ATH(0|1)
Pick up (1), or Hang up (0)
ATZ
Reset the modem
Table 15 AT Command Set
S-registers
Like all other intelligent devices for PCs, the BIOS in the modem can be configured
differently from the factory defaults. These settings are known as the S-Registers.
Suppose your PBX at the office could get an outside line in less than one second. By
changing the value of Register S8, you control the amount of time a comma pauses.
Modems may have dozens or even hundreds of S-register settings.
The most recent scrap was the 56K Flex verses US Robotics X2. Of course, the two
standards didn't interoperate. The battle was ended with the release of the standard
set by the International Telecommunications Union (ITU). This example was just the
latest of a 20-year battle of different standards.
V.xx Standards
The V.xx standards were set by CCITT (Comite Consultatif International
de Telegraphique ed Telephonique), now known as ITU-T. The higher the
number after the V. the newer the standard. Sometimes this meant full
duplex was added; sometimes it meant extending a standard to include higher data rates.
V.42 is an error-detection standard, which competes with the various Microcom
Networking Protocol (MNP) versions.
V.90
This standard ended the battle between Rockwell/Lucent's K56flex and US
Robotics' (now 3COM) entry with X2. The idea behind 56k is simple.
Instead of converting the returning signal back to analog, require the
service provider to have a direct digital connection, and leave the returning to the V.90 a
signal digital.
Line quality and distance will limit the actual throughput of the returning digital data.
V.90
V.90 modems under ideal conditions are 53kbps to the user
(downstream), and 33kpbs from the user (upstream).
Printers
Next to monitors, printers are the second most used output devices for PCs. There are
many varieties of printers used by individuals and businesses today.
They range from noisy impact printers used to print multi-formed invoices, to highresolution color inkjet printers used for printing photographs.
All of these various types of printers can be placed into one of three categories:
Impact printers
Inkjet printers
Laser printers
Impact Printers
An impact printer works by hitting an inked ribbon with a print head onto the paper.
Although they are slow, noisy, and considered antiquated technology, they're still
widely used by many businesses that generate multi-formed documents.
The two types of impact printers still in use today are:
Dot-Matrix
Daisy Wheel
Dot-Matrix
A dot-matrix printer works by using a print head that is made up of 9 to 24 pins. The
more pins that are used the better the quality of printing can be achieved.
Each of the pins is controlled individually by an electromagnetic solenoid. Signals
from the printer controller controls when each of the pins is fired.
When electromagnetic pulses are sent to a pin, the pin is forced out of the print head
and impacts the printer ribbon, creating the image on paper.
The print head on dot-matrix printer can become extremely hot due to the
electromagnetic nature of the solenoid. Care should be taken when servicing a dotmatrix printer as the print head can easily burn your skin.
To prevent the printer from overheating, dot-matrix printers use a thermal resistor, or
thermistor, that is designed to pause or shut the printer down when it gets too hot. If
this resistor goes bad the printer may constantly shut down and re-power.
The performance of a dot-matrix printer is measured in the number of characters per
second (cps) it can generate. Dot-matrix printers have an advantage over their daisy
wheel cousins in that they are the only type of impact printer that can produce graphic
images.
Dot Matrix
Some Dot Matrix printers handle paper using a tractor feed system that
requires continuous form paper.
Daisy Wheel
Daisy wheel printers get their name from the way the print head is designed. The print
head has a series of 'petals' arranged radically. At the end of each petal is a letter or
number.
A small pin behind the wheel that is controlled by an electromagnetic solenoid
engages a single petal and it impacts it into the print ribbon. The printer controller
makes sure that the correct letter is in front of the pin before it engages the petal.
Due to the fact that they use preset letters, daisy wheel print heads must be swapped
out in order to change typestyles or font sizes. This also limits daisy wheel printers to
letters and numbers, but no graphics. Like dot-matrix printers, daisy wheel printer's
performance is measured in characters per second (CPS).
Inkjet Printers
Inkjet printers are the most commonly used type of non-impact printers among home
users and small businesses. This has been aided by the fact that many inkjet
manufacturers have taken the "Gillette approach" to marketing their printers (give
them the razor, sell them the blades). Of course nobody's giving away inkjet printers but
their low cost compared to other printing technologies, like laser printers, is very
attractive for users with small budgets.
Inkjet printer technology has been distilled down to two commonly used methods for
transferring ink to paper:
Bubble Jet
Piezo-Electric Inkjet
Bubble Jet
Canon, Inc. developed Bubble Jet technology in the late '70s. The ink cartridge of a
Bubble Jet printer houses the ink as well as the print head in a replaceable cartridge.
The print head itself is made up of 300 to 600 nozzles that are less than the diameter of
a human hair. The more nozzles in the print head the better the resolution of the
printer.
The printing process itself involves three main steps. First, a heating resistor behind
each print nozzle heats up when it receives a signal to place a dot on the page from the
print controller. This heat creates a bubble behind the ink reservoir and forces a small
drop of ink out of the nozzle on to the paper.
As the resistor cools down, the bubble collapses, and a vacuum forms. This draws ink
from a reservoir to replace the ink that was ejected. This heating and cooling process
requires special inks that are heat resistant.
It also means increased printing time due to the overhead of these two processes.
Leaking Ink
Pump pressure prevents ink from leaking out of an ink jet printer
Laser Printers
The most common type of printers used by mid to large size businesses and some and
home users is laser printers. Because laser printers represent a large percentage of
calls to PC technicians, CompTIA has made them a significant focus on their A+ Core
Technologies test.
The entire printing process revolves around a special device contained within the print
cartridge called a photosensitive drum. The process in which a page is printed from a
laser printer involves six stages:
Cleaning
Conditioning
Writing
Developing
Transferring
Fusing
Note: This process is specifically the HP printing process. It is the only laser printing
process covered on the A+ Core Technologies exam.
Cleaning
The HP printing process begins by cleaning the photosensitive drum of any excess
toner leftover from the last print.
A small plastic blade, called a cleaning blade (go figure), scrapes off the leftover toner
into a small waste bin called a hopper
Conditioning
During this phase the Primary Corona Wire conditions the photosensitive drum.
This conditioning involves applying a uniform negative charge to the entire surface of
the drum. The charge is supplied by the High Voltage Power Supply (HVPS).
Writing
The writing phase is where the laser is actually used. By exposing the photosensitive
drum to light, it loses any charge it is holding.
The laser paints a mirror image of the print on the drum. Any areas that are hit by a
laser drop down to 0V charge.
Developing
In this phase the photosensitive drum passes by a developing roller that is coated with
toner.
The toner itself has a small negative charge to it. Any areas on the photosensitive
drum that were exposed by the laser attract the negatively charged toner to the drum.
Transferring
The transferring phase is where the paper finally shows up. The paper passes over a
Transfer Corona Wire that applies a uniform positive charge to the paper.
The toner that is stuck to the photosensitive drum is pulled onto the paper by its
opposite electro-magnetic charge.
The paper is then discharged buy a static eliminator strip to prevent it from becoming
hung up in any of the other assemblies.
Fusing
Fusing the toner into the paper finishes the process. This is done by a set of fusing
rollers that apply heat and uniform pressure to the paper and toner.
The toner itself contains small particles of plastic that allow for better fusing into the
paper. This is the reason printed pages are warm when they come out of a laser printer.
WARNING: Fuser rollers get extremely hot. You should always let a laser printer cool
down before putting your hands in it.
Knowing the six steps and the components involved in the process allows you to
determine a myriad of challenges that can occur with laser printers.
For example, when you pick up a printed page from a laser printer and the toner falls
off of it, you may deduce that the fusing rollers are not applying enough heat or
pressure to fuse the toner into the paper.
Many of these components used in the laser printing process are contained within the
toner cartridge itself.
The cleaning blade, toner, developing roller, and the photosensitive drum are all
contained within this cartridge.
The cartridge can be easily replaced should any of the components fail or if the toner
runs out.
This makes toner cartridges one of the most environmentally friendly components
used in the computer industry.
California Cows
An easy way to remember the six steps in the HP Printing Process is
California Cows Wont Dance The Fandango (Cleaning, Conditioning,
Writing, Developing, Transferring, Fusing).
Sticky Sheets
Damp paper, improper media or a worn separation pad could cause a
Laser print to feed more than one sheet at a time
Troubleshooting Tip
A contaminated fuser roller could make the backside of the copy dirty.
Printer Cable
The printer cable for ECP use is parallel cable IEEE 1284
Test Page
After installing a printer always print a test page
Initiate a test page using the operator panel on the printer or through the properties page
in the printer settings.
To print a test page:
, settings
Go to start
printer you wish to test.
, and printers
IV
Summary
We began this chapter with a discussion of the various methods of transferring data from
the PC to the peripheral devices. We discussed the most basic form of data transfer, serial
communications. You learned about asynchronous and synchronous communications, as
well as the difference between half-duplex and full-duplex communications.
We learned about parallel communications. You were shown the three different variants
of parallel communications; Standard Parallel Port mode, Enhanced Parallel Port mode,
and Enhanced Capabilities Port mode.
You then learned about the most exciting advancements in peripheral interfaces, USB
and IEEE 1394. You were shown that these interfaces are not only truly plug and play,
but that a single interface can be shared with 127 or 63 devices respectively.
Next, you learned about wireless data communications. You were shown the two most
common methods, infrared and radio, and their respective advantages and shortcomings.
We discussed the various ports and connectors that are used by each of these data
transmission technology.
We discussed the various peripheral devices and how they make use of these interfaces
and communications methods. We began with input devices. You learned the difference
between switched and capacitive keyboards. You also learned about the differences
between mechanical and optical mice.
We then concluded with a discussion of output devices. First, you learned about how
modems allow two computers to communicate using a phone line. You then learned
about the three types of printers used by PCs; impact printers, inkjet printers, and laser
printers.
A. USB A
B. USB B
C. DB-15
D. DB-25
8. How many IRQs would a USB chain with a mouse, keyboard, and printer use?
A. 1
B. 2
C. 3
D. 4
11. What type of printer is commonly used for printing multi-part forms?
A. Inkjet
B. Laser
C. Daisy wheel
D. Dot Matrix
12. You are servicing a laser printer. The pages come out of the printer with ghost images
of the previous page. Which component do you suspect?
A. Primary corona
B. Transfer corona
C. Developing roller
D. HVPS
13. You are servicing a dot matrix printer that is shutting down sporadically. What should
you replace first?
A. Print head
B. Ribbon
C. Paper feed rollers
D. Thermistor
14. What is the command for resetting a modem to its default settings?
Reset COM1
A. ATR
B. AT0
C. ATZ
15. You wish to dial 555-1212 with your modem using pulse dialing. You are in a
building with a PBX that requires a 9 be dialed to reach an outside line. What
command do you use?
A. ATD 5551212,9
B. ATDP 9,5551212
C. Dial 9,5551212
D. ATH1 9,5551212
16. What 'standard' for modems was named after its creator?
A. Baud
B. Hayes
C. V.90
D. Kflex
17. What step comes after the conditioning phase is a laser printer?
A. Cleaning
B. Developing
C. Writing
D. Transferring
18. What methods are used to transfer ink to paper in Inkjet technology? Choose two
A. Boiling ink
B. Vibrating crystals
C. Compressed air
D. Piston
E. Vacuum
20. You find a cable that has a DB-25 male connector on one end, and a Centronics 36
male connector on the other. What type of cable is it?
A. IEEE 1284
B. IEEE 1394
C. Null modem cable
D. RS-232C cable
A. USB A
B. USB B
*C. DB-15
D. DB-25
Explanation: This is a DB-15 female connector. Specifically, it is a VGA connector,
because the pins are arranged in 3 rows. If it were a two row DB-15 female, it would
be a Gameport / MIDI connector.
8. How many IRQs would a USB chain with a mouse, keyboard, and printer use?
*A. 1
B. 2
C. 3
D. 4
Explanation: Although a single chain can support 127 devices, the chain and all its
devices only require 1 IRQ.
11. What type of printer is commonly used for printing multi-part forms?
A. Inkjet
B. Laser
C. Daisy wheel
*D. Dot Matrix
Explanation: Impact printers are the only printers that can output to multi-part forms.
Although a daisy wheel would work, they are rarely used anymore.
12. You are servicing a laser printer. The pages come out of the printer with ghost images
of the previous page. Which component do you suspect?
*A. Primary corona
B. Transfer corona
C. Development roller
D. HVPS
Explanation: Ghost images are commonly an effect of the photosensitive drum not being
conditioned thoroughly after the last print. The Primary Corona Wire does this.
13. You are servicing a dot matrix printer that is shutting down sporadically. What should
you replace first?
A. Print head
B. Ribbon
C. Paper feed rollers
*D. Thermistor
Explanation: The thermistor in a dot matrix printer is responsible for slowing down or
shutting off the printer should it get too hot. If it goes bad, it can turn the printer off
sporadically, even if it is not hot.
14. What is the command for resetting a modem to its default settings?
Reset COM1
A. ATR
B. AT0
*C. ATZ
Explanation: To reset a modem to its factory default settings, ATZ is typed in the
HyperTerminal program.
15. You wish to dial 555-1212 with your modem using pulse dialing. You are in a
building with a PBX that requires a 9 be dialed to reach an outside line. What
command do you use?
A. ATD 5551212,9
*B. ATDP 9,5551212
C. Dial 9,5551212
D. ATH1 9,5551212
Explanation: To dial using pulse, use the ATDP command, followed by the phone
number. Because of the PBX, you preface the number with 9, which reaches the
outside line and pauses for 2 seconds before dialing the number.
16. What standard for modems was named after its creator?
A. Baud
*B. Hayes
C. V.90
D. KFlex
Explanation: Denny Hayes is credited with creating the first modem. Hayes
compatible modems have long been a standard.
17. What step comes after the conditioning phase is a laser printer?
A. Cleaning
B. Developing
*C. Writing
D. Transferring
Explanation: The writing phase follows the conditioning phase. It is here that the laser
paints an image of what is to be printed on the photosensitive drum.
18. What methods are used to transfer ink to paper in Inkjet technology? Choose two.
*A. Boiling ink
*B. Vibrating crystals
C. Compressed air
D. Piston
E. Vacuum
Explanation: The two most common methods, Bubble Jet and Piezo-electric, use boiling
ink and vibrating crystals respectively.
20. You find a cable that has a DB-25 male connector on one end, and a Centronics 36
male connector on the other. What type of cable is it?
*A. IEEE 1284
B. IEEE 1394
C. Null modem cable
D. RS-232C cable
Explanation: A cable with a DB-25 male connector on one end and a Centronics 36
connector on the other is a printer cable. The official name for it is an IEEE 1284
cable.
Multimedia 383
2.1 Identify common symptoms and problems associated with each module and
have to troubleshoot and isolate the problems.
DVD
Sound
Sound card/audio
Monitor/video
USB
Cables
Peripherals
4.2 Identify the categories of RAM (random access memory) terminology, their
locations, and physical characteristics.
VRAM (video RAM)
WRAM (Windows accelerator card RAM)
Multimedia 385
Introduction
The multimedia experience has revolutionized the way PC users experience their
computers and has redefined the way technology has progressed since the early 1990's.
Simply put, multimedia is the combination of audio and video elements to present
information. In this chapter, we will discuss the various components that make up both
the audio and video elements of the multimedia experience in a PC.
Multimedia 387
II
Multimedia
We will begin by discussing how images are displayed on a monitor. You will learn how
the video card in a PC converts the digital instructions from the processor into analog
signals for the monitor. We will discuss the different technologies involved and how they
affect the quality of the images displayed on the screen.
You will learn how the monitor used on a PC affects the image quality. We will discuss
the various factors and technologies that separate a good monitor from a bad one, and it is
not just cost.
You will then learn how a PC creates audio sounds. We will begin by discussing how
analog sounds are sampled and converted into digital information for the processor.
Finally, we will conclude by discussing MIDI (Musical Instrument Digital Interface)
technology that lets users plug digital music instruments into a PC.
Video Cards
The video card in your PC is responsible for outputting the images you see displayed
on your monitor.
A primary function of the video card is to take the digital information given to it by
the processor and convert it into an analog signal that is sent to the monitor. Of all the
other components in a PC the video card is definitely one of the hardest working as it is
constantly sending a new signal to the monitor for display.
Image Display
Before we continue on discussing the various ins and outs of how a video card works,
let's take some time to discuss how an image is displayed on your monitor.
The most basic element of any graphic display is the pixel (short for picture element).
The number of pixels used to make up the entire display is referred to as the
resolution of the display.
This is commonly expressed by writing the number of horizontal pixels by the number
of vertical pixels. Commonly used image resolutions are 640x480, 800x600, and
1,024x768.
By simply multiplying these numbers together you have the total number of pixels
that make up the image display.
These pixels are actually made up of even smaller red, green, and blue dots.
By varying the intensity (brightness) of these dots the color of a single pixel can be
changed into any color that is visible by the human eye.
For example changing the brightness of all three dots to their maximum intensity
produces a white pixel.
By completely turning off all three of these dots, a black pixel is produced.
Color Depth
The number of variants of intensity you can produce in any one of these three dots is
known as the color depth of the image display.
The color depth is determined by the number of bits per pixel that can be used to
create these variants of intensity.
For example, standard VGA mode has a 16-color palette, which requires 4 bits (2^4 =
16) of memory per pixel.
Multimedia 389
Some video cards can support True Color mode, which uses 24 or 32 bits per pixel.
In 24-bit True Color mode, each pixel can display 16,777,216 different colors.
In 32-bit True Color mode, each pixel can display 4,294,967,296 colors.
This is approaching the ridiculous as the human eye can only differentiate slightly more
than 1,000,000,000 different colors. 32-bit True Color mode is simply a waste of
memory.
This was accomplished with the use of a special circuit called a Digital to Analog
Converter, or DAC. This of course was perfectly fine when all you are displaying on the
monitor was simple text.
With the advent of the Windows operating system, the poor CPU found itself bearing
the extra burden of moving and re-sizing windows, dragging and dropping icons, and
managing a large palette of colors.
This quickly became a major bottleneck in system performance and was a contributing
factor in many people's perception of the Windows operating system as a "resource pig
and a performance nightmare."
To relieve the burden from the processor, video card manufacturers introduced what
came to be known as accelerated video cards (as a matter of fact some companies even
marketed them as Windows accelerators).
These new video cards took over the responsibility of re-calculating any changes in the
display. Now if the user moved an icon on their desktop, the processor would simply
send an instruction to the video card to move that particular graphic element from
point A to point B.
The video card, using its own on-board processor and memory, would crunch all the
numbers necessary to make that interaction occur on the screen.
Virtually all in modern PCs come equipped with the accelerated video cards.
This technology has been developed even further with the advent of 3D Accelerated
video cards. These video cards take the enormous burden of performing 3D to 2D
conversion mathematics.
Video Chipset
As you learned in Chapter 0001, the motherboard of the PC contains a special logic
circuit called a chipset that controls the interaction of all the various components in
the computer and the processor.
Accelerated video cards have a similar logic circuit called the video chipset, or the
video coprocessor. The capability of the video coprocessor has a direct impact on the
performance of the video card.
Some video card manufacturers, like Matrox, design every component of their video
cards internally from the ground up.
Multimedia 391
These companies have a greater control over the performance of their product because
they're able to write more efficient software interfaces and drivers for their video
cards.
Other manufacturers, like Elsa, purchase third-party chipsets that are then integrated
into their card design. This isn't necessarily a bad thing, but it can lead to poorly
written and problematic driver software.
Video BIOS
It should be becoming readily apparent the statement made earlier about video cards
being "mini-computers inside your PC," is beginning to take shape.
Like PCs, video cards have their own BIOS programs that control the interface of the
video chipset and video memory.
Of course BIOS programs are simpler versions of the ROM BIOS found on your
motherboard, as it only needs a set of instructions related specifically to video
functions.
Like BIOS programs found on motherboards, video BIOS programs can differ on
different video cards that use the same chipset.
In fact many video cards differ primarily in software, with not much difference in
hardware.
The frame buffer is made up of various types of RAM (discussed later). The
advantage of having video memory integrated into the video card itself is that it can be
designed and tuned for the specific task of displaying an image.
The more video memory contained on the video card, the larger the size of the frame
buffer.
This will dictate the maximum screen resolution and color depth.
For example, a 1024x768 screen resolution in 32-bit True Color requires 3MB of video
memory.
A 1600x1200 screen resolution in 32-bit True Color requires slightly more than 7MB
of video memory.
You will notice that both of these values are just shy of the PC friendly 4MB and 8MB
values that PC memory comes in. A video card that would support these settings
would need 4MB or 8MB of video memory respectively. As you can see, almost a full
megabyte of memory goes unused.
RAMDAC
The information about a video display is stored in memory, digitally, as ones and
zeros.
In order to communicate this information to the monitor, which needs an analog
signal, a special logic circuit called a RAM Digital to Analog Converter, or RAMDAC,
is used to take the digital information from memory and convert it into an analog
signal that is then sent to the monitor.
The speed and quality of the RAMDAC has a direct effect on the maximum resolution,
color depth, and refresh rate supported by the video card.
Video Memory
As with all types of technology, the video memory for the frame buffer can come in
many different flavors. Some of the more common types found on video cards today
are:
Multimedia 393
One critical limitation of using standard DRAM on a video card is its lack of
available bandwidth.
The issue stems from the fact that the video co processor and the RAMDAC are
competing for access to the video memory frame buffer.
VRAM addresses this issue by using a technology called dual porting.
Simply put, this means that there are two separate access channels to the memory, one
for the video co processor and one for the RAMDAC. This allows both devices to
access the memory simultaneously.
Multimedia 395
It does not share any of its bandwidth with any other devices. AGP has been
developed further and now offers a 2X mode that doubles its bandwidth to 508.6
MB/s. This doubling of the bandwidth is achieved by sending data on the rising and
falling edge of a single clock cycle, much like the technology used in DDR SDRAM.
Display Modes
With all the different ways available to display video images, the PC industry greatly
needed a set of standards that could insure that a compatible video card could be used
with a compatible monitor.
At first all the manufacturers that had a large stake in their proprietary technologies
decided it would set their own standards. But standards aren't really worth anything
unless everybody can agree on them.
Therefore the Video Electronics Standard Association (VESA) was formed to define a
new set of standards that would be industry recognized.
Display Mode
VGA
SVGA
XGA
8514/A
Table 16 Display Modes
Resolution
640x480
320x200
800x600
1024x768
1280x1024
640x480
1024x480
1024x768
Color Depth
16 colors
256 colors
True color
65,536 colors
256 colors
65,536 colors
256 colors
256 colors (interlaced)
Multimedia 397
III
Monitors
To this point we have discussed how the digital information for the display of images is
translated on the video card by the RAMDAC and sent to the monitor.
The second component of displaying eye-popping graphics from your computer is the
monitor itself. A number of factors go into determining the quality of a monitor. In a
general sense a monitor's capabilities can be determined by simply looking at the
resolution modes that it will support. However there is a handful of other lesser-known
factors that separate a beautiful monitor from the less than useful ones.
Dot Pitch
One of the most critical characteristics to look for when determining the quality of a
monitor is the dot pitch.
As you learned earlier a pixel is made up of red, green, and blue dots. The combination
of these three colors can create any color visible to the human eye. It would stand to
reason that the closer that you can pack these dots together, the better the resolution
your monitor will support.
Dot pitch is the measurement of space between two like-colored dots. The
measurement is specified in millimeters. The lower the number, the better the
resolution of the monitor. Common dot-pitch settings found in new monitors range
from 0.18 mm to 0.30 mm.
Refresh Rate
The second most critical characteristic of a monitor is the refresh rate that it will
support.
You should also note that the refresh rate is a function of the video card and the
monitor.
What it comes down to is finding the highest common denominator of the refresh
rates supported between your video card and the monitor.
Simply put, the refresh rate of a monitor is the number of times the screen is painted
per second.
This value is measured in Hz, or cycles per second.
This means a monitor set to 60 Hz will be repainted 60 times per second.
A rule of thumb is that the higher the refresh rate the less eyestrain the user will
experience.
60 Hz Refresh Rate
While we are speaking about eyestrain, if you use artificial lighting
(florescent tubes), never set your refresh rate to 60 Hz. In the United
States, electrical power cycles at 60 Hz. This means the light from the
fluorescent tubes is actually flickering 60 times per second. When viewing a monitor that
is set to 60 Hz refresh rate under artificial lighting, eyestrain is at it's worst. The most
comfortable settings for refresh rate under any lighting conditions are around 80- 85 Hz.
Note: Never set the refresh rate higher than what is supported by the monitor. This can
cause irreparable damage to the monitor, so be careful.
Multimedia 399
Multimedia 401
2. less
3. 800x600
4. interlacing
Multimedia 403
IV
Sound
The final component of the multimedia experience in a PC is sound. If you recall your
physics classes, you'll remember that sound is created by vibrations of a material, like
your vocal cords.
The energy from these vibrations vibrates the air molecules around it.
The air carries this vibration energy to another material, say somebody's eardrum,
which is vibrated.
The brain interprets these vibrations and processes them as sound. Sound is by its
very nature analog.
It can be visually represented by a sine wave. You remember from previous discussions
that sine wave frequencies are measured by the number of cycles per second, or Hz.
Lower pitched sounds have a low frequency and high pitch sounds have a high
frequency.
You know by now that computers communicate digitally.
So in order to interpret an analog sound, the computer must sample it for both
changes in frequency and amplitude.
This is done through a process called sampling. Let's refer to the graphic to see how this
really works.
Sample Rate
The computer samples for changes in frequency along the x-axis of the wave, at given
intervals. This is known as a sample rate. Its frequency is measured in hertz (Hz).
If a sound had a frequency of 1 Hz, it would take a minimum of two samples per
cycle, or a 2 Hz sample rate to interpret. The human voice can range from 50 Hz up to
about 4 KHz (KHz Kilohertz 1000 Hz).
This would require the ability to sample at a minimum of 8 KHz to capture the sound
of a human voice. Musical instruments can reach frequencies up to 20 KHz. That is
why standard sampling rate for CD-Audio is 44.1 KHz, just enough to sample music.
Sample Resolution
Now that we have a point along the x-axis, the computer must assign a value to
describe its position on the y-axis of the wave.
This is called sample resolution, and is measured in bits. An 8-bit sample resolution
will allow for a possible 256 different positions along the y-axis.
This really isn't enough information to reproduce high quality sound. Therefore
modern digital sound devices typically have a 24-bit sample resolution, allowing for
16,777,216 variants in amplitude.
MIDI
The 15-pin connector on the back of a standard sound card can be used as a Musical
Instrument Device Interface, or MIDI port.
This port is designed to take digital signals produced by MIDI compatible instruments,
such as a synthesizer, and record or play back the sounds through the computer.
Multimedia 405
Using the MIDI format, sounds can be created in one of two ways:
FM Synthesis
Wavetable Synthesis
FM Synthesis
Using this technique, instrument sounds are created by small hardware devices called
oscillators.
Synthesizers often utilize this technique. Because the sounds are produced entirely
through electronics, they sound noticeably artificial.
Wavetable Synthesis
This technique uses recordings of actual sounds that are stored in a ROM chip on the
sound card.
When one of the sounds is applied to the digital signals produced by the MIDI
instrument, the resulting sound is a near-perfect reproduction of real sound.
Most sound cards that support Wavetable Synthesis come with Wavetable RAM that
stores predefined, and in some cases user defined, sound samples in the memory.
In this chapter, you learned how the multimedia experience is created within a PC. We
began by discussing how the video card of a PC is responsible for converting digital
signals from the processor into analog signals that are sent to the monitor using a
special chip called a DAC. You then learned how the screen resolution and color depth
defines image quality.
We discussed the various components that make up a modern video card. You learned
how the video processor, video chipset, video BIOS, frame buffer, and RAMDAC
work together to create the images that are displayed on the monitor. You then learned
about the evolution of video card interfaces from VLB to PCI to AGP and beyond.
We discussed the second component of the multimedia experience, sound.
You learned how the DAC and ADC convert analog sounds into digital signals and vice
versa.
We discussed how the quality of digitally recorded sound is affected by the sample rate
and sample resolution of the sound card. Finally, you learned how to interface a digital
musical instrument with a PC using MIDI technology.
We discussed the two methods used in MIDI, FM synthesis and Wavetable synthesis.
There is much more to learn about multimedia and we encourage you to explore and
continue your adventures.
Multimedia 407
VI
2. True or False: Today's video cards use the CPU's massive processing power to handle
video signal conversion and calculations.
A. True
B. False
5. Pixels are made up of even small _________, ________ and _________ dots.
A. Red, green, and blue
B. Red, white and blue
C. Red, yellow and blue
D. Black, blue and red
6. What video card milestone was achieved to relieve the burden of video processing
from the processor?
A. Accelerated video cards
B. AGP
C. 32-bit True Color
D. Digital Audio Converter
Multimedia 409
7. What item has a direct impact on the performance of the video card?
A. Video RAM
B. CPU Mhz
C. Video chipset
D. Video Driver
8. What agency or organization was responsible for creating a standard for controlling
video output?
A. IEEE
B. ISO
C. W3C
D. The Video Electronics Standard Association, VESA
10. Which of the following memory types can be found on video cards?
A. Video RAM (VRAM)
B. Window RAM (WRAM)
C. Synchronous Dynamic Random Access Memory (SDRAM)
D. Multi-bank DRAM (MDRAM)
E. All of the above
11. Which of the following are valid video card architectures? [Choose all that apply]
A. VGA
B. 8-bit ISA
C. 16-bit ISA
D. MCA
12. True or false: The AGP bus can run at the same frequency as the system bus.
A. True
B. False
Multimedia 411
13. What agency was responsible for insuring that a compatible video card could be used
with a compatible monitor?
A. VESA
B. IEEE
C. CompTIA
D. OSI sub-committee
15. How should you set the refresh rate on your monitor?
A. Above 60Mhz
B. Through the start-up menu
C. 70 Mhz
D. Any rate
17. What does the acronym ADC stand for, and what is its purpose? [Choose two]
A. Analog to Digital Converter
B. Alternating Direct Current
C. Converts analog sounds to WAV files
D. Converts analog sounds to digital signals
18. In which two ways can sounds be created using the MIDI format? [Choose two]
A. AGP
B. FM Synthesis
C. Wavetable Synthesis
D. DAC
Multimedia 413
2. True or False: Todays video cards use the CPUs massive processing power to handle
video signal conversion and calculations.
A. True
*B. False
Explanation: Video cards started as simple devices with relatively little intelligence and
simply did the video signal conversion and nothing else, which left the processor to
handle all of the calculations needed to display all of the objects in a particular area
of the screen.
5. Pixels are made up of even small _________, ________ and _________ dots.
*A. Red, green, and blue
B. PRed, white and blue
C. 6Red, yellow and blue
D. Black, blue and red
Explanation: A pixel is made of red, green, and blue dots.
Multimedia 415
6. What video card milestone was achieved to relieve the burden of video processing
from the processor?
*A. Accelerated video cards
B. AGP
C. 32-bit True Color
D. Digital Audio Converter
Explanation: Video card makers created what came to be known as accelerated video
cards, which took over the responsibility of re-calculating any changes in the display.
7. What item has a direct impact on the performance of the video card?
A. Video RAM
B. CPU Mhz
*C. Video chipset
D. Video Driver
Explanation: Just as the chipset on a motherboard contains a special logic circuit called a
chipset that controls the interaction of all the various components in the computer
and the processor, accelerated video cards have a similar logic circuit called the
video chipset, or the video coprocessor, which has a direct impact on the
performance of the video card. See heading Video Cards, subheading Video Chipset.
8. What agency or organization was responsible for creating a standard for controlling
video output?
A. IEEE
B. ISO
C. W3C
*D. VESA
Explanation: The Video Electronics Standard Association, VESA, introduced a standard
for high resolution BIOS code called the VESA BIOS Extensions, or VBE, because
of the overwhelming number of different video card technologies, chipsets, and
video BIOS program, and programming software. See History tip: Video Electronics
Standards Association (VESA).
Multimedia 417
10. Which of the following memory types can be found on video cards? [Choose all that
apply]
*A. Video RAM (VRAM)
*B. Window RAM (WRAM)
C. Synchronous Dynamic Random Access Memory (SDRAM)
*D. Multi-bank DRAM (MDRAM)
E. All of the above
Explanation: Video memory for the frame buffer has many types Here are some of the
more common ones: VRAM, WRAM, SGRAM and MDRAM.
11. Which of the following are valid video card architectures? [Choose all that apply]
A. VGA
*B. P8-bit ISA
*C. 616-bit ISA
*D. MCA
None of the above
Explanation: Video cards can use any expansion bus architecture. VGA is a standard,
not an architecture.
12. True or false: The AGP bus can run at the same frequency as the system bus.
*A. True
B. False
Explanation: In theory, the AGP bus was designed to run at the same speed as the
system bus, but that doesn't always seem to be the case.
13. What agency was responsible for insuring that a compatible video card could be used
with a compatible monitor?
*A. VESA
B. IEEE
C. CompTIA
D. OSI sub-committee
Explanation: The Video Electronics Standard Association (VESA) was formed to define
a new set of standards that would be recognized by the industry because of all the
different ways available to display video images.
15. How should you set the refresh rate on your monitor?
*A. Above 60Mhz
B. Through the start-up menu
C. 70 Mhz
D. Any rate
Explanation: Because the American standard for electricity is set at 60 MHz is best to
set your monitor have a refresh rate above that to avoid eyestrain. When the full
refresh rate of your monitor is close to the electricity standard, the to attend to
synchronize, thus creating eyestrain.
Multimedia 419
17. What does the acronym ADC stand for, and what is its purpose? [Choose two]
*A. Analog to Digital Converter
B. Alternating Direct Current
C. Converts analog sounds to WAV files
*D. Converts analog sounds to digital signals
Explanation: A device such as a microphone converts analog sounds into digital signals
with an Analog to Digital Converter (ADC) for the very purpose its name implies.
18. In which two ways can sounds be created using the MIDI format? [Choose two]
A. AGP
*B. FM Synthesis
C. Wavetable Synthesis
*D. DAC
Explanation: Using the MIDI format, sounds can be created in FM synthesis or
Wavetable synthesis. The 15-pin connector on the back of a standard sound card can
be used as a MIDI port, which is designed to take digital signals produced by MIDI
compatible instruments, such as a synthesizer, and record or play back the sounds
through the computer.
Notebooks 421
Introduction
This chapter examines notebooks. While we may think we get rest from our labors, the
advent of the notebook makes us more portable and we still are working albeit in a more
comfortable home environment or while on the road.
Notebooks 423
II
Notebooks
Lucky You. This is an amazingly small chapter. Which is appropriate because we have an
amazingly small subject. That is the notebook computer.
Here you will discover the specifics of PC Card Types, and learn why some batteries
have a 'memory effect'. In addition, the pros and cons of different types of batteries are
explained.
The details of active matrix and passive matrix along with the give and take of each
choice become clear.
While the notebook computer comes in wide variety of form factors, for the most part
they share a common feature set. That feature set includes:
Running on batteries
Low power and sensitive display screens
PC Card (PCMCIA)
Cost more; get less
Limited expandability
The largest notebooks are virtually replacements to the desktop PC. All components are
built-in to the unit; they include everything you would expect to find in a desktop, well,
almost everything.
While they don't usually offer multiple monitors, everything else is there. This includes
network connection, CD-ROM and a display as big as the viewable display on a
standard CRT.
Notebooks 425
The new USB hubs are small and portable making this a feasible travel or desktop
option. Often the USB capability is referred to as hot plugging or hot swapping.
Sub-Notebooks are the smallest units that still have some sort of keyboard large enough
to be able to use your fingers (as opposed to a stylus or pencil) to type. The hard drive is
internal, and a cable or other external connector (USB port) is used to connect any other
type of drive. Typically, this would be any USB device for portability. The network
card and modem are generally built in.
PDA
Personal Digital Assistant (PDA) computers are not PCs, which is probably why they're
called PDA and not PCs. However they are computers.
They don't have any sort of built in keyboard, and anything pretending to be a keyboard
on the display is pecked at with a stylus. There are portable folding keyboards available
to add on to the PDA.
Generally, PDAs are to-do list machines with contact data for your life. It is possible to
get email on them, via a wireless modem, however these modems (typically) are much
slower than desktop connections. RAM and other upgrades are available for these devices
just as with any other computer.
Docking
Some notebooks and even some PDAs use Port Replication or Docking Stations.
These docking stations sit at the desktop, waiting for the return of the notebook. This
makes it simpler to hook up an external monitor, printer, mouse, etc. once at 'home
port'.
The docking station is usually a long rectangular device with all the ports necessary to
add a full size keyboard, monitor, or mouse. The notebook or laptop computer simply
plugs (docks) into the station, and then un-plugs and goes with you when you travel.
This is very convenient for the person who wishes to use one computer and leave all the
peripherals behind.
As laptops or notebooks have become full featured and powerful in terms of CPU speed,
RAM and hard drive size this is a practical option for many.
Charged Up
One critical issue with the notebook is the battery. Once the battery goes dead, you
might as well be hauling around a brick.
The first batteries for laptops (I wont call them notebooks) were the lead-acid battery.
This was replaced by the Nickel-Cadmium (NiCd) battery. While weighing less than the
lead-acid, they have less total power between charges than lead-acid.
Notebooks 427
NiMH
Nickel-metal hydride (NiMH) generally replaced the NiCd, since they have more energy
storage, and are less likely to have the memory effect.
Li
Lithium-ion (Li) is the latest in battery technology. They recharge quickly, have much
lower weight (being the least dense of all the metals in the periodic table), without the
issue of memory effect.
The next part of the notebook to consider is how does it look (when it has power). Let's
take a look at the elements of notebook display.
Displays
Oh say, can you see...
The Liquid Crystal Display (LCD) can be the blessing or bane of any notebook.
Generally, when we say LCD we are referring to a flat panel display screen.
It should be noted that the basic LCD display is monochrome, (really grayscale) and
utilizes backlighting to enhance the image.
A simple example of this type of display would be found in digital watches. The watch is
displaying time and date in a dark gray over a light gray background with the lighting to
make it appear black (grayscale) and thus more visible.
This type of display is also low on power consumption. The backlighting is helpful but
can be difficult to see under some surrounding light conditions.
The LCD display is created when an electrically reactive material is placed between
electrodes. When current is applied to and removed from the electrodes, the display will
become darker or lighter as needed.
If the liquid crystals are close together then you have something approximating a pixel.
Have you ever noticed a blue spot on a LCD screen? This is likely a failure of a crystal.
Similarly, when the display becomes distorted on a watch or even a cell phone (maybe
you dropped it?) then the crystals are dead.
Active-Matrix
The Active-Matrix LCD is generally the most preferred. Each pixel (or cell) is
controlled by a transistor, which is used to control the display.
Powering each transistor takes power. This display type uses a grid of wires with diodes
over it. When current hits the diode at points on the grid, you see a display, again much
as pixels are found on a traditional monitor.
This display type is expensive, because to achieve the wonderful color and clarity of
image; transistors must be used to turn the diodes on or off. This is a display type
considered easy on the eyes as the refresh rate is controlled actively.
The diodes are actually liquid crystal. Have you ever touched an active matrix flat
panel display screen? No?
Notebooks 429
Well if you do the screen appears to float like liquid at the point of contact, what you
are really doing is disturbing the current to the crystal.
One or more transistors, allowing refined control over the resulting display, control the
current that is applied to the crystal diodes.
This presents the sharper, brighter and clearer colors associated with this display type
and the resulting additional expense.
Watch for small spots on the screen where there is either a defective crystal (diode,) or if
the whole screen is black then it is likely the transistors are gone.
TFT
Thin Film Transistor (TFT) is a specialized and desired form of display.
With TFT three discrete transistors, for the colors, red, blue and green, control each cell
or pixel.
The thin film is often a magnetic material on a base material. The term TFT is typically
referring to active matrix displays, as there is a controlled refresh rate (current) to the
transistors.
Passive-Matrix
Passive-Matrix LCD uses one transistor to control an entire row of pixels. Like the
active matrix the diodes are liquid crystal over a very fine grid of wires.
When the current is applied to the diodes the crystal lights and you get a display again
much like a pixel.
The display is constant because the current applied to the diode is at a persistent
(passive) or specific rate. This rate is known as the refresh rate.
The lack of individual transistor controls to the diodes makes this a less expensive display
option.
There is also dual scan passive matrix display option where the screen is divided into
half and scanned at the same time with current making the screen appear brighter with
higher resolution.
Notebooks 431
Figure 89 A Type III Hard Drive PC Card compared to a standard IDE Hard Drive
Uses
RAM
NIC, Modem
HDD
Options
1 Type I
1 Type II
1 Type II & 1 Type I
Hot Swapping
It is possible to exchange PC Cards 'on the fly'. This means no rebooting.
Another term for this is Hot Swapping. The Operating System needs to
support this operation.
Notebooks 433
IrDA
Before closing this chapter, a brief look at the possibilities presented by the Infrared
Data Association (IrDA). This is a standard for transmitting data using infrared light,
much like a remote control for home electronics.
If a notebook or PDA has an IrDA port, it can be used to transmit contact information
between two PDAs or if a printer has an IrDA port, the ability to print almost
automagically is given. Refer back to Chapter 0111 for a refresher on infrared.
III
You discovered that notebooks come with different capabilities, and the price to be paid
beyond the financial outlay is the weight and size. It became clear that each portable
device needs power and choices in battery technology impact usability.
The tradeoffs of getting a brighter and more colorful screen with an Active Matrix
display are paid with more power consumption because each pixel needs a transistor.
It became clear that the latest version of Active Matrix technology is the Thin Film
Transistor. You know that the Passive Matrix screen only has one transistor per row of
pixels, which saves power at the expense of display quality.
You learned that most PDAs have an IrDA port for communications with another PDA
or printers that have an IrDA port.
It was revealed to you that the friendliest Operating System from Microsoft for notebooks
is Windows 2000.
You discovered the different types of PC Cards and what each type is typically used for.
Notebooks 435
IV
2. PC Cards are available in what data paths. (Choose All That Apply)
A. 8-bit
B. 16-bit
C. 32-bit
D. 64-bit
6. True or False. It is possible to have two physical PC Cards if they do not exceed the
specifications of the notebook.
A. True
B. False
Notebooks 437
9. PC Cards are available in what data paths. (Choose All That Apply)
A. 8-bit
B. 16-bit
C. 32-bit
D. 64-bit
Notebooks 439
13. True or False. It is possible to have two physical PC Cards if they do not exceed the
specifications of the notebook.
A. True
B. False
2. PC Cards are available in what data paths. (Choose All That Apply)
*A. 8-bit
B. 16-bit
*C. 32-bit
D. 64-bit
Explanation: PC Cards are available in 16-bit and 32-bit varieties.
Notebooks 441
6. True or False. It is possible to have two physical PC Cards if they do not exceed the
specifications of the notebook.
*A. True
B. False
Explanation: A notebook with a Type III compatible port can accept 1 Type I and 1
Type II card at the same time.
9. PC Cards are available in what data paths. (Choose All That Apply)
A. 8-bit
B. 16-bit
C. 32-bit
D. 64-bit
Notebooks 443
13. True or False. It is possible to have two physical PC Cards if they do not exceed the
specifications of the notebook.
A. True
B. False
No matter how far you have gone on a wrong road, turn back.
_Turkish Proverb
Chapter 1010:
Safety and the Environment
The objective of this chapter is to provide the reader with an understanding
of the following:
This domain requires the knowledge of safety and preventative maintenance.
With regard to safety, it includes the potential hazards to personnel and
equipment and working with lasers, high-voltage equipment, ESD, and items that
require special disposal procedures to comply with environmental guidelines.
With regard to preventive maintenance, this includes knowledge of preventive
maintenance products, procedures, environmental hazards, and precautions
when working on microcomputer systems.
3.1 Identify the purpose of various types of preventive maintenance products and
procedures and when to use them.
Liquid cleaning compounds
Types of materials to clean contacts and connections
Non-static vacuums (chassis, Power supplies, fans)
3.2 Identify issues, procedures and devices for protection within the computing
environment, including people, hardware and the surrounding workplace.
UPS (uninterruptible power supply) and suppressors
Determining the signs of power issues
Proper methods of storage of components for future use
Potential hazards of public safety procedures relating lasers
High-voltage equipment
Power supply
CRT
Introduction
This chapter examines how to control the 3 Demons, EMI, RFI and static. You will
discover that dust is a major contributor to overheating and how to safely clean PC
components. It will be revealed to you why some parts require special disposal handling.
Government regulations regarding both interference and pollution become clear. And you
will learn how to safely control a fire. Finally, details on how surge suppressors are rated
are unveiled.
II
While this chapter may be brief in pages, you are advised to pay close attention.
Failing to heed material in this chapter may lead to:
Substantial fines or penalties
You and/or equipment winning a Darwin award.
The old saying is, "Forewarned is forearmed." While passing the A+ test is great, we
offer you information for the more important tests of life as well.
3 Demons
Without power, a PC is nothing more than a box of metal, plastic, and sand. Lurking
about waiting to destroy your data or your equipment is what we call the three demons.
They are:
Electromagnetic Interference (EMI)
Radio Frequency Interference (RFI)
Electrostatic Discharge (ESD)
Lets examine each of these demons to learn how to avoid them.
EMI
Electromagnetic Interference can be an elusive creature because it cannot be seen with
the eyes. Almost every device that uses electricity is capable of producing EMI.
Electrical motors, TVs and monitors are common examples of devices that both produce
EMI and are sensitive to EMI issues.
PCs usually do not cause EMI because the manufacturer must follow government
guidelines regarding the generation of EMI. In the United States, the agency responsible
is known as the Federal Communication Commission (FCC).
Equipment used in Canada is the province of the Canadian Radio-Television and
Telecommunications Commission (CRTC)26.
26
(5) The a similar of a personal computer system may be required to test the system and/or
make necessary modifications a system is known to cause harmful interference how or
two be non-compliant with the appropriate standards the configuration in which is
marketed (see Sections 2.909, 15.1, 15.27 (d) and 15.101 (e) of this Chapter).
We have included this section as direct quote from the FCC, should anyone question the
legality of building a PC yourself. It used to be a big issue.
Note that when a PC has been certified to the limitation of EMI production, it is done
with the case cover on. This both protects the PC from EMI as well as preventing EMI
from the PC interfering with other devices. Tracking down EMI issues can be incredibly
challenging. Your authors have found a very simple and inexpensive device to assist in
tracking down EMI challenges. That device is a compass.
Avoiding EMI
Use dedicated circuits. Avoid plugging in a PC to a circuit that is
powering devices with motors, such as refrigerators or air conditioners.
Physically isolate the PC from radios, televisions, appliances, fluorescent
lights, and stereo speakers that have not been shielded.
Use an Uninterruptible Power Supply (UPS), line conditioner, or at the very least a
surge suppressor.
Refer back to the diagram on how a power supply works in Chapter 0000. Notice that the
capacitor filters out the last bits of a wave after the bridge rectifier.
Now lets think about a power surge. The transformer will simply step down the excess
voltage by a proportional amount.
The rectifier will convert this to too much voltage with all positive waves.
From there, the capacitor will become saturated, passing on an excessive amount of
voltage in positive pulses, which are fed to all the parts of your PC. In other words, zap,
your PC is dead.
Surge Protection
A surge protector can remove excessive voltages as well as RFI. Without
a power source, a surge protector cannot protect against a power drop or
under voltage condition known as power sag.
Stay Grounded
Always plug computer equipment into a three prong grounded outlet.
UL 1449
When looking at surge suppressors, be sure it carries the Underwriters
Laboratories (UL) rating 1449.ULtests for 1449 consists of Grades A, B
and C, Classes 1, 2 and 3, and Modes 1 and 2. Within this classification,
the best rating is A-1-1.
Dedicated Circuits
Looking for an AC outlet that is orange color can Identify dedicated
circuits for AC power. The triangle on an AC outlet indicates dedicated
ground.
UPS
The UPS typically combines surge suppression with batteries. The surge suppressor
portion protects against over voltage conditions along with EMI and RFI.
Under voltage conditions or no power, then the power is either boosted or replaced from
the batteries. Unfortunately, batteries do not last forever. Most UPS system are designed
to keep a PC running, and do not have the ability to keep a device with heavy power
demands running.
Laser Warning
Do not plug a laser printer into an UPS.
RFI
Radio Frequency Interference (RFI) is our second demon, and like its evil brother,
EMI, can be a real challenge to track down.
On occasion, for example a PC near a radio or television station transmitter can be
overwhelmed by the sheer amount of energy.
The only solution in this type of situation is to make a Faraday shield. This is a metal
shield that is attached to ground.
ESD
The nature of electricity was revealed in Chapter 00000. In the PC, the non-motor
components operate at 5 Volts or less.
The smallest amount of static you can feel (as you touch a grounded object) has about
30,000 Volts. An electronic component such as an IC doesn't stand a chance of living if it
is subjected to a charge of that dimension. Static does not have to be at the level you can
feel to kill an electronic component. Any time static electricity is released it is called
Electro Static Discharge (ESD).
Static Alert
The probability of creating static is highest with cold temperatures and
low humidity.
Because electronic components are sensitive to ESD, they arrive in anti-static bags.
Never remove an electronic part from this bag until you are wearing an anti-static strap
that is attached to ground. The work area should have an anti-static mat that is attached
to ground.
Grounding Strap
Anti-static strap should be worn and attached to ground before removing
any components.
Contamination
Even when you're wearing an anti-static strap, handle all components by the edges. Not
only does this diminish the possibility of damaging component with ESD, it keeps body
oils off of metal contact points.
Erasing Contamination
If you think you have (or can actually see contamination) on an expansion
card or SEC based CPU, wiping gently with a pencil eraser will clean the
surface.
Before closing on the topic of contamination, we have one more consideration. Dust does
a very poor job of transferring heat.
Contact Care
Think like Mr. Clean Dont touch contacts. Dust is your enemy for
many reasons from data loss to improper cooling.
After ESD, heat is the next villain a PC must contend with. According to
www.engineeringtalk.com, "Excessive temperature causes more failures than vibration,
humidity and dust in industrial computer systems and accounts for 55% of all electronic
equipment failure."
Given this data, vacuuming out a PC from time to time might seem like a good idea. Dust
is such a pervasive enemy many certification exams test for knowledge on topics ranging
from clean tape heads HVAC and server filters.
Keep it clean
Use a computer vacuum (these have long power cords) or compressed air
on a regular basis to remove dust from inside the PC case. Try using a
small brush to dust.
Of course dust isn't fussy, and will collect everywhere. Keyboards should get a regular
dose of 'canned air' or gone over with a vacuum.
Dust Bunnies
Anti-static spray can be used to reduce dust attraction
Alcohol
If the challenge is more grime and dirt then dust a small amount of denatured alcohol and
a lint free cloth should be used for keyboards, mice, etc.
Safety
This chapter opened with tips on avoiding a Darwin award.
While it is true that everyone is allowed one fatal mistake, your authors wish to help you
avoid this type of mistake.
To that end, we offer the following Success Tip that will not only improve your chances
of passing the A+ test, but will also improve your chances of living to tell someone you
passed.
High Voltage
Both monitors and laser printers require high voltages (backed with watts)
to operate correctly. Do NOT even consider touching any area within
these components marked High Voltage, or having a symbol inside a
triangle.
Chemicals
Generally speaking the computer industry has a good reputation for being a 'clean'
business. Sorry to say, parts of it are not, and can be hazardous to you and the
environment. Consider the simple battery.
Even a simple battery is known as a lead-acid battery. The acid part could hurt you now,
the lead, if it gets into the water supply will get you later. The metal nickel is what makes
a nickel metal hydride battery. Or consider today's lithium ion batteries. Did you know
lithium violently explodes if it comes into contact with water?
Even a dead monitor can make other things, well, dead.
Transformers may contain Polychlorinated Biphenyls (PCB). The US Congress in 1976
enacted the Toxic Substance Control Act (TSCA) to deal with PCB management.
Even if a monitor doesn't have PCB in the transformer, every Cathode Ray Tube
contains phosphorus, and "the radiated tube and phosphor are too toxic for landfills.
quoting Will Mitchell at the University of Michigan Business School.
Toxic Waste
Monitors and batteries contain toxic chemicals. Proper disposal requires
taking them to a hazardous waste site.
Chemical issues and HAZMAT sites are not the only safety issues concerning the PC
environment. Have you ever seen a monitor or even a laptop catch fire?
Fire Safety
Yes, malfunctioning batteries have set notebook computers into what would appear to be
spontaneous combustion. A transformer going seriously awry in a monitor can ignite the
plastic case. What do you need?
Why a fire extinguisher.
However, not any fire extinguisher will do. Some fire extinguishers are nothing more
than pressurized water tanks. Since computer equipment has electricity, using water on a
flaming PC live only exasperate the challenge.
Class C
In North America, Class C fire extinguishers are required for electrical
fires.
III
Fires are not the only hot issue on the topic of safety. Several types of printers can burn
you as well. When reviewing the chapter on printers, pay close attention to how laser
printers work. In particular the part about fusing the toner to the paper. Notice this
process happens with heat and pressure. The fuser roller will give you a scorch you won't
forget for some time. Ditto with the print head on a dot matrix printer.
It is good to keep in mind the three demons of computers: EMI, RFI, and static. It
wouldn't hurt either, to bear in mind that opening a computer monitor and sticking your
hand in there can cause death.
That having a high quality surge protector or UPS installed can be the difference between
you being a hero or a zero, due to data loss and hardware damage.
IV
1. Which governing body declared that, if certain conditions are met, assemblers of PC
systems employing modular CPU boards and/or power supplies are not required to
test the resulting system?
A. FCC
B. EPA
C. IEEE
D. IRS
4. Which device is commonly used to eliminate the presence of static while working on a
PC?
A. Anti-static pad
B. Anti-static wrist strap
C. Anti-static pad and strap
D. Anti-static bag
5. What device is the most effective in eliminating the potential of power loss or power
sags?
A. EMI box
B. UPS
C. DVM
D. UL 1149 approved suppressor
6. Which governing body is responsible for the proper management and disposal of toxic
chemicals such as lithium acid and lead?
A. TSCA
B. EPA
C. Congress
D. UL
7. What type of fire suppression device is required for extinguishing electrical fires in
North America?
A. Class D fire extinguisher
B. A garden hose
C. A wet blanket
D. Class C fire extinguisher
4. Which device is commonly used to eliminate the presence static while working on a
PC?
A. Anti-static pad
B. Anti-static wrist strap
*C. Anti-static pad and strap
D. Anti-static bag
Explanation: Using an anti-static pad and wrist strap combination is the most effective
way of eliminating potential static. See Heading 3 Demons, and sub-heading Static
for details.
5. What device is the most effective in eliminating the potential of power loss or power
sags?
A. EMI box
*B. UPS
C. DVM
D. UL 1149 approved suppressor
Explanation: An Uninterruptible Power Supply (UPS) uses a battery that can augment
power sags and total losses of power
6. Which governing body is responsible for the proper management and disposal of toxic
chemicals such as lithium acid and lead?
A. TSCA
*B. EPA
C. Congress
D. UL
Explanation: The U.S. Congress enacted the Toxic Substances Control Act in 1976 to
deal with PCB and management.
7. What type of fire suppression device is required for extinguishing electrical fires in
North America?
A. Class D fire extinguisher
B. A garden hose
C. A wet blanket
*D. Class C fire extinguisher
Explanation: Class C fires are classified as electrical fires in North America.
Networking 469
6.1 Identify basic networking concepts, including how network works and the
ramifications of repairs on the network.
Installing and configuring network cards
Network access
Full-duplex, half-duplex
Cabling-twisted pair, coaxial, fiber-optic, RS-232
Ways to network a PC
Physical network topologies
Increasing bandwidth
Loss of data
Network slowdown
Infrared
Hardware protocols
Networking 471
Introduction
The certification test for A+ for 2001 has doubled the number of Network questions.
While some think that isn't enough, CompTIA had to leave something for the Network+
test
Networking 473
II
Networking
This chapter begins networking with the PC itself. For a PC to talk to a network, it either
needs Dial-Up Networking (if you are using Windows) and a modem, or a Network
Interface Card (NIC). Dial-Up Networking is covered in the Operating System test (and
our book for that part of A+) so here the focus begins with the NIC, and the PC.
Plug n Play
As with virtually every other expansion card in a PC, the NIC needs an interrupt and a
memory I/O address. Back in the ISA days, this had to be done manually. No fun at all.
And it went downhill from there, typically having to add monolithic drivers, such as
NDIS (Network Driver Interface Specification). Fortunately for us, those days are over.
Today, dropping a NIC in really is Plug and Play.
Since the PCI slot only talks to PCI cards, and a PCI NIC has a BIOS telling the PC and
the Operating System who it is, the biggest challenge is making sure that you have the
correct driver for the NIC. If you are using a popular NIC, when the BIOS identifies the
NIC, chances are the driver will already be waiting in the source files on the Operating
System.
If not, be sure to have the CD or diskette ready from the NIC manufacturer. It doesn't
really help to have it on a server that you cannot network to, without a network card
running.
Ethernet
While there are several different topologies available to networking, by far the most
popular today is Ethernet. Ethernet comes in two basic types, the one simply called
Ethernet which operates at 10Mbps (Megabits per second) and Fast Ethernet which
operates at 100 Mbs.
Duplex/Simplex
With the card and driver installed, the next question that may need to be addressed goes
something like, "is the card operating at full duplex?" The real question to be answered
here is what the heck is duplex versus simplex.
Chances are you have communicated by voice using both. When you make a phone call,
and you and the other person you are chatting with can both talk and listen at the same
time it is known as Duplex.
If you have ever used (or at least heard) a conversation over radio, where the person
transmitting signifies they've completed their thought by saying "over", then the other
person responds, that is Simplex or half duplex.
When a NIC can use duplex, email for example can be both coming in and going out at
the same time. With simplex, the NIC must switch back and forth between sending and
receiving. It may do this so quickly that it appears simultaneous, however it is not.
Cables, etc.
So we have a bunch of PCs with NICs that a ready to go. How do we connect them
together so that they can share resources, like files and printers? The answer for most
networks is by cables. As with everything else in the PC world, there are a myriad of
various cable types that can be used to wire a network. For the purposes of this book and
the A+ exam, we will focus on cabling used in Ethernet networks. We will begin our
discussion on network cabling with a history lesson about the early days of Ethernet
cabling.
Networking 475
Thinnet
The first option to appear that didn't involve wrestling and building upper body strength
faster than a Bow flex exercise machine was a thinner version of the frozen garden hose,
called Thinnet.
Still a copper core and with a plastic cladding, then the metal braided sheath to control
interference, coated with a plastic wrapper. Only this entry was as thin as the cable for
TV. They look identical, and the only difference is the Ohm ratings.
TV is 75 Ohm, while thinnet is 50 Ohm. Thinnet is called RG-58 U and it even comes
in a version known as RG-58 AU, which has a braided copper center instead of the solid
copper center found in RG-58 U.
Networking 477
STP
One idea that looked better than the frozen garden hose was Shielded Twisted Pair
(STP). This looks like telephone wire, only with a braided metal shield to keep the
interference demons at bay. IBM used this with Token Ring, calling it Type 1 cable.
UTP
So, what is cheaper to make and faster to work with then STP? Unshielded Twisted Pair
(UTP).
The goal has always been to make everything faster and cheaper, and cable is no
exception. As UTP cable got more capable, it needed ratings to Identify which UTP cable
could do what.
The system for doing that is known as Category X. Theis a variable, which indicated the
capacity of the cable. For example the wiring for the telephone system is called Category
1 (sometimes referred to as POTS for Plain Old Telephone Stuff).
After a time, it became clear that Ethernet was gaining ground over the alternatives, and
out came Category 3 cable, which supported Ethernet at 10Mpbs over UTP! This
advance secured Ethernet as the defacto standard for the office Local Area Network
(LAN). It didn't matter that the distance limitation was only 100 meters (333 ft). This
was enough for most any office building.
As nature hates a void, network traffic rapidly filled the LAN. The next step was
Category 5 cable, capable of 100 Mbps, and Ethernet morphed into Fast Ethernet.
100BaseT
100BaseT (Fast Ethernet) is only supported using UTP Category 5 in a
Star Topology, and has the 100-meter limit of Ethernet over UTP. The
limitation includes the cable from the wall to the PC.
UTP
Uses the RJ-45 connector
Gigabit Ethernet
Nature still hates a void. The next round is Gigabit Ethernet (1000
Mbps). Gigabit Ethernet uses 4 pairs of wires, not two pair as in Ethernet
and Fast Ethernet. If you must install wiring today, carefully follow all the
rules for supporting, bending and terminating the cable and install all 4 pairs. We cannot
promise your Cat 5 will work as Gigabit, and if you use good quality components,
installed very carefully, you may well get Gigabit out of your Cat 5 wiring.
Networking 479
CABLE TABLE
Number of workstations/Transmission Speed
Cable Type
Speed
Length
Nodes
# of Pairs
Category
10BaseT
10Mbit
100 meters
1024
100BaseT
100Mbit
100 meters
1024
1000BaseT
1024
Length
Taps/Nodes
5-4-3 Rule
10Base5
Thicknet
500 meters
100
Yes
10Base2
Thinnet
185 meters
30
Yes
Cable Type
Maximum Length
Data Rate
10Base5
Coax
10 Mb/s
10Base2
Coax
10 Mb/s
10BaseT
UTP
10 Mb/s
10BaseF
Fiber
4 Km
10 Mb/s
100BaseT
UTP
100 Mb/s
100Base
UTP/STP
VGAnyLan
100BaseT4
100 Mb/s
UTP
100BaseTX
2 pair Cat 5 UTP or100 m between the hub and100 Mb/s Fast Ethernet
Cat 1 STP
network node
100BaseFX
Fiber
2 Km
Category 3
UTP
Token
Ring
and
VGAnyLan at 16 MHz
Category 5
UTP
100 MHz
Wireless LAN
Another type of wireless is for the LAN. This technology has been sputtering along for
almost 10 years.
With advances and a real vendor neutral standard, (802.11x) 11 to 54Mpbs network
cards are available as PCI and PC Card and are now in retail stores.
Wireless networking for virtually everybody may well become a reality by the time the
A+ test sees its next refresh.
Networking 481
Old Reliable
Using the serial port between two PCs is still a network option. Direct cable networking
is a part of Windows 9.x.
While limited to 115Mpbs, yes point 115 or about a tenth of a Megabit, means that serial
networking is not for the inpatient.
With the speeds covered, it is time to review the feeds to fill out the slang term, 'speeds
and feeds'. That is covered next.
Bus
Did you review the Geek Tip about how to remove the BNC connector from a PC listed
above?
The mention of breaking the backbone is the key to a Bus topology network.
Just as with any human, messing with a backbone is a sure recipe to bad things.
In a Bus network, each computer is in line to another.
The ends of the network get termination for the same reason that was revealed in the
SCSI chapter.
Ring
An alternative to the bus topology is the ring topology.
Don't confuse a physical ring with a logical ring. IBM's Token Ring is an example of a
network system that operates as a logical ring, but doesn't look like a ring.
This is due to the fact that all of the PCs in the ring plug into a central device called a
Media Access Unit (MAU).
MAUs are similar looking to hubs (discussed later).
The difference is that each computer in the ring takes turns transmitting or receiving
data.
This is accomplished using a token. The token is passed from one computer to the next.
The PC with the token can transmit or receive data only when it has the token.
Networking 483
Mesh
Suppose your network has to behave almost as the motto of Federal Express? It
absolutely, positively, has to be up, all the time.
Then you need a mesh topology. As the picture shows, if one line breaks, another one is
available. The down side to a true mesh is the amount of cable and NICs to each
machine.
The slightly simpler alternative is to create a hybrid mesh. For example, routing
network data through a second machine as a way to get from PC 1 to PC 3.
The Internet is the largest hybrid mesh network in the world.
Star
Today, for a variety of reasons (all coming down to some sort of cost) the star topology
is where most network installations in an office occur.
All the wiring comes to a central point, not unlike the spokes of a wagon wheel.
When data needs to get from one spoke to another, it heads off to the center and back out.
The good news about a star topology is if a PC goes down, it doesn't affect the other
computers, like the Bus does.
The bad news is there is a single point of failure at that center point that can kill the
ability of every PC to send data on the network.
Protocols
When someone calls you on a telephone, a common response would sound something
like 'Hello?'
That is how you inform the caller that you heard the request to talk, and you are ready for
them to say something. In other words, that is protocol. Networking has all sorts of
protocols. The ones we are concerned about at this point are transport protocols. This
means some standard way of how data moves.
If I called you and the response was 'hola,' I might repeat, "hello," you repeat 'hola,' and
within a few moments we would most likely both give up, and hang up the phones. This
is because while I was following protocol in English, you were following protocol in
Spanish. We are both attempting to talk, but we had a failure to agree on a protocol (of
language).
The same thing can happen in network Transport Protocols. We have to agree on a
common way of communicating, or no communication happens. There are several
transport protocols that are in use in networking. Lets look at a few popular network
transport protocols next.
IPX/SPX
Many games played in a multi-user mode (multiple game players playing each other) use
Internetwork Packet eXchange (IPX).
Networking 485
The other half of the title is Sequence Packet eXchange (SPX). Most folks just call it
IPX, but in the interest of being complete and accurate, we show both.
IPX is a pretty nice protocol, and was born from the early PARC days, which used Xerox
Network System (XNS). A Network Operating System (NOS) company by the name
of Novell hawked IPX as its choice for network transport protocol for years with its
product Netware. When Version 5 of Netware was released, they left behind IPX for
another choice, which may sound familiar, and is seen next.
TCP/IP
Transmission Control Protocol/Internet Protocol (TCP/IP) has almost become a
household word with The Internet becoming part of daily life.
TCP/IP works with numbers, not names. Since numbers are more difficult to remember,
we have tools in TCP/IP to make it easier for humans to work with, but when the rubber
meets the road, it's about numbers.
To make the numbers somewhat simpler for humans to work with, they are broken down
into four groups of numbers called Octets. If you have ever setup an Internet
connection, you may have had to enter these numbers. They look something like
192.168.10.2. This is known as the IP Address. Every computer connected to a TCP/IP
network must have a unique IP Address. If two computers have the same address, one or
both will be kicked off the network. A computer on a TCP/IP network is referred to as
host.
The four octets have to be some combination of 32 zeros and ones. This string of binary
numbers actually represents two things.
The first is the 'name' of the network; the second is the 'name' of the PC or other device
on the same network. How the division of the 32 zeros and ones happens is the role of
another set of four octets, known as the SubNet Mask.
The Gateway
You will see a third option, again using octets. When network traffic
happens and it isn't for your network, the data has to leave your network.
That third octet entry is known as the Gateway. The job of the gateway is
to be a place marker for the device that acts as a doorway for data that must enter and
leave your network. Once again, it is a binary thing. The data is or isn't for your network.
If the data isn't for your network, it throws it to the gateway. It sort of works like "I don't
know what to do with this data traffic, Gateway, you handle it and let me know when you
got something for me."
Networking 487
One more thought before moving on to network hardware. Remember we said that
everything is number based? Yet, you probably know that you enter human talk, like
TotalRecallPress.com, to see what is going on there. The Internet has a big database in
the sky that makes the conversion from names to numbers. This conversion is done by
Domain Name Servers (DNS). DNS servers have static IP addresses so everything
knows where to get the conversion from so the binary devices can talk.
We could get into DNS with more depth, but if we did, you wouldn't have a reason to buy
our stuff that covers Network+ ;-) And we need to save some trees for this book to tell
you more about network hardware, the next topic.
NIC
Be it notebook or desktop PC, nothing talks without a Network Interface Card (NIC).
As with virtually every expansion card in a PC, the NIC needs an I/O address and IRQ to
interface. While today, Plug and Play makes that assignment automagic; one area that
may need your manual attention is the driver. Updating drivers when upgrading an
Operating System with a Service Pak is common.
As you may have figured out by now it is the actual NIC that gets the IP address that we
discussed earlier in this chapter. However, the IP address gets a close association with an
identity that is ingrained into the cards creation as part of creating a network.
MAC Address
Every network card ever created has a unique number assigned to it (theoretically at
least) called a Media Access Control (MAC) address. We say that this theoretical because
some overseas manufacturers have been known to use false MAC addresses on their NIC
cards. Like IP addresses, no two computers on a network can have the same MAC
address.
The MAC address is a hexadecimal number, separated into two halfs. The first half is a
number assigned to the maker of the card, and the second half is the unique number the
maker assigns that particular NIC. The theory goes that no two cards will have the same
MAC address.
MAC address
The MAC address is a permanent part of the card, and is part of its ROM.
An IP number is bound to the MAC address. To see the MAC address that is tied to an
IP address, ask the computer by issuing a request using the command Address
Resolution Protocol or ARP.
If your computer says no entries, create some network activity, and type ARP again. The
MAC address will be revealed.
The MAC address can be used by other hardware, such as a bridge or a switch, which
will be discussed shortly.
Switch
That better way is known as a switch. A switch looks like a hub, but it is a bit smarter. It
takes all the connections in, but doesn't actually connect them all together. Instead, it
waits until it hears something from one device intended for another, and then makes the
connection. This means that they have a private conversation. As soon as the
conversation is done, it connects another two devices. And this process repeats, over and
over. This makes for faster network connections and actually allows for more network
traffic than with a hub.
Networking 489
When a switch is first turned on, it doesn't know whos who in the digital zoo. So the first
time a packet of traffic comes down the pike it sends it to all the devices. When one
responds, it makes a note in RAM of the MAC address for the device that responded and
what port the response came from. Using this data, it learns which device is where, so the
next time it just builds the connection.
This works great in LANs but they are not big enough or bright enough to run the entire
Internet.
Crossover Cable
The purpose of a crossover cable is to connect two computers together
with out a hub or a switch.
Router
The router is much like a switch, but bigger and smarter. A router works with the IP
number of the devices. And since TCP or UDP breaks up network traffic into little bite
sized clumps called Packets, it can round robin to different incoming and outgoing lanes
(wires) very quickly. Without routers, the Internet wouldn't exist today.
Routers
A router provides Physical, electrical and electronic connection to a
network
Networking 491
III
In this chapter, you learned how two or more computers could make a network, or
LAN. We began this with a discussion of Network Interface Cards, NICs. You learned
how each NIC is given a unique number called a MAC address.
Next, you were shown the various types of cable that are used to connect PC NIC cards
into a network. You learned about the most common type of network, an Ethernet
network, and the types of cables it uses.
You learned about how computers are arranged in networks using topologies. You
learned to Identify bus, star, ring, and mesh network topologies.
Next, you were shown how computers talk to each other over their network cables using
protocols. You learned about IPX/SPX and TCP/IP protocols. You were then shown
how TCP/IP is actually of suite of protocols that is made up of transmission protocols,
networking protocols, and utilities.
Dissecting TCP/IP further, you learned how each computer on a TCP/IP network is
given a unique address called an IP address. You were also shown that each computer
must be assigned a subnet mask and a gateway address to complete its configuration.
We concluded our discussion on TCP/IP with the Dynamic Host Configuration
Protocol (DHCP). You learned how this application is used to lease configuration data
to computers that wish to join a TCP/IP network.
Finally, you learned about the various pieces of hardware that can be found in a
network.
First, you learned about hubs and their intelligent cousin, the switch. You then learned
about even more intelligent devices known as routers and bridges.
IV
2. What two flavors does Ethernet come in, and what are the respective data speeds?
[Choose two]
A. Real Ethernet, 10Mbps
B. Fast Ethernet, 100Mbps
C. Ethernet, 10Mbps
D. Ethernet, 100Mbps
3. Duplex data transmission is the most efficient method for sending and receiving
information.
A. True
B. False
Networking 493
4. What is the maximum distance data can be transmitted over Thick Ethernet network
cabling without signal attenuation?
A. 500 feet
B. 5000 feet
C. 500 meters
D. 5000 meters
5. What is the maximum distance data can be transmitted over Thinnet cabling without
signal attenuation?
A. 185 feet
B. 185 meters
C. 500 feet
D. 500 meters
Networking 495
10. It is possible to transmit data between two or more PCs without being physically
connected.
A. True
B. False
11. What is the approximate maximum data speed for two PCs using serial port
connections on a network?
A. 100Mbps
B. 100Kbps
C. 115Mbps
D. 115Mbps
13. Which of the following network topologies offers the most reliability in terms of
redundancy?
A. Star
B. Mesh
C. Bus
D. Token Ring
14. Which of the following network protocols, or suite of protocols, is the most
commonly used on The Internet?
A. TCP/IP
B. IPX/SPX
C. Banyan Vines
D. UDP/STP
15. What is it that allows you to break network IDs into multiple networks
A. TCP/IP
B. Octet
C. FQDN Identifier
D. Subnet Mask
Networking 497
16. What is it that allows data to exit from your LAN, or PC, onto The Internet?
A. Default GUI
B. Default gateway
C. LAN segments
D. Hub
17. If you were the owner of a thriving ISP (Internet Service Provider) and wanted to
make the most efficient usage of your IP addresses, what tool or service would you
use?
A. DNS
B. WINS
C. UDP
D. DHCP
18. What type of server or protocol is responsible for matching domain names to IP
addresses?
A. UDP
B. WINS
C. DNS
D. ISP
Networking 499
2. What two flavors does Ethernet come in, and what are the respective data speeds?
[Choose two]
A. Real Ethernet, 10Mbps
*B. Fast Ethernet, 100Mbps
*C. Ethernet, 10Mbps
D. Ethernet, 100Mbps
Explanation: The two types are simply referred to as Ethernet, which operates at
10Mbps; and Fast Ethernet, which operates at 100Mbps.
3. True or False: Duplex data transmission is the most efficient method for sending and
receiving information.
*A. True
B. False
Explanation: The most efficient method for sending and receiving data is Duplex, and
the least efficient is Simplex.
4. What is the maximum distance data can be transmitted over Thick Ethernet network
cabling without signal attenuation?
A. 500 feet
B. 5000 feet
*C. 500 meters
D. 5000 meters
Explanation: Thick Ethernet, or what was more commonly referred to as Frozen
Garden Hose, has a maximum distance of 500 meters before the signal strength
begins to drop off, and a repeater becomes necessary.
5. What is the maximum distance data can be transmitted over Thinnet cabling without
signal attenuation?
A. 185 feet
*B. 185 meters
C. 500 feet
D. 500 meters
Explanation: Technically, the maximum distance is 185 meters, but most technicians
refer to the limit as 200 meters.
Networking 501
9. True or False: Fiber Optic and Cat 5 cabling are prone to the 3 demons?
True
*False
Explanation: Cat 5 is susceptible, but Fiber is not.
10. True or false: It is possible to transmit data between two or more PCs without them
being physically connected.
*True
False
Explanation: Wireless technology basically consists of using a wireless modem on a PC,
which bounces signals off of a fixed position in the sky, like a satellite.
11. What is the approximate maximum data speed for two PC's using serial port
connections on a network?
A. 100Mbps
B. 100Kbps
*C. 115Mbps
D. 150 Mbps
Explanation: It may not be pretty, but it is reliable.
Networking 503
13. Which of the following network topologies offers the most reliability in terms of
redundancy?
A. Star
*B. Mesh
C. Bus
D. Token Ring
Explanation: With a mesh topology, a company could have multiple breaks in the
system and still be able to communicate with every other PC on the network. These
are somewhat expensive to make, and are usually found in institutions like banks
where everything is mission critical.
14. Which of the following network protocols, or suite of protocols, is the most
commonly used on The Internet?
*A. TCP/IP
B. IPX/SPX
C. Banyan Vines
D. UDP/STP
Explanation: TCP/IP is a suite of protocols that almost completely dominates all Internet
traffic
15. What is it that allows you to break network IDs into multiple networks
A. TCP/IP
B. Octet
C. FQDN Identifier
*D. Subnet Mask
Explanation: Depending on your subnet mask, you can sub-divide your network into
multiple sub-networks.
16. What is it that allows data to exit from your LAN, or PC, onto The Internet?
A. Default GUI
*B. Default gateway
C. LAN segments
D. Hub
Explanation: The default gateway is what identifies the far side of the router that allows
you access to The Internet.
17. If you were the owner of a thriving ISP (Internet Service Provider) and wanted to
make the most efficient usage of your IP addresses, what tool or service would you
use?
A. DNS
B. WINS
C. UDP
*D. DHCP
Explanation: DHCP (Dynamic Host Configuration Protocol) allows you to pool your IP
addresses and dole them out on an as-needed basis.
18. What type of server or protocol is responsible for matching domain names to IP
addresses?
A. UDP
B. WINS
*C. DNS
D. ISP
Explanation: Domain Name Servers (DNS) hold databases of IP addresses and domain
names, much like your telephone
Networking 505
"We often discover what will do, by finding out what will not
do; and probably he who never made a mistake never made a
discovery."
--Samuel Smiles
Introduction
The In this chapter you will learn how-to Identify most challenges both great and small.
You will learn how to keep challenges in perspective as well as how to use two powerful
tools; so amazing you may think of them as secret weapons. The secret to successful
troubleshooting is about knowing where to find the answers and taking things one step at
a time.
II
Troubleshooting
It will be revealed to you the process of taking a monumental challenge and turning into
a molehill.
The process of performing upgrades without falling into an abyss is explained.
Vexing issues, such as intermittent challenges (historically the bane of any technician) is
revealed to have simple solutions.
Little known issues regarding viruses are unveiled.
Youll see how to avoid what first appear to be simple projects that turn into sand traps.
In this chapter you will discover 'tricks of the trade' that have been accumulated over
decades of experience.
Stay Focused
Dont let users swamp you with "could you help me with" When you
have a trouble ticket, stay on task. This does not mean be impolite.
If I could pass on only one thought on the topic of Troubleshooting Challenges it would
be that in the words of Darcy E. Gibbons, "success is just a matter of attitude."
Think
Our definition of insanity is doing the same thing over and over, when it isn't working,
and expecting different results. Don't get an idea in your head, and refuse to let go of your
assumptions. Track your changes, and make only one change at a time.
Your authors are big fans of two secret weapons. They are using your right brain
(subconscious) and asking someone on an email list.
I maintain a great variety of emails lists, including A+. To subscribe, send a plain text
email to AsktheAuthor@TotalRecallPress.com Enter in the subject line and body (both to
account for different email systems) SUB A+. Success will send you a response asking
you to confirm. Be sure to remove yourself before changing email accounts. We
encourage all our students to use subliminal audio to re-enforce concepts, regardless of
the topic. We didnt dream this up ourselves. The U.S. Army learned this, almost 100
years ago.
Upgrades
Getting the mind set that installingwill only take a few minutes is a recipe for disaster.
Ideally, you would consult a manufacturer's web site before proceeding on any sort of
upgrade. And the world is less than ideal. If you find yourself wishing you had checked
first, at least go check second. Read documentation, and look for Read.me, Readme.1st,
read me files on the driver or installation disk, NOW, or on web sites Frequently Asked
Questions (FAQ).
It is amazing how many answers are on there, once we start looking for them. If you just
made a change, and an issue arises, look around.
So your mouse doesn't work (all of sudden). Did you change keyboards for some reason?
Maybe you bumped the mouse cable.
Virus Scan
Viruses can be the root issue for almost any sort of bizarre behavior. There is discussion
of viruses that will re-program the EEPROM of BIOS. Some system boards now come
with two BIOSs to address this. Update your virus scanning software and scan for
viruses.
Rube Goldburg
The name Rube Goldburg is associated with inefficient contraptions. Partial upgrades of
computers, such as keeping an ISA based video card for a new system board is just
asking for an all nighter at the Full-Circle Ranch. The same thing applies to keeping
every software program/utility/game open.
KISS
The old saying Keep It Simple Silly applies here. Before looking for esoteric issues,
check what may be so obvious it is overlooked. One particularly vexing challenge we
faced several times was a PC that operated fine on the bench, and not in the field.
In these cases, the solution was with the power supplying the PC. Installing a UPS
worked like magic.
Intermittent Challenges
The challenge with intermittent challenges is determining if the issue is something that
occurs only in a particular set of circumstances or is the intermittent more vague to nail
down. When our graphic artist was attempting to work on 'Owl' for indicating historical
data, PhotoShop would lock up, but only on this graphic. It turned out that it was using
more RAM than other graphics. That meant that this graphic spilled over to the next
memory chip, which happened to be a different brand of ram. When the two different
manufacturers of ram attempted to work together, the difference caused a lockup, both
under NT 4 and Windows 2000.
Inverse Cause
So a system is locking up, corrupting files, and generally giving the user grief. It may
appear that an issue is causing corrupt files. Maybe corrupt files are causing the freeze.
Diagnostic Relief
In many ways, you are the Digital MD. Humans tend to get frustrated with Medical
Doctors who don't talk to the patent for 5 minutes, write out a prescription for some
magic pill, and make it all go away. Like the Medical Doctor, it may be difficult to find a
simple answer. That is why a doctor orders tests to help determine the root issue. We
have to do the same thing with our tools.
Some of them include ScanDisk, which has been built into Microsoft Operating Systems
for the past 10 years. Maybe you should invest in hardware diagnostic tools such as
POSTcard. This is a tool that plugs into a slot on the PC and works with the BIOS to
display where a system in hanging up in POST. Or maybe a software diagnostic program
like Quick Tech Pro, which allows you to boot to a DOS floppy disk and run tests on the
hardware.
Barring the investment of a POST-card, have a spare PC handy that can be used to test
suspect components, one piece at a time.
Cable Issues
I was a very young man coming from electronics to computers, when a very old man told
me "90% of your problems are cable related." I was astonished, and he had a very serious
look on his face. Well, 30 years later, if anything, I think that advice was understated.
Take this simple example:
You install new RAM. Before you put the case on, you test, thoroughly. Everything is AOK. You put the case back on, and test again. Everything is cool, life is good, and you
leave.
Two days later you get a call telling you that when the user changes floppies, it still is
showing the directory of the previous floppy. The user swears it wasn't doing that before,
and it only happened after you installed the new ram. They want the ram swapped out
with ram that "works". Does the ram have anything to do with the challenge? Of course
not. What happened is you either pinched the edge of the floppy cable on the case,
opposite the pin 1 conductor, or the cable got caught and almost pulled off the floppy.
Since pin 34 indicated the diskette change, and it isn't making contact, it won't tell the PC
that the user made a change, so it reads the directory of the first floppy, and will continue
to do so until the PC is restarted.
Pin 34
Pin 34 on a floppy carries the signal for a Diskette change.
Beep, Beep
Maybe you remember that cartoon character, Road Runner, letting Wiley E. Coyote lose
again, with a "Beep-Beep", and taking off.
If during the POST the computer finds an issue, it tells you with a Beep-Beep.
POST
POST is performed when a computer is first powered on. It cannot help
you after a computer has successfully booted. Some troubleshooting
issues may be caused by environmental factors such as 'dirty' power or
excessive temperatures.
BIOS Corruption
A continuous short beep is a good indicator that the BIOS is corrupt
There are many different BIOS manufacturers, both great and small, American
Megatrends Incorporated (AMI) is one of the most popular BIOS manufacturers in use
today.
In particular there are two different beep codes that would be good for you to keep in
memory, to save time.
RAM Failure
AMI uses 1 long and 3 short beeps to indicate a RAM failure.
(Typically in the first 64K).
Video Error
AMI uses 1 long and 8 short beeps to indicate a Video Display error. (System may still
boot).
Networking Challenges
As was pointed out before, 90% of all network challenges are cable related. In a bus
network, not terminating the ends of the cable results in signal reflection or bounce,
which can be the cause of all sorts of havoc. If a terminator is defective, which could
happen easily if moisture got into the resistor, it could look just like no termination.
VOM testing
If a network is not behaving test the cable and terminator with a VOM
(Voltage Ohm Meter).
Somewhat more rare, but still quite vexing is not all Ethernet networks are created equal.
In fact there are different types of Ethernet! This happened because Ethernet has been
around quite a while.
IEEE
Originally, the Institute Of Electrical And Electronic Engineers (IEEE) took a subsection of the work done by the committee formed by the Industry Standards
Organization (ISO) known as the Open Standards Interconnect (OSI).
The whole point of OSI was to give some organization to networking among different
manufacturers. While the OSI committee did great work, the IEEE felt some vagueness
in the famous 7-layer OSI model. So, they took the second layer (Data Link) and
created committees to further refine this layer.
The first results were released in February 1980. Since February is the second (2) month
of the year, their work became known as 802.
Different groups took on different definitions and each group took a number. So we have
things like 802.2 that tightly define Logical Link Control (LLC). Another committee
defined 802.3, Carrier Sense Multiple Access with Collision Detection. (CSMA/CD).
CSMA/CD Frames
While CSMA/CD is the most popular type of Ethernet in use today, it is not the only one.
For reasons beyond the scope of this work, Novell used 802.2 up through Version 3.11.
When Version 3.12 came out, they switched to 802.3 Technically, 802.3 is NOT
Ethernet. And this is what most people mean when they refer to Ethernet. To compound
the issue, both standards can co-exist however don't have to inter-operate!
For example, NT4 by default runs in Auto. It will accept anything looking like a signal.
And once it finds a 'type of Ethernet' it locks on, ignoring all others. An administrator
must specifically configure NT to accept the other types of signals. The actual signal
from many standpoints is identical. It is only when looking at the logical order within a
hunk of data (known as a frame) that the differences are seen.
Frame mismatch
If a computer running IPX will not communicate with a network, (but
others can) and it operates properly when using another transport
protocol, you have a frame mismatch.
Driver Issues
Drivers will make or break a PC. Printer printing garbage, video displays
not behaving nice, mice behaving erratically, or NICs that are suddenly
not communicating. Suspect drivers. Check Device Manager in Win 9.x
or Windows 2000 (Not in NT 4).
CMOS check
Check the CMOS for Plug-and Play if Win 9x fails to automatically recognize a new
device
BIOS updates
As it was revealed in chapter 0100 on hard drives, older BIOSs don't know about larger
hard drives. Installing a new hard drive in an older computer either needs Disk
Management software or new BIOS. The new BIOS is the preferred choice.
BIOS MB
Depending on the specific drive geometry, older BIOS will "top out"
between 512 MB and 540 MB.
Printers
Printers fail for a variety of reasons. The most common is, Out Of Paper (no kidding).
With network printing becoming common, the user can't see the printer in front of their
face. They print, expecting to see the work. Surprise!
It is common, when printing in Windows, to spool the print job to a hard drive either
remotely or locally, then have the printer begin printing. Unless the Network Operating
System is setup to notify the user the printer has an issue -- such as no paper, how would
the user know?
Other printer issues include:
Permissions
Permissions.
Has the network administrator allowed printing? Yes/No. How about to a specific
printer? Yes/No. Are there restrictions to printing to certain times of day? Yes/No.
Drill Down!
The example above shows the drill down successful troubleshooting
really entails.
Don't stop with the first question!
Keep drilling to find the root issue.
Meltdown
Dot Matrix printers will slow down or stop before going into thermal
meltdown, damaging the printer.
Sensors
Dot Matrix printers depend on micro-switches to control behavior. A trouble report
requesting a service call on a dot-matrix printer was one of the strangest ones I had ever
read. The report said, "Missing platen".
The platen is the round rubber roller behind the paper, to both help the paper advance
and give the pins something to push against. Getting to the site, sure enough -- no platen.
The paper out sensor had filled with paper dust. It couldn't operate to shut off the
printing.
The paper thickness sensor simply 'thought' it now had thinner paper, moving the print
head closer. With no paper the pins started beating up the platen. As the platen wore
down from the beating, the same paper thickness sensor moved the print head closer, so it
could continue to eat the platen.
Based on when the office started the job, I figured it beat away at the print job, sans
paper, from 7 PM to 9 AM. By then, the platen was a bunch of chewed up bits of rubber
on the bottom of the printer and the floor.
Clean Sensors
Paper sensors not cleaned will give all sorts of grief.
Operator Error
Operator Error, or as we like to say, "error between seat and keyboard", has the potential
to be most expensive in terms of repair. For example: a project calls for transparencies.
There are several types of transparency materials. One type for ink jet printers, another
for monochrome lasers, and a third for color lasers. The transparency material for ink jets
is very porous to accept the ink. The transparency material for lasers is specially
formulated to take the heat from the fuser. The third has a paper strip for the paper
gripper.
Suppose what happens when in a rush, a worker grabs the transparency material for ink
jets and runs it though a laser? The transparency material melts and becomes one with the
fuser and the rollers around the fuser.
Environment
Related to operator error is not observing environmental conditions. Environmental
conditions play a huge part in printing. Ever wonder why some days print jobs go fine,
and other days a sheet fed printer will take three sheets at a time? Anyone who has
worked in a printing shop for more than two weeks will tell you its all about humidity.
Quoting www.photocopiers.com.
Other Influences Regarding Paper Jamming:
"One of the most common causes of paper jamming, other than poor quality paper, is
humidity. Humidity disrupts the integrity of the paper, reduces its strength, and strongly
inhibits copy quality. Paper can be ruined in as little as two hours if the copier is not
located in a climate controlled office or if you live in such a place as the Pacific
Northwest where rain and humidity are commonplace."
Before you say lasers are not copiers, well, it really doesn't matter whether the Organic
Photo Conducting (OPC) drum is written to by a laser or by high-intensity light. The
operations of an analog copier and a laser printer are almost identical.
Drivers
This chapter has already discussed drivers. Here it is again. Once I made the mistake of
installing the first color laser printer. To keep the service fees down, once the printer
'talked' and was shared, I left. Big Mistake. Some people had taken the printer drivers
from the network share point (as I had asked them to). Others used the CD that came with
the printer. When they attempted to bring together all the sub-jobs of a huge report from
different workers, some had one set of shading colors, while other groups had a different
set. It depended on who used the CD based drivers, and who used the latest drivers I
downloaded and left on a network share. Both drivers 'worked', but the outcome was, err,
colorful.
III
In this chapter you learned about the techniques and tools used to successfully
troubleshoot a malfunctioning PC. First, you learned about the mindset that must be taken
by any PC technician to successfully troubleshoot a PC. Then, you were shown various
methods of critical thinking to help diagnose challenges.
Next, you learned about the various tools that should be handy anytime you are
troubleshooting PCs. You were then shown the AMI beep code set that can be used to
troubleshoot PCs during the POST phase.
Following that, you learned about different networking challenges, including cable
issues, NIC issues, and protocol issues. You learned about the IEEE CSMA/CD Ethernet
protocol and how it can present challenges for network connectivity.
Next, you were shown various issues that can arise with printers. This included a
discussion about printers, paper, ink, and sensors. Finally, you were shown how the
environment that a computer is in could affect its performance.
IV
1. Your company has recently added a new workstation to the network. The network is
using both TCP/IP and IPX/SPX protocols. The servers on the network are running
IPX/SPX only. The new workstation can connect to the other workstations using
TCP/IP, but it cannot connect to the servers. What should you check first?
A. IP Address
B. DHCP settings
C. Subnet Mask
D. Frame Type
2. The computer you just built gives you a series of beep when it is powered on. What
should you do first?
A. There is nothing wrong. Boot again
B. Count the beeps. Look up the beep code in the motherboard manual or look it up
on the BIOS manufacturers website.
C. Replace the motherboard
D. Replace the RAM
E. Replace the video card.
3. A user complains that she cannot see anything on here monitor. The monitor is
powered on and has a green light showing. The computer successfully POSTs and
appears to be loading the OS. What should you check first?
A. Reseat the video card
B. Replace the VGA cable.
C. Check the brightness setting
D. Replace the monitor
4. A workstation on your network is not connecting to any other computers. You verified
that the domain and IP settings are correct. You can successfully PING 127.0.0.1 and
the computers IP address. The back of the NIC has no light illuminated. The port on
the hub that the workstation is plugged into has no light illuminated either. What
should you check first?
A. Replace the NIC
B. Reinstall TCP/IP
C. Check the DHCP settings
D. Replace the drop cable with a known working cable.
5. You are trying to read a floppy disk in you OS, but the OS is showing the contents of
the last floppy disk that was in the machine. What should you do first?
A. Replace the floppy cable
B. Flip-flop the ribbon cable
C. Replace the FDD
D. Get a new floppy disk
6. A printer you are servicing is reporting that it is out of paper even though plenty is
loaded. What should you check first?
A. Replace the printer cable
B. Reinstall the printer driver
C. Try different paper.
D. Clean the inside of the printer with canned air
7. A computer you are servicing will not allow you to set a display mode higher than
640480 with 16 colors. What should you do first?
A. Replace the video card
B. Reload the video drivers
C. Replace the VGA cable
D. Replace the RAM
8. You are trying to connect to a shared folder on your network. The computer gives you
an error of Access Denied. What should you check first?
A. Reinstall TCP/IP
B. Reset your login password
C. Check the permissions of the folder
D. Replace the drop cable
9. A computer that you are working has an external serial modem that connects to the
Internet just fine in HyperTerminal. However, when you move the mouse to launch
your browser, the computer locks up. What is the issue?
A. Replace the mouse
B. Replace the serial cable to the modem
C. Dont use the mouse
D. Check the IRQ settings for the modem and mouse
10. You are troubleshooting a PC. You reboot and are immediately asked to enter a
password to continue with POST. What should you do?
A. Replace the motherboard
B. Set the Clear CMOS jumper
C. Replace the BIOS chip
D. The system is trashed. Get a new one
2. The computer you just built gives you a series of beep when it is powered on. What
should you do first?
A. There is nothing wrong. Boot again
*B. Count the beeps. Look up the beep code in the motherboard manual or look it
upon the BIOS manufacturers website.
C. Replace the motherboard
D. Replace the RAM
E. Replace the video card.
Explanation: Beep code can vary depending on the BIOS being used. Always consult
your documentation before replacing any parts.
3. A user complains that she cannot see anything on here monitor. The monitor is
powered on and has a green light showing. The computer successfully POSTs and
appears to be loading the OS. What should you check first?
A. Reseat the video card
B. Replace the VGA cable.
*C. Check the brightness setting
D. Replace the monitor
Explanation: It is not too uncommon for janitorial service to wipe down a monitor and
roll the brightness knob down to zero. Always check the obvious first.
4. A workstation on your network is not connecting to any other computers. You verified
that the domain and IP settings are correct. You can successfully PING 127.0.0.1 and
the computers IP address. The back of the NIC has no light illuminated. The port on
the hub that the workstation is plugged into has no light illuminated either. What
should you check first?
A. Replace the NIC
B. Reinstall TCP/IP
C. Check the DHCP settings
*D. Replace the drop cable with a known working cable.
Explanation: Remember that 90% of networking issues are cable related. In this
scenario, TCP/IP is successfully loaded and bound to the NIC, given that you can
PING the MS Loopback Adapter (127.0.0.1) and the IP address of the NIC. The
lights on the NIC and the hub not illuminating are a telltale sign of a bad connection.
5. You are trying to read a floppy disk in you OS, but the OS is showing the contents of
the last floppy disk that was in the machine. What should you do first?
*A. Replace the floppy cable
B. Flip-flop the ribbon cable
C. Replace the FDD
D. Get a new floppy disk
Explanation: This is a sign that pin-34 on the cable, controller or FDD is bad. Replace
the cheapest thing first.
6. A printer you are servicing is reporting that it is out of paper even though plenty is
loaded. What should you check first?
A. Replace the printer cable
B. Reinstall the printer driver
C. Try different paper.
*D. Clean the inside of the printer with canned air
Explanation: Many printers have sensors that tell the printer when it is out of paper. If
they become dirty, they can provide false errors.
7. A computer you are servicing will not allow you to set a display mode higher than
640480 with 16 colors. What should you do first?
A. Replace the video card
*B. Reload the video drivers
C. Replace the VGA cable
D. Replace the RAM
Explanation: 64048016 is the default VGA setting for most video cards. Many times,
reinstalling the drivers for the video card will resolve this issue.
8. You are trying to connect to a shared folder on your network. The computer gives you
an error of Access Denied. What should you check first?
A. Reinstall TCP/IP
B. Reset your login password
*C. Check the permissions of the folder
D. Replace the drop cable
Explanation: The error message says it all. You have network connectivity, you can see
the network, and you are communicating with the network. Without sufficient
permissions, you will not be able to access the share.
9. A computer that you are working has an external serial modem that connects to the
Internet just fine in HyperTerminal. However, when you move the mouse to launch
your browser, the computer locks up. What is the issue?
A. Replace the mouse
B. Replace the serial cable to the modem
C. Dont use the mouse
*D. Check the IRQ settings for the modem and mouse
Explanation: Older serial mice used a COM port, and hence an IRQ. The modem is
using a COM port as well. If they are using the same IRQ, they cannot be used at the
same time. Option C may work but is not very practical.
10. You are troubleshooting a PC. You reboot and are immediately asked to enter a
password to continue with POST. What should you do?
A. Replace the motherboard
*B. Set the Clear CMOS jumper
C. Replace the BIOS chip
D. The system is trashed. Get a new one
Explanation: Someone has set a password in the CMOS setup. You must set the Clear
CMOS jumper or remove the CMOS battery.
Appendix A 535
Base 10
Our everyday lives typically use some math that we call "checkbook" math. The
scientific name for this is known as Base 10 math. Let's look at why it is called that.
Begin by looking at the value of zero, or nothing. It is displayed as a 0. While zero is
empty of a positive value, it still is a value. It isn't positive, and it isn't negative. It is used
to either indicate the value of neither positive nor negative, or can be a placeholder that
contains nothing. To make the point, let's look at the following example, and use US
dollars as currency. $1,000,000.00
That is a base 10 expression of one million United States dollars. Not one penny more or
less. The six zeros before the decimal point indicates no amounts of hundred(s)thousands, ten(s) thousand(s), thousand(s), hundred(s), ten(s), or single dollar(s). The
zeros after the period indicate no fractions of a US dollar. Now visualize yourself having
$1,000,000.00
OK, now mentally give it to your authors. Now tell us the zero has NO value. Great, will
just remove the items with no value. Now you have $1.00. "Oops". No value with the two
zeros after the decimal point. Now you have $.01, or a US copper penny. Still want to say
zero has no value? As you can see, it is more accurate to see the zero as neither a positive
or negative value. It does indeed have a value, even if it is as a place holder.
536 Appendix A
OK. Let's do some counting with checkbook math. Starting with zero.
0
1
2
3
4
5
6
7
8
9
Now count the number of digits used. See? Ten, including the zero. Viola. Base 10 math.
We're willing to bet that you don't have to be as skilled as an NSA cryptographer to guess
the next highest value is 10. Let's pick this apart a bit more. The zero is a placeholder
meaning no positive values are contained in the units column. And the value in the
second column is indicated with a 1 for one unit of tens. If you continue counting you get
11. This creates one value in the unit column, and one in the tens column. This continues
until you get to twenty, or 20. This is because you have exhausted all the possibilities in
the units column using base ten, and have to notch up the ten column. Of course this
continues until you exhaust both columns with ninety-nine or 99. Once again, the two
columns zero out and a third column gets a one or 1, and becomes one-hundred or 100.
In math theory, this can go on forever.
Now as clever as computers seem to be, they really are only made up of a form of an
earth element known as silicon. In other words, sand. Now, how smart is sand? Its about
as smart as a rock. In fact, computers are nothing more than "sand with an attitude". Lets
take a look at how dumb computers really are in the next section.
Appendix A 537
Base 2
At its base form, a computer only knows two things. On or off. Put in a math expression,
that looks like zero or 0 and one, expressed as 1. OK. Lets start counting.
0
1
Oops!
Your computer just ran out of fingers!
So, in base 2 math, the same trick used in base 10 is applied.
Base 2
Base 10
10
11
100
101
And just like base 10, can in theory, go on forever. Reading base 2 math to base 10 math
is not difficult with a little practice. Since there are eight bits in a byte, lets look at what
8 bits can do
128
64
32
16
Each value can be expressed in a column, just like base 10, only this is base 2.
128
64
32
16
= 3
538 Appendix A
128
64
32
16
=4
128
64
32
16
=5
64
32
16
+ = 255
Of course you can get any combination between 0 and 255 by selecting which value you
want to turn on or off.
Maybe now you realize why you so often see subnet masks that contain the values 255
and 0.
If you are wondering about subnet masks, read on in Appendix C.
Appendix A 539
Base 16
No law prevents us from counting in values other than base 2 or base 10. I'll bet that if we
were born like some creatures from Star Trek and had 7 fingers and 1 thumb on each
hand, we would balance our checkbooks in base 16 math. Base 16 looks like this:
1=1
2=2
3=3
4=4
5=5
6=6
7=7
8=8
9=9
A = 10
B = 11
C = 12
D = 13
E = 14
F = 15
Notice that a single column can be used for higher values than with base 2 or base 10.
There is a bit of an issue with using base 16. Consider the value 278. In base 10 that is 2
hundreds, 7 tens, and 8 units. In base 16, that would not be the case, as the middle value
is counted against 16, not 10. In base 10, the value is 70 (7x10). In base 16 the base 10
value is 112 (7x16). So, when writing values in base 16, it is considered best practice to
put a little h after the value, such as 278h. Sometimes it is obvious it is in hex or
hexadecimal when a value contains an alpha character between A and F, for example:
3F8(h).
540 Appendix A
Base 16 and there values are seen with memory values. The value 278h is the in/out (I/O)
starting memory address for the second printer port (LPT2). The value 2F8h is the
starting memory address for Com 1. All Network Interface Cards (NIC) need a memory
address as well. Common NIC memory addresses are 300h, 320h, 340h, and 360h.
If after reading this several times it still sounds like some alien language, perhaps you
could consider taking a trip to a local Radio Shack. Ask for their book that teaches basic
electronics. It is inexpensive, and has a ton of easy to create experiments. With some fun
experiments you will learn more about electronic components as well as the math and
physics of electronics.
ExamInsight Network+
TotalRecall The IT Insight Book Series
ExamWise Network+
TotalRecall The IT Question Book Series
InsideScoop to Security+
TotalRecall Certification System
ExamInsight Security+
TotalRecall The IT Insight Book Series
ExamWise Security+
TotalRecall The IT Question Book Series
http://www.TotalRecallPress.com
281-482-5390 Fax
http://www.bfq.com
It's a Passing day here at the BeachFront.
Thank you for using the TotalRecall Publications ExamInsight Guaranteed Success
Program.
Bruce Moran President
Call: 281-992-3131