Troubleshooting the IA Platform

Troubleshooting the IA Platform
Troubleshooting for
the IA Platform
Revision 1.00
August 2001
Pat R.
Product Support Engineer
Intel® Customer Support
Revision History
Rev
Comments
Date
1.00
Initial Release
31 August 01
DISCLAIMER
Information in this document is provided in connection with Intel® products. No license,
express or implied, by estoppel or otherwise, to any intellectual property rights is
granted by this document. Except as provided in Intel's Terms and Conditions of Sale
for such products, Intel assumes no liability whatsoever, and Intel disclaims any express
or implied warranty, relating to sale and/or use of Intel products including liability or
warranties relating to fitness for a particular purpose, merchantability, or infringement of
any patent, copyright or other intellectual property right. Intel products are not intended
for use in medical, life saving, or life-sustaining applications.
Intel may make changes to specifications and product descriptions at any time, without
notice.
Contact your local Intel sales office or your distributor to obtain the latest specifications
and before placing your product order.
*Other names and brands may be claimed as the property of others.
Copyright © 2001, Intel Corporation
Intel Customer Support (ICS)
Troubleshooting the IA Platform
2
Table of Contents
INTRODUCTION
1 TROUBLESHOOTING OVERVIEW
1.1
1.2
1.3
1.4
Types of Problems
Troubleshooting Approach
Field Replaceable Units
Electrostatic Discharge (ESD)
2 WHERE DO ERRORS COME FROM
3 PROCESSOR OVERVIEW
3.1
3.2
3.3
3.4
3.5
3.6
3.7
Processor Characteristics
Thermal Solutions
Monitoring the Processor’s Temperature
Installing Processors
Slot-to-Socket Adapters (SSA)
Multiprocessor
Frequency ID Utility
4 MOTHERBOARDS
4.1
4.2
4.3
4.4
7
8
9
9
11
11
11
14
15
16
16
17
20
23
Chipsets
CMOS/BIOS Setup Program
Common BIOS setup option
Intel Chipset Software Installation Utility
5 RAM
23
23
26
27
28
5.1 Technologies (SDRAM, DDR, RDRAM)
5.1.1 SDRAM
5.1.2 DDR SDRAM
5.1.3 RDRAM
5.2 Troubleshooting RAM Problems
6 POWER SUPPLIES AND CHASSIS
6.1
6.2
6.3
6.4
6.5
6.6
6.7
6
7
28
29
30
30
31
34
Power Supplies and Power
On/Off Switch
Shorts and Overloads
AC Power Outlet
Power Irregularities
EMI
Thermal Management
34
35
35
36
37
37
38
7 EXPANSION CARDS AND SLOTS
8 DISK DRIVES
38
40
8.1
8.2
8.3
8.4
Floppy Disk Drives
IDE Hard Drives
Intel® Ultra ATA Storage Driver
SCSI Hard Drives
8.4.1 SCSI Encoding
Intel Customer Support (ICS)
Troubleshooting the IA Platform
41
41
43
43
44
3
8.4.2 SCSI Cables
8.4.3 SCSI ID Numbers
8.4.4 SCSI Termination
8.4.5 SCSI BIOS Programs
8.5 CD Drives and DVDs
9 FILE SYSTEMS
9.1
9.2
9.3
9.4
51
File System Characteristics
FAT32
NTFS
Troubleshooting File System Problems
10 VIDEO SYSTEM
51
52
52
53
55
10.1 Video Characteristics
10.2 PCI and AGP
10.3 Video Driver
56
56
57
11 PERIPHERALS
58
11.1 Serial and Parallel Port
11.2 USB
11.3 Infrared
58
58
60
12 NETWORK CARDS AND NETWORKING
12.1
12.2
12.3
12.4
12.5
45
47
47
49
50
Network Cards and Cables
Cable Modems and DSL Modems
TCP/IP
Sharing Resources
Internet and Browsers
13 WINDOWS* 95, WINDOWS* 98 and WINDOWS* MILLENIUM*
60
60
62
63
64
65
66
13.1 Problems with Installation
13.2 Boot Problems
13.3 Windows Device Manager
13.4 System Information Utility
13.5 Terminate a Specific Application
13.6 General Protection Faults and Exception Errors
13.7 Performance and System Monitor
13.7.1 Optimizing the Swap File
13.7.2 Optimizing File System Performance
13.7.3 Optimizing CD-ROM File System Performance
13.7.4 Using File System Troubleshooting Options
13.7.5 Graphics Compatibility Options
13.7.6 Optimizing Printing
13.7.7 Windows 9X Resources
13.7.8 Using System Monitor
66
69
73
74
77
77
80
80
81
83
83
85
86
86
87
14 WINDOWS NT, WINDOWS 2000 AND WINDOWS XP
90
14.1
14.2
14.3
14.4
Windows 2000 and Windows XP Boot Sequence
Boot Options for Windows NT, Windows 2000 and Windows XP
Event Viewer and Services Console
Changing the HAL
Intel Customer Support (ICS)
Troubleshooting the IA Platform
90
93
97
99
4
14.5 Stop Errors and Exception Errors
14.6 System Performance
15 DIRECTX
16 DATA
17 VIRUSES
100
102
109
110
111
APPENDIX A: COMMON PROBLEMS AND RECOMMENDED
TROUBLESHOOTING PROCEDURES.
A.1 Computer Does Not Complete Boot
A.2 BIOS Beep Codes
AMI BIOS
Award BIOS
Phoenix BIOS
A.3 Operating Problems
A.4 Processor Problems
A.5 Memory Problems
A.6 Disk Problems
Hard Drives
Floppy Disk Drives
Compact Disk Drives
A.7 General Expansion Card Problems
A.8 Video Problems
A.9 Sound Problems
A.10 Problems with Input Devices
Keyboards
Joysticks
A.11 Serial Ports, Parallel Ports and Modems
Serial Port
Parallel Port
Modem
A.12 Problems Shutting Down
Problems Shutting Down Windows 95/98/Me
A.13 Windows NT/2000/XP Stop Errors
Intel Customer Support (ICS)
Troubleshooting the IA Platform
112
118
118
119
120
122
124
126
129
129
136
137
139
140
141
142
142
142
143
143
143
143
144
144
151
5
INTRODUCTION
This guide is designed to give an overview of basic PC troubleshooting and to be used
as a quick reference guide. To use this manual effectively, you should already be
familiar with using and troubleshooting PCs. The first section will give an overview of the
PC and describe the individual components as they relate to troubleshooting the PC.
The second part will list common problems and will discuss on how to isolate these
types of problems. This guide is designed for quick access of a lot of information.
Important Note: Disassembly, assembling, upgrading and troubleshooting
computers should be performed only by a computer professional since the
electronic devices may cause serious damage to the installer, the system and its
components if it is done improperly. Before attempting to disassemble or
assemble computers, install components in a computer or to troubleshoot
computers, carefully review the documentation specific for the computer and its
related components. Make sure that you will not be voiding the warranty of the
system by opening the computer or replacing any of the components in the
computer. Lastly, make sure to follow Electrostatic Discharge (ESD) procedures.
Intel Customer Support (ICS)
Troubleshooting the IA Platform
6
1 TROUBLESHOOTING OVERVIEW
When troubleshooting any PC, there are certain guidelines that you should follow to
make yourself more effective. First, your mind must be clear and rested so that you are
able to concentrate on the problem. If not, you may overlook something that was
obvious such as the power cord not being plugged in.
Next, don’t panic, don’t get frustrated and allow enough time to do the job right. If you
panic, you may do something that will make the situation worse. If you start to get
frustrated, take a break. You will be amazed how five or ten minutes away from the
problem will clear your mind and will allow you to look at the problem a little differently
when you come back. Lastly, make sure that you have enough time to properly analyze
the problem, fix the problem and properly test the system after the repair. Again, if you
rush a job, you may make the problem worse or you may overlook something simple.
1.1 Types of Problems
When troubleshooting computer problems, you must keep an open mind on the cause
of the problem. The problem could be a hardware failure, compatibility problem,
improper configuration, software glitch, environment factor or user error. See Table 1.
Many of these may appear as hard problems but are actually something else.
Therefore, instead of only checking for faulty components (such as hard drives, floppy
drives, power supplies, cables or expansion cards), problems can also be caused by a
virus, software that isn’t compatible with a screen saver, CMOS setup program settings,
power management features, Control Panel settings, software drivers, hardware
settings, power fluctuations or electromagnetic interference.
Table 1: Computer Problem Classifications
Reason for Failure
Description:
This is when one or more components fail inside the computer.
Hardware Failure
This error is when a component is not compatible with another
Hardware
component.
Compatibility
Improper Hardware This error is when the hardware has not been installed or configured
properly.
Configuration
This error is when the software (Operating System or Application
Improper Software
Software) is not installed or configured properly.
Configuration
This error is when there is a glitch in the software. This can range
Software Failure
from corrupted data to a flaw in the programming.
The software may not be compatible with the hardware or other
Software
software.
Compatibility
The location of the computer and its environment (temperature, air
Environment
flow, electromagnetic interference, magnetic fields may affect the
reliability of the PC and has a direct impact on the PCs life.
A very common situation where the user hits the wrong keys or is not
User Error
familiar with the computer and/or software. It could be something as
simple as the user hitting the zero (0) key rather than the letter O.
Intel Customer Support (ICS)
Troubleshooting the IA Platform
7
1.2 Troubleshooting Approach
Before trying to fix the problem, you need to gather information. First, make sure that
you can duplicate the problem and that the user is not part of the problem. In addition,
determine if the problem is always repeatable or is an intermittent problem. If it is an
intermittent problem, does the problem follows a certain pattern (such as when the
computer is on for a while) or does it occur completely randomly.
The more difficult problems are the intermittent problems. Since the intermittent errors
do not happen on a regular basis, they may work fine when you test them, but, as soon
the customer takes it home, it fails again. When dealing with intermittent problems, you
make a change that might fix the system. You must then thoroughly test the system
over a period of time to see if the problem actually goes away.
You can gather additional information by trying to use software utilities to test your
system and by using a digital multimeter (DMM). Some of the utilities include software
to test the computer components, check for viruses, look for formatting errors on a disk
or check software configuration. In addition, find out if the computer was serviced or
changed recently. Many times servicing or changes can cause other problems.
After you have gathered as much information as you can, you are now ready to make
the repair or fix. Sometimes, you will know exactly what to change or replace. Other
times, you will have it narrowed down to several causes, which would require you to
isolate the problem. If you suspect a faulty component, you can replace the component
with a known good component. If the system works with the new component, the
problem was the item that you just removed. Of course, while the component could be
faulty, the problem could also be a compatibility problem, a BIOS problem or a driver
problem. For example, if you replace a processor and the new processor works, the
other processor could still be good. To determine if it is good, it would have to tested in
another known working machine.
Note: when a new item is taken off the shelf, it does not mean that the item is always
good. In addition, you can try “reverse swapping” which is trying a suspected
component into a second working system.
Other solutions including reconfiguring the software or hardware, reloading the
operating system, application software or drivers, making changes to the CMOS Setup
program or reconfigure the software. Whatever course of action that you choose, you
should only make one change at a time. If the problem still exists, you will then make
another change until the problem no longer exist. When determining which item to
check or swap, you should first try to check items that are likely to cause the problem
and are the easiest and quickest to check.
After you fixed the problem, you should always thoroughly test the computer before
returning it to the customer or client. This will make sure that the problem did go away
and that you did not cause another problem when fixing the first problem.
Intel Customer Support (ICS)
Troubleshooting the IA Platform
8
1.3 Field Replaceable Units
A field replaceable unit, also commonly referred as a FRU, is any PC component that
can be replaced without any special skills or tools. For the PC, the most common FRUs
are:
• expansion cards
• RAM
• power supplies
• keyboards
• monitors
• CMOS batteries
• processors
• motherboards
• floppy disk drives
• hard drives
• CD/DVD drives
• drive cables
Before replacing any FRUs, you should do the following:
1.
2.
3.
4.
Confirm that all cables and connectors are connected properly and securely
Verify CMOS setup programs
Update the BIOS
Verify that all drivers are installed properly and that you have loaded the newest
drivers
5. Look for updated device drivers
1.4 Electrostatic Discharge (ESD)
Electrostatic electricity is an electrical charge at rest created by friction and separation.
If the proper steps are not taken to avoid electrostatic discharge (ESD), it could damage
the electronic components of a PC or cause the PC to loose data and programs.
To give you an idea on how much charge can be generated from ESD, on a dry day,
your body can generate up to 35,000 volts. To feel a voltage discharge such as being
shocked when touching a doorknob is at least 3,000 volts. Lifting a foot or moving a
chair can easily generate 1,000 volts. Today, computer components can be destroyed
or degraded as low as 10 to 20 volts depending on the type of chip. These voltages are
not lethal to a person because the discharge occurs within a fraction of a second.
As a technician, it is imperative to avoid electrostatic discharge. You will have fewer
difficult-to-trace intermittent problems and fewer unnecessary service visits to make.
This will lead to fewer disgruntled customers.
The best way to avoid electrostatic discharge is to properly wear a wrist strap that is
grounded. It can be grounded by attaching to the metal frame of the PC (preferable an
area free of paint). The wrist strap must make good contact (metal part within the strap)
with the skin. It is you whom is being grounded not your clothing. It should be kept
clean, in good condition and should be tested daily to verify that there is continuity
Intel Customer Support (ICS)
Troubleshooting the IA Platform
9
(good connection) between you and ground. In addition, you should roll up your sleeves
and secure your neckties so that they don’t make contact with the electronic device.
Some books will recommend an alternative method if you do not have a wrist strap. You
can touch the power supply/metal case before working within the computer. You should
then touch the power supply/metal case often while you are handling computer
components to reduce any charge build up. This method is not as effective as using a
wrist strap. Therefore, this method is not recommended.
Warning: When working on high voltage devices such as monitors or power supplies,
a wrist strap should not be used since the voltage stored in these devices can kill a
person. Note: Monitors and power supplies should only be serviced by trained
experienced electronic technicians.
Another device that you can use is an electrostatic mat. The computer that you are
repairing sits on top of the mat and the mat absorbs static electricity. Of course, the mat
is grounded much like the electrostatic wrist strap.
When transporting and storing electronic devices, electrostatic bags should be used.
An electrostatic shielded bag is usually silver, smoke-gray or grid-lined. The static
shielded bag must be securely closed to be effective and should not have holes. You
should always keep spare static shielded bags.
When handling integrated chips, you should avoid touching any of the metal parts.
When handling circuit boards, you should avoid touching any of the chips, electronic
components, metal traces and edge connectors. When inserting or removing any
integrated chips or expansion cards, you should make sure that the computer is off and
unplugged. If you need to hand a component to a person, make sure it is sealed in an
electrostatic bag or you need to touch the person’s hand with your other hand before
handing the component over.
Figure 1: Warning for Electricity.
Intel Customer Support (ICS)
Safety Guidelines – Electricity
• Never use both hands to handle an electronic
device.
• Wear rubber bottom shoes.
• Don't wear any jewelry or other articles that could
accidentally contact circuitry and conduct current, or
get caught in moving parts.
• Connect/disconnect any test leads or connections
with the equipment off and unplugged.
• Never assume anything without checking it out for
yourself.
• Don't take shortcuts.
Troubleshooting the IA Platform
10
2 WHERE DO ERRORS COME FROM
Knowing where errors are generated are very useful in trying to isolate a problem to fix
a computer. When you first turn on the PC, the computer performs a quick power check.
When the microprocessor receives a power good signal, the processor starts reading
and executing the instructions in the System ROM BIOS on how to boot the computer.
The processor will then initialize and test the essential PC components (RAM and
motherboards) and perform an inventory of the hardware devices installed in the
computer. When a component does not respond correctly to the test, the test failure will
be identified as a series of beeps (audio codes) or a number code/message (video
code).
During this time, the system ROM BIOS has instructions to go out and search for
additional ROM BIOS chips. Since the system ROM BIOS cannot possibly hold
instructions for every piece of hardware or could not include instructions for new pieces
of hardware, which are introduced every day, there was a need for additional
instructions to control hardware. The additional hardware instructions are located in the
other ROM BIOS chips found on the expansion cards.
The last thing that the system ROM BIOS chip does during boot up is to find a boot
device to load the operating system. If all goes well, the system will finish with a prompt
waiting to input a command (DOS or UNIX/Linux) or will display a GUI interface
(Windows* 95/98/Me or Windows NT/2000/XP).
Therefore, if the monitor has not been activated and you get a series of beeps, the
errors are the audio beep codes that are generated by the system ROM BIOS. After the
video system is initiated and before the operating system is loaded, the error message
will be text messages or numeric codes displayed on the screen generated by the
system ROM BIOS. If the error message occurs when the operating system starts to
load then the error message is generated by the operating system or a software
application.
3 PROCESSOR OVERVIEW
The computer is built around an integrated chip called the processor. It is considered
the brain of the computer since all of the instructions it performs are mathematical
calculations and logical comparisons. It is the main component that the rest of the
computer is built around.
3.1 Processor Characteristics
All processors are described by the following characteristics:
• Clock Speed – The speed at which the microprocessor executes instructions.
Every computer contains an internal clock that regulates the rate at which
Intel Customer Support (ICS)
Troubleshooting the IA Platform
11
•
•
•
•
•
•
instructions are executed. It is expressed in Megahertz (MHz), which is 1 million
cycles per second or Gigahertz (GHz), which is 1 billion cycles per second.
Bus Speed – The bus that connects the CPU to main memory (RAM). As
microprocessors have become faster and faster, the system bus has become one
of the chief bottlenecks in modern PCs. Today, typical bus speeds are 66 MHz,
100 MHz, 133 MHz or 400 MHz.
Architecture – The architecture determines how the processor will execute
instructions. The Intel® Pentium® II, Intel® Pentium® III and Intel® Celeron™
processors use the P6 architecture while the Intel® Pentium® 4 processor uses
the Intel NetBurst™ Micro-Architecture. In addition, newer processors have used
additional instruction sets (MMX™ technology and Streaming SIMD) designed to
perform the same type of calculation on several values at the same time.
Cache – An ultra fast memory typically included as part of the processor used to
increase memory performance. By keeping as much of this information as possible
in the cache, the computer avoids accessing the slower RAM. Most processors
several levels of cache, referred to as L1 cache and L2 cache.
Technology – The Pentium II processors, early Pentium III processors and some
Intel Celeron processors were produced with .25 micron technology, while later
Pentium III processors and the Pentium 4 processors are produced with .18 and
.13 micron technology. The .25 micron, .18 micron and .13 micron indicate how big
the transistors are. As the transistors get smaller, allows the processors to get
faster, to run cooler and to run with less power.
Voltages – The processor will need to run with a specific voltage to run properly.
All current Intel processors provide a Voltage Identify signals to tell motherboards
that can automatically detect the voltage on what the correct voltage is. Voltages
are displayed on the label on a boxed processor.
Package Type – The form and number of pins of a processor. See table 2.
Currently, the package types include:
Table 2: Package Types Used in Processors
Package Type
SECC2
Number of Pins or Contacts
242-contact slot connector (Slot 1)
SECC (330)
Flip Chip PGA (FC-PGA)
330-contact slot connector (previously
called Slot 2)
370-pin socket
423 Pin Package
478 Pin Package
603-pin micro- PGA
423-pin socket
478-pin socket
603-pin package
Notes:
Pentium II and Pentium
Pentium III processors
Pentium II and III Xeon™
processors
Pentium III processor
Intel Celeron processor
Pentium 4 processor
Pentium 4 processor
Intel Xeon procecessor
• Boxed processor versus OEM processor – A boxed Intel processor, sometimes
referred to as a retail box, comes in a sealed box, which comes with a 3-year
warranty from Intel. An OEM processor, sometimes referred as a tray processor, is
a processor sold to an OEM manufacturer or distributor intended for installation.
The warranty varies in length and is provided by the place of purchase. For more
information on Intel’s warranty policy, go to:
Intel Customer Support (ICS)
Troubleshooting the IA Platform
12
http://support.intel.com/support/processors/warranty.htm
• Steppings – Throughout the life of a processor life cycle, the processor may go
through several steppings or versions. Newer steppings typically have some type
of improvement over previous steppings of the processor. Note: Steppings can be
identified by the CPUID number, also shown in the support website, the
specification update guide and the Frequency ID Utility.
Figure 2: Sample Processor Specification Guide and Intel S-Spec Support Website
Processor can be identified with their S-Spec Number (A 5 character string starting with
SL), which is printed on the box of a boxed processor or on the processor itself. By
looking up the S-Spec Number at the processor support website or by using the
Specification Update Guide, the characteristics of the processor can be identified
including the stepping, cpuid, bus speed and cache speed. See Figure 2.
When viewing the specification update guide, each specification is assigned a CPUID
string, which is four characters in length. The first character is typically a 0 and the
second and third characters represent the family and model number. The fourth
character represents the processor stepping. For example, if you find one processor
with a CPUID of 0672 and a second processor with a CPUID of 0673, both processors
have the same family and model number of 67, but with different steppings. You can
identify the stepping by reference the CPUID with the Specification Update Guide or the
S-Spec Support website. Note: using the Intel frequency ID utility, you can view the
processors CPUID.
Intel Customer Support (ICS)
Troubleshooting the IA Platform
13
Processor Weblinks
Intel Pentium® III processor
Specification Update Guide
http://developer.intel.com/design/pentiumiii/specupdt/
S-Spec Information (Support)
http://www.intel.com/support/processors/sspec/p3p.htm
Known Issues & Solutions (Support)
http://www.intel.com/support/processors/pentiumiii/
Intel Pentium® III Xeon processor
Specification Update Guide
http://developer.intel.com/design/pentiumiii/xeon/specupdt/244460.htm
S-Spec Information (Support)
http://www.intel.com/support/processors/sspec/p3xp.htm
Known Issues & Solutions
http://www.intel.com/support/processors/pentiumiii/xeon/
Intel Pentium 4 processor
Specification Update Guide
http://developer.intel.com/design/pentium4/specupdt/
S-Spec Information (Support)
http://www.intel.com/support/processors/sspec/p4p.htm
Known Issues & Solutions
http://www.intel.com/support/processors/pentium4/
Intel Celeron processor
Specification Update Guide
http://developer.intel.com/design/celeron/specupdt/
S-Spec Information (Support)
http://www.intel.com/support/processors/sspec/icp.htm
Known Issues & Solutions
http://www.intel.com/support/processors/celeron/
Intel Xeon processor
Specification Update Guide
http://developer.intel.com/design/Xeon/specupdt/
S-Spec Information (Support)
http://www.intel.com/support/processors/sspec/p4xp.htm
Known Issues & Solutions
http://www.intel.com/support/processors/xeon/
3.2 Thermal Solutions
Since the processor (the processor generating the bulk of the heat) and other
components produce heat, all systems require thermal management. Without it, the
system will act erratic, generate errors, degrade performance or fail to operate. In this
case, “thermal management” includes three major elements:
1. a heatsink properly mounted to the processor
2. local airflow to transfer the heat to the chassis air
3. airflow to evacuate the heated air from the chassis.
The power supply fan and sometimes other chassis fans provide the local airflow in the
case and evacuate the heated air from the chassis. See Section 6.6 for more
information on chassis and power supplies.
Intel Customer Support (ICS)
Troubleshooting the IA Platform
14
The ultimate goal of thermal management is to keep the processor at or below its
maximum operating temperature. The maximum operating temperature is measured at
the center of the surface of the processor’s thermal plate and varies depending on the
particular frequency and stepping of the processor.
Proper thermal management is achieved when heat is transferred from the processor to
heatsink, from the heatsink to the chassis air, and from the inside the chassis to the
outside. Boxed desktop processors are shipped with an attached high-quality heatsink
and most come with a high-quality fan, can effectively transfer processor heat to the
system air. It is the responsibility of the system integrator to ensure good system airflow
to remove the heat from the heatsink, and from the chassis to the outside air. For OEM
processors, you need to add a high quality heat sink/fan to the processor.
The heatsink design allows heat to transfer from the processor, through the thermal
interface material, through the heatsink base, and up each of the heatsink fins. Airflow
around the fins carries the heat off the fins and into the chassis interior. The heatsink is
designed for maximum efficiency when air flows either horizontally or vertically across
the heatsink. In some cases, air can be blown directly into the middle of the heatsink,
provided that the exiting air is eventually removed form the chassis. The system
integrator is responsible for removing the heat from the heatsink fins with localized
airflow.
The fan power cable must be connected to the fan power connector and the
motherboard power header. The motherboard 3-pin power header uses two pins to
supply +12V and Ground (GND). The fan uses the third pin to transmit fan-speed
information to motherboards that support fan-speed detection. Your motherboard must
have a 3-pin fan power header located close to the slot or socket. Refer to your
motherboard manual for the location of the power header.
3.3 Monitoring the Processor’s Temperature
If a motherboard has a thermal sensor, the temperature of the processor can be
monitored. If the motherboard thinks that the processor is running too hot, it will throttle
the processor in attempt to cool it or it could shut down the entire system. These
motherboards may allow you to view the temperature using the CMOS/BIOS setup
program or by running a software package such as the Intel Active Monitor that comes
with Intel boxed motherboards.
Sometimes, in particularly when a new family or stepping of a processor is released, the
motherboard may misread the temperature of the processor. If this occurs with the
different processors on the same motherboard, you should question is it the processor
or is it the motherboard. If you feel little heat coming from the processor and/or heatsink,
it is most likely the motherboard misreading the temperature. Note: Since a heat sink
can get very hot, you should be careful not to burn yourself. To get an accurate
reading, you will have to use a thermal measuring device that you will have to put on top
Intel Customer Support (ICS)
Troubleshooting the IA Platform
15
of the processor die. Note: If motherboard misreads the temperature, it can usually be
corrected with a BIOS update.
3.4 Installing Processors
Motherboards used with an Intel processor must specifically support the processor type,
core speed and stepping. Some older motherboards may not support all speeds and/or
all steppings of newer processors. Using a processor in a motherboard that does not
support its current electrical requirements can cause permanent damage to the
processor or motherboard. Some older motherboards will require BIOS upgrades and
microcode updates (for more information on the microcode update, see section 3.4) in
order to properly recognize newer processors. In addition, using a processor with the
incorrect BIOS revision may cause a system to be unstable. Verify that the specific
motherboard and BIOS revision will support the processor that you want to install.
3.5 Slot-to-Socket Adapters (SSA)
Intel does not recommend using slot-to-socket adapters (SSA). See figure 3. If you
decide to use a SSA, be sure to follow these critical recommendations:
1. The SSA must be recommended by the motherboard manufacturer for use with
the specific processor and your selected motherboard. Integrating the Pentium III
processor in the FC-PGA package into SSAs that are designed for the Intel
Celeron processor may damage the Pentium III processor. For specific validation
of any SSA with your motherboard and processor, please check with your
motherboard vendor.
2. The SSA must operate the processor within Intel’s electrical, mechanical, and
thermal specifications.
3. Intel does not validate SSAs for use with Intel branded boards or third-party
motherboards.
Note: Severe mechanical and electrical damage may occur to the processor,
motherboard, or both if these recommendations and considerations are not carefully
followed.
Be aware of thermal considerations and mechanical requirements when installing any
SSA. The fan heatsink included in the boxed processor package will provide sufficient
cooling. However, this solution violates the maximum mechanical volumetric dimensions
and weight restrictions that are recommended for motherboards with the 242-contact
slot connector. Violating this recommendation will result in the SSA "leaning" to one
side, which may cause broken electrical connections and interference with motherboard
components. Therefore, system integrators using an SSA must follow two requirements
when installing a Pentium III processor.
1. The SSA is electrically connected and mechanically held secure in the 242contact slot connector (consider a SSA with a retention mechanism interface).
Intel Customer Support (ICS)
Troubleshooting the IA Platform
16
2. No motherboard components are adversely affected by installing the processor
on the SSA.
Figure 3. Slot-to-Socket Adapter (SSA)
Installed into a Motherboard
3.6 Multiprocessor
Symmetric MultiProcessing (SMP) involves a computer that uses two or more
microprocessors that share the same memory. If software is written to use the multiple
microprocessors, several programs can be executed at the same time or multithreaded
applications can be executed faster. A multithreaded application is an application that is
broken into several smaller parts and executed simultaneously. For example, Microsoft*
Word* uses a thread to respond to keys typed on the keyboard by the user to place
characters in a document. Other threads are used to check spelling, paginate the
document as you type and to spool a document to the printer in the background. The
ability for an OS to use additional microprocessors is known as multiprocessing
scalability.
When verifying if processors will work for your multiprocessing system, you must first
determine if the processors will support multiprocessing and if the processors will work
with each other. Therefore, you must check the following:
1. The only processors that are multiprocessing capable are Intel Pentium II, Intel
Pentium III, Intel Pentium II Xeon and Pentium III Xeon processors. Note: not all
of these processors support multiprocessing. The Intel Celeron, Pentium 4 and
mobile processors do not support multiprocessing.
2. Make sure the processor supports multiprocessing. This can be determined by
using the respective specification update guide. To determine if the processor
supports multiprocessing, find your S-Spec in the Processor Identification and
Package Information tables and read the notes for the particular processor to
determine if the processor is multiprocessing capable. An S-Spec is the number
starts with an "S" and is followed by 4 characters (for example, "SL3CC"). If there
Intel Customer Support (ICS)
Troubleshooting the IA Platform
17
are no notes specifying if the processor supports multiprocessing or not, assume
that it supports multiprocessing. See figure 4.
3. The processors must have identical frequency, bus speed and cache size.
Processors used in a multiprocessing system must share the same family and
model number. This will ensure that processors used in a multiprocessing system
will have the same speed, same cache size and same bus speed. In many
cases, you may mix processor steppings in a multiprocessor system.
4. Processors must have compatible steppings. By using the specification update
guide, you can determine the stepping by referring to the S-Spec in the
Processor Identification and Package Information tables. Then you need to
cross-reference both steppings, processor speed and bus speed with the DP
Platform Population Matrix tables to see if there are any issues and if there are
any issues, what are the issues. NI indicates no issues. See Figure 5. Note:
Identical silicon steppings may sometimes be shipped with different test
specification numbers, depending on whether they are for OEMs (tray
processors) or for system integrators (boxed processors).
Even though Intel recommends using identical processor steppings in multiprocessor
systems whenever possible (as this is the only configuration which receives Intel's full
testing), Intel supports mixing processor steppings, and does not actively prevent
various steppings of the Pentium III processor from working together in dual-processor
systems. However, since Intel cannot test every possible combination of devices, each
new stepping of a device is fully tested only against the latest steppings of other
processors and chipset components. In addition, when partially populated systems are
shipped, a customer returning for additional processors may have difficulty locating a
processor with the same stepping. If this occurs, the customer may have to replace the
original processor to obtain a system with two processors of the same stepping.
To make sure that the processors will work in a multiprocessing system, you must make
sure that the motherboard will fully support multiprocessing and it will support the
specific processors you want to use. Therefore, you should check the following:
1. Due to the variety of motherboard vendors in the market and the number of system
BIOS revisions, some system-level issues may occur that lie outside the realm of
any mixed stepping evaluation performed by Intel. Intel recommends the following:
• Choose a motherboard vendor with a history of dual-processor experience.
• Contact your motherboard vendor for information on mixed stepping validation
performed on your particular dual processor motherboard.
Note: Although Intel includes a multiprocessing "known issues" table, only the
motherboard vendor can say definitively if a particular pair of mixed steppings has
been validated in that particular board.
2. The power supply must be able to provide sufficient power to the motherboard with
multiprocessors and all peripherals connected to the motherboard.
3. If you use two processors that have different voltages, the motherboard must have
two voltage regulators to operate both processors properly.
Intel Customer Support (ICS)
Troubleshooting the IA Platform
18
Figure 4: By looking up the S-Spec number in the specification update
guide, you can look up the core stepping, speed, bus speed, cache size,
package type and notes for the processor.
Figure 5: The DP Platform Matrix used in the Specification Update Guides
can determine if steppings are compatible. In this example, a Pentium III
processor running at 600 MHz with a 133 MHz bus and 256 KB cache
have no issues between processors using the cB0 stepping and the cC0
stepping.
Intel Customer Support (ICS)
Troubleshooting the IA Platform
19
4. Intel does not recommend using slot-to-socket adapters (SSA) in dual processing
systems. If you decide to use a SSA, be sure to that the motherboard manufacturer
has validated the specific SSA with the specific motherboard and processor.
Recommendations for Integration of Dual Processor Systems:
1. In dual processor systems, the processor with the lowest feature-set, as determined
by the CPUID Feature Bytes, must be the Bootstrap Processor (BSP). The BSP is
the processor that starts the boot process. In the event of a tie in feature-set, the tie
should be resolved by selecting the BSP as the processor with the lowest stepping
as determined by the CPUID instruction. Example of mixing b0 and c0
2. Be sure to use an operating system that supports multiprocessing. Currently,
Microsoft Windows NT 4.0, Windows 2000 and Windows XP (excluding Windows XP
Home) support multiprocessor and some versions of Linux. Microsoft Windows 95,
Windows 98, Windows ME and Windows XP Home edition do not. If you only have
one processor installed when you install Windows NT, Windows 2000 or Windows
XP, you will need to change the HAL to a multiprocessing HAL when you start using
more than one processor. For Linux to support multiple processors, you may have to
recompile the kernel. Check with the documentation that comes with the OS.
3. If you use a multiprocessing motherboard and you don’t use all of the processors,
some motherboards require having a special terminator in the empty processor slot.
If the motherboard is self-terminating and you are having problems booting, you may
try inserting a special terminator or a matched processor in the empty slot or socket
to verify that the self-termination is actually working properly.
4. Check the CMOS Settings. Some boards may have options to retest or
enable/disable processors.
3.7 Frequency ID Utility
The Intel® Processor Frequency ID Utility was developed by Intel Corporation to enable
consumers the ability to identify and, in some circumstances, determine if their Intel
processor is operating at the correct and rated frequency intended by Intel Corporation.
Beginning with the Pentium® processor, this utility provides consumers with the ability
to determine standard CPU identification of Intel processors. In addition, supported
processors can utilize the Frequency Test feature of the utility to help determine if a
processor has been overclocked.
Version 3.0 of the Intel Processor Frequency ID Utility has been updated to support the
new Pentium® 4, Intel® Xeon™, Pentium III, Mobile Pentium III with Intel SpeedStep™
technology, Pentium III Xeon™ and Intel Celeron™ processors with 66, 100, and 133
MHz system memory bus products, as well as adding new processor identification
functionality (CPUID).
There are two versions of the utility available. The Windows version of the utility can be
used with systems that support the Windows operating system environment. The
Intel Customer Support (ICS)
Troubleshooting the IA Platform
20
"Bootable" version of the utility does not require an operating system. By installing the
tool on a bootable device (e.g. floppy drive) the tool is run as the system is booted, but
before the PC's operating system is loaded. This is normally done from a PC's floppy
"A:" drive.
There are two separate testing & reporting capabilities in the Intel® Processor
Frequency ID Utility represented by unique tabs. Each tab has specific features and
acts independent of the other. See figure 6.
Frequency Test Tab (Windows Version) - On supported processors, the utility runs a
speed test algorithm to determine the processor's operating frequency and compares it
to the processor's expected frequency. Both the reported frequency and expected
frequency are displayed on the Frequency Test Tab. If the processor is determined to
be running as expected, the utility shows positive test with a "pass" icon. If the utility
determines that the processor is running above it's expected frequency, it cautions the
user that the tested processor appears to be running above its expected frequency - a
condition commonly referred to as "overclocked". Please note that the utility will
automatically enable the Frequency Testing Tab only if you are using a system with a
supported processor.
CPUID Data Tab (Windows Version) - Some operating systems can misidentify new
processors. The CPUID Data Tab of the Intel® Processor Frequency ID Utility properly
identifies Intel products beginning with the Pentium processor. It also displays the brand
name for the Pentium III processor family.
For Intel processors starting from the Pentium processor generation, the utility identifies
the product’s brand designation (product name) and displays information about the
processor’s cache size, multimedia instructions, and Intel manufacturing designations.
This information is occasionally helpful when communicating with Intel about a
customer's specific processor.
For more information on the Processor Frequency ID Utility:
http://www.intel.com/support/processors/tools/frequencyid/
To download the Processor Frequency ID Utility:
http://www.intel.com/support/processors/tools/frequencyid/download.htm
Intel Customer Support (ICS)
Troubleshooting the IA Platform
21
Figure 6: The Intel Processor Frequency ID Utility
Intel Customer Support (ICS)
Troubleshooting the IA Platform
22
4 MOTHERBOARDS
The motherboard, also referred to as the main board or the system board, is the primary
printed circuit board located within the PC. It includes connectors for attaching
additional boards (expansion slots) and additional devices (ports). In addition, it
contains the microprocessor, the RAM chips, the RAM cache, several ROM BIOS chips,
CMOS RAM, the real-time clock and several support chips.
4.1 Chipsets
A chipset consists of chips and other components on the motherboard, which allow
different PC components to communicate with each other including the microprocessor.
It consists of the bus controllers, peripheral controllers, memory controllers, cache
controllers, clocks and timers. See table 3 and 4 to identify Intel chipsets.
Table 3: Intel Chipset with Integrated Graphics
Intel Chipset
I/O Controller Hub
(ICH)
Intel 810 chipset
82801AA or 82801AB
Intel 810E chipset
82801AA or 82801AB
Intel 810E2 chipset
82801AA or 82801AB
Intel 815 chipset
Intel 815E chipset
Intel 815EM chipset
82801AA or 82801AB
82801BA
82801BA
Graphics Memory
Controller Hub
(GMCH)
FW82810 or
FW82810DC100
FW82810E or
FW82810DC100
FW82810E or
FW82810DC100
FW82815
FW82815E
FS82815EM
Firmware Hub (FWH)
82802AB or 82802AC
82802AB or 82802AC
82802AB or 82802AC
82802AB or 82802AC
82802AB or 82802AC
82802AB or 82802AC
Table 4: Intel Chipset without Integrated Graphics
Intel Chipset
I/O Controller Hub
(ICH)
Intel 815EP chipset
Intel 815P chipset
Intel 820 chipset
Intel 820E chipset
Intel 840 chipset
Intel 850 chipset
82801BA
82801AA or 82801AB
82801AA or 82801AB
82801BA
82801AA or 82801AB
82801BA
Graphics Memory
Controller Hub
(GMCH)
82815EP
82815P
82820
82820
82840
82850
Firmware Hub (FWH)
82802AB or 82802AC
82802AB or 82802AC
82802AB or 82802AC
82802AB or 82802AC
82802AB or 82802AC
82802AB or 82802AC
4.2 CMOS/BIOS Setup Program
The configuration information is stored in CMOS, a special form of RAM. When you
disconnect the power from the RAM, it will forget its content. Yet, when you shut off your
computer and you turn on the system later, the PC still remembers the configuration
information and the clock still has the correct date and time. This is because the CMOS
chip is connected to a battery, which keeps the CMOS circuit powered while the PC is
off. This battery is often referred to as the CMOS battery.
Intel Customer Support (ICS)
Troubleshooting the IA Platform
23
The CMOS Setup program (also referred to as the BIOS setup program), which is
stored in the system ROM BIOS, varies greatly from computer to computer. Even PCs
from the same manufacturer containing the same CPU and system ROM BIOS from the
same manufacturer may also have major differences.
Most Intel motherboards have a single jumper block. When the jumper specifies
configure mode, the BIOS/CMOS setup program will run, which also includes a
maintenance menu that is not available during normal access of the program. See
figure 7. CAUTION: Do not move the jumper with the power on. Always turn off the
power and unplug the power cord from the computer before changing the jumper.
Figure 7: CMOS setup program
Table 5: BIOS Configuration Jumper Settings
Function
Jumper Configuration
Normal
1-2
Normal operation and changing standard BIOS options
Configure
2-3
After the POST runs, Setup runs automatically. In this mode, you can access the
maintenance menu, modify passwords, configure memory latencies and IRQs.
Recovery
none
The BIOS attempts to recover the BIOS configuration. A recovery diskette is
required. Note: SPD system memory must be used to complete a BIOS recovery.
BIOS Recovery:
If you are sure that you have integrated the PC components correctly but the system is
not booting properly, you cannot enter the BIOS/CMOS Setup program, or reflash the
BIOS, then you may need to perform a BIOS recovery. To do this for Intel
motherboards:
1.
2.
3.
4.
5.
6.
Place the latest BIOS revision on a bootable floppy disk
Make sure the system power is off
Ensure that you have known-good SPD PC100 or PC133 memory installed
Put the BIOS disk in the floppy disk drive
Remove the BIOS configuration jumper
Turn on the PC
Intel Customer Support (ICS)
Troubleshooting the IA Platform
24
7. Wait several minutes while the BIOS is reinstalled into the desktop board flash
memory. There will not be any video during the BIOS recovery.
8. After you hear two beeps and the floppy disk drive light turns off, remove the
BIOS disk.
9. Turn off the PC
10. Replace the BIOS configuration jumper
11. Turn the PC back on
To update a BIOS
Before Upgrading to the Latest BIOS, it is very important that you enter the “BIOS Setup
Utility” and write down all of your current CMOS settings. For Intel boards, you can also
go to the Exit menu and choose the “Save Custom Defaults” options. Choosing ‘Yes’
will cause the current BIOS settings to be saved as a custom default setting. When the
CMOS memory is cleared during the BIOS upgrade, the BIOS will default to these
custom default settings. If you do not choose to save the defaults as custom defaults,
you will need to manually reenter these settings after you have upgraded to the latest
BIOS.
For Intel boards, the easiest way to update the BIOS is to use the Intel Express BIOS
Update program specific to the motherboard that you have. By executing the
downloaded file within Windows* 98, Windows 98SE, Windows ME, Windows NT*, or
Windows 2000, the executable file will runs a Windows InstallShield* program, which
automatically updates the BIOS.
Installation procedure:
1. Go to this Intel Web site: http://developer.intel.com/design/motherbd/
2. Click the Express BIOS Update utility for the board's BIOS you want to update.
3. Either download the file to your hard drive or select Run from Current Location.
(If you opt to Run from Current Location, skip step 5). Note: Make a note of the
path the file is saved to on your hard drive. In addition, You can save this utility
onto a floppy disk. This is useful if you're updating the BIOS for multiple identical
systems.
4. Close all other applications. Note: This is a required step. Your system will be
rebooted at the last Express BIOS Update window.
5. Double-click the executable file from the location on your hard drive where it was
saved. This runs the update program.
6. Follow the instructions provided in the dialog boxes to complete the BIOS
update.
Intel also provides a DOS based Iflash utility to update your BIOS on all other operating
systems. For more information about the Iflash utility, go to
http://developer.intel.com/design/motherbd/standardbios.htm
Intel Customer Support (ICS)
Troubleshooting the IA Platform
25
4.3 Common BIOS setup option
For the most point, it will be pointless to cover every single BIOS settings. Therefore, I
will not cover common well known BIOS setup options in this manual. Instead, this
manual will discuss options that are important in troubleshooting a system. Instead of
covering each of the BIOS parameters in this section, I will cover a couple of options
(see table 6) that don’t fit elsewhere in the manual or that are important to be aware of
now and cover other parameters throughout the rest of the manual as needed.
Table 6: A couple of BIOS Setup Options
Options: Enabled, Disabled
Flash BIOS
This function protects the BIOS from accidental corruption by
Protection
unauthorized users or computer viruses. When enabled, the BIOS’
data cannot be changed when attempting to update the BIOS with a
Flash utility. To successfully update the BIOS, you'll need to disable
this Flash BIOS Protection function. You should enable this function
at all times. The only time when you need to disable it is when you
want to update the BIOS. After updating the BIOS, you should
immediately re-enable it to protect it against viruses.
Hardware Reset Options: Enabled, Disabled
This function is useful for file servers and routers, etc., which need
Protect
to be running 24 hours a day. When enabled, the system’s hardware
reset button will not function. This prevents the possibility of any
accidental resets. When set as Disabled, the reset button will
function as normal. It is recommended that you leave it as Disabled
unless you are running a server and you have kids who just love to
press that little red button running around.
This setting is shown on only some BIOSes. If present, enabling this
Plug and Play
tells the BIOS that you are using an operating system that supports
Aware OS
the Plug and Play specification (such as Windows 95). When
enabled, the BIOS will look for and initialize any Plug and Play cards
in the system. Enable the setting if using Windows 95 or another
Plug and Play compatible OS. The default is normally "Disabled".
Virus Warning/
Anti-Virus
Protection
Note: Some BIOSes will perform the initialization of Plug and Play
cards automatically regardless of the operating system being used,
and will thus not have this setting. Some will work fine with Plug and
Play regardless of how this option is set. However, on some
systems that have this setting, Plug and Play may not function
properly if the setting is disabled.
Options : Enabled, Disabled, ChipAway
When Virus Warning is enabled, the BIOS will flash a warning
message whenever there's an attempt to access the boot sector or
the partition table. You should leave this feature enabled if possible.
Note that this only protects the boot sector and the partition table,
Intel Customer Support (ICS)
Troubleshooting the IA Platform
26
not the entire hard disk. However, this feature will cause problems
with the installation of certain software. One good example is the
installation routine of Win95/98. When enabled, this feature will
cause Win95/98's installation routine to fail. Disable it before
installing such software.
Also, many disk diagnostic utilities that access the boot sector can
trigger the error message as well. You should first disable this option
before using such utilities.
Some motherboards will have their own rule-based anti-virus code
(ChipAway) incorporated into the BIOS. Enabling it will provide
additional anti-virus protection for the system as it will be able to
detect boot viruses before they have a chance to infect the
boot sector of the hard disk. Again, this is useless if the hard disk is
on a separate controller with its own BIOS.
4.4 Intel Chipset Software Installation Utility
The Intel® Chipset Software Installation Utility was developed for updating Windows
Operating Systems with the Plug and Play feature on Intel chipsets. This utility allows
the operating system to correctly identify the Intel chipset components and properly
configure the system. The Intel Chipset Software Installation Utility installs specific
Windows INF files. The INF files inform the operating system how to properly configure
the chipset for specific functionality, such as AGP, USB, Core PCI, and ISA PnP
services. Since these components are necessary for optimum stability and
performance, the Intel Chipset Software Installation Utility should be the first driver
loaded after the Operating System is installed. The Intel Chipset Software Installation
Utility is not required or compatible with Operating Systems without the Plug and Play
feature, such as Windows NT* 4.0.
The Intel Chipset Software Installation Utility for Intel chipsets can be downloaded from
the following Intel web site:
http://developer.intel.com/design/software/drivers/platforms/inf.htm
After installing the Intel Chipset Software Installation Utility, Device Manager should
indicate that your system is now using the Intel Chipset Devices. An example of an Intel
850 Chipset is shown in figure 8. Note: After the Intel® Chipset Software Installation
Utility is installed, there should not be any yellow exclamation points ( ! ) associated with
any of the Intel chipset devices.
For more information about the Intel Chipset Software Installation Utility:
http://developer.intel.com/design/software/drivers/platform/inf.htm
To find the Intel Chipset Software Installation Utilities User’s Manual:
http://support.intel.com/support/chipsets/chip_man.htm
Intel Customer Support (ICS)
Troubleshooting the IA Platform
27
Note: The User’s Manual contains a table that indicates if the Intel Chipset Software
Installation Utility is needed for a specific chipset and a specific version of Windows.
Figure 8: Before you load the Intel Chipset Software Installation utility (INF), you have some
unknown devices and after they are identified.
5 RAM
The RAM is considered the main memory or the primary memory because RAM is the
memory that the microprocessor accesses directly. This means the program instruction
that the processor is executing must be first copied to the RAM from a device such as a
disk. In addition, the data, such as letters, reports and charts, which are generated by
programs, must be in the RAM to be manipulated. Therefore, it is the amount of RAM,
which determines how many programs can be executed at one time and how much data
can be available to a program. In addition, the amount of memory is a major factor in
determining how fast your computer will operate.
5.1 Technologies (SDRAM, DDR, RDRAM)
The most common RAM chips are Dynamic RAM (DRAM) chips, which use a storage
cell consisting of a tiny solid-state capacitor and a MOS transistor. A capacitor is a
simple electrical device, similar to a battery, which is capable of storing a charge of
electrons. The charge or lack of charge represents a single bit of data. If it is charged, it
has a logic state of 1. If it is discharged, it has a logic state of 0. The storage cells are
organized into a large two-dimensional array or table made up of rows and columns.
Intel Customer Support (ICS)
Troubleshooting the IA Platform
28
5.1.1 SDRAM
Synchronous DRAM or SDRAM differs from earlier types in that it does not run
asynchronously to the system clock the way older, conventional types of memory do.
SDRAM is tied to the system clock and is designed to be able to read or write from
memory in burst mode (after the initial read or write latency) at 1 clock cycle per access
(zero wait states) at memory bus speeds up to 133 MHz. SDRAM supports 5-1-1-1
system timing when used with a supporting chipset. The 5-1-1-1 means that it takes the
RAM five clock cycles to access the first piece of data in a row. After that row is
activated, it only takes one clock cycles each for the next three pieces of data contained
within the same row.
SDRAMs are still DRAMs, and therefore still have latency. The fast 12, 10 and 8
nanosecond numbers that everyone talks about refer only to the second, third and
fourth accesses in a four-access burst. The first access is still a relatively slow 5 cycles,
just as it is for conventional EDO (Extended Data Out) and FPM (Fast Page Memory)
memory.
SDRAM modules are generally speed-rated in two different ways: First, they have a
“nanosecond” rating such as 8, 10 or 12 nanonseconds. A lower number indicates a
faster speed. The second way is the MHz rating such as 100 MHz or 133 MHz. Second,
they have a "MHz" rating, so they are called “83 MHz” or “100 MHz” SDRAMs for
example. Because SDRAMs are, well, synchronous, they must be fast enough for the
system in which they are being used. With SDRAM however, the whole point of the
technology is to be able to run with zero wait states. In order to do this, the memory
must be fast enough for the bus speed of the system. One place where people run into
trouble in this regard is that they take the reciprocal of the "nanosecond" rating of the
module and conclude that the module can run at that speed. For example, the
reciprocal of 10 ns is 100 MHz, so people assume that 10 ns modules will definitely be
able to run on a 100 MHz system. The problem is that this allows absolutely no room for
slack. In practice, you really want memory rated slightly higher than what is required, so
10 ns modules are really intended for 83 MHz operation. 100 MHz systems require
faster memory, which is why the PC100 specification was developed. Of course, PC66
is designed to run on a 66 MHz bus, PC100 SDRAM is designed to run on a 100 MHz
bus and PC133 SDRAM is designed to run on a 133 MHz bus.
Most motherboards are now being created that requires the use of special SDRAM
modules that include a technology called a Serial Presence Detect (SPD). This is an
EEPROM that contains speed and design information about the module. The
motherboard queries the chip for information about the module and makes adjustments
to system operation based on what it finds.
As memory components get faster, chips become more dense and more circuits get
squeezed onto smaller boards. Dissipation of excess heat becomes more of an issue.
Newer memory module designs use heat sinks or heat spreaders to maintain safe
operating temperatures.
Intel Customer Support (ICS)
Troubleshooting the IA Platform
29
5.1.2 DDR SDRAM
DDR SDRAM is similar in function to regular SDRAM, but doubles the bandwidth of the
memory by transferring data twice per cycle--on both the rising and falling edges of the
clock signal. The clock signal transitions from "0" to "1" and back to "0" each cycle; the
first is called the "rising edge" and the second the "falling edge".
There are two current forms of DDR system memory running around, each with a
variety of interesting names. The first is DDR1600 or DDR200, which runs at 100MHz
DDR for an equivalent of 200MHz of bandwidth. On a 64-bit wide memory bus, the
memory bandwidth is 1,600 MBps, hence the name DDR1600. The common one used
today is DDR2100/DDR266, which runs on a 133MHz giving an equivalent of 266MHz
of bandwidth. On a 64-bit wide memory bus, the theoretical peak is about 2,100 MBps
of memory bandwidth, hence the name DDR2100.
5.1.3 RDRAM
One of the two main competing standards to replace SDRAM is called Direct Rambus
DRAM or DRDRAM (formerly called “Rambus DRAM” or “RDRAM”). DRDRAM works
more like an internal bus than a conventional memory subsystem. It is based around
what is called the Direct Rambus Channel, a high-speed 16-bit bus running at a clock
rate of 400 MHz. As with DDR SDRAM, transfers are accomplished on the rising and
falling edges of the clock, yielding an effective theoretical bandwidth of approximately
1.6 Gbytes/second. This is an entirely different approach to the SDRAM and DDR
SDRAM, which use a wide 64-bit memory bus. It may seem counterproductive to
narrow the channel since that reduces bandwidth, however the channel is then capable
of running at much higher speeds than would be possible if the bus were wide. As with
SDRAM, DRDRAM makes use of a serial presence detect (SPD) chip to tell the
motherboard certain characteristics of the DRDRAM module when the system is
booted.
The Rambus Inline Memory Module (RIMM) conforms to the standard DIMM form
factor, but it is not pin-compatible. Its architecture is based on the electrical
requirements of the Direct Rambus Channel. Since the RAM runs at a faster speed than
the DIMMs, RIMMs typically include a heat spreader to help dissipate heat faster. The
motherboards support RIMMs will support two or more RIMMs. Since the RIMMs fit into
a channel, low-cost continuity modules must be used to maintain channel integrity in
systems that have RIMM modules in every memory slot or socket.
For example, for the Intel D850GB Desktop board, all RIMM sockets must be populated
to achieve continuity for termination at the Rambus interface. Continuity RIMMs (or
“pass-through” modules) must be installed in the second RDRAM bank if memory is not
installed. If any of the RIMM sockets are not populated, the computer will not complete
the Power-On Self-Test (POST) and the BIOS beep codes will not be heard. In addition,
the Bank 0 (labeled on the board as RIMM1 and RIMM2) must be populated first
Intel Customer Support (ICS)
Troubleshooting the IA Platform
30
ensuring that the RDRAM installed in RIMM1 and RIMM2 is identical in speed, size, and
density. If the desired memory configuration has been achieved by populating Bank 0,
then Bank 1 should be filled with two Continuity RIMMs. If memory is to be installed in
Bank 1, the RIMM modules installed in RIMM3 and RIMM4 must be identical in size and
density to each other, and match the speed of the RIMM modules in Bank 0. The RIMM
modules do not, however, need to match those in Bank 0 in size and density. If ECC
functionality is required, all installed RIMM modules must be ECC-compliant. Some
motherboards might have Bank 0 set as RIMM1 and RIMM3 and Bank 1 as RIMM2 and
RIMM4. Therefore, you need to check your motherboard documentation.
5.2 Troubleshooting RAM Problems
The following are common ways for the computer to tell you it is having problems with
memory:
1. The computer won’t boot, merely beeps.
2. The computer boots but doesn’t recognize all the installed memory.
3. The computer boots but the screen is blank.
4. The computer reports a memory error.
5. Memory mismatch error
6. Memory parity interrupt at xxxxx
7. Memory address error at xxxxx
8. Memory failure at xxxxx, read xxxxx, expecting xxxxx
9. Memory verify error at xxxxx
10. The computer has other problems caused by memory.
11. The computer intermittently reports errors, crashes frequently, or spontaneously
reboots.
12. Registry Errors
13. General-protection faults, page faults, and exception errors
14. The server system manager reports a memory error
15. Serial Presence Detect (SPD) not detected
When you have a problem with memory, the cause is usually one of three things:
• Improper Configuration including having the wrong part for your computer or did
not follow the configuration rules.
• Incompatibilities with memory modules from different manufacturers, different part
numbers or different speeds.
• Improper Installation including the memory may not be seated correctly may need
cleaning.
• Defective Hardware including a faulty motherboard or memory module.
• For RDRAM systems, all memory slots or sockets for a channel do not contain a
RIMM or a Continuity RIMMs or the RIMM or Continuity RIMM is not installed
properly. Note: You will have to check the motherboard documentation to
determine which memory slots or sockets are assigned to which channels.
Intel Customer Support (ICS)
Troubleshooting the IA Platform
31
The fact that many computer problems manifest themselves as memory
problems makes troubleshooting difficult. For example, a problem with the
motherboard or software may produce a memory error message.
The following basic steps apply to almost all situations:
• Make sure you have the right memory part for your computer. At the
manufacturer’s Web site you can look up the part number. Many memory
manufacturers have configurators, which indicate the compatibilities of your
module. If not, phone the memory manufacturer, consult your computer manual,
or phone the computer manufacturer. Note: On the Intel Support website, Intel
list the tested memory list for the Intel motherboards.
• Confirm that you configured the memory correctly. Many computers require
module installation in banks of equal-capacity modules. Some computers require
the highest capacity module to be in the lowest labeled bank. Other computers
require that all sockets be filled; still others require single-banked memory.
• Re-install the module. Push the module firmly into the socket. In most cases
you hear a click when the module is in position. To make sure you have a
module all the way in the socket, compare the height of the module to the height
of other modules in neighboring sockets.
• Swap modules. Remove the new memory and see whether the problem
disappears. Remove the old memory, reinstall the new, and see whether the
problem persists. Try the memory in different sockets. Swapping reveals whether
the problem is a particular memory module or socket, or whether two types of
memory aren’t compatible.
• Clean the socket and pins on the memory module. Use a soft cloth to wipe the
pins on the module. Use a PC vacuum or compressed air to blow dust off the
socket. Do NOT use solvent, which may corrode the metal or prevent the leads
from making full contact. Flux Off* is a cleaner used specifically for contacts. You
can purchase it at electronics or computer equipment stores.
• Update the BIOS. Computer manufacturers update BIOS information frequently
and post revisions on their Web sites. Make sure you have the most recent BIOS
for your computer. This applies especially when you have recently installed new
software or you are significantly upgrading memory.
• Check the BIOS Settings.
6 POWER SUPPLIES AND CHASSIS
The chassis (also sometimes referred to as the case) of the PC is a large metal or
plastic box, which is designed to hold and protect the motherboard, the drives and the
power supply. Like motherboards and microprocessors, cases come in many different
configurations. The cases can be characterized by the orientation of the box, how many
drives it can hold, the number of the expansion slots and size of the expansion cards.
Typically, when you buy a case, the case includes the power supply. The primary
function of the PC power supply is to convert AC power to clean DC power (+3.3 volts,
±5 volts and ±12 volts). The secondary function of the power supply is to provide
cooling to the system using a fan.
6.1 Power Supplies and Power
Today, PCs will use either an ATX or ATX12V power supply. It contains software control
of the power on/off signal so that it can shut down the system. Since the ATX/ATX12V
power supplies are software activated/deactivated, you need to connect the “Power
SW” cable from the chassis to the motherboard. If not, you will have to temporary short
the two “Power SW” pins on the motherboard. In addition, if you decide to turn on a
power supply. Most power supplies require to have a load connected to the power
supply. In other words, you must have at least one component such as a drive or
motherboard connected to the power supply.
The ATX12V power supply provides increased 12 V, 3.3 V, and 5 V current and
provides additional cooling capability. An ATX12V power supply can be easily identified
by the addition of an additional new 2x2 pin connector and an optional 1X6 pin
connector. For the Pentium 4 and Intel Xeon processors, you should use an ATX12V
power supply. See figure 9.
Note: System damage and poor system operation may result if your motherboard is not
using the correct power supply. Consult the motherboard documentation to determine
power supply requirements.
As already stated, most power supplies designed to be used in the United States
operate at 120 volts with a frequency of 60 Hz. In other nations, the supply voltage and
frequency may be different. In Europe, you will find 230 volt with a 50 Hz frequency as
the standard.
Today, most PC power supplies are willing to operate at either voltage. Some can
automatically switch over to the proper voltage while most are done by using a small
switch on the rear of the power supply. Ensure when plugging in your PC and turning it
on, the correct voltage is selected. If you have a power supply switched over to 230 V
and the voltage is 120 V, the PC will not boot up. Unfortunately, if the power supply is
set to 120 V and it is connected to a 230 V outlet, it will seriously damage your power
supply and other important components.
Intel Customer Support (ICS)
Troubleshooting the IA Platform
34
Figure 9: ATX/ATX12V Pinouts
6.2 On/Off Switch
Different from the AT power supply, the ATX power supply uses the PS_ON signal to
power up the system. A +5 volt signal is constantly sent through pin 14 (PS_ON) of the
ATX power connector. When the PS_ON is shorted tells the power supply to turn on
and start the boot process. A push button contact switch is connected to two pins on the
motherboard that link to the PS_ON signal to ground. When the push button is pushed,
it connects the PS_ON signal to ground. When the push button is pushed, it connects
the PS_ON signal with ground, shorting it out. Therefore, when you are installing an
ATX motherboard, you need to connect the push button wires (usually labeled PWR
SW) to the motherboard. If you decide to test a motherboard without physically installing
it into an ATX case, you can start the system by either connecting a push button switch
to the motherboard and pressing the button or by taking a small screw driver and make
contact with the two pins that make up the power switch connector. Since the switch
only toggles the on/off status, the switch carries only +5 V of DC power, rather than the
full 110 V AC current used in the T power supplies.
6.3 Shorts and Overloads
Besides supplying the power to the PC components, the power supply also provides the
power-good signal. During boot up, the processor tells the computer to constantly reset.
As soon as the power supply performs a self-test, testing if all voltage and current levels
are acceptable, the power supply will send a power good signal (+5 volts) to the
microprocessor. When the power good signal is sent, the computer will finish the boot
process.
If the power supply detects a short or overload, the power supply will stop sending the
power good signal and the system will reboot. To check for shorts and overloads, you
Intel Customer Support (ICS)
Troubleshooting the IA Platform
35
need to use isolation. First take out all of the expansion cards except the video card and
the floppy drive/hard drive controller card and disconnect any drives except your floppy
disk drive and your primary hard drive. If the system powers on with the minimum
devices, one of the components which you removed or disconnected is causing a short
or overload or all of the components together is too much for the power supply. To find
out which one is causing the problem, put one expansion card or connect one drive at a
time and turn on the system to find out if that device causes the power supply to go into
idle mode.
If it still does not work after removing all of the extra devices, it must be the
motherboard/RAM, video card, floppy drive/hard drive controller card, floppy drive or the
hard drive, which is causing the problem. In this case, you must then replace one device
at a time until you find out which one is causing the problem.
If you have the computer working when it is lying on its side but does not work when it is
standing erect, or vise versa, usually indicates a short. Check to see if any foreign
objects fell into the chassis such as paper clips or screws. In addition, make sure that
you have the correct type of spacers located in the proper place (refer to your
motherboard documentation) and that the screws that hold the motherboard in place are
not too tight.
6.4 AC Power Outlet
The AC power outlet has three wires leading to three connectors. A power plug
connects to the AC power outlet and will have two blades and a ground. The smaller
blade is the hot wire while the other blade is the return. The wire leading to the slot used
for the small blade is usually black while the wire leading to the slot used by the large
blade is usually white. The ground is usually green.
To test a power outlet, the voltage between the two slots should be between 110 and
125 volts AC. The voltage between the smaller slot and ground should be the same
voltage as measured between the two slots. Lastly, the voltage between the larger
blade and the ground should be .5 volts AC or less. See figure 10. Note: an inexpensive
outlet wire tester can be bought at any hardware store.
<.5 V
Same as
110 -125 Volts AC
Figure 10: The correct voltage measurements when checking an AC outlet.
Intel Customer Support (ICS)
Troubleshooting the IA Platform
36
It is important to have and use the ground connector. It makes electronic equipment
(including computers) safer to work with, prevents continuous electrical shock and will
minimize errors on the PC.
6.5 Power Irregularities
When you turn on your PC, you expect the power to be there. Unfortunately, the power
that you get from the power company is not always 120-volts AC. The voltage level may
drop or increase. While the power supply can handle many of these power fluctuations,
other power fluctuations may shut down or damage your computer, corrupt your data
and/or loose any unsaved work.
Studies have shown a typical computer is subject to more than 120 power problems per
month. The most common of these are voltage sags. Obvious power problems such as
blackouts and lightning make up only 12% of the power problems. American Power
Conversion states that data loss caused by power problems occur 45.3% of the time
making it the largest cause. Symptoms of bad power can cause frozen
computers/keyboards, errors in data transmissions, corrupt or lost data, frequently
aborted modem transfers and total failure of a computer or computer component. To
protect the computer from over voltages and under voltages, several devices can be
used. They are surge protectors, line conditioners, standby power supplies and
uninterruptible power supplies.
When troubleshooting the PC, you should always be concerned with power related
problems. As you have seen, the power problems may be caused by the power supply,
bad power provided by the AC power outlet or a PC component, which causes a short
or overload. Power problems can cause errors and may cause failure in RAM chips,
microprocessors and hard drives. Power related problems may be harder to diagnose
because they are sometimes not obviously related to the power and they are often
intermittent.
6.6 EMI
Protecting the PC components is not the only function of the case. It also limits the flow
of RFI radiation from your computer and into your computer. Since some of the radiation
is broadcast as radio and television signals, an electronic component can interrupt a
radio, a television set or the navigation equipment used in an aircraft while RFI that
enters your system can cause data corruption, lockups and reboots.
To minimize the RFI radiation, computer manufacturers use several techniques. A steel
case does a fair job of limiting RFI. If the computer uses a plastic case, the case needs
to be treated with conductive paint made with silver to minimize radiation. As the
frequencies of the PC increase, so does the radiation. Radio signals can leak out
through any cracks in the case or if parts of the chassis are not electrically connected to
the case. Lastly, cables attached to the computer can act as an antenna, sending out
radio signals. To reduce the emissions, special metal fingers on the edge of the case
Intel Customer Support (ICS)
Troubleshooting the IA Platform
37
and its lid ensure that the two pieces have good electrical contact and the cables are
shielded.
6.7 Thermal Management
To provide a complete thermal solution, you must use a chassis that provides sufficient
airflow to keep the processor, RAM and other components under its maximum operating
temperature in the warmest user environment. Note: This may require extra fans in
large tower cases. Running these components above their maximum temperature
specification may void the warranty and can lead to functional and performance
degradation. The combination of a good fan heatsink, an ATX or microATX form-factor
motherboard, and an ATX or microATX-compliant chassis can be a good thermal
management solution. If you use Baby AT or 1U chassis, be aware that the airflow
varies significantly in the different chassis, depending on venting, internal brackets, and
other factors. Chassis with low airflow can cause components to exceed their maximum
thermal specification.
7 EXPANSION CARDS AND SLOTS
The expansion slot, also known as the I/O bus, extends the reach of the processor so
that it can communicate with peripheral devices. They are called expansion slots
because they allow you to expand the system by inserting circuit boards called
expansion cards. They are essential for the computer because a basic system cannot
satisfy everyone’s needs. In addition, the expansion cards allow the system to use new
technology as it becomes available.
By 1987, the Institute of Electrical and Electronic Engineers (IEEE) committee approved
the AT bus as the Industry Standard Architecture (ISA). Although the ISA slot is
considered legacy technology, even the most advanced motherboards contain some
ISA slots. Other names include the classic bus, the AT bus or Legacy bus.
The ISA bus has not had a significant change until 1993 when Intel and Microsoft*
introduced the Plug and Play (PNP) ISA. The PNP ISA bus allows a PNP ISA card to
be inserted and its resources automatically assigned without any need to use jumpers
or DIP switches to configure the card. Note: ISA cards, which are not plug-and-play,
are referred to as Legacy cards.
Instead of physically changing the ISA slot, a plug and play manager (software)
identifies the card and assigns its needed resources. This is made possible because the
PNP cards contain a ROM chip, which contains a number. Since the number identifies
the card, not its model number, several of the same cards can be used within the same
system, such as two network cards or two hard drive controllers. After the card is
identified, the Plug-and-Play manager assigns the needed resources (I/O and memory
addresses, the IRQs and the DMAs) to the card.
Intel Customer Support (ICS)
Troubleshooting the IA Platform
38
Today, when you hear the term local bus, it probably refers to Peripheral Component
Interconnect (PCI) local bus. The PCI slot was developed by Intel and was designed to
eventually replace the older bus designs. The PCI slot is mounted on the system board
offset from the normal ISA, MCA or EISA connector. The PCI design provides between
3 to 5 slots.
PCI supports plug and play technology. Different from other expansion bus designs, the
expansion cards have their own storage space to store the cards configuration
information. Therefore, most PCI cards are not configured using jumpers or DIP
switches, but are configured using software or a true plug and play system.
When an expansion card is inserted into a system, it must be configured to use the
proper resources. The resources include I/O addresses (including COMx/LTPx), IRQs,
DMAs and Memory addresses. When configuring a card, one general rule should apply.
No two devices can use the same resource. Therefore, two expansion cards should not
be set to use the same DMA channel or the same I/O address. If two devices are using
the same setting, the devices will not work properly or not work at all. To determine
what the available resources are, use software diagnostic software such as Dr.
Hardware*, or Checkit Pro* or use utilities, which come with the operating system such
as Windows Device Manager located within the control panel. Note: It is possible to
share an IRQ among more than one device, but only under limited conditions. If you
have two devices that are seldom used, and that you never use them simultaneously,
you may be able to have them share an IRQ. Unfortunately, this can lead to more
problems if you are not too careful.
After determining which resources the card can use, the card itself can be configured
one of several ways:
1. DIP switches and/or Jumpers
2. Electronically using a software setup program
3. Using Plug and Play (PnP)
Normally, the BIOS assigns unique IRQs to PCI devices. If your system supports PCI
IRQ Steering, and it is enabled, the BIOS will assign the IRQs to the PCI devices and
Windows will typically only change the IRQ settings if it detects a conflict. If there are
free IRQs to go around, IRQ Steering allows Windows to assign multiple PCI devices to
a single IRQ, thus enabling all the devices in the system to function properly. Without
IRQ Steering, Windows will begin to disable devices once it runs out of free IRQs to
assign.
In Windows 95 (OSR2), Windows 98 and Windows Me, to determine if your system is
using IRQ Steering, you can access the Windows’ Device Manager, double-click the
PCI Bus, click the IRQ Steering tab and look at the check box for IRQ Steering. See
figure 10.
Intel Customer Support (ICS)
Troubleshooting the IA Platform
39
In Windows 2000, some or all of the devices on your ACPI motherboard may be listed
on the Resources tab in Device Manager as
using the same IRQ (IRQ 9). You cannot
change the IRQ setting because the setting is
unavailable. This occurs because Windows
2000 takes advantage of the ACPI features of
the motherboard, including advanced PCI
sharing. IRQ 9 is used by the PCI bus for IRQ
steering. This feature lets you add more
devices without generating IRQ conflicts.
The Plug and Play operating system settings
in the computer's BIOS should not affect how
Windows 2000 handles the hardware in
general. However, Microsoft recommends
that you set this setting to "No" or "Disabled"
in the computer's BIOS. In addition, manually
assigning IRQs to PCI slots in the System
BIOS as a troubleshooting method may work
on some non-ACPI systems when using a
standard PC HAL, but these settings are
ignored by Plug and Play in Windows 2000 if
Figure 10: Enabling IRQ Steering in
ACPI support is enabled. If you need to
Windows
manually assign IRQ addresses through the
BIOS to a device on an ACPI motherboard, you will have to disable ACPI support by
installing the standard PC HAL, which usually means you will have to reinstall the
operating system.
Note: after installing the expansion card, many expansion cards require software drivers
to function. Windows drivers are typically loaded by running the Add New Hardware
Driver icon or by using the installation disks that come with the expansion card. Lastly,
make sure that you loaded the chipset software that comes with the motherboard
immediately after the operating system and any service packs. See section 4.4. Lastly,
if you install an expansion card such as a sound card that duplicates a feature that is
built into a motherboard, you must make sure that the motherboard device is disabled.
8 DISK DRIVES
Disk drives are half electronic/half mechanical devices, which use magnetic patterns to
store information onto rotating platters. They are considered long-term storage device
because they do not forget their information when power is disconnected. They are also
considered the primary mass storage system because they hold all of your programs
and data files, which are fed into RAM. Disk drives communicate with the rest of the
computer with a cable connected to a controller card. The controller card could be an
Intel Customer Support (ICS)
Troubleshooting the IA Platform
40
expansion card or could be built on the motherboard. Today, hard drives can be 50 GB
or larger.
8.1 Floppy Disk Drives
The drive and controller (Almost always built into the motherboard) are connected
together with a gray 34-pin ribbon cable. When connecting the gray ribbon cable, the
cable must be connected properly. Pin 1 of the cable is designated with a red or blue
stripe. The controller card and drive will use either a small number 1 or 2 to designate
which end has pin 1 or use a square solder (other pins use a round solder). Note: Most
floppy drives will have pin 1 next to the drive’s power connector. Since most computers
only have one drive connected to the floppy disk drive, Intel motherboards only have a
cable with two connectors, one that connects to the motherboard and one that connects
to the 3 ½ inch connector. The one end of the cable with pin 6 filled in connects to the
motherboard while the end with all 34 holes connects to the floppy disk drive.
8.2 IDE Hard Drives
Today, there are two main types of hard drives, IDE and SCSI. The IDE (integrated
drive electronics) interface The IDE hard drive connects to a controller card
(expansion card or built-into the motherboard) using a 40-pin cable. Note: the Ultra
DMA-66 and Ultra DMA-100 mode requires a special 40-pin cable that has 80 wires, 40
pins for the signal and 40 ground pins to reduce noise. This type of cable is referred to
as a 40-pin, 80 conductor. Today, the IDE interface supports up to four IDE devices by
using two IDE cables. In addition to hard drives, it supports non-disk peripherals, which
follow the ATAPI (ATA Packet Interface). The ATAPI is the protocol used on the
Enhanced IDE devices (IDE CD-ROM drives, IDE tape backup drives and other IDE
storage devices).
When connecting the ribbon cable, you must make sure the cable has the correct
orientation. Pin 1 of the cable (marked with a red or blue stripe along one edge) must be
connected to pin 1 of the drive and pin 1 of the cable connector on the motherboard or
expansion card. Pin 1 of the drive is designated by a small 1 or 2 printed on the drives
circuit board. Pin 1 of the expansion card/motherboard connector can be identified with
a small number 1 or 2 printed on the circuit board or will use a square solder (other pins
use a round solder). Some drives and cables will have a notch, which will prevent the
cable from being inserted incorrectly. If the cable is connected incorrectly, the hard drive
may cause a short causing the computer not to boot.
Up to two hard drives can be connected to one IDE cable. If one drive is installed, the
drive is known as a stand-alone drive. If two hard drives are installed, the first physical
drive is known as the master drive (drive C) while the second physical drive is known as
the slave drive (drive D). The master drive got its name because the controlling
electronics on the logic board of the master drive control both the master drive and the
slave drive. The stand-alone, master, and slave drives are determined with jumpers on
the hard drive. Note: it does not matter where the drive and controller card is connected
Intel Customer Support (ICS)
Troubleshooting the IA Platform
41
to the cable. The most common jumpers used to determine a drive as a stand-alone,
master and slave device are the Master (M/S) jumper and the Slave Present (SP).
Different from older IDE drives, the connectors on the cable define which connector goes with
which device. The blue connector attaches to the controller. The gray connector is in the middle
of the cable and goes to any slave drive (if present). The black connector is at the opposite end
from the controller connector and goes to the master drive, or a single drive if only one is used.
This was done because if you connect a single drive to the middle connector of the ribbon cable,
the unconnected cable can cause signaling problems.
The two methods in transferring data are Processor Input/Output (PIO) and Direct
Memory Access (DMA). Depending on the transfer mode, hard drives can transfer rate
up to 100 Megabytes/sec. PIO mode has the transfers directly controlled by the
processor while the Direct Memory Access drives use a form of bus mastering. See
table 8 and 9.
Table 8: PIO modes used with EIDE devices
PIO
Average
Flow
Specification
Mode
Transfer Rate Controlled
3.3 MB/s
No
ATA
0
5.2 MB/s
No
ATA
1
8.3 MB/s
No
ATA
2
11.1 MB/s
Yes
ATA-2
3
16.6 MB/s
Yes
ATA-2
4
Table 9: DMA Mode used with EIDE devices
DMA Mode
Average
Requirements
Transfer Rate
2.08 MB/s
0 – Single Word
4.16 MB/s
1 – Single Word
8.33 MB/s
2 – Single Word
4.16 MB/s
Local-bus controller
0 – Multi-Word
13.33 MB/s
Local-bus controller
1 – Multi-Word
16.6 MB/s
Local-bus controller
2 – Multi-Word
33.3 MB/s
Local-bus controller
3 – Multi-Word
66.6 MB/s
Local-bus controller and 404 – Multi-Word
pin/80 conductor cable
100.0 MB/s
Local-bus controller and 405 – Multi-Word
pin/80 conductor cable
Standard
ATA
ATA
ATA
ATA
ATA-2
ATA-3
Ultra DMA-33
Ultra DMA-66
Ultra DMA-100
After the hard drive is configured and physically installed, the next step is to enter the
CMOS Setup program and input the proper hard drive parameters for the installed hard
drive (such as the number of cylinders, the number of read/write heads and the number
of sectors/track or between various modes such as Large or LBA).
Intel Customer Support (ICS)
Troubleshooting the IA Platform
42
8.3 Intel® Ultra ATA Storage Driver
The Intel Ultra ATA Storage Driver is a driver that is designed to take advantage of
devices supporting ATA DMA/Ultra DMA technologies and improve system
performance. It is not a requirement for your operating system to work properly.
Currently, the Intel Ultra ATA Storage Driver supports Intel 810, 815, 820, 840, 850 and
860 chipsets. In addition, you must have either Windows* 98, Windows 98 Second
Edition (SE), Windows Millennium Edition (Me), Windows NT* 4.0, or Windows* 2000.
To install the Intel Ultra ATA Storage Driver:
1. Install the Intel® Chipset Software Installation Utility before installing the Intel
Ultra ATA Storage Driver so that the Intel chipset is properly recognized by the
operating system.
2. Download the Intel® Ultra ATA Storage Driver from the Intel website
(http://support.intel.com/support/chipsets/storagedrivers/ultraATA/incon.htm).
Save the file to a known location on your computer's hard drive
3. The installation program will automatically launch from this location
4. Click "Next" at the Welcome screen
5. Read the license agreement and click "Yes" to continue
6. Click "Next" to install the driver in the default folder location
7. Click "Next" to create the default Program Folder
8. The driver files will now be installed. When finished installing, choose the "Yes"
to reboot option and click "Finish" to restart your computer. The driver should
now be loaded
8.4 SCSI Hard Drives
Another very popular hard drive and system interface is the small computer system
interface (SCSI), which is ideal for high-end computers, including network servers. It is
cable of support multiple devices in and out of the computer. SCSI interfaces that have
an 8-bit data bus, referred to as narrow, allow up to seven devices (hard drives, tape
drives, CD-ROM drives, removable drives/disks and scanners) connected to one SCSI
adapter/controller. It would be eight devices if you count the controller card. SCSI
interfaces with 16-bit data bus, referred to as wide, interfaces allow up to 15 devices (16
devices including the controller card). See Table 10.
Table 10: Types of SCSI devices and interfaces
Transfer Mode
"Regular" SCSI (SCSI-1)
Wide SCSI
SCSI
Spec
SCSI-1/
SCSI-2
SCSI-2
Fast SCSI
(also known as Fast Narrow)
SCSI-2
Fast Wide SCSI
SCSI-2
Intel Customer Support (ICS)
Bus
Width1
(bits)
8
16
8
16
Signalin Maximum Maximum
Throughpu
g
Devices
Cable
t (MB/s) Cabling Method Per Bus Length (m)
5
10
50-pin
50-pin
10
50-pin
20
68-pin
SE
8
6
HVD
8
25
SE
16
6
HVD
16
25
SE
8
3
HVD
8
25
SE
16
3
Troubleshooting the IA Platform
43
(also known as Fast SCSI)
Ultra SCSI
(Also known as Ultra Narrow
or Fast-20 SCSI)
Wide Ultra SCSI
(Also known as Fast Wide
20)
Ultra2 SCSI
HVD
SCSI-3/
SP1
8
SE
50-pin
HVD
SCSI-3/
SP1
16
SCSI-3/
SPI-2
8
40
SE
68-pin
HVD
40
LVD
50-pin
HVD
SCSI-3/
SPI-2
16
Ultra3 SCSI
(also known as Ultra 160
SCSI)
SCSI-3/
SPI-3
16
Ultra320 SCSI
SCSI-3
Serial SCSI
(fiber channel SCSI I)
SCSI-3
Serial SCSI
SCSI-3
(fiber channel SCSI II)
1
Narrow=8 bits and Wide=16 bits
Wide Ultra2 SCSI
20
80
LVD
68-pin
HVD
16
25
8
1.5
4
3
8
25
8
1.5
4
3
16
25
8
12
2
25
8
25
16
12
2
25
16
25
16
12
2
25
16
12
160
68-pin
LVD
320
68-pin
LVD
2
25
Serial
100 MB/s
126
NA
10,000
N/A
Serial
200 MB/s
126
NA
10,000
N/A
16
8.4.1 SCSI Encoding
Traditional SCSI uses single-ended signaling (SE), which is very similar to that used for
most other interfaces and buses within the PC. A positive voltage is a one and a zero
voltage is a zero. It was relatively simple to today’s more advanced methods and
inexpensive to implement.
As SCSI became faster, problems arose with bouncing signals, interference,
degradation over distance and crosstalk from adjacent signals. In addition, the length of
a single-ended SCSI cable is rather limited. To get around these problems, differential
signaling, later renamed to high differential signaling, was used.
As SE could not handle the faster speeds and the cost of differential was considered to
high, another form of differential signaling was created. Low voltage differential (LVD),
which was defined in SPI-2, took the best attributes of both SE and HVD signaling. It
used a lower voltage, which reduced cost and power requirements. It also allowed LVD
devices to function on a single-ended SCSI bus. LVD signaling is required for Ultra2 or
Wide Ultra2 SCSI (unless HVD) is used) and is required for any SCSI method faster
than Ultra2. As you might imagine, LVD has become very popular.
Note: Since single-ended and HVD SCSI use very different voltage levels, they are
incompatible at the electrical level. You should not mix single-ended or low voltage
Intel Customer Support (ICS)
Troubleshooting the IA Platform
44
differential devices with high voltage differential SCSI devices on the same bus. If you
do, you may damage the single-ended or LVD devices.
Some LVD devices can automatically switch between LVD and single-ended operation
by detecting whether the other devices on the chain are running in SE or LVD mode.
These devices are called multimode LVD devices and are abbreviated LVD/SE or
LVD/MSE (the M is for multimode). Note: As soon as multimode LVD devices begin
running as single-ended, all the rules and restrictions of single-ended operations apply,
including cable length.
In addition, to the usual SCSI rules, LVD operation requires the following:
•
•
•
•
All devices on the chain must be LVD-capable; if even one device is only SE, all
devices "drop down" and run as single-ended.
All devices must not be set to run in SE mode; some multimode devices have a
jumper to force SE operation, which will cause the entire SCSI chain to not work
in LVD.
LVD (or multimode LVD/SE) terminators must be used.
Since single-ended singling does not support bus speeds over 20 MHz, if you
use a faster LVD/MSE device on a single-ended chain, it will run at only a
maximum of 40 MB/s.
8.4.2 SCSI Cables
Currently, there are many types of SCSI cables available. Since each SCSI cable must
meet the specific electrical requirements associated with the SCSI signaling speeds and
signaling methods, the design of any SCSI cable is based on a combination of different
attributes chosen to implement a particular kind of SCSI bus. Some of characteristics of
the cable include the number of pins, the type of cable (internal or external), the
connector type, the length of the cable, the number of connectors, spacing of the
connectors on the cable, and the overall quality of the cable. In addition, some cables
have a built-in terminator at the end of the cable while others require the addition of a
separate terminator.
Narrow cables, officially called “A” cables, are used for all narrow (8-bit) SCSI transfer
modes. These are also sometimes called 50-conductor or 50-pin cables after the
number of wires in the cable and pins in its connectors. They are organized into 25 pairs
of two wires each. Traditionally, each pair generally consists of a signal and a signal
return, which is the same as a ground line. Wide cables, officially called “P” cables are
used for 16-bit SCSI transfer modes. These are also sometimes called 68-conductor or
68-pin cables after the number of wires in the cable or pins in its connectors. Similar to
the narrow cable, the wide cables are also organized into 34 pairs.
Lastly, cables are classified into single-ended and low voltage differential. LVD narrow
cables are similar to the SE cables except on each pair consists of a positive and
negative complement of the signal. There is also a change to the function of one cable
Intel Customer Support (ICS)
Troubleshooting the IA Platform
45
conductor is used to control the differential signaling. While you can the SE and LVD
cables use the same connectors, the LVE are typically a much higher quality to
accommodate the faster speeds used in LVD signaling. Therefore, it is not wise to use
an SE cable for an LVD implementation. Of course, to determine if a cable is an SE
cable or an LVD cable, you would have to look at the connectors for the appropriate
icons or text stating the type of cable or to look on the cable itself.
For external cables, you want to use cabling that is specifically rated for LVD use; these
cables have been designed and tested for LVD applications, even if they look the same
as their single-ended cousins. As with regular wide cables, they may be found with
either high density or very high-density connectors. Internal LVD cables are actually
very different from SE cables. The reason is that to improve signal integrity, internal
LVD cables typically don't use regular flat ribbon cabling. Instead, they use so-called
"Twist-N-Flat" cabling, where adjacent pairs of wires are twisted between the
connectors, and the wires "flatten out" where the connectors attach.
SCSI cables come in two distinct varieties, external and internal. External cables are
used to connect SCSI devices that do not reside inside the PC, but rather have their
own enclosures and power supplies. Internal cables connect SCSI devices installed
within the PC case. Internal cables are flat ribbon cables like IDE or floppy disk drive
cables.
Since external cases are more affected by external environment factors such as
electromagnetic interface, the external cables are shielded to protect the data from
being corrupted. First, they use twisted pair wiring where the two wires are twisted
around each other to reduce the affects of crosstalk. Crosstalk is if you have two copper
wires next to each other, the signal will induct (law of induction) or transfer from one
wire to the other. Therefore, by twisting the wires around each other, the
electromagnetic fields cancel each other, greatly reducing the crosstalk. So an external
narrow cable with 50 wires actually contains 25 pairs and a 68-wire cable will have 34
pairs. Second, to reduce the noise and interference, the wires are shielded or wrapped
with a metallic shield such as aluminum or copper foil or braid. Since these cables are
not simple to manufacture, external SCSI cables quire expensive compared to the
internal cables.
Even with internal cables, there are differences in construction such as the thickness of
the wire and the insulation that goes over the wires. Better cables generally use Teflon
as a wire insulation material, while cheaper ones may use polyvinyl chloride or vinyl.
Regular flat cables are typically used for single-ended SCSI applications up to Ultra
speeds (20 MHz). For Ultra2 or faster internal cables using LVD signaling, newer wires
also use twisted pairs, but overall still keep the shape of a flat ribbon where the
connectors go so that you can connect them easily.
Intel Customer Support (ICS)
Troubleshooting the IA Platform
46
8.4.3 SCSI ID Numbers
Each device, including the adapter/controller card, is numbered with a SCSI ID number
from 0 to 7. The numbers are selected with jumpers, DIP switches, or a thumb wheel.
The SCSI adapter is usually set to ID #7. The primary SCSI hard drive or any other boot
device is set to #0. The SCSI ID number do not have to be in order, nor do they have to
be sequential. Of course, no two devices within the same chain can use the same SCSI
ID number.
The priority that a device has on the SCSI bus is based on its ID number. For the first 8
IDs, higher numbers have higher priority, so 7 is the highest and 0 the lowest.
Therefore, if two devices need to be communicating on the SCSI bus, the device with
the highest priority will go first and the lower-priority device will have to wait. For Wide
SCSI, the additional IDs from 8 to 15 again have the highest number as the highest
priority, but the entire sequence is lower priority than the numbers from 0 to 7. So the
overall priority sequence for wide SCSI is 7, 6, 5, 4, 3, 2, 1, 0, 15, 14, 13, 12, 11, 10, 9,
8.
In high-traffic settings, you will typically want to set the slower devices (scanners, tape
drives) to the higher-priority IDs, to ensure that they are not crowded off the bus by the
faster devices like hard disks. For best performance, you might consider moving the
slow devices to their own SCSI controller card as they may significantly degrade the
performance of the bus. Lastly, if you have devices that absolutely cannot tolerate
delays in receiving their stream of data such as CD recording drive or a video encoder,
you should also give these devices top priority on the bus.
Today, most SCSI devices use jumpers to configure the SCSI ID numbers. For a narrow
bus, three jumpers are required to set the SCSI ID. To represent the SCSI ID number
would actually be derived from the binary representation of the jumpers. For example,
setting all three ID jumpers off represents a binary number of 0 0 0, which would
translate to an ID of 0. A binary setting 0f 0 0 1 equals ID 1, 0 1 0 equals 2, 0 1 1 equals
3, and so on. Of course, 1 1 1 would be 7. Unfortunately, the jumper can appear either
forward or backward, depending on how the manufacturer set them up.
8.4.4 SCSI Termination
The two SCSI devices at either end of the chain must be terminated; the other devices
should not be terminated. If only an internal cable is used, the adapter card and the
device at the end of the cable need to be terminated, but the other devices in between
should not be terminated. See Figure 11, 12, and 13. If using an internal cable and an
external cable, the two devices located at the end of each cable should be terminated.
To terminate or not terminate a device, the terminators are inserted/removed or
enabled/disabled with jumpers or DIP switches. Some SCSI devices can be
automatically terminated.
Intel Customer Support (ICS)
Troubleshooting the IA Platform
47
To terminate/not terminate a device, you would either insert or remove the terminators
on the end device or enabled/disabled them with jumpers or DIP switches. Some SCSI
devices have automatic termination. Today, most have a jumper to enable or disable
the termination.
There are several different kinds of termination used on SCSI buses. They differ in the
electrical circuitry that is used to terminate the bus. Better forms of termination make for
more reliable SCSI chains; the better the termination, the fewer problems (all else being
equal) with the bus, though cost is generally higher as well. In general terms, slower
buses are less particular about the kind of termination used, while faster ones have
more demanding requirements. In addition, buses using differential signaling (either
HVD or LVD) require special termination.
•
•
•
•
•
Passive terminator – The classic/standard terminator is the passive terminator.
The passive terminator uses special electrical resistors to act as voltage dividers.
Since they help ensure that the chain has the correct impedance load, they
prevent signals from reflecting or echoing when the signal reaches the end of the
chain. Passive terminating resistors work well for chains of short distances (2-3
feet) and slower speeds (SCSI-1 specification). The chain should never exceed 6
meters. It is for single-ended SCSI buses only that should only be used in narrow
(8-bit) SCSI busses running at 5 MHz.
Active Termination – Active termination acts as voltage regulator to maintain a
stable voltage through the chain by utilizing the termination power lines to
compensate for voltage drops. Since the active termination helps reduce noise, it
allows for longer cable lengths and faster speeds. In fact, active termination is
the minimum required for any of the faster-speed single-ended SCSI buses. The
chain should never exceed 18 meters. It is for single-ended SCSI buses only.
Forced Perfect Terminator (FPT) – The Forced Perfect Terminator diode clamps
are added to the circuitry to automatically match the line impedance by forcing
the termination to the correct voltage thus allowing “perfect” termination. It is for
single-ended SCSI buses only and should be used for single-ended applications
that experience high levels of electrical noise.
High Voltage Differential (HVD) – Buses using high voltage differential signaling
require the use of special HVD terminators.
Low Voltage Differential (LVD) – Newer buses using low voltage differential
signaling also require their own special type of terminators. In addition, there are
special LVD/SE terminators designed for use with multimode LVD devices that
can function in either LVD or SE modes; when the bus is running single-ended
these behave like active terminators. Note: Many internal cables have a built-in
LVD terminator at the end of the cable.
Note: LVD/SE multimode terminators are available. They will function as LVD types on
an LVD bus, and as active types on a SE bus. Note that if there any SE devices on the
bus, it will function in SE mode and never use LVD mode, severely limiting bus length
and performance. If any SE-only terminators or SE devices are on the bus, the bus will
default into SE mode.
Intel Customer Support (ICS)
Troubleshooting the IA Platform
48
Figure 11: Internal daisy chain
Figure 12: External daisy chain
Figure 13: Internal and external daisy chain
8.4.5 SCSI BIOS Programs
While IDE drives are controlled by the system ROM BIOS, the SCSI devices are
controlled by a SCSI ROM chip that is on the SCSI controller card or the motherboard if
the SCSI is built into the motherboard. During boot up, the SCSI BIOS will search and
activate the individual SCSI devices, which is shown during boot up before the
operating system is loaded.
As the SCSI BIOS activates the individual SCSI devices, the SCSI BIOS will usually
display a message to press a key or key combination to enter a SCSI setup program.
Most of these setup programs will allow you to configure the SCSI card (setting the host
adapters SCSI ID, enable/disable SCSI parity checking and host adapter SCSI
termination), configure the SCSI devices (setup sync negotiation, maximum sync
transfer rate, enable disconnection and so on) and SCSI disk utilities (low-level format
SCSI hard drives and verify disk media).
Intel Customer Support (ICS)
Troubleshooting the IA Platform
49
The hard drive can consume three or four times more power during the first few
seconds after power-on than during normal operations. If several drives are drawing all
this power at the same time, the power supply may be overloaded, which can cause the
system to hand, have intermittent start problems or to constantly reboot. Therefore,
nearly all SCSI drives provide a way to delay drive spinning so that this problem does
not occur. When enabled, the SCSI drive will wait until it gets a command from the SCSI
host adapter before spinning. It actually does this in order from highest to lower priority.
Note: this is not necessary for external hard drives since they get their power from an
external power source.
The terminators at each end of the SCSI bus require power from at least one device on
the bus. In most cases, the host adapter supplies this terminator power. If no device
supplies terminator power, the bus will not be terminated correctly and will not function
properly. To ensure that this is not a problem, some people configure all devices to
supply terminator power.
8.5 CD Drives and DVDs
The compact disk (CD) is a 4.72-inch encoded platter, which is read by laser. To read
a CD, you need a CD drive. Since they can hold large amounts of information (680 MB)
and are very inexpensive, CDs are used to distribute programs and many people even
use them to backup data. Through the years, there have been different kinds of
compact disks available in the market. The more common ones include the audio CD,
CD-ROM, CD-R and CD-RW and DVD.
Internal CD drives are half-height devices designed to fit into a standard 5.25" open
drive bay. Like other drives, the drive is typically held in place with four screws. External
drive bays of course sit outside of the computer. Today, most CD drives use an IDE or
SCSI interface. Therefore, if it uses an IDE interface, you must configure the drive as
either a stand-alone, master or slave and if it uses the SCSI interface, you must
configure the drive’s SCSI ID number and its terminating resistor.
After the drives are physically installed and configured, the final step to activating the
CD drive is to load the device drivers and to configure the operating system so it
recognizes the CD drive. This is usually done with an installation disk that comes with
the CD drive and/or controller card. Some operating systems will install the proper
drivers automatically during installation.
Today, some System ROM BIOS chips support a bootable CD drive (El Torito standard)
without loading any CD drivers. Therefore, you can insert a bootable compact disk in
the drive and if the CMOS Setup program is set to search the CD drive for a bootable
and disk and boot just like the A drive or the C drive.
For DOS and Windows 9X machines, EIDE CD drives that don’t follow the El Torito
standard, to activate the drive, it requires a device driver to be loaded in the
CONFIG.SYS while SCSI drives usually require a device driver for the controller card
Intel Customer Support (ICS)
Troubleshooting the IA Platform
50
and a device driver for the CD-ROM drive. The device drivers allow the operating
system to control the CD-ROM drive. Of course, since there are many drives on the
market, you must make sure that the driver matches the CD-ROM drive.
Another driver needed to activate the CD drive is a file system extension. The file
system extension enables the operating system to identify and use data from CDs
attached to the system. For DOS, the file system extension is the Microsoft Compact
Disk Extension (MSCDEX.EXE), which is loaded in the AUTOEXEC.BAT. For Windows
9X, the file system extension is built into the operating system. Note: if you boot a
Windows 95 or Windows 98 system to Windows DOS mode, the CONFIG.SYS will have
necessary device drivers but will not have the file system extension (MSCDEX.EXE)
loaded. Therefore, it needs to be executed at the command prompt.
Note: In Windows 9X, you can create a startup disk by double clicking the
Add/Remove Programs applet in the Control Panel and clicking on the Startup Disk
tab. The startup disk will have the common drivers for most CD drives and includes the
MSCDEX.EXE file.
Many read/write errors are caused by dirty or damaged disks. These disks can be
cleaned with a soft dry cloth. Make sure you wipe from the inside of the disk to the
outside of the disk in a straight line. Scrubbing in circular motions can damage the
surface. Never use a disk that has deep scratches, is cracked, or appears to be warped.
Such a disk can damage the drive beyond repair.
9 FILE SYSTEMS
Through out the life of the PC, there have been a few different file systems. Today, the
most common include:
1. FAT
3. FAT32
4. NTFS
Used by DOS, Windows 95, Windows 98, Windows Me, Windows NT
and Windows 2000
Used by Windows 95B, Windows 98, Windows Me and Windows 2000
Used by Windows NT and Windows 2000
9.1 File System Characteristics
The Master Boot Record (MBR) is always found on the first sector of a hard drive
(Master Boot Sector). It tells the system ROM BIOS how the hard drive is divided and
which part to boot from. The first 466 bytes consist of a special program code called the
Bootstrap Loader. The Bootstrap Loader is the program that locates the first active or
bootable partition on the disk. The last 44 bytes contain the Master Partition Table,
which list all partitions on the hard drive.
The Volume Boot Sector (VBS) is the first sector of any partition or the first sector of a
floppy disk. It is created by a high-level format (DOS’s FORMAT program or equivalent).
Within the VBS, there is the media parameter block or disk parameter block. The
Intel Customer Support (ICS)
Troubleshooting the IA Platform
51
operating system uses this information to verify the capacity of the disk volume as well
as the location of the file allocation table. It contains the partitions boot program. The
partition boot program checks and execute the necessary boot files
(IO.SYS/MSDOS.SYS, WINBOOT.SYS or NTLDR).
Disks are divided into a circular grid consisting of tracks and sectors. Tracks are
concentric circles. Concentric means circles that share the same center much like rings
in a tree. Tracks are numbered starting with the outside track as track 0. The tracks are
then further divided into sectors where each sector is 512 bytes of useable data.
Sectors are numbered on a track starting with 1.
For operating systems, the most basic storage unit is not a sector, but a cluster. Another
name for cluster is allocation unit. It consists of one or more sectors (usually more). The
size of the cluster depends on the operating system, the version of the operating
system, the file system the operating system is using and the size of the volume
(partition or floppy disks).
The file allocation table is an index used to remember which file is located in which
cluster. It lists each cluster, if the cluster is being used by a file, the name of the file
using the cluster and the next cluster if the file doesn’t fit within the cluster.
9.2 FAT32
FAT32, which has 32-bit FAT entries, was introduced in the second major release of
Windows 95 (OSR2/Windows 95B) and is an enhancement of the FAT file system. It
supports hard drives up to two terabytes. It uses space more efficiently, such as 4 KB
clusters for drives up to 8 GB, which resulted in 15% more efficient use of disk space
relative to large FAT drives.
The root directory is an ordinary cluster chain. Therefore, it can be located anywhere in
the drive. In addition, it allows dynamic re-sizing of FAT32 partitions (without loosing
data) and allows the FAT mirroring to be disabled, which allows a copy of the FAT other
than the first to be active. Consequently, FAT32 drives are less prone to failure of
critical data areas such as the FAT.
When running newer versions of FDISK on hard drives greater than 512 MB, it will ask
whether to enable large disk support. If you answer yes, any partition you create that is
greater than 512 MB will be marked as a FAT32 partition.
9.3 NTFS
NTFS is a new file system for Windows NT. Since it was designed for both the server
and workstation, it has many enhancements and feature built into it. It supports long file
names, yet maintain an 8.3 name for DOS. Since the NTFS is a 64-bit architecture,
NTFS is designed to support up to 264=18,446,744,073,709,551,616 bytes=16
exabytes.
Intel Customer Support (ICS)
Troubleshooting the IA Platform
52
Since Windows NT includes enhanced security, it supports a variety of multi-user
security models and allows computers running other operating systems to save files to
the NTFS volume on an NT server. This includes DOS, Windows 9X, Windows NT/2000
Workstation, UNIX (including Linux), and even Macintosh computers. It does not allow
DOS to access an NTFS volume directory directly, only through the network (assuming
you have the proper permissions or rights to access the volume). It can compress
individual files or directories, including infrequently used files or directories. To make a
NTFS volume more resistant to failure, NTFS writes updates to a log area. If a system
crash occurs, the log area can be used to quickly cleanup problems.
While FAT is simpler and smaller than NTFS, NTFS is generally faster because it
minimizes the number of disk accesses required to find a file, which makes access to
the file faster, especially if it is a large folder. In addition, it tries to keep the hard drive
unfragmented.
9.4 Troubleshooting File System Problems
Since the data on the disk is the most important part of the computer, it is necessary to
protect the integrity of the disk. Disk problems and errors can occur without warning and
may easily result in a data loss. Some causes include power fluctuations, viruses, parts
of the hard drive wearing out, poorly written software, user wear and tear/errors. Some
of the errors include:
•
•
•
•
•
•
•
Invalid entries in the file allocation table
Invalid directory entries
Bad sectors (unreadable areas on the disk)
Corrupt MBR or corrupt volume boot sector
Compressed header, compressed file structure, compression structure and
signatures
Lost clusters
Cross-linked files
The file allocation table and directory entries are databases. Therefore, if part of the
database becomes corrupted, part or the entire database may be unusable. Of course,
without these two databases, the operating system would not be able to find the clusters
belonging to the file. Consequently, the data would be lost.
Bad sectors are areas on the disk, which cannot reliably store the magnetic patterns
that make up data. If an area of the disk becomes unreliable, the operating system
needs to indicate that those sectors are marked bad on the file allocation table. This
way, the operating system will know not to use these areas in the future, which could
lead to data loss.
The master boot record consists of the startup program to boot from the disk. If it
becomes unreadable, the volume boot sector will not be found during bootup. Of
Intel Customer Support (ICS)
Troubleshooting the IA Platform
53
course, if the volume boot sector becomes corrupt, the computer cannot load the
operating systems boot files. In either case, the computer does not boot.
Lost clusters are clusters that get “lost” or detached from a file when a program
unexpectedly stops running or isn’t shut down properly, power fluctuations or disk
system problems. Unfortunately, the lost clusters may be a very important part of the file
(an executable file or a data file). Therefore, when the cluster is no longer part of the
file, it may cause the computer to “lock up” or act erratic when the system tries to
execute the missing instructions or it may make a data file inaccessible. If the disk
system is going bad or the system is experiencing power problems, the system may
generate lost clusters. When lost clusters are found on a system, the most likely
suspect would be that the user may be shutting off the computer without closing all
programs and performing the proper shut down procedure (File, Exit in Windows 3.XX
Program Manager or Start, Shutdown in Windows 95/98.
Some utilities that search for and retrieve lost clusters will make them available as a file
on the disk (converting lost clusters into chains). Unfortunately, since most files are in
machine code, you could not reattach them back to the file in which they came from not
to mention, you would have to figure which file it came from and exactly which part of
the file the lost cluster is. Therefore, it is best to delete the file. Of course, if the lost
cluster was vital for the program to run properly, the program would have to be
reinstalled or the data file would have to be restored from a backup.
A cross-link file is the opposite of a lost cluster. Instead of the cluster not belonging to a
file, the cross-linked errors occur when two or more files use the same cluster. Typically,
the cluster belongs to only one of the files. Therefore, repairing cross-linked files usually
results in only one file remaining usable.
To fix disk errors, there are several utilities, which can help. Some example of these
utilities include:
FDISK
SCANDISK
DEFRAG
The FDISK is the standard utility used to partition a hard drive. When the hard drive
does not boot, FDISK causes the computer to lock up or the changes to the partition
table cannot be saved may indicate that the Master Boot Record (MBR) may be bad.
What most people who use FDISK don’t know is that it can also be used to recreate the
Master Boot Record on Windows 9X systems without changing the partition table. To
recreate the MBR, you just boot with a bootable floppy containing the FDISK.EXE
program, change to the drive with the corrupt or damaged MBR and perform the
following command:
FDISK /MBR
Intel Customer Support (ICS)
Troubleshooting the IA Platform
54
If you are using a Windows NT, Windows 2000 or Windows XP and you need to repair
the master boot record, refer to the Emergency Repair Disk explained in section 14.2.
If this corrects the problem, you should be concerned with what caused the corruption of
the master boot record. It could be caused by a virus, power fluctuations when the drive
is being access or a failing hard drive. If the problem still exists, try a low-level format.
ScanDisk, Norton* Disk* Doctor* and similar utilities, are design to try to fix disk errors.
This includes invalid entries in the file allocation table, invalid directory entries, lost
clusters, cross-linked files, problems with the compressed volume, files left in an open
status and bad sectors. Unlike the previous two utilities, Scandisk is a menu-based
utility. As Scandisk checks a hard drive, it can be divided into two parts. First, it checks
disk structure followed by completing a surface scan.
With SCANDISK, you can choose to save the lost clusters to files or to delete them to
free up the disk space. If you choose to save them, they will be saved with a CHK
extension (examples FILE0000.CHK, FILE0001.CHK, FILE0002.CHK and so on) in the
root directory. If you analyze the files with the converted files and find that you do not
need them, they can be deleted at any time.
Figure 14: Windows 95's ScanDisk Utility
The surface scan test identifies any sector that may be failing. It accomplishes this by
reading the data of a sector, write a predetermined pattern to the sector and reread the
sector to make sure that it is the same. If the data is the same, the original information is
placed back onto the sector. If it finds the sector unreliable, it marks the sector as bad
and move the data to another sector.
Whichever disk utility is used, make sure that it is made for your operating system and
file system. If not, the long file names maybe truncated or corrupted, the file allocation
table and directories may be corrupted or files may be inaccessible.
Intel Customer Support (ICS)
Troubleshooting the IA Platform
55
Disk Defragmenter and other similar utilities reorganize the files on the hard drive to
optimize disk performance. It does this by gathering all parts of a file and placing them
in continuous sectors. If the drive is badly fragmented, the program may take several
hours to finish. Defrag cannot move files that are marked as system (system file
attribute) or hidden (hidden file attribute) since some of these are placed in a particular
position on the disk.
10 VIDEO SYSTEM
The video card is the component that takes the visual output from the computer and
sends it to the monitor. It tells the monitor which pixels to light up, what color the pixel
should be and at what intensity. The video card is an expansion card or is built-into the
motherboard.
10.1 Video Characteristics
The image (text, lines, boxes and pictures) on the monitor is made of many dots called
picture elements or pixels. The number of pixels that can be displayed on the screen at
one time is called the resolution of the screen. The resolution will consist of two
numbers, the number of pixels going from left to right and the number of pixels going
from top to bottom.
Color depth or bit depth is the amount of information that determines the color of a pixel.
If more bits are used to define the number of colors for each pixel, the screen can show
more colors. If you have more colors, you can have more shades of the same color,
which can lead to a more realistic, detailed picture. Unfortunately, if each pixel has a
higher color depth, the video system requires more video memory to store the pixels
information and requires more processing by the computer.
For the electron guns to draw the entire screen before the phosphorous material fades,
the electron guns must be fast. The number of times that the screen is redrawn in a
second is referred to as the refresh rate, which is measured in Hertz (Hz). If the refresh
rate is too low, the electron guns do not recharge the phosphorous material before it
fades, which causes the monitor to flicker. Although the flicker is hardly noticeable, it
can lead to eyestrain. To avoid flickering, you use a refresh rate with at least 70 Hz. The
Video Electronics Standards Association (VESA) has established that an 85 Hz refresh
rate is the standard flicker-free monitor.
10.2 PCI and AGP
Today, when you hear the term local bus, it probably refers to Peripheral Component
Interconnect (PCI) local bus. It is usually found on Pentium systems. The PCI slot was
developed by Intel and other companies (see http://www.pcisig.com) was designed to
eventually replace the older bus designs. PCI boasts a 32-bit data path, 33 MHz clock
speed and a maximum data transfer rate of 132 MB/sec.
Intel Customer Support (ICS)
Troubleshooting the IA Platform
56
AGP, short for Accelerated Graphics
Port, is based on PCI, but is designed
especially for the throughput demands
of 3-D graphics. It is called a port
because it introduces a dedicated
point-to-point channel so that the
graphics controller can directly access
main memory.
AGP 1x and 2x are part of the original
AGP Specification 1.0; AGP 4X is part
of the AGP Specification 2.0. AGP 1X
is the original version of AGP, running
at 66 MHz clock speed, a 64-bit data
bus and a maximum transfer rate of
266 MB/second. AGP 2X runs at 133
MHz and offers a transfer rate up to
Figure 15: Various Types of AGP Cards
533 MB/second. AGP 4X cards offers
1.07 GB/seconds, but will only run on
boards that are AGP version 2.0 compliant. If you do run an AGP 4X video card onto a
board that is only AGP version 1.0 compliant, it will run at 2X.
The AGP 2X connectors uses 3.3 volts while the AGP 4X connector uses 1.5 volts. To
prevent a AGP 2X card to be inserted into an AGP 4X connector, and vise versa, keys
have been added on most cards. Eventually, an AGP Universal connector was
introduced that can accommodate both types of cards and support both voltages. Note:
The D850GB only supports 1.5 AGP cards.
10.3 Video Driver
The software driver enables your software to communicate with the video adapter. All
video adapters including those built into the motherboard come equipped with drivers
supplied by the manufacturer. Most manufacturers of video adapters and chipsets
maintain Web sites from which you can obtain the latest driver. These new versions
may provide better performance or resolve a particular problem. Note: For
motherboards with the Intel chipset, unless the operating system has the Intel Chipset
Software Drivers built into the operating system, you should load Intel Chipset Software
Installation Utility immediately after loading the operating system and service packs,
even before the video driver is loaded. See section 4.4 for more information.
The video driver also provides the interface that you can use to configure the display
produced by your adapter. On a Windows 9x system, the Settings page of the Control
Panel’s Display applet identifies the monitors and video adapter installed on your
system and enables you to select the color depth and screen resolution that you prefer.
By choosing the correct video card and monitor, the driver does not allow you to choose
parameters that are not supported by the hardware.
Intel Customer Support (ICS)
Troubleshooting the IA Platform
57
If you click the Advanced button on the Settings page, you see the Properties dialog box
for your particular video display adapter. The contents of this dialog box can vary,
depending on the driver and the capabilities of the hardware. The Adapter tab displays
detailed information about your adapter and the drivers installed on the system. It also
enables you to set the Refresh Rate for your display.
Since every modern video card has a graphics accelerator, the Performance tab
contains a Hardware Acceleration slider that you can use to control the degree of
graphic display assistance provided by the adapter hardware. When troubleshooting, if
you slide the bar away from full and the problem goes away will indicate a buggy video
or mouse driver. Download and install updated video and mouse drivers and try to
revert to full acceleration.
If you choose the wrong video driver or you select a resolution, color depth or refresh
rate that your video system does not support and prevents you from viewing anything
on the monitor, you can then reboot the computer and enter either VGA mode or Safe
mode depending on which operating system you have.
11 PERIPHERALS
So far, we have talked about the basic devices that make up the PC. Now this section
discusses how to expand your system by connecting peripherals to your system using
the serial ports, parallel port, USB and infrared.
11.1 Serial and Parallel Port
The serial port is used for mice and external modems. Since today’s mice are either
PS/2 mice or USB, the serial port is not use as much as it used to be. While Windows
will automatically have a driver for the serial port, the serial port driver is rarely the
problem. Instead, most problems are either in the CMOS setup program. For example,
the serial port can be enabled or disabled and the resources (IRQ and I/O addresses)
can be assigned. Other problems may be a resource conflict with Windows and a
defective cable.
The parallel port is used for printers and some external drives such as zip drives. While
Windows will automatically have a driver for the parallel port, the parallel port driver is
rarely the problem. Instead, most problems are either in the CMOS setup program. For
example, the parallel port can be enabled or disabled, the resources (IRQ and I/O
addresses) can be assigned and the transfer mode (ECP or EPP). Other problems may
be a resource conflict with Windows and a defect cable.
11.2 USB
The universal serial bus (USB) is an external port allowing connection of external PC
peripherals in series. The USB connector will accept any USB peripheral including mice,
Intel Customer Support (ICS)
Troubleshooting the IA Platform
58
keyboards, printers, modems and external disk drives. Seven devices can be connected
directly using the standard four-pin connector. These seven devices can then be
connected to other, to a total of 127 devices, by connecting external hub (each hub
accommodates another seven devices) in a chain (hence, the term daisy chain for this
type of serial connection). One of the biggest advantages of the USB interface is that it
only requires a single interrupt from the PC for all 127 devices. Note: while USB can
support up to 127 devices, it is usually limited by bus speed and power limitations.
Lastly, Systems with Windows 98 or newer support hot swappable USB device where
you can connect the USB devices while Windows is running and it will automatically be
detected (assuming INF is installed) and you will not need to reboot the system.
A USB port or hub can provide power to the attached peripherals. A hub recognizes the
dynamic attachment of a peripheral and provides at least 0.5 W of per peripheral during
initialization. Under control of the host PC driver software, the hub may provide more
device power, up to a maximum of 2.5 W, for peripheral operations. Typically, if you
have a device more than 0.5 W, the device must have an external power source such
as an AC adapter.
When you attach a Universal Serial Bus (USB) device on a Window system that
supports USB, you may receive a "USB Hub Power Exceeded" error message
containing the following text:
The hub does not have enough power available to
operate the device driver name. Would you like
assistance in solving this problem? If you click No,
the device may not function properly.
When this happens, the USB device mentioned in the error message does not work
properly and the device listing (in the system properties of the Device Manager
properties window for the device) appears with the following error message:
A USB device has requested more power than the hub can
provide. Windows cannot enable the device. Remove the
device and restart your computer.
To resolve this issue, replace the passive USB hub with an active (powered) hub or
disconnect the affected device from the passive hub and connect it directly to the USB
host controller's hub (in the back of the computer) or attach an AC adapter if one is
available.
USB 1.1 offers a data transfer rate of up to 12 megabits/second. For low-speed
peripherals such as pointing devices and keyboards, the USB also has a slower 1.5
Mbit/sec subchannel. Maximum cable length between two full-speed (12 Mbit/sec)
devices or a device and a hub is five meters using twisted pair shielded cable with 20gauge wire. Maximum cable length for low speed (1.5 Mbit/sec) devices using nontwisted pair wire is three meters. These distance limits are shorter if smaller gauge wire
is used.
Intel Customer Support (ICS)
Troubleshooting the IA Platform
59
You can view how much bandwidth a USB device uses by right-clicking My Computer
and selecting Properties. Click the Device Manager tab, followed by View Devices by
Type. Double-click Universal Serial Bus Controllers, followed by the entry for the USB
Controller. Next, choose the Advanced tab and click the Bandwidth Usage button.
11.3 Infrared
The IrDA port, sometimes known as Ir port, was developed by the Infrared Developers
Association (IrDA), aimed at using wireless technology for the notebook and laptop
computers. Today, you can find keyboards, serial port devices (mice and other pointing
devices) and parallel port devices, which use IrDA.
Like the name indicates, IrDA ports use infrared light (same as your TV or VCR remote
controls) to send data. Infrared is an invisible electromagnetic radiation that has a
wavelength longer than that of visible light. So far, the IrDA data rate is four megabits
per second. Unlike other wireless technology, Infrared does not use any radio waves.
Infrared signals are generated by a tiny light emitting diode (LED), which focuses its
beam into a 30 degree cone. To avoid interfering with other infrared devices, IrDA
devices don’t use much power. Consequently, the two communicating devices must be
within a few feet of each other and they must have a clear line of sight between them. In
addition, bright sunlight may drown out the signal.
12 NETWORK CARDS AND NETWORKING
As networks have grown in popularity, it is becoming more important to understand how
to connect PCs together to form a network. Today, most networks are Ethernet
networks on a 10BaseT/100BaseT using unshielded twisted pair cabling.
12.1 Network Cards and Cables
To communicate on LANS, network cards, or network interface card (NIC), found as an
expansion card or are integrated into the motherboard, uses transceivers (devices that
both transmit and receive analog or digital signals). Today, most network cards are plug
and play, so you just have to insert into an empty expansion card while the computer is
off and turn the computer on so the appropriate resources such as IRQ can be
assigned. If the driver is not automatically loaded, you will have to load it using the
Add/Remove Hardware in the Windows Control Panel.
Installing a network card is not much different from other expansion cards. First, you
must configure the card’s I/O address, IRQ, DMA and memory addresses and any
options that are specific to the type of network card that you selected. This would be
done with jumpers, DIP switches, configuration software or by plug and play. After the
card is physically installed and the network cable is attached, you must then install the
appropriate driver (included with the Operating System or a disk that comes with the
network card or it will have to be downloaded from the Internet) and client software. In
Intel Customer Support (ICS)
Troubleshooting the IA Platform
60
addition, you must bind the protocol stack to the network card and install any additional
network software needed. Depending on the type of the card, the driver may need to
have the hardware resources specified (I/O address, IRQ, DMA and memory address).
Of course, this needs to match the settings of the cards. Lastly, you should make sure
that the LED for the expansion card is active.
A Media Access Control (MAC) address is a hardware address identifying a node on the
network, much like a street address identifies a house or building within a city. Much like
a street address within a city, you cannot have two network cards or nodes with the
same MAC address on the same network. You can have the same MAC address on
two separate networks. For Ethernet and Token Ring cards, the MAC address is
embedded onto the ROM chip on the card.
When building a 10BaseT/100BaseT network, you must follow these rules:
•
•
•
The minimum unshielded cable segment length is 2 feet or 0.6 meters.
The maximum unshielded cable segment length is 328 feet or 100 meters.
To avoid EMI, you must route UTP cable no closer than 5 feet to any highvoltage (power) wiring or fluorescent lighting.
One advantage of using 10BaseT/100BaseT is that if you need to add another
workstation, you just run an additional cable to the hub and plug it in. If there is a break
in the cable, the hub has its own intelligence that will route traffic around the defective
cable segment. As a result, the entire network is not affected.
Unshielded twisted pair (UTP) is the same type of cable that is used with telephones
except it consists of four pairs of wires in each cable. Each pair of wires are twisted
around each other and used together to make a connection using an RJ-45 connector
(phones connectors use an RJ-11 connector and are a little smaller than RJ-45
connectors). The biggest disadvantages of UTP are its limited bandwidth of 100 meters
and it is quite susceptible to interference and noise. Traditional, UTP has had a limited
network speed, but more recently UTP can be used in networks running between 4
Mbps and 1 Gbps.
A hub is a multiported connection point used to connect network devices via a cable
segment. When a PC needs to communicate with another computer, it sends a data
packet to the port, which the device is connected to. When the packet arrives at the
port, it is forwarded or copied to the other ports so that all network devices can see the
packet. A switch, which is similar to a hub, will only send the packet to port that the
packets needs to go, greatly reducing traffic. Most hubs and switches are active hubs,
which mean they require power and they act as a multiported repeater for the signal.
Note: when installing an active hub, should make sure the fan (which provides cooling)
is operating.
Most hubs have a crossover connection that allows you to use a normal cable to
connect one hub to another hub. Some hubs will have a separate crossover connection,
Intel Customer Support (ICS)
Troubleshooting the IA Platform
61
some will have a crossover connection and a straight-through connection, which only
one can be active at a time and some will have a port that can toggle between a
crossover connection and a straight-through connection.
Cables cause the most problems in networks. Since you have an understanding of how
the cables are connected together and how the signals travel through the cable system,
this will give you a solid foundation in isolating and correcting cable problems.
If the conductor such as a wire has a break in it (known as an open) and there is no
other pathway for the electrons, the current will not flow. A short is when a circuit has
zero or abnormally low resistance path between two points, resulting in excessive
current. A short on a network cable is usually when one wire connects to another wire
when it is not supposed to.
To help troubleshoot cables, there are several tools that you can use. The most
common are Digital multimeters (DMMs) and a cable testers. Digital multimeters contain
at least a voltmeter and ohmmeter. An ohmmeter can be used to measure the
resistance of a wire. If the wire is good, the ohmmeter should measure 0 ohms. If you
measure infinite ohms, the wire will usually have a break in it. Some of the multimeters
will also contain a continuity checker. A continuity check will beep when it measures 0
ohms (indicating a good connection) and will not beep when it measures more than 0
ohms.
A cable tester is a device specifically made to test a cable. An inexpensive cable tester
will typically test for shorts and opens. More expensive cable testers (sometimes
referred to as an advanced cable tester) can display additional information about the
condition of the physical cable as well as message frame counts, error frame counts
excess collisions, late collisions, congestion errors and beaconing.
When experiencing a network failure, you must first determine what is affected by the
network failure. If only one computer is affected, it is most likely the cable that connects
to the individual computer, the defective network card or the driver and client software is
installed or configured incorrectly. If multiple computers have failed, you will need to
look for a network item that is common to all of the computers. If you are using Ethernet
using 10BaseT/100BaseT, look for a hub or a single cable that connects the hub to the
rest of the network.
12.2 Cable Modems and DSL Modems
Cable Modem and DSL Modems will be expansion cards, or a small boxes that connect
to the USB or a network card. If it were an expansion card or a USB device, you would
have to load the appropriate driver. If is a box that connects to a network card, you
install the driver for the network card. In either case, you then configure the TCP/IP
properties such as the IP address, subnet mask and so on.
Intel Customer Support (ICS)
Troubleshooting the IA Platform
62
12.3 TCP/IP
Each connection on a TCP/IP address is called a host (including server and client) and
is assigned an IP address. The format of the IP address is four 8-bit numbers divided by
a period (.). Each number can be zero to 255. For example, a TCP/IP could be
131.107.3.1 or 2.0.0.1. Since the address is used to identify the computer, no two
connections can use the same IP address.
To make TCP/IP more flexible, a network can be subnetted into smaller networks. To
identify which bits are used to define the network number and which bits are used to
define the host address, a subnet mask must be assigned. For a network, the network
number and the subnet mask must be the same for all computers. The TCP/IP address
and the subnet mask are assigned automatically when a computer connects to a
network by a Dynamic Host Configuration Protocol (DHCP) server or are assigned by
the network administrator. Note: if you are using DHCP and you have manually
specified a TCP/IP address, subnet mask, gateway, DNS server or WINS server, the
manual settings will always overwrite the DHCP settings.
When a single local area network is connected to other networks, the computers need
to know the default gateway, which specifies the local address of a router. If the default
gateway is not specified, you will not be able to communicate with computers on other
networks.
In Windows 95, Windows 98 and Windows Me, the network components is installed and
configured using the network dialog box, which is accessed by clicking on the network
applet in the control panel or by accessing the short cut menu of the Network
Neighborhood icon and selecting the Properties option. Note: the Network
Neighborhood icon does not show on the desktop unless you have some network
components installed.
The Identification tab is used to assign the NetBIOS computer name. Much like the
MAC addresses, no two computers can have the same computer name on the same
workgroup or domain. You can then use utilities such as the Network Neighborhood to
access the computer and its network resources by using its computer name. Note: The
Computer name is up to 15 characters.
While IP addresses are 32 bits in length, most users do not memorize the numeric
addresses of the hosts to which they attach. Instead, people are more comfortable with
host names or NetBIOS (Computer) names. Most IP hosts, then, have both a numeric
IP address and a host name. While this is convenient for people, however, the name
must be translated back to a numeric address for routing purposes. Name resolution is
typically done by DNS and WINS. DNS, which stands for Domain Naming System,
translates Fully Qualified Domain Names (FQDN) or Internet names to IP addresses.
WINS translates NetBIOS name or Computer name to the IP address. The WINS
Configuration and DNS Configuration are used to assign the addresses of the DNS and
WINS servers, which are used for name resolution.
Intel Customer Support (ICS)
Troubleshooting the IA Platform
63
To show your TCP Settings, type either IPCONFIG /ALL at the prompt if you are using
Windows NT4.0, Windows 2000 or Windows XP and WINIPCFG for Windows 9X.
Another very useful command to verify a network connection is the PING.EXE
command. If you ping the loopback address of 127.0.0.1 and the ping fails, you need to
verify that the TCP/IP software is installed correctly. If you ping your own IP address
and it is unsuccessful, you need to check the IP address. If you can’t ping the IP
address of the default gateway/router, you need to verify the IP address (make sure the
address is in the correct subnet) and subnet mask. If you can’t ping a remote computer
on a different subnet or network, you need to verify the IP address of the default
gateway/router, make sure the remote host/computer is functional and verify the link
between routers.
The PING command can also be used to ping a host/computer by NetBIOS (computer)
name or host/DNS name. If you ping by address but not by name tells you that the
TCP/IP is running fine but the name resolution is not working properly. Therefore, you
must check the WINS server to resolve computer names and the DNS server to resolve
domain names.
Another useful command is the TRACERT.EXE command, which sends out a packet of
information to each hop (gateway/router) individually, and does it three times.
Therefore, the TRACERT command can help determine where the break is in a
network.
12.4 Sharing Resources
File and Print Sharing for Microsoft networks gives you the ability to share your files or
printers with Windows. Note: just because you enable file and print sharing for an
individual computer, you must still share a drive, directory or printer before it can be
accessed through the network.
If you enabled file sharing on a Windows computer, you can share any drive or directory
by right-clicking the drive or directory and selecting the Sharing option from the File
menu or by selecting the Sharing option from the shortcut menu of the drive or directory.
You would then provide a Share Name, the name seen by other clients and type of
access that users can have. A shared drive and directory will be indicated with an image
of a hand under the drive or directory.
The shared drive or directory can then be accessed in one of three ways:
1. Using Network Neighborhood and accessing the resources under the server.
2. Specifying the Universal Naming Convention (UNC) using the Run… option
under the Start button. The UNC format is specified as
\\computername\sharename.
Intel Customer Support (ICS)
Troubleshooting the IA Platform
64
3. By selecting the Map Drive button on the toolbar of Network Neighborhood or
Microsoft Explorer on machines using the Active Desktop.
To share a printer that is connected to a Windows machine, you would share the printer
to make it available to other people. To share a printer is very similar to sharing files.
You would first install the print driver using the Add Printer icon in the Printer folder. You
would then select the printer and you would then choose the Sharing option. After
assigning the printer share name and option password, you would then click on the OK
button and a hand will appear under the printer to indicate that it is shared.
You would then go to the clients computers, run the Add Printer wizard. Instead of
choosing the local printer option, you will be choosing the network printer option. You
will then be prompted for the UNC name (//computername/sharename) of the printer.
When you print to this printer, the print job will then be automatically redirected to the
Windows 95/98 machine and sent to the printer.
12.5 Internet and Browsers
To connect to the Internet, you have to have a direct connection to the Internet, or you
have to go through an Internet Service Provider (ISP). If you are using an analog
modem to connect, you will typically use an Internet Server Provider. The most common
two browsers are Microsoft* Internet Explorer* and Netscape* Navigator*.
If problems do occur when attempting to access the Internet, you must verify the TCP/IP
settings including the IP address, subnet mask, gateway and DNS Server addresses. In
addition, if you are behind a firewall or go through a proxy server, you must configure
your browser. These configuration settings include automatic configuration, configuring
through scripts or by manually specifying the settings.
To configure the proxy selection and proxy bypass settings in Internet Explorer:
1.
2.
3.
4.
Open the Tools menu, and then click Internet Options.
Click the Connections tab, and then click LAN Settings.
In the Proxy Server area, select the Use a proxy server check box.
Click Advanced, and then fill in the proxy location and port number for each
Internet protocol that is supported.
Note: In most cases, only a single proxy server is used for all protocols. In those cases,
enter the proxy location and port number for the HTTP setting, and then select the Use
the same proxy server for all protocols check box. If you want to manual set the
addresses, then enable the Use a proxy server and specify the address and port
number of the proxy server. If you need to specify different addresses and/or port
numbers for the various Internet services, click on the Advanced button. Note: if you
don’t know your proxy settings, contact your network administrator or ISP.
For Netscape, the same basic options are available if you:
Intel Customer Support (ICS)
Troubleshooting the IA Platform
65
1. Open the Edit menu and select the Preferences option.
2. Find and open the Advanced option
3. Click on the Proxy option.
Figure 16: Changing the Proxy settings in
Internet Explorer.
13 WINDOWS 95, WINDOWS 98 and WINDOWS MILLENIUM
13.1 Problems with Installation
Windows installation is usually successful; but on occasion problems do occur. These
problems are often traceable to legacy hardware or poorly documented off-brand
equipment. Windows Setup includes a Safe Recovery feature that helps you recover
from failed installations. If Setup fails before hardware detection, when you restart, you'll
be asked whether you want to use Safe Recovery to resume the installation. If you
choose to use Safe Detection, Setup will inspect the log file setuplog.txt to determine
where the installation terminated. Setup will then skip the step that caused the failure
and resume the installation. If you choose not to use Safe Recovery, Setup will start
over with the installation.
If Setup fails during hardware detection, a log file called detcrash.log will be created.
detcrash.log gives Setup the information it needs to resume the hardware detection and
bypass the step that caused the failure. If detcrash.log is present in the root directory of
the boot partition, Setup will automatically switch to Safe mode. As you'll learn later in
this section, a text-file equivalent of detcrash.log called detlog.txt is also created so that
users can review hardware detection information. See Table 11.
Intel Customer Support (ICS)
Troubleshooting the IA Platform
66
Table 11: Various log files for Windows 9X.
Windows
Setup Files
setuplog.txt
detcrash.log
detlog.txt
netlog.txt.
bootlog.txt.
Description
setuplog.txt is an ASCII text file that contains a log of the installation process. As
Windows 9X installation progresses, entries are written into setuplog.txt for each
step in sequence. This file is used by Windows 9X's Safe Recovery feature in
case of setup failure, and you can use it to look for the source of Setup errors.
setuplog.txt is stored as a hidden file on the computer's root directory. Because
new entries are added to setuplog.txt chronologically, start at the end of the file if
you're trying to determine what caused the Windows 9X Setup program to fail.
detcrash.log is a binary file containing information on Windows 9X Setup's
hardware detection process. When Setup fails during hardware detection,
detcrash.log is created as a hidden file in the root directory of the installation
drive. When you reboot, Setup will use the information in detcrash.log to restart
the installation. An ASCII equivalent of detcrash.log, called detlog.txt, is also
created.
This is an ASCII text file that contains a record of all devices found during the
hardware detection phase of installation. If a device is found, the detected
parameters are identified and recorded. detlog.txt is stored as a hidden file on
the computer's root directory. Information is added to this file in the same order
as the hardware detection phase. If you need to determine what caused the
Windows 9X Setup program to fail or lock up, refer to the entries at the bottom of
this file before restarting the system.
This is an ASCII text file that contains a record of all detected network
components found during installation. The network detection phase consists of
four parts. These correspond with the four class types of network configuration:
network clients, network protocols, network adapters, and network services
(such as file and print sharing).
netlog.txt is stored as a non-hidden file on the computer's root directory.
Information is added to this file in the same order as the network detection
phase. If you need to determine what caused the Windows 9X Setup program's
failure to communicate across the network, refer to the entries in this file. You
will see where Windows 9X found the network adapter and identified which
protocols, clients, and services to bind to the card. At the end of each line, you
should see OK. If you see a line such as Couldn't determine..., or some other
failure notice, you have found your problem.
This is an ASCII text file that contains a record of the current startup process
when starting Windows 9X. When Windows 9X is started for the first time,
bootlog.txt is created automatically. For subsequent reboots, you can create a
bootlog.txt file by pressing F8 at startup (to invoke the Boot menu) and choosing
the Boot menu option Logged (\bootlog.txt). You can also create a Boot log by
running win.com from the command line and including the /b switch. bootlog.txt
records the Windows 9X components and drivers as they are loaded and
initialized, and records the status of each step. The information in bootlog.txt is
written in sequence during startup. You might need to examine the information
closely to determine which error occurred. The Windows 98 Resource Kit has a
good description of the sections within this file. bootlog.txt is stored as a hidden
file on the computer's root directory. Information is added to this file during the
Windows 9X startup process. If you need to determine what caused Windows 9X
to fail or lock up, refer to the entries within this file before restarting again.
bootlog.txt seems to write everything twice. The first line will read something like
Loading VXD=..., and the next line will read something like LoadSuccess
VXD=.... In troubleshooting, it is important to see what loaded successfully, as
well as what loaded unsuccessfully. bootlog.txt provides that information.
Intel Customer Support (ICS)
Troubleshooting the IA Platform
67
Given Setup’s elaborate recovery mechanisms, the best thing to do if Setup fails is to
restart your system and let Setup try to work around the problem. Microsoft
recommends the following action (in this order) if Setup stops unexpectedly:
1. Press F3 or click on the Exit button.
2. Press Ctrl + Alt + Del.
3. Turn off your computer, wait 15 seconds, then restart.
If Safe Recovery does not solve your Setup problems, you can use Setup’s log files to
guide your troubleshooting. The Windows Setup program creates the log files
setuplog.txt and detlog.txt (if Setup fails during hardware detection). Another pair of files
(netlog.txt and bootlog.txt) is created as Windows starts up the first time.
The first step in troubleshooting Windows 9X setup issues is to get back to the basics.
When installation fails early in the setup process, it is often related to one or more of the
following issues:
1. Is a disk management utility installed? If it is, can you safely remove it? This might
entail a BIOS upgrade (if available) and quite possibly require using FDISK and
repartitioning the drive. If this is the case, do you have a good backup?
2. Is the disk compressed with a third-party utility? If so, can you safely uncompress
the drive?
3. Does Setup hang during the ScanDisk phase of installation? (An application is
considered to be hung if it is still running, but not responding to any messages from
the system.) If Setup does hang during the ScanDisk phase, you can manually run
ScanDisk, and force a thorough inspection of the drive prior to Windows 98
installation.
4. Have you disabled virus protection? Remember, it may be loading automatically
from your previous OS, or it may be enabled in the computer's BIOS. Some antivirus programs prevent writing to the boot sector of the hard disk. If this is the case,
Windows 98 Setup will either fail to install, or will not load correctly. To correct this
situation, you must disable the virus protection. After Windows 98 is properly
installed, you can re-enable virus protection.
5. Have you turned off power management? At times power management can cause
problems (particularly if your computer decides to take a nap during a lengthy phase
of Windows 98 Setup).
6. Have you made certain that your computer does not have a virus? Run a current
anti-virus program and check for viruses after booting from a known clean disk.
(Don't forget to disable virus protection before restarting Setup.)
7. SmartDrive can cause problems with some SCSI hard disks. By default, during
Windows 9X Setup, Smart Drive is loaded, but double-buffering is turned off. To
work around this situation, simply use the /c switch to run Setup without SmartDrive.
8. What is being loaded from the autoexec.bat and config.sys? To see if autoexec.bat
or config.sys may be causing the problem, perform the following steps:
a. Rename autoexec.bat by typing the following from the command line: ren
c:\autoexec.bat autoexec.aaa.
Intel Customer Support (ICS)
Troubleshooting the IA Platform
68
b. Rename config.sys by typing the following from the command line: ren
c:\config.sys config.aaa.
c. Restart the computer and run Windows 98 Setup again.
If none of these helped, check for incompatible hardware, incompatible TSR or driver,
faulty hardware and viruses
13.2 Boot Problems
A large majority of the technical problems that arise under Windows 98 can be traced
back to the configuration files and how these files control the boot process. Isolating
which of these settings are incompatible with Windows 98 can be difficult. Windows 98
includes a number of features that will help you restart your system after a boot failure
and find the source of the problem:
1.
2.
3.
4.
The Startup menu
The Startup disk
Win.com switches
System Configuration Utility
Pressing the Ctrl key when you boot your Windows 98 computer invokes the Windows
98 Startup menu, which provides a number of different modes in which Windows 98 can
be booted. If the system fails to start normally, you may still be able to reboot into one of
these alternative modes in order to troubleshoot the problem. The Startup menu options
depend in part on the parameters specified in the msdos.sys file, but generally consists
of those listed in table 12:
Table 12: Windows 9X Startup Menu
Startup Menu
Normal mode
Logged mode
Safe mode
Description
This is the normal operation mode of Windows 98. If you boot to the Startup menu but
then decide to complete the boot process and start Windows 98 under normal
conditions, select this mode.
When you select Logged mode, the entire boot process is logged to a file called
bootlog.txt, which catalogs VxD initializations, driver loads, and various other bootrelated events. bootlog.txt was described earlier in this chapter. You can use the boot
log to determine where the boot failure occurs and what the system is doing at the
time. Aside from the logging, the Logged mode performs a normal boot procedure (of
course it will be a bit slower because it writes to the bootlog.txt file). bootlog.txt will
normally be found in the root directory. You can load bootlog.txt into a text editor such
as Notepad to examine the contents.
Safe mode is likely the single most important troubleshooting tool available in
Windows 98. In this mode, a number of key Windows 98 components and settings are
disabled, including the following:
•
•
•
•
•
•
config.sys and autoexec.bat
The [Boot] and [386Enh] sections of system.ini
The Load= and Run= parameters of win.ini
The Startup program group
The Registry
All device drivers except the keyboard, mouse, and standard VGA video drivers
Intel Customer Support (ICS)
Troubleshooting the IA Platform
69
Disabling these items allows the separation of fundamental operating system
problems from those caused by a combination of software factors. In a situation in
which the display is not functioning properly in Normal mode, for example, if the
problem does not appear in Safe mode, the problem probably is video driver-related
and is not due to a defective video card.
Similarly, you can use Safe mode to troubleshoot scenarios such as the following:
Step-by-Step
Confirmation
mode
CommandPrompt-Only
mode
Safe Mode
CommandPrompt-Only
mode
Previous Version
of MS-DOS
• GPFs (General Protection Faults)
• Application hangs
• A hang during the boot process
• A blank screen at boot time
This boot mode is similar to the F8 function of previous versions of MS-DOS; it
permits the user to step through the various stages of the boot process and specify
whether each should or should not be completed. This mode can be very useful when
you are trying to isolate boot stages to determine which may be causing a given
problem. You can also use it to view system responses to various parameters in
config.sys and autoexec.bat, which otherwise are displayed far too quickly to read.
Command-Prompt-Only boot mode is similar to a normal boot of MS-DOS. Only
config.sys, autoexec.bat, command.com, and the Registry are processed (along with
any necessary disk compression drivers). This mode is useful in troubleshooting
problems running MS-DOS applications in a VM under Windows 98. If the application
functions in this mode but not inside Windows 98, the problem is likely due to a
compatibility issue. If the application does not function in Command-Prompt-Only
mode, the problem is likely a configuration problem in config.sys or autoexec.bat, or
the application may be corrupt.
Safe Mode Command-Prompt-Only mode is similar to a Safe mode, except that
command.com is processed. Also, Startup does not load himem.sys or ifshlp.sys and
does not execute win.com to start the Windows interface. This mode is useful if your
computer fails to boot properly in Safe mode. Safe Mode Command-Prompt-Only
mode offers slightly different options from Safe mode, so you can use it for slightly
different situations. For instance, you can use this mode if you don't want Windows to
process win.com or himem.sys.
Although the Previous Version of MS-DOS boot mode is not intended for
troubleshooting, it can be used in situations in which particular MS-DOS–related
functions worked in previous versions of MS-DOS but do not seem to function
properly under Windows 98. Of course, you can boot to a previous version only if you
upgraded the computer from a previous version.
You can create the Startup disk at installation time or later through the Startup Disk tab
of the Add/Remove Programs Control Panel.
The disk serves as an emergency boot disk should the operating system fail to load. In
addition to the boot files necessary to start Window 98 in Command-Prompt-Only mode,
the Startup disk contains several tools you can use to troubleshoot boot problems.
Some of those tools are as follows:
If you can boot to the Windows 98 Startup disk, you can navigate through your hard
drive to find a file that is interfering with the boot process. You can also troubleshoot
your hard drive using ScanDisk, or even reformat and repartition your hard drive (start
over) using format and fdisk. The CD-ROM drivers on the Startup disk will provide you
Intel Customer Support (ICS)
Troubleshooting the IA Platform
70
with access to the Windows 98 installation CD, in case you'd like to reinstall Windows
98 or copy files that are located on the CD.
Microsoft Windows 95 Startup Menu
===========================
1. Normal
2. Logged (\BOOTLOG.TXT)
3. Safe mode
4. Safe mode with network support
5. Step-by-step confirmation
6. Command prompt only
7. Safe mode command prompt only
8. Previous version of MS-DOS
Enter a choice: 1
F5-Safe mode Shift+F5=Command prompt Shift+F8=Step-by-step Confirmation [N]
Figure 17: Windows Startup Menu
Win.com includes support for a number of error-isolation switches. Although some are
available from within Windows 98, you may have to specify them from the command
prompt in situations in which Windows 98 fails to load. These switches are specified in
the following format:
win /d:[f] [m] [n] [s] [v] [x]
Table 13: WIN switches
Switch
Function
F
Disables 32-bit file system drivers
M
Starts Windows 98 in Safe mode
N
Starts Safe mode with networking
S
Excludes the ROM address space between FOOO and 1MB from use by Windows
98
V
Tells Windows 98 that the ROM routine should handle disk interrupts
X
Excludes the adapter area from the area that Windows 98 scans when looking for
unused disk space
You can use these switches independently or together as part of a single command.
If Windows won’t start normally, you may be able to boot to a system disk and run
win.com using one or more of these switches to isolate the problem.
Boot failures are often caused by drivers or settings invoked through one of the
Windows Startup configuration files.
Intel Customer Support (ICS)
Troubleshooting the IA Platform
71
System Configuration Utility is an ingenious tool that lets you turn off or turn on specific
entries in the Startup files autoexec.bat, config.sys, system.ini, or win.ini. You can also
make other changes to the startup process that may help with diagnosing startup
problems. The System Configuration Utility General tab is shown in Figure 18.
Note that, in addition to a normal startup, you can choose Diagnostic startup, or you can
choose to selectively disable files. The tabs named for the Startup files (autoexec.bat,
config.sys, system.ini, and win.ini) let you enable or disable specific statements within
the file.
You can also change the order of the statements within the file by using the Move Up
and Move Down buttons. Or, you can add a new statement to the file or edit a statement
by using the New and Edit buttons. The Startup tab lets you decide whether to load
certain items at startup.
Figure 18: System Configuration Utility allows you to selectively load certain components.
Windows system problems often occur at startup. These problems are, of course,
impossible to troubleshoot if you can't start your system. Windows includes several
methods for starting Windows that may help you diagnose a problem, or at least, start
the system so you can pursue other troubleshooting remedies. The preceding sections
discussed some of these options. The first step is usually to boot your system using the
option boot modes of the Startup menu such as Safe mode or Step-by-Step
Confirmation mode. The win.com switches are usually for subtler problems. System
Configuration Utility is often useful for troubleshooting startup problems related to older
devices and applications.
Intel Customer Support (ICS)
Troubleshooting the IA Platform
72
Figure 19: The WIN.INI an SYSTEM.INI options in the System Configuration Utility.
13.3 Windows Device Manager
The Windows Device Manager is a powerful resource that is designed to help you
inspect the configuration and settings of almost every device in your system. By
opening the System applet in the Control Panel, it will display the System Properties
dialog box. Within the dialog box, you will see five tabs, General, Network Identification,
Hardware, User Profiles and Advanced. The hardware tabs allow access the hardware
wizard, the device manager and hardware profiles. The hardware wizard helps you
install, uninstall, repair, eject and configure your hardware.
The Device Manager lists all of the hardware devices on your computer and allows you
to change the properties of any device. If you open the View menu and select
Resources by type, you will see the IRQ, DMA, I/O address and memory areas used by
the different hardware components.
If you open the view menu and select Devices by type or Devices by connection, you
see all hardware devices organized in a tree structure. A red X appears through an icon
means the hardware device has been disabled. A yellow circled exclamation point
through the icon means that the hardware device has a problem. The problem could be
a resource conflict (IRQ, DMA, I/O address or memory address) or the drivers are not
installed properly.
To see hardware in a category, click on the plus sign next to the hardware type. To see
information about a piece of hardware, double-click on the hardware device icon or
select the hardware device and click on the properties button. The properties will usually
contain several tabs indicating if the device is working properly, what drivers are
installed and gives you an option to update them and to view and change the device
resources (if it is a plug and play device).
Intel Customer Support (ICS)
Troubleshooting the IA Platform
73
When viewing device information on your Windows 2000-based computer using Device
Manager, you may see an unknown device listed next to a yellow question mark.
Determining the cause of this unknown device can be difficult, because there are few
indications of what could be creating it. The unknown device may also cause a conflict
when trying to install a driver for another device.
The most common reasons Device Manager may list a device as unknown are as
follows:
•
•
•
•
The Device Manager does not have a device driver.
Windows does not recognize the device ID.
The device is created by software.
The system has faulty hardware or firmware.
When a device driver for a device is not available. Device Manager displays the device
as unknown and places it in the Other Devices folder. This commonly occurs when
universal serial bus (USB) and IEEE 1394 composite devices. A status of “Error Code
1” or “Error Code 10” also may be displayed when you view the properties of the device
in Device Manager.
Don’t forget that most motherboards today need to have their chipset drivers loaded to
recognize all of the drivers. For example, for Intel motherboards in the 8xx series, there
is the Intel Chipset Software Installation Utility. This program stalls specific Windows
INF files, which inform Windows how to properly configure the chipset for specific
function, such as AGP, USB, Core PCI and ISA PnP services. Since these components
are necessary for optimum stability and performance, the Intel Chipset Software
Installation Utility should be the first loaded after the operating system
The unknown device can also be caused by faulty hardware or firmware. To isolate the
correct hardware device, remove the hardware devices from the Windows computer
one a time until the unknown device is not longer listed in Device Manager.
13.4 System Information Utility
Starting with Windows 98, Microsoft includes the System Information Utility, which can
be used to analyze your system for problems. System Information collects and displays
your system configuration information. For example, you can view:
•
•
•
•
Operating system information, such as the version number and system boot
options, plus process, system, and user environment variables
Hardware details such as BIOS information, video resolution, processor type,
and processor steppings
Physical memory, paging file information, and direct memory access (DMA)
usage
The current state of each driver and service on the computer
Intel Customer Support (ICS)
Troubleshooting the IA Platform
74
•
•
•
Drives and devices installed on the computer, plus related interrupt (IRQ) and
port information
Network information, including transports, configuration settings, and statistics
Printer settings, font settings, and system processes that are running
You can access the System Information utility by going to the Start menu and clicking
Programs, clicking Accessories, clicking System Tools and selecting System
Information.
System Information displays a comprehensive view of your hardware, system
components, and software environment. The displayed system information is organized
into three top-level categories that correspond to the Resources, Components, and
Software Environment nodes.
•
•
•
The Resources node displays hardware-specific settings, namely DMA, IRQs,
I/O addresses, and memory addresses. The Conflicts/Sharing node identifies
devices that are sharing resources or are in conflict. This can help identify
problems with a device.
The Components node displays information about your Windows configuration
and is used to determine the status of your device drivers, networking, and
multimedia software. In addition, there is a comprehensive driver history, which
shows changes made to your components over time.
The Software Environment node displays a snapshot of the software loaded in
computer memory. This information can be used to see if a process is still
running or to check version information.
Other applications may add nodes to System Information that display information
specific to the application.
You can use the View menu to switch between the display of Basic and Advanced
information. The Advanced view shows all of the information in the Basic view plus
additional information that may be of interest to the more advanced user or to Microsoft
Product Support Services.
While the first step to solving complex problems is to get more information about the
problem, information alone doesn’t fix what's wrong. For this reason, the System
Information utility includes ten tools that you can use to possibly fix a malfunctioning
system. In the following sections, we'll cover each of these tools in detail. To access any
of these tools, simply choose the appropriate tool from the System Information utility's
Tools menu. See table 14.
Table 14: Window 9X Support Utilities
Tools
Windows
Report Tool
Description
The Windows Report Tool will help solve problems that you just can't seem to figure out
yourself. This tool asks you to describe your problem, what you're trying to accomplish,
and the steps you've taken to reproduce the problem. When you have answered these
questions, simply click Next and enter a filename. Windows 9X will build a CAB file
Intel Customer Support (ICS)
Troubleshooting the IA Platform
75
Update
Wizard
Uninstall
System File
Checker
Signature
Verification
Tool
containing answers to your questions and a summary of your system's configuration. You
can now send this file to Microsoft's technical support staff for assistance with your
problem.
As you're probably aware, Windows 98 contains a Windows Update utility you can use to
install software patches from the Internet. Unfortunately, sometimes a patch can do more
harm than good. If you install a patch and it causes problems, you can use the Update
Wizard Uninstall utility to remove the patch quickly and easily.
Corrupted system files cause many of the most serious Windows problems. Until now, the
easiest way to correct such problems was to reinstall Windows and replace all files.
Unfortunately, since files get overwritten, you're forced to reload software (such as
Exchange Client) that updates system files. Windows 98's System File Checker utility can
examine Windows for changed or corrupted files and restore the original Windows 9X
version if necessary. To do so, click Scan For Altered Files and click the Start button.
As you're probably aware, every piece of hardware in your computer requires a driver.
Unfortunately, although several drivers come with Windows, many drivers are supplied by
third parties—and aren't tested by Microsoft. A poorly written driver can affect more than
just the device it controls: It can cause other devices to malfunction or result in General
Protection Faults.
With the release of Windows 98, Microsoft has begun digitally signing device drivers. The
digital signature means that the driver has been tested by Microsoft and is certified to work
with Windows 9X. The digital signature also proves that the driver hasn't been tampered
with. Third-party vendors can now submit their drivers to Microsoft for testing and
application of a digital signature.
Registry
Checker
Automatic
Skip Driver
Agent
Dr. Watson
The Signature Verification Tool will check your system for files that don't contain a digital
signature. It's important to point out that Microsoft doesn't sign every file—only driver files
contain digital signatures.
If a registry gets corrupted, it is very difficult to fix them. Many times, you end up losing
Windows completely and have to reinstall it from scratch. However, Windows 9X is much
more forgiving when it comes to registry problems. Each time Windows 9X successfully
boots, it creates a backup copy of the registry. Windows 9X stores several backup
registries in a compressed format.
When Windows 9X starts, the Registry Checker utility runs in the background. It scans for
corruption and, if no corruption is detected, creates a backup copy of your registry.
If a problem is detected, the utility reboots your computer and launches an MS-DOS
version of the utility, which automatically restores the most recent backup of your registry.
If no backup is available, the utility will attempt to correct the error it has detected. Since
the utility runs automatically, you may wonder why it's included in the System Information
tool. Keep in mind that the utility runs automatically at boot up. Since many people don't
shut down their computers at night, you may want to manually run the utility if you suspect
a problem.
If you have the boot fail because of a malfunctioning driver or hardware components, you
would typically boot to safe mode and guess which device may be causing the problem.
Then you must disable the device, reboot, and hope that you guessed right. Unfortunately,
this can be a time consuming activity. Windows 9X solves this problem with the Automatic
Skip Driver Agent. This utility monitors the boot process. If a boot fails because of a driver
or hardware problem, the agent makes a note of it. If a boot failure occurs twice in a row as
a result of the same problem, the agent will disable the driver on the third attempt. You can
then boot successfully and use the Automatic Skip Driver Agent to determine which driver
is causing your problem.
At one time or another, you've probably seen a cryptic Application Fault screen. If this
errors occurs on a regular basis, then you want to use Dr. Watson. When you launch Dr.
Watson, it runs minimized on your toolbar. Dr. Watson captures information about the error
when it occurs. It can then be analyzed by Microsoft technical support.
Intel Customer Support (ICS)
Troubleshooting the IA Platform
76
System
Configuration
Utility
ScanDisk
Version
Conflict
Manager
The System Configuration Utility allows you to perform a selective startup.
ScanDisk utility repairs hard disk errors and correct problems such as cross-linked files or
damaged file allocation tables. It can also flag bad hard disk sectors to prevent Windows
from writing information to them.
The Version Conflict Manager will back up all altered files before switching to the Windows
98 version so that you can revert to the previous version if you have problems.
13.5 Terminate a Specific Application
Windows 98 lets you terminate a specific application without affecting other currently
running processes. To terminate an application, press Ctrl+Alt+Del, which opens a
Close Program dialog box. Listed in this dialog box are all currently running tasks
(including system processes not otherwise listed on the taskbar). You must then select
a process (Not Responding usually is indicated in brackets next to the process name)
and click on the End Task button. The operating system then attempts to terminate the
process (which might take several seconds). Depending on the reason why the
application is hung, you also might be presented with the option to wait a few seconds
for the application to respond, and then to terminate the application if no response is
received.
13.6 General Protection Faults and Exception Errors
Although applications should normally run without interruption, situations do arise when,
due to either programming errors or incompatibilities, applications cease to function
properly. Knowing how to handle application problems is important in being able to
administer the operating system properly. The two main problems that occur with
applications are General Protection Faults (GPFs) and application hangs.
A General Protection Fault (GPF) typically is caused by an application that attempts to
violate system integrity in one of a number of ways:
1. By making a request to read or write to a memory address space owned by
another application.
2. By attempting to access the system hardware directly.
3. By attempting to interact with a failing hardware driver. (Drivers operate at Ring
0, and so can seriously impact the operating system.)
The GPF is generated when the operating system shuts down an offending application
to prevent a system integrity violation. How the offending application is specifically
handled depends on its application type. Because MS-DOS applications reside in their
own virtual memory and have their own message queue, if they cause a GPF, a
message is displayed and the application is terminated without impacting the rest of the
operating system.
In the case of WIN16 (Windows 3.XX) applications, the procedure is somewhat more
complex. Because WIN16 applications share both a common address space and a
Intel Customer Support (ICS)
Troubleshooting the IA Platform
77
common message queue, when one application creates a GPF, all others are
suspended until the offending application is terminated. After this is done, the remaining
applications resume processing.
Finally, with 32-bit applications, the procedure is quite straightforward. Because 32-bit
applications exist in their own separate address space, and each has a separate
message queue, a GPF in one 32-bit application in no way affects any other 16- or 32bit programs. Only the offending program is terminated.
Dr. Watson is a utility that tabulates information about the system at the time of an
application fault. In order to tabulate this system information, Dr. Watson must be
running at the time of the fault. If you know how to trigger the fault (for instance, if the
fault always occurs when you open a specific application), start the Dr. Watson
application, then trigger the fault. Dr. Watson will record the state of the software
environment at the time of the failure. After the fault, the Details button in the error
message dialog box will provide the information collected by Dr. Watson.
To start Dr. Watson, do the following:
1. Click on the Start menu and choose Programs, Accessories, System Tools.
2. Choose System Information from the System Tools menu.
3. In Microsoft System Information Utility, pull down the Tools menu and choose Dr.
Watson.
Dr. Watson will start minimized in the system tray (next to the clock). When you click on
the Dr. Watson icon, it will generate a snapshot of your system. After Dr. Watson
records the system information, the Dr. Watson main window appears on your screen.
The default view is the standard view showing the Diagnosis tab, which provides
relevant diagnostic information based on the state of the system at the time the
snapshot was recorded.
Note that the Diagnosis tab includes a space for the user to write a note to the support
technician describing what was happening at the time of the failure. For additional
information on the state of the system at the time of the snapshot, choose Advanced
View in the View menu. Dr. Watson’s Advanced view provides several tabs describing
different aspects of the system at the time of the snapshot. See figure 20.
When an application fault occurs, Dr. Watson saves the logged information to a file
called watsonxx.wlg in the \Windows\Drwatson directory. xx (in watsonxx.wlg) is an
incremented number. You can also save a snapshot to a file using the Save and Save
As options in the Dr. Watson File menu. You can view the contents of a Dr. Watson log
file by opening it using the Open Log File option in the Dr. Watson File menu. You can
also open a Dr. Watson in Microsoft System Information Utility.
Intel Customer Support (ICS)
Troubleshooting the IA Platform
78
Figure 20: Dr. Watson Utility
If you know an application fault will occur again but you don’t know how to reproduce it,
Microsoft suggests that you add a shortcut to Dr. Watson to the Start, Programs,
Startup folder for your computer or user profile. Putting a shortcut to Dr. Watson in the
Startup folder will cause Dr. Watson to start automatically when the system starts.
If a General Protection Fault occurs repeatedly, isolating a GPF can be a long and
lengthy process. If the General Protection Fault message indicates a GPF with the
GDI.EXE, USER.EXE or Video driver, the problem is most likely caused by the video
system and driver. Try to use a new video driver from the manufacturer or try to use
Windows 95 generic video drivers (the generic SVGA or VGA drivers). In addition, you
can use the Video Performance tab and move the slide bar a notch or two to the left.
The Video Performance tab can be accessed by starting the Control Panel, doubleclicking on the System applet, clicking on the Performance tab followed by clicking on
the Graphics button.
If the General Protection Fault occurs with the same application, check for lost clusters
with ScanDisk, check for viruses and delete any temporary files from the temporary
directory (defined with the SET TEMP line in the AUTOEXEC.BAT file). Remember that
both lost clusters and many temporary files in the temporary directory often indicate that
a user has been shutting of the computer without properly exiting Window. You should
also check to see if the software that you are using has a history of causing general
protection faults and to see if there is a patch or fix that will correct the problem.
After checking for lost clusters, viruses and temporary files, try to reinstall the program
so that any incomplete files will be restored. If the problem still exist, try to reinstall
Windows.
Intel Customer Support (ICS)
Troubleshooting the IA Platform
79
One cause of a general protection fault is an invalid page fault (PFE). An invalid page
fault occurs when the operating system cannot find specific data that is stored in the
RAM or virtual memory. It occurs more often when the system has little RAM or the hard
drive is badly fragmented. Therefore, systems with more RAM and lots of free disk
space on an unfragmented hard drive are more resistant against general protection
faults.
Lastly, GPFs can be caused by faulty RAM cache or faulty RAM chips. To determine if
you have faulty RAM cache, you can disable it usually using the CMOS Setup program
or by using jumpers/DIP switches on the motherboard. If the problem goes away, you
know it is the RAM cache. If the problem still exist, you can enable the RAM cache and
check elsewhere. To isolate the RAM, remove a bank of RAM chips at a time (if you
have a lot of RAM) or swap a bank with a known good RAM chips.
13.7 Performance and System Monitor
Wherever possible, Windows 9X is self-tuning, adjusting cache sizes or other elements
of the system environment to provide the best performance for the current configuration.
Windows 9X can also detect when the loaded drivers or other performance-related
components are not providing the optimal performance. In Control Panel, double-click
System, and then click the Performance tab. Windows 9X reports the current
performance status, including whether 32-bit, protected-mode components are being
used.
13.7.1 Optimizing the Swap File
Virtual memory is disk pretending to be RAM. For Windows 9X virtual memory is known
as a swap file. By default, the Windows 9X swap file is dynamic. If hard disk space is
running low, the size of the swap file is decreased. If hard disk space is available and
more memory is required, the size of the swap file is increased.
The single best way you can ensure high swap file performance is to make sure that the
disk containing the swap file has ample free space so that the swap file size can shrink
and grow as needed. If you completely disabling virtual memory might cause the
computer to stop operating properly. You might not be able to restart the computer, or
system performance might be degraded. If you have less than 10 MB of free disk space
on the drive where the swap file is contained, may cause numerous read/write errors,
general protection faults and drastically slow performance.
The most common reason for changing the default swap file is to move it to some drive
other than the C drive. Many systems tend to fill up the C drive, so little or no room is
left for the swap file. The swap file can use only the free space on a drive. When the
space is filled, the swap file cannot get any larger, resulting in the “Not Enough Memory”
error. The solution is to move the swap file to another drive. To move the file in
Windows 9X, click on the Let me specify my own virtual memory settings and select
another drive.
Intel Customer Support (ICS)
Troubleshooting the IA Platform
80
Although the system defaults usually provide the best performance, you can adjust the
parameters used to define the swap file. For example, to optimize swap file
performance on a computer with multiple hard disk drives, you might want to override
the default location of the Windows swap file. The swap file should be placed on the
drive with the fastest performance, unless that disk is overused. If a user usually loads
all software from the same drive in a computer that has multiple drives, placing the
swap file on one of the drives that is not as busy might boost performance.
You can manually set the minimum and maximum file sizes by selecting the Let me
specify my own virtual memory settings button in the Virtual Memory dialogue box.
Entering the same number for Minimum and Maximum will fix the file's size. This may
improve system performance because Windows no longer has to adjust the file size.
However, you risk a memory shortage if the file size isn't large enough.
To adjust the virtual memory swap file
1. In Control Panel, double-click System, click the Performance tab, and then click
Virtual Memory.
2. To specify a different hard disk, click the Let me specify my own virtual memory
settings option. Then specify the new disk in the Hard disk box. Or type values
(in kilobytes) in the Minimum or Maximum box. Then click OK
Windows 9x allows the swap file to become fragmented on the disk—that is, not stored
in a single, contiguous block. In systems that make heavy use of virtual memory, this
fragmentation can take a steep toll on performance. The easiest way to defragment the
swap file is to use Norton Utilities' Speed Disk utility. You can also try Windows 98's
Disk Defragmenter utility (under the System Tools menu). But since Disk Defragmenter
treats the swap file as immovable, you'll first have to move the swap file to a second
disk or partition—using the swap file "Hard disk" setting—then defragment the disk and
finally move the file back. If you have just one disk, check Disable virtual memory,
reboot, defrag the disk, enable virtual memory and reboot again. But beware: If your
system doesn't have enough RAM to run without virtual memory, it may not boot.
13.7.2 Optimizing File System Performance
In Windows, the disk cache is dynamic. You do not need to configure its size as part of
system configuration. Changing the cache size is not a good method of limiting paging.
Paging through the cache would quickly overwhelm it and make it useless for other file
I/O. Although swap file I/O operations do not go through the cache, memory-mapped
files and executable files do. The cache, however, is designed to make sure it cannot be
overwhelmed by such I/O operations.
Changing the cache size (even if you could) probably would not have much effect on
paging. The cache grows and shrinks as needed. If the system begins to page a lot, the
Intel Customer Support (ICS)
Troubleshooting the IA Platform
81
cache shrinks automatically. However, people often think they are seeing a lot of
paging, but they are really seeing other disk activity.
A common reason for excessive paging is that the working set of the applications you
are running is greater than the amount of physical RAM available. If the amount of
paging is extreme, to the point where system performance is poor, a real-mode driver
for the hard disk may be the cause, and should be replaced with a protected mode
driver. If Windows 9X needs to use real-mode for its disk I/O operations, a lot of code
has to be locked down that would otherwise be pageable, and your working set
increases significantly. Note: Paging through a real-mode driver increases paging, and
on a computer with 16 MB of memory, it can cause unacceptable performance.
In Windows, file system and disk performance can be controlled based on how the
computer is used in most situations. If you double-click the system applet in the Control
Panel, click on the Performance tab and click the File System button, you can select the
typical role of the computer. This setting determines how much space is set aside in
main memory to handle path and filename caching. The default is Desktop Computer,
allocates space for 32 paths and 677 filenames, whereas the Network Server allocates
space for 64 paths and 2,729 filenames. Even if the computer is not a network server, if
there is more than 24 MB of RAM, it should be set to Network Server for better
performance. (Note: Network Server is not an available option on the original release of
Windows 95. To make this option available to the original version, the changes to the
Registry shown in table 15 would have to be made.
Table 15 Computer role descriptions
Role
Desktop computer
Mobile or docking
system
Network server
Description
A normal computer acting primarily as a network client, or an individual
computer with no networking. This configuration assumes that there is more
than the minimum required RAM, and that the computer is running on
power (rather than battery).
Any computer with limited memory. This configuration assumes that RAM is
limited and that the computer is running on battery, so the disk cache
should be flushed frequently.
A computer used primarily as a peer server for file or printer sharing. This
configuration assumes that the computer has adequate RAM and frequent
disk activity, so the system is optimized for a high amount of disk access.
The 32-bit disk access feature is always turned on in Windows 98 unless Windows 98
detects a real-mode disk driver that does not have a protected-mode replacement. This
could be, for example, an older Stacker driver or a hard-disk security or encryption
driver for a disk drive. To prevent the performance loss that occurs when Windows 98 is
forced to use a real-mode disk driver, upgrade to a protected-mode replacement for that
driver. If you need to determine why a Windows 98 real-mode disk driver was installed,
check the Ios.log file. The Ios.log file in the Windows directory is created when a
protected-mode driver is not available or the operating system detects that an unknown
device driver is controlling a device. In most cases, the first line in Ios.log states why the
protected-mode driver was not loaded. If the first line mentions Mbrint13.sys, the
Intel Customer Support (ICS)
Troubleshooting the IA Platform
82
problem most likely is a virus (unless you are using a driver that replaces the master
boot record).
13.7.3 Optimizing CD-ROM File System Performance
The CD-ROM cache is separate from the cache used for disk file and network access
because the performance characteristics of the CD-ROM are different. This cache can
be paged to disk (the file and network cache cannot), reducing the working set for
Windows 9X but still allowing for better CD-ROM performance. When Windows 9X is
retrieving data from a compact disc, it is still faster to read a record from the cache even
if it has been paged to disk, because the disk-access time is much faster than the
compact disc–access time. Note: A small CD-ROM cache makes a big difference in
streaming performance, but a much larger cache does not pay off as significantly,
unless the cache is large enough to contain entire multimedia streams. In addition,
many multimedia programs perform better with a smaller cache, because they tend not
to reuse data.
To set the supplemental cache size for CDFS:
1. In Control Panel, double-click System, click the Performance tab, and then click
File System.
2. Click the CD-ROM tab, and drag the slider to set the Supplemental cache size.
Move the Supplemental Cache Size slider to the right to allocate more RAM for
caching data from the CD-ROM drive or to the left to allocate less RAM for
caching data.
3. In the Optimize access pattern for box, select a setting based on your computer's
CD-ROM drive speed. Table 16 shows the size of the cache created for each
CD-ROM drive speed setting.
Table 16 Optimizing cache size
Cache
CD-ROM drive speed
size
No read-ahead
1088 KB
Single-speed drives
1088 KB
Double-speed drives
1138 KB
Triple-speed drives
1188 KB
Quad-speed or higher
1238 KB
4. Click OK, and then shut down and restart the computer.
13.7.4 Using File System Troubleshooting Options
The System option in Control Panel presents a set of options for changing file system
performance. You can use these options when you experience rare hardware or
software compatibility problems. It is important to note that enabling any of the file
Intel Customer Support (ICS)
Troubleshooting the IA Platform
83
system troubleshooting options will seriously degrade system performance. Typically,
you want to enable these options only if instructed to do so by a product support
representative. Otherwise, these options should rarely, if ever, be used.
To display the file system troubleshooting options
1. In Control Panel, double-click System, and then click the Performance tab.
2. Click File System, and then click the Troubleshooting tab.
Table 17 summarizes the settings in Troubleshooting properties. Each option sets a
value in the following registry key.
HKEY_LOCAL_MACHINE \System \CurrentControlSet \Control \FileSystem
Table 17: File system troubleshooting options
File system
Description
option
Disable new file This option alters the internal rules for file sharing and locking on
hard disks, governing whether certain processes can have access to
sharing and
open files in certain share modes that guarantee a file will not be
locking
modified. This option should be checked only in the rare case that an
semantics.
MS-DOS-based application has problems with sharing under
Windows 98. This sets SoftCompatMode=0 in the registry.
This option turns off the tunneling feature, which preserves long file
Disable long
names when files are opened and saved by applications that do not
name
preservation for recognize long file names. This option should be checked in the rare
case that an important legacy application is not compatible with long
old programs.
file names. This sets PreserveLongNames=0 in the registry.
This option prevents Windows 98 from terminating interrupts from the
Disable
protected-mode hard disk controller and bypassing the read-only memory (ROM)
routine that handles these interrupts. Some hard disk drives might
hard disk
require this option to be checked in order for interrupts to be
interrupt
processed correctly. If this option is checked, the ROM routine
handling.
handles the interrupts, slowing system performance. This sets
VirtualHDIRQ=1 in the registry. This setting is off by default in
Windows 98.
The file commit API is used to guarantee integrity of user data that is
Disable
being written by an application to a disk. Normally, the file commit
synchronous
buffer commits. API is used by applications to ensure that critical data that the
application is writing is written to the disk before returning from a call
made to the file commit API. Choosing this option disables this
feature. Data is still written to disk, but it is written to disk in the
background at the discretion of the file system. Choosing this option
can compromise data written to disk by an application should the
system crash before the data is actually written. This option was
added to allow adequate performance of a defective database
application that used the file commit API incorrectly and excessively.
Intel Customer Support (ICS)
Troubleshooting the IA Platform
84
Disable all 32-bit This option ensures that no 32-bit disk drivers are loaded in the
protected-mode system, except the floppy driver. Typically, you would check this
option if the computer does not start because of disk peripheral I/O
disk drivers.
problems. If this option is enabled, all I/O will go through real-mode
drivers or the BIOS. Notice that in this case, all disk drives that are
visible only in protected mode will no longer be visible. This sets
ForceRMIO=1 in the registry.
This option ensures that all data is flushed continually to the hard
Disable writebehind caching disk, removing any performance benefits gained from disk caching.
This option should be checked only in the rare case that you are
for all drives.
performing risky operations and must ensure prevention of data loss.
For example, a software developer who is debugging data at Ring 0
while creating a VxD would check this option. This sets
DriveWriteBehind=0 in the registry.
13.7.5 Graphics Compatibility Options
In Windows 9X, graphics hardware acceleration features can be turned off when system
performance indicates incompatibility problems. Specifically, problems can occur when
Windows 9X assumes a display adapter can support certain functionality that it cannot.
In such cases, the side effects might be anything from small irregularities on the screen
to system failure. You can disable hardware acceleration features of the display adapter
so that the computer can still be used if there is a problem with the display adapter. If
changing these settings fixes otherwise unexplained system crashes or performance
problems, the source of the problem is probably the computer's display adapter.
To change graphics performance settings, double-click the System applet in the Control
Panel, click the Performance tab, and then click Graphics. Drag the slider to change the
Hardware acceleration setting, as summarized in the following list. Then click OK.
The default setting is Full, which turns on all graphics hardware acceleration features
available in the display driver. The first notch from the right can be set to correct mouse
pointer display problems. This setting disables hardware cursor support in the display
driver by adding SwCursor=1 to the [Display] section of System.ini. The second notch
from the right can be set to correct certain display errors. This setting prevents some bit
block transfers from being performed on the display card and disables memory-mapped
I/O for some display drivers. This setting adds SwCursor=1 and Mmio=0 to the [Display]
section of System.ini, and SafeMode=1 to the [Windows] section of Win.ini. The last
notch from the right (None) can be set to correct problems if your computer frequently
stops responding to input, or has other severe problems. This setting adds SafeMode=2
to the [Windows] section of Win.ini, which removes all driver acceleration support and
causes Windows 9X to use only the device-independent bitmap (DIB) engine rather
than bit block transfers for displaying images.
Intel Customer Support (ICS)
Troubleshooting the IA Platform
85
13.7.6 Optimizing Printing
The way printing to a printer attached to a file or print server occurs depends on your
server's operating system. If you print to a server running Windows 9X, the rendering
from the Enhanced Metafile graphics (EMF) format to the printer-specific language
happens on the server. This means that less work is performed on the client computer,
giving the user better performance.
When you print to NetWare or Windows NT servers, the rendering from EMF to the
printer-specific format happens on the client computer. Although this happens in the
background, it still means more work is performed on the client computer. Printing to a
printer attached locally causes both the EMF rendering and the device-specific
rendering to happen on the computer.
To define spool settings for print performance, double-click Printers in the Control
Panel, right-click a printer icon, and then click Properties. Then click the Details tab, and
then click Spool Settings. Select Spool print jobs so program finishes printing faster, and
then click one of the following options:
•
•
Click Start printing after last page is spooled if you want the return-to-application
time to be faster. This requires more disk space and increases the total print
time. The second rendering does not start until the entire file is written to the
EMF file, decreasing the amount of work performed on the computer as you print,
but increasing the disk space, because the entire file has to be written before the
second rendering starts.
Click Start printing after first page is spooled if you want the second rendering to
take place simultaneously with the writing of the EMF file. This reduces the total
print time and disk space required, but it increases the return-to-application time.
13.7.7 Windows 9X Resources
The kernel loads and runs Windows applications and handles their memory
management, the GDI manages graphics and printing, and USER controls user input
and output devices, including the keyboard and the mouse.
The GDI and USER use storage areas known as the local heap. The GDI heap contains
information about graphical objects, such as pens, brushes, cursors, fonts, icons,
buttons, scrollbars, and so on. The USER heap contains information about windows,
icons, menus, and dialog boxes. Windows 9X has larger heaps than Windows 3.XX
applications, including a larger 32-bit GDI heap, but it retains the same 16-bit GDI heap
as Windows 3.XX so that it can support older programs. If any heap drops too low, the
system can become unstable and cause problems even through there is a lot of free
RAM. To view the amount of free system resources (the percentage of the heap, that
has the lowest amount of free memory), use the Performance tab in the System applet
in the Control Panel. To see the mount of free system resources for the System, USER,
and GDI, use a utility called Resource Meter. See figure 21.
Intel Customer Support (ICS)
Troubleshooting the IA Platform
86
Figure 21: Windows 95 Resource Meter
13.7.8 Using System Monitor
System Monitor is a useful tool to help measure the performance of hardware, software,
services, and applications to determine what is acting as a bottleneck and slowing the
system. By using real-time graphs, it can provide a snapshot of key performance
statistics on memory, processor and other settings. Running System Monitor before
making a change to the system and then running it after making a change shows how
much the change has affected performance. See figure 22.
Figure 22: Windows 95 System Monitor
Intel Customer Support (ICS)
Troubleshooting the IA Platform
87
To install System Monitor:
1.
2.
3.
4.
In Control Panel, double-click Add/Remove Programs.
Click the Windows Setup tab.
Click System Tools, and then click Details.
Click System Monitor, and then click OK.
To run System Monitor:
On the Start menu, point to Programs, point to Accessories, point to System Tools,
and then click System Monitor.
To start logging:
1. In System Monitor, click File, and then click Start Logging.
2. Type a file name for the log file, and then click Save.
3. On the File menu, click Stop Logging to stop logging.
To edit an item in a chart:
1. In System Monitor, click Edit, and then click Edit Item.
2. Click Add Item, Remove Item, or Edit Item, and then make the changes you
want. Note To view a definition of an item in the Item list, click the item, and then
click Explain.
3. Click OK. Note You can edit only a chart that you are currently viewing.
To change the look of a chart:
In System Monitor, click View, and then click Line Charts, Bar Charts, or Numeric
Charts. Note: You can edit only a chart that you are currently viewing.
To use System Monitor to track performance problems:
1. In System Monitor, click the Edit menu, and then click Add Item.
2. In the Category list, click the resource that you want to monitor.
3. In the Item list, select one or more resources that you want to monitor. To select
more than one item, press CTRL while clicking the items that you want to select.
To select several items in a row, click the first item, and then press and hold
down SHIFT while clicking the last item.
4. Click Explain for more information about a selected resource.
5. Click OK. A performance chart of the resource is added to System Monitor.
6. To change the view of the data from a line chart to a bar chart or a numeric
listing, click the related button on the toolbar.
System Monitor offers menu commands for configuring the charts:
Intel Customer Support (ICS)
Troubleshooting the IA Platform
88
•
•
•
To change the update interval, click the Options menu, and then click Chart.
To configure the color and scaling for a selected item, click the Edit menu, and
then click Edit Item.
To control the display of the toolbar, status bar, and title bar, click the View
menu, and then click Toolbar, Status Bar, or Hide Title Bar, respectively.
If you browse System Monitor, you will find that there are far more statistics available
than you would ever want to put on a single screen. Most of them are too technical for
the average user. But a handful of statistics are useful to just to just about every PC
user. See table 18.
Table 18: Useful Statistics for System Monitor
Statistic
Description
Statistic Usage
Kernel:Processor
Usage (%)
Specifies the percentage
of the time that the
processor is doing work.
Kernel:Threads
Specifies the Active
threads, small pieces of
software that occupy
RAM.
Memory
Manager:Unused
physical memory
Specifies the amount of
physical RAM that is still
free for use.
Memory
Manager:Swapfile size
Windows uses a swap
file (also called virtual
memory) as a temporary
holding area for RAM
data not currently being
used.
This statistic indicates
the amount of RAM data
that is actually stored in
the swap file at any given
moment.
These indicate how
much the paging file has
to be accessed.
If your processor usage consistently runs at over 80%, you're
overtaxing your PC. The underlying cause may be insufficient
memory, too many programs running at once or a corrupt
program that won’t release the processor. In some cases
adding memory will fix the problem. In others you need to get a
more powerful system.
The right number of threads for your PC depends on the
software you happen to be running. For this to be useful, you
need to perform a baseline, a measurement of normal use. A
non-networked Windows 98 machine desktop system usually
has 50 to 70 active threads with no other applications open.
If you run a program, it will typically increase the number of
threads when open but does not release those threads when
closed (memory leak). In Windows 9x, you can usually release
stranded threads by closing the offending application. If a
newer 32-bit program (one written for Windows 9x) is causing a
memory leak, the source of trouble may very well be a
corrupted file. If this is the case, you need to reinstall the
application.
You may be surprised at how little RAM remains available
under Windows, even when only a few applications are
running. This is because Windows constantly moves data in
and out of the swap file on the hard disk. The "Unused physical
memory" setting is most useful when viewed in conjunction with
the following six memory statistics.
If your system has a limited amount of hard-disk space, you
can use this statistic to balance your hard-disk storage needs
against Windows' memory needs.
Memory
Manager:Swapfile in
use
Memory Manager:Page
faults
Memory Manager:Pageouts
Memory
This identifies the total
Intel Customer Support (ICS)
This space intentionally left blank
For this to be useful, you need to perform a baseline, a
measurement of normal use. If either of these two statistics
jumps to higher-than-normal levels, Windows may be relying
too heavily on the swap file. If the increase coincides with
sluggish performance, you need to add more physical RAM to
your system.
To determine exactly how much RAM a given program
Troubleshooting the IA Platform
89
Manager:Allocated
memory
amount of data that
Windows is manipulating
in memory.
Memory
Manager:Locked
memory
Locked memory refers to
the amount of data that
must remain in physical
RAM and can't be
swapped to the hard
disk.
This statistic reports the
amount of RAM allocated
to caching hard-disk
data.
Memory Manager:Disk
cache size
requires, subtract the "Memory Manager: Disk cache size"
value from the amount of "Allocated memory", with and without
the program running. The difference is the amount of RAM
used by that program.
If an application forces a high percentage of data to be locked
in physical RAM, other applications' performance can slow
because a large amount of data must be shuttled on and off the
hard disk.
On systems running the original version of Windows 95 with the
FAT16 file system, you may be able to rescue a few megabytes
of RAM by lowering this setting. Watch your disk cache values
to determine your system's maximum cache needs. If you see
more than a megabyte of difference between what your disk
cache uses and its fixed maximum value — which is
determined by the MaxFile Cache= setting listed under [vcache]
in your system.ini file — you can recover some of the wasted
RAM by lowering the MaxFileCache= setting. Of course, the
memory you save may not be worth the effort if your system
has more than 32MB of RAM. But if you're working with 32MB
of memory or less, your applications will benefit from the extra
RAM.
14 WINDOWS NT, WINDOWS 2000 and Windows XP
14.1 Windows 2000 and Windows XP Boot Sequence
During boot up, the first Windows 2000/Windows XP file read is NTLDR, which switches
the processor from real mode to protected mode and starts the appropriate minifile
system drivers (built into NTLDR) so that it can read the VFAT/FAT16, FAT32 or NTFS
file systems.
Next, NTLDR reads the BOOT.INI (if one is available) and displays the Boot Loader
Operating System Selection menu. NTLDR then loads the operating system (such as
Windows NT 4.0, Windows 2000 and Windows XP, Windows 9X or DOS) selected from
the menu. If you do not select an entry before the timer reaches zero, NTLDR loads the
default operating system specified in the BOOT.INI file. Note: if the BOOT.INI file is not
present NTLDR tries to load Windows 2000/Windows XP from the Windows
2000/Windows XP directory (typically C:\WINNT). In addition, the menu will not be
displayed if you only have the one operating system or if you have an operating system
that cannot read the NTFS partition and Windows 2000/Windows XP.
NTLDR runs NTDETECT.COM, which attempts to detect the bus/adapter type, serial
ports, floating-point coprocessor, floppy disks, keyboard, mouse/pointing device, parallel
ports, SCSI adapters and video adapters. If you have a system that has a SCSI hard
disk, for which the BIOS on the SCSI adapter is disabled, NTLDR will load
NTBOOTDD.SYS to access the SCSI devices during boot up.
After NTLDR collects the hardware information, it then gives you an option to press the
Spacebar to invoke the Hardware Profile/Configuration Recovery menu. This menu will
Intel Customer Support (ICS)
Troubleshooting the IA Platform
90
list the hardware profiles that you have saved on your drive and list the Last Known
Good configuration option.
The NTLDR finally loads the NTOSKRNL followed by the hardware abstraction layer
(HAL.DLL). It will then load the HKEY_LOCAL_MACHINE\SYSTEM registry key from
systemroot\system32\CONFIG\SYSTEM directory). The SYSTEM hive specifies which
device drivers to load during boot up. When the device drivers are loaded, a hardware
list is made and stored in the Registry. Lastly, the Session Manager (SMSS.EXE) is
loaded which loads the appropriate services needed for Windows to function. Lastly, the
NTLDR initializes the NT kernel (NTOSKRNL.EXE) and takes control of the boot
process.
The active partition that contains the NTLDR and BOOT.INI file is known as the system
partition. The partition that contains the Windows 2000 operating system files is called
the boot partition. If a system has only one partition, which contains the initial boot files
and the Windows 2000 directory, the partition is both the system partition and the boot
partition.
DOS or Windows 9X uses a different volume boot sector. In addition, Windows NT 4.0,
Windows 2000 or Windows XP starts with NTLDR while DOS, Windows 9X start with
IO.SYS. If you have already DOS or Windows 9X boot files on a partition when you
install Windows 2000 or Windows XP, the old boot sector gets copied into a file called
BOOTSECT.DOS. Therefore, when you select the old operating system from the boot
menu, NTLDR loads BOOTSECT.DOS and passes control to it. The operating system
then starts up as normal. Note: if you are dual booting between Windows
2000/Windows XP and DOS or Windows 9X, DOS, Windows 9X cannot read NTFS
partitions.
During boot up, the BOOT.INI file provides a Boot Loader Operating System Selection
menu, which allows the selection between multiple operating systems. The BOOT.INI
file is a read-only, hidden system text file located in the root directory of the system
partition. See figure 23.
[boot loader]
timeout=30
default=multi(0)disk(0)rdisk(0)partition(1)\WINNT
[operating systems]
multi(0)disk(0)rdisk(0)partition(1)\WINNT="Microsoft Windows 2000 Advanced Server" /fastdetect
multi(0)disk(0)rdisk(1)partition(1)\WINNT="Microsoft Windows 2000 Professional" /fastdetect
C:\=“Windows 98”
Figure 23: Typical BOOT.INI file
The BOOT.INI file is divided into two sections, [boot loader] (see table 19) and
[operating system]. The entries in the [boot loader] section configures the number the
seconds that the Boot Loader Operating System Selection menu appears on the screen
and the default operating system loaded.
Intel Customer Support (ICS)
Troubleshooting the IA Platform
91
Table 19: The Settings in the [boot loader] section of the BOOT.INI file
Timeout=XX
Specifies the number of seconds the user has to select an
operating system from the boot loader screen before
NTLDR loads the default operating system. If the value is
0, NTLDR immediately starts the default operating system
without displaying the boot loader screen.
Default=
The ARC path to the default operating system
The [operating systems] section contains the list of available operating systems. Each
entry includes an ARC path to the boot partition for the operating system, the string to
display in the boot loader screen, and optional parameters. The optional parameters are
shown in table 20.
Table 20: Optional Parameters found in the BOOT.INI file
/3GB
Changes the standard of allocating 2 GB for each process and 2
GB for the system to 3 GB for each process and 1 GB for the
system.
/BASEVIDEO
Specifies that Windows NT uses the standard VGA video driver.
/fastdetect=[comx|
Disables serial mouse detection. Without a port specification, this
comx,y,z]
switch disables peripherals detection on all COM ports.
/maxmem:n
Specifies the amount of RAM that Windows 2000 uses. Use this
switch if you suspect is bad.
/noguiboot
Boots the computer without displaying the graphical boot status
screen.
/SOS
Displays the device driver names while they are being loaded.
An ARC (Advanced RISC Computing) path is used to specify the location (partition) of
an operating system. It follows the format:
multi(x)disk(y)rdisk(z)partition(a).
or
scsi(x)disk(y)rdisk(x)partition(a)
SCSI is used for a SCSI disk with its BIOS disabled. MULTI is used for disks other than
SCSI or a SCSI disk with its BIOS enabled. The number after MULTI is the ordinal
number of the hardware adapter card starting from 0. The number after DISK is the
SCSI bus number and will always be 0 for a non-SCSI disk or for a SCSI disk with its
BIOS enabled. The number after RDISK is the ordinal number of the disk starting from
0. The number after PARTITION is the ordinal number of the partition. Unlike the other
values, it starts at 1. Note: Primary partitions are numbered first followed by the logical
drivers. Therefore, if you have a primary partition and a logical partition in the extended
partition, the primary partition will be 1 and the logical partition will be 2. If you then add
Intel Customer Support (ICS)
Troubleshooting the IA Platform
92
another primary partition to the same drive, the first primary partition will be 1, the new
primary partition will be 2 and the logical driver will change to 3.
If you boot from a computer that does not use the correct ARC path, you could get one
of the following error messages:
BOOT: Couldn’t find NTLDR. Please insert another disk
NT could not start because the following file is missing or corrupt: \winntroot\
system32\ntoskrnl.exe
NTDETECT V1.0 Checking Hardware...
NTDETECT failed/missing
14.2 Boot Options for Windows NT, Windows 2000 and Windows XP
If you have problems where Windows 2000 cannot load or start, there are several tools
that you can use. They are the Advance boot menu, Windows Recovery Console and
Emergency Repair Disk.
If you press F8 while the Windows 2000/XP boot menu is displayed, it will display the
advance boot menu. Safe mode lets you start your system with a minimal set of device
drivers and services (PS/2 mouse, monitor, keyboard, mass storage, base video,
default system services) with no network connection. Typically, you would use safe
mode if you installed device driver or software that prevents the computer from starting,
you might be able to start the computer is safe mode in the hopes to remove the
software or device drivers from your system.
You can also use the Enable VGA mode option if the wrong video driver is loaded and
you need to force the video system into 640x480 with 16 colors and the Last Known
Good Configuration to start Windows 2000 using the last saved configuration stored in
the registry.
The Directory Service Restore Mode is a special version of safe mode that loads all
drivers and service and performs a CHKDSK on all volumes. After logging in, it allows
you to restore the Active Directory of the domain controller from backup media. After
getting into this mode, it will automatically set the computer to have the Active Directory
check all indices next time you boot the domain controller normally.
Lastly, the Debugging Mode has Windows 2000/XP send debug information through a
serial cable to another computer. This allows you to monitor the process of a server’s
boot from another server.
Note: All options except the Last Known good Configuration creates the
NTBTLOG.TXT file located in the \WINNT directory and uses the VGA driver
(resolution of 640X480 at 16 colors). The boot log can then be checked to see
what was the last thing loaded.
Intel Customer Support (ICS)
Troubleshooting the IA Platform
93
Windows XP also includes a System Configuration Utility (MSCONFIG.EXE) and that
lets you turn off or turn on specific entries in the Startup files system.ini, win.ini, boot.ini,
services and startup. You can also make other changes to the startup process that may
help with diagnosing startup problems. Note that, in addition to a normal startup, you
can choose Diagnostic startup, or you can choose to selectively disable files. The tabs
named for the Startup files let you enable or disable specific statements within the file.
See figure 24.
Figure 24: System Configuration Utility for Windows XP
If your computer doesn’t boot at all, even in safe mode, the best thing to use is the
Recovery Console. The Recovery Console provides a command-line interface that will
let you repair system problems using a limited set of command-line commands,
including enabling or disabling services, repairing a corrupted master boot record or
read and write data on a local drive (FAT, FAT32 or NTFS). Note: since the Recovery
Console is a very powerful tool, it should only be used by advanced users who have a
thorough knowledge of Windows 2000. In addition, you must be an administrator to use
the Recovery Console.
There are two ways to start the Recovery Console. You can run the Recovery Console
from your Windows 2000 Setup disks or from the Windows 2000 Professional CD or
you can install the Recovery Console on your computer to make it available in case you
are unable to restart Windows 2000.
To start the computer and use the Recovery Console:
1 Insert the Windows 2000 Setup compact disc (CD), or the first floppy disk you
created from the CD, in the appropriate drive: Of course, if you cannot boot from
the CD drive, you must use floppy disk. Restart the computer. If you are using
floppy disks, you will be asked to switch disks.
Intel Customer Support (ICS)
Troubleshooting the IA Platform
94
2 When the text-based part of Setup begins, follow the prompts and choose the
repair or recover option by pressing R. When prompted, choose the Recovery
Console by pressing C. Again, if you are using floppy disks, you will be asked to
switch disks.
3 If you have a dual-boot or multiple-boot system, choose the Windows 2000
installation that you need to access from the Recovery Console.
4 When prompted, type the Administrator password.
5 At the system prompt, type the Recovery Console commands.
6 To exit the Recovery Console and restart the computer, type EXIT.
To install the Recovery Console as a Startup menu option:
•
•
•
•
•
•
•
Insert the Windows 2000 Setup CD into your CD-ROM drive.
Click No when prompted to upgrade to Windows 2000.
At the command prompt, switch to your CD-ROM drive, and then type the
following:
\I386\WINNT32.EXE /CMDCONS
When it ask if you want to install the Recovery Console, click on the Yes button.
Click OK when the Setup is complete.
To run the Recovery Console, restart your computer and select the Recovery
Console option from the boot menu.
To get a list of the commands that are available in the Recovery Console, you just have
type in help while in the Recovery Console. To help about a specific command, type in
HELP commandname. See table 21.
Table 21: List of Commands Used in the Recovery Console
Command
Description
ATTRIB
Changes or displays attributes of files or directories
CD
Changes directory
CHKDSK
Executes a consistency check of the specified disk
CLS
Clears the screen
COPY
Copies a file
DEL
Deletes a file
DIR
List Directory contents
DISKPART
Adds and Deletes partitions.
ENABLE
Starts or enables a system service or a device driver
DISABLE
Stops or disables a system service or device driver
EXTRACT
Extracts a file from a compressed file
FIXBOOT
Writes a new partition boot sector onto the system partition
FIXMRB
Repairs the master boot record of the partition boot sector
FORMAT
Formats a disk
LISTSVC
Lists the services and device drivers available on the computer.
LOGON
Logs on to a Windows 2000 computer
MAP
Displays the drive letter mappings
Intel Customer Support (ICS)
Troubleshooting the IA Platform
95
MD
TYPE
RMDIR
REPAIR
Creates a directory
Displays a text file
Deletes a directory
Updates an installation with files with the Windows 2000 installation
CD
REN
Renames a file
SYSTEMROOT Changes to the installation system root directory (typically \WINNT)
The Emergency Repair Disk (ERD) is a disk that contains information about your
current Windows system settings. You can use this disk to repair your computer if it will
not start or your system files are damaged or erased.
To create an Emergency Repair Disk:
1. Open the Backup program.
2. On the Tools menu, click Create an Emergency Repair Disk. Note: you could
also use the Emergency Repair Disk wizard.
3. Insert a blank formatted floppy disk into the drive. Check the checkbox if you
want a backup copy of the registry copied to the REPAIR folder (highly
recommended) and click on the OK button.
4. Click on the OK button when the disk has been created.
5. Label the Emergency Repair Disk and put it in a safe place.
Note: When using the Emergency Repair Disk, it relies on information that has been
saved to the %systemroot%\REPAIR folder. Therefore, do not change or delete this
folder. In addition, you should recreate the ERD after each service pack, system change
or updated driver.
To use an Emergency Repair Disk for system repairs:
1. Insert the Windows 2000 Setup compact disc (CD), or the first floppy disk you
created from the CD, in the appropriate drive.
2. When the text-based part of Setup begins, follow the prompts; choose the repair
or recover option by pressing R.
3. When prompted, insert the Windows 2000 Setup CD in the appropriate drive.
4. When prompted, choose the emergency repair process by pressing R.
5. When prompted, choose between the following:
a. Manual Repair (press M): This should be used only by advanced users or
administrators. Use this option to choose whether you want to repair system
files, partition-boot sector problems, or startup environment problems.
b. Fast Repair (press F): This is the easiest option, and does not require input.
This option will attempt to repair problems related to system files, the partition
boot sector on your system disk, and your startup environment (if you have a
dual-boot or multiple-boot system).
6. Follow the instructions on the screen and, when prompted, insert the Emergency
Repair Disk in the appropriate drive.
Intel Customer Support (ICS)
Troubleshooting the IA Platform
96
7. During the repair process, missing or corrupted files are replaced with files from
the Windows 2000 CD or from the systemroot\Repair folder on the system
partition. Note: Follow the instructions on the screen; you might want to write
down the names of files that are detected as faulty or incorrect, to help you
diagnose how the system was damaged.
8. If the repair was successful, allow the process to complete; it will restart the
computer. The restarting of the computer indicates that replacement files were
successfully copied to the hard disk.
Windows XP includes a System Restore feature that enables users and administrators
to restore a computer to a previous state without losing data. System Restore
automatically creates easily identifiable restore points, which allow you to restore the
system to a previous time. Therefore, if you load a driver or software program which
causes the computer to act erratic or become unstable, you can use the System
Restore program to put Windows back to an earlier state before the problem occurred.
14.3 Event Viewer and Services Console
The event viewer is a very useful utility used to view and manage logs of system,
program, and security events on your computer. Event viewer gathers information about
hardware and software problems, and monitor Windows 2000 security events. For
example, if you get a “At least one service or driver failed during system startup” error
message during boot up, you should look at the event viewer to find out which service
or services have stopped. The event viewer can be executed by clicking on the Start
button, clicking on Programs, clicking on Administrative Tools and clicking on Event
Viewer or by adding it to the MMC console.
Windows 2000 event viewer starts with three kinds of logs:
Application log - The application log contains events logged by programs. For
example, a database program might record a file error in the programs log. Program
developers decide which events to monitor. The application log can be viewed by all
users.
Security log - The security log contains valid and invalid logon attempts as well as
events related to resource use, such as creating, opening, or deleting files or other
objects. For example, if you have enabled logon and logoff auditing, attempts to log
on to the system are recorded in the security log. By default, security logging is
turned off. To enable security logging, use Group Policy to set the Audit policy or by
changing the registry. To audit files and folders, you must be logged on as a
member of the Administrators group or have been granted the Manage auditing and
security log right in Group Policy. See Chapter 12 for policies and rights.
System log - The system log contains events logged by the Windows 2000 system
components. For example, the failure of a driver or other system component to load
during startup is recorded in the system log. The event types logged by system
Intel Customer Support (ICS)
Troubleshooting the IA Platform
97
components are predetermined by Windows 2000. The application log can be
viewed by all users.
All BackOffice family applications (Microsoft applications designed to work on Windows
NT or Windows 2000 servers) can post security events to the Windows 2000 event log.
In addition, several servers also make their own logs, which can also be viewed by the
Windows 2000 Event Viewer.
There are five types of Event Types, error, warning, information, success audit and
failure audit. They are shown table 22.
Table 22: Event Viewer Event Types
Event type
Description
A significant problem, such as loss of data or loss of functionality such
Error
as when a service fails during startup.
An event that is not necessarily significant, but may indicate a possible
future problem. For example, when disk space is low, a warning will be
Warning
logged.
An event that describes the successful operation of an application,
Information
driver, or service. For example, when a network driver loads
successfully, an Information event will be logged.
An audited security access attempt that succeeds. For example, a
Success Audit user's successful attempt to log on the system will be logged as a
Success Audit event.
An audited security access attempt that fails. For example, if a user
Failure Audit
tries to access a network drive and fails, the attempt will be logged as
a Failure Audit event.
When you double-click on an event, the Event Properties window will appear. The Event
Properties can be divided into two parts, event header and event description. The event
header information is shown in table 23.
Table 23: Event Header Information
Information
Meaning
The date the event occurred.
Date
Time
The local time the event occurred.
User
The user name of the user on whose behalf the event occurred.
Computer
The name of the computer where the event occurred. The computer name
is usually your own, unless you are viewing an event log on another
Windows 2000 computer.
Event ID
A number identifying the particular event type. The first line of the
description usually contains the name of the event type.
Source
The software that logged the event, which can be either a program name,
such as "SQL Server," or a component of the system or of a large
program, such as a driver name.
Type
A classification of the event severity: Error, Information, or Warning in the
Intel Customer Support (ICS)
Troubleshooting the IA Platform
98
Category
system and application logs; Success Audit or Failure Audit in the security
log. In Event Viewer's normal list view, these are represented by a symbol.
A classification of the event by the event source. This information is
primarily used in the security log. For example, for security audits, this
corresponds to one of the event types for which success or failure auditing
can be enabled in Group Policy.
The description of the event is the most important information within the Event
Properties window and will usually indicate what happened or the significance of the
event.
A service is a program, routine or process that performs a specific system function to
support other programs. To manage the services, you would use the Services console
(located under administrative tools) or the MMC with the Services snap-in. To start,
stop, pause, resume or restart services, you would right-click on the service and click on
the desired option. On the left of the service name, you can find a description.
To configure a service, you would right-click the service and click on the Properties
option, On the General tab, under the Startup type pull-down option, you can set the
following options:
•
•
•
Automatic – specifies that the service should start automatically when the system
starts
Manual – specifies that a user or a dependent service can start the service.
Services with manual startup do not start automatically when the system starts.
Disable – prevents the service from being started by the system, a user, or any
dependent service.
The workstation service allows a user sitting at the computer to access the network to
access network resources such as shared folders. The workstation is also known as the
Redirector.
The server allows a computer to provide network resources. When you pause the
Server service, only users in the computer's Administrators and Server Operators
groups will be able to make new connections to the computer. When you stop the
Server service, all users who are connected over the network to the computer will be
disconnected.
Of course, if you stop a Server service, the affected computer can no longer be
administered remotely and you must start the Server service locally. Of course, it is a
good idea to warn connected users before stopping the Server service.
14.4 Changing the HAL
If you get an error message stating that the Hardware Abstraction Layer is incorrect for
the machine, you will need to change the HAL. This usually occurs when one machine
Intel Customer Support (ICS)
Troubleshooting the IA Platform
99
mixes single and dual processor configuration files. In Windows NT, restart the install
and when the message “Windows NT is examining the hardware” appears, press F5.
Choose your computer type from the list that appears and continue with the installation.
In Windows 2000, you should begin the installation process again and install a clean
copy of OS.
If Windows 2000 was originally installed on a computer with a single processor and you
change your system to use multiple processors, then the HAL on your computer must
be updated so that it will recognize and use the multiple processors. To change the
HAL, take the following steps:
1.
2.
3.
4.
Click on the Start Button. Select Control Panel and then select System.
Click the Hardware tab and click on the Device Manager button.
Double-click the Computer branch to expand it.
Double-click the computer type listed under the Computer branch, click the
Drivers tab, click the Update Driver and then click Next.
5. Click “Display a list of known drivers for this device,” and then click “Show all
hardware for this device class.”
6. Click the appropriate computer type (one that matches your current type,
except for multiple processors), click Next, and then click Finish.
14.5 Stop Errors and Exception Errors
Applications run in the User mode layer (Intel protection model – ring 3). When an
application causes an error, Windows NT 4.0, Windows 2000 and Windows XP halts the
process and generates Illegal Operation error. Because every Win32 application has its
own virtual protected memory space, this error condition doesn’t affect any other Win32
program running. If the application tries to access the hardware without going through
the correct methods, Windows notices this and generates an exception error. In these
cases, if the application faults, Windows or you can close the offending program and
resume work.
The Windows NT kernel runs in the Kernel mode layer (Intel protection model – ring 0).
When the kernel encounters a fatal error such as a hardware problem, inconsistencies
within data necessary for its operation or similar error, Windows can displays a stop
error, sometimes known as a “blue screen of death”. These errors would be similar to a
General Protection Fault (GPF) for Windows 95/98. These errors can be caused by
hardware errors, corrupted files, corrupted file system or software glitch.
Even though a stop screen may look intimidating, only a small amount of data on the
stop screen is important in determining the cause of the error. At the top of the screen,
you will find the error code and parameters. In the middle of the screen you will find the
list of modules that have successfully loaded and initialized and at the bottom of the
screen you will find a list of modules that are currently on the stack.
Intel Customer Support (ICS)
Troubleshooting the IA Platform
100
You can configure Windows to write a memory dump file each time it generates a kernel
STOP error. This file contains all the information needed by the DUMPEXAM* utility to
troubleshoot the kernel STOP error. To make a memory dump file, you must have
sufficient space on a hard disk partition for the resulting memory dump file, which will be
as large as your RAM memory. Therefore, if you system has 256 MB of RAM, you must
have 256 MB of free disk space.
To configure the Windows to save STOP information to a memory dump file, you would
double-click on the System applet in the control panel. Within the Startup/Shutdown tab,
you would put a check in the Write debugging information under the Recovery section.
You can then specify the file location and name. Note: if the Overwrite any existing file
option does not have a check and there is a file with that name, Windows NT* will not
overwrite the file.
The DUMPEXAM utility is included on the Windows NT Server CD. It is a command-line
utility that examines a memory dump file (MEMORY.DMP), extracts information from it,
and writes it to a text file (MEMORY.TXT). The text file can then be used by support
personnel to determine the cause of the kernel STOP error. Note: the MEMORY.TXT
file is significantly smaller than the MEMORY.DMP. In many cases, the DUMPEXAM
analysis provides enough information for support personnel to determine the cause of
the error without directly accessing the memory dump file. See the Windows NT
Resource Kit for more information on the DUMPEXAM utility.
For Windows 2000 and Windows XP, you can create file dump by opening open the
Control Panel and double-clicking on the System applet. Then in the Advanced Tab,
click on the Startup and Recovery button. From there, you can select complete memory
dump, kernel memory dump or small memory dump.
When isolating any stop error, you should check the following:
•
•
•
•
•
•
•
•
•
•
Make sure that you have the most up-to-date BIOS with the most recent
microcode.
Using a recent anti-virus software package, check for viruses on your computer.
Check the Event Viewer, specifically the System Log to help determine the
device or driver causing the problem.
Disable any newly installed drivers and remove any newly added programs. Try
Known Last Good Configuration and Safe mode.
Make sure that you have updated drivers for your hardware device.
Remove any newly installed hardware.
Verify that all hardware is on the Microsoft Hardware Compatibility List (HCL). It
can be found at http://www.microsoft.com/hcl/default.asp.
Run a PC hardware diagnostic package on your system. While you should
perform a basic system that includes motherboard, you should do an intensive
check on the memory.
Disable memory caching in the BIOS, typically done in the CMOS setup program.
Be sure that you are using the newest Windows service pack.
Intel Customer Support (ICS)
Troubleshooting the IA Platform
101
When a stop error appears, search Microsoft’s web site for possible solutions. In any
case, make sure that the system has the most updated BIOS and Windows
NT/Windows 2000/Windows XP Service Pack. Also refer to section A10 in the Appendix
for specific Stop error messages.
The virtual device drivers (VxDs) is software that helps Windows communicate with
system hardware. If you get an error message that contains “Windows protection error”
indicates that one of the VxDs is corrupt, missing, or inaccessible. You may need to
reinstall Windows to fix this problem.
Runtime errors are errors that occur while an application is running. You may get a
runtime error if the application runs out of memory, hits some buggy code, or hiccups at
an inopportune time. Fortunately, runtime errors rarely affect the entire system. You can
resolve these problems by closing the affected application. The text of a runtime error
message should identify the affected application.
14.6 System Performance
The Performance option allows you to adjust how the processors resources are
distributed between running programs. Selecting applications assign more resources to
the foreground applications, while the background services assigns an equal amount of
resources to all programs. You should select the background services if it is a server
with network resources/services or an Applications server if you want the server to act
more like a desktop machine.
The Virtual memory area allows you to configure the paging file size (virtual memory)
and the maximum size of the registry. The minimum paging file is 2 MB. For Windows
2000 Professional, the default size of the file is equal to the total amount of RAM plus
12 MB, not to exceed the amount of available disk space. Usually the size of the paging
file can stay at the default value assigned during installation, but the recommend size
for the paging file should be 1.5 times the amount of RAM available on your system.
Note: if you are changing the virtual memory settings, you must make sure to click on
the OK button and not the cancel button.
As mentioned in earlier sections, virtual memory is much slower than physical RAM
because the memory is on the hard drive, a mechanical device. If you want to enhance
system performance and you have several physical hard drives (not necessary
partitions/logical drives), you can create a paging file on each disk and by moving the
paging file off the boot partition (the partition with the WINNT directory). This is because
the hard disk controller can read and write to multiple hard disks simultaneously and the
virtual memory management (VMM) tries to write the page data to the paging file on the
disk that is the least busy.
Lastly, you can enhance system performance by setting the initial size of the paging file
to the value displayed in the Maximum Size box in the Virtual Memory dialog. This is
Intel Customer Support (ICS)
Troubleshooting the IA Platform
102
because when the system needs more RAM beyond what it already has, it will increase
the size of the paging file. If the file is already set to its maximum size, it eliminates the
time required to enlarge the file from the initial size to the maximum size. Note: the
maximum size of the file is determined the largest contiguous (continuous) space of disk
space. Note: The disk defragmenter utility in the Accessories system tools is important
because it ensures that a maximum amount of the drive is contiguous and available for
page file use.
Note: For Pentium 4 systems, it is recommended to set the initial size and the maximum
size of the virtual memory to twice the amount of RAM, assuming that you have enough
free disk space. This way the paging file will not be fragmented and it does not have to
grown and shrink.
If an error code has been generated, the code appears in the Device Status box on the
General tab. In some cases, there will be a solution.
When you press the Ctrl+Alt+Delete keys in Windows NT 4.0, Windows 2000 or
Windows XP, you bring up the Windows Security dialog box. To bring up the Task
Manager button, you then click the Task Manager button, which will can be used to
bring a list of the current programs that are running and the overall CPU (including the
number of threads and processes) and memory usage. If the computer suddenly
appears slow, use the Task Manager to see which threads are taking up most of the
processor’s utilization.
Figure 25: The Task Manager can be used to display what programs are running and how much
processor utilization the program is using.
Intel Customer Support (ICS)
Troubleshooting the IA Platform
103
Performance is the overall effectiveness of how data moves through the system. To be
able to improve performance, you must determine the part of the system that is slowing
down the throughput; it could be the speed of the processor, the amount of RAM in the
system, the speed of the disk system, the speed of your network adapter card, or some
other factor. This limiting factor is referred to as the bottleneck of the system. With
System monitor/Performance Monitor, you can measure the performance of your own
computer or other computers on the network.
In Windows 2000, to open Performance Monitor, click the Start button, select the
Setting Option and click the Control Panel. In the Control Panel, double-click the
Administrative Tools and then double-click Performance.
When you first start System Monitor/Performance Monitor, all you see is a blank screen.
You must select the objects, instances and counters that you want to monitor. An object
is any Windows 2000 system component that possesses a set of measurable property.
It can be a physical part of the system such as the processor, RAM, disk subsystem and
network interface, a logical component such as a disk volume or software element such
as a process or a thread. An Instance shows how many occurrences of an object are
available in the system. A Counter represents one measurable characteristic of an
object. See figure 26.
Figure 26: Adding counters to performance monitor.
The performance of the computer is greatly affected by the performance of the
processor.
Processor Counter
Processor:% Processor Time
Intel Customer Support (ICS)
Description
% Processor Time measures how busy the processor
is. While the processor may jump up to 100%
processor usage, you need to look at the overall
average. If the processor is at 80% all the time, you
Troubleshooting the IA Platform
104
Processor:Interrupts/Second
System:Processor Queue
Length
should upgrade the processor (faster processor or
adding the number of processors) or move some of the
services to other systems. If you want to see what % of
the processor utilization each process is using, use the
task manager. Note: Do not use 3D screen savers
since they can consume 90% of the processor
utilization. Instead, use a blank screen saver.
Interrupts/Second measures how many hardware
interrupts per second are occurring. A server running
100 interrupts/second can be normal. If the counter is
increasing without a corresponding increase in the
server load may be due to a hardware problem. You
should always view this counter when you load a new
device driver to make sure that it functions properly. A
poorly written device driver can cause huge increases
in interrupt activity.
The number of threads indicated by the processor
queue length is a significant indicator of system
performance, because each thread requires a certain
number of processor cycles. If demand exceeds
supply, long processor queues develop and system
response suffers. Therefore, a sustained processor
queue length greater than 2 on a single processor
generally indicates that the processor is a bottleneck.
This counter is always 0 unless you are monitoring a
thread counter as well.
RAM is one of the biggest factors in PC performance. You can typically increase the PC
performance by adding more RAM. The best six RAM counters are:
Memory Counter
Memory:Available Bytes
Memory:Page Faults/Sec
Intel Customer Support (ICS)
Description
Available Bytes measures the amount of available
virtual memory. It is calculated by summing space on
the zeroed, free and standby memory lists. Free
memory is ready for use. Zeroed memory are pages of
memory filled with zeros to prevent later processes
from seeing data used by a previous process. Standby
memory is memory removed from a process’ working
set (physical memory) on route to disk, but is still
available to be recalled. If it is less than 4 MB, consider
adding more RAM.
A page fault occurs when a process attempts to access
a virtual memory page that is not available in its
working set in RAM. Hard page faults must be retrieved
from disk, which greatly slows performance. Soft page
faults are those that can be retrieved from the standby
Troubleshooting the IA Platform
105
Memory:Cache Bytes
Memory:Pages/sec
Paging File:% Usage
Paging File:%Usage Peak
list, and therefore do not require disk I/O. If this value is
high (20) or increasing, it is sign that you need to add
physical RAM to your server.
The Memory\Cache Bytes monitor the size of the File
System Cache. When memory is scarce the system
trims the cache, and when memory is ample the
system enlarges the cache. Note: you should compare
this to general memory availability. In addition, you
should keep track of how small the cache gets, and
how often that happens. This data is useful when
associating the size of the cache with its performance.
The pages/second indicates the number of requested
pages that were not immediately available in RAM and
had to be read from the disk or written to the disk to
make room in RAM for other pages. If your system
experiences a high rate of hard page faults, the value
for Pages/sec can be high.
% Usage of the Paging File is the percentage of space
allocated to the page file (virtual memory) that is
actually in use. To calculate how much more RAM you
would need to add to a server to minimize paging, you
would multiple the percentage by the size of the page
file.
The %Usage Peak is the highest percentage usage of
the page file. If this value frequently exceeds 90%,
allocate more space to the page file and possibly more
RAM. A 100% usage peak indicates that the server
has, at least momentarily, run out of both physical and
virtual memory.
Since one of the task of a server is to provide file access. Therefore, a hard drive
system is an important factor in server performance. The important hard drive counters
include:
Disk Counter
Physical Disk:% Disk Time
Physical Disk:% Avg. Disk
Queue Length
LogicalDisk:% Free Disk
Space
Intel Customer Support (ICS)
Description
The % Disk Time measures the percentage of elapsed
time that a disk drive is actually occupied in reading
data from and writing data to the disk. A value greater
than 90% indicates that the disk is the bottleneck.
The % Avg. Disk Queue Length is the average
number of read requests and write queued for the disk
in question. A sustained average higher than 2
indicates that the disk is being over utilized.
Free Disk Space reports the percentage of
unallocated disk space to the total usable space on
the logical volume.
Troubleshooting the IA Platform
106
LogicalDisk:Avg. Disk
Bytes/Transfer and
PhysicalDisk:Avg. Disk
Bytes/Transfer
LogicalDisk:Avg. Disk
sec/Transfer and
PhysicalDisk:Avg. Disk
sec/Transfer
LogicalDisk:Disk Bytes/sec
and PhysicalDisk:Disk
Bytes/sec
LogicalDisk:Disk
Transfers/sec and
PhysicalDisk:Disk
Transfers/sec
The Avg. Disk Bytes/Transfer measures the size of I/O
operations. The disk is efficient if it transfers large
amounts of data relatively quickly.
Avg. Disk sec/Transfer indicates how fast data is
being moved (in seconds). It measures the average
time of each data transfer, regardless of the number of
bytes read or written. A high value for this counter
might mean that the system is retrying requests due to
lengthy queuing or, less commonly, disk failures.
Disk Bytes/sec indicates the rate at which bytes are
transferred and is the primary measure of disk
throughput.
Disk Transfers/sec measures disk utilization by
indicating the number of read and writes completed
per second, regardless of how much data they involve.
If value exceeds 50 (per physical disk in the case of a
stripe set), then a bottleneck might be developing.
Note: By default, the system is set to collect physical drive data. For Windows NT 4.0
and Windows 2000, Logical drive data is not collected by default, you must enable it
specifically. To enable the disk counters, you would execute the DISKPERF -Y
command at the command prompt and restart the computer.
The Network Interface counters measure the performance of the network interface
cards. The more important counters include:
Network Interface Counter
Network Interface:Output
Queue Length
Network Interface:Bytes
Total/Sec
Network Interface:Packets
Outbound Errors
Network Interface:Packet
Received Errors
Intel Customer Support (ICS)
Description
Output Queue Length indicates the length of the
output packet queue. The value should be low.
Queues that are one or two items long constitute
satisfactory performance; longer queues mean that
the adapter is waiting for the network and cannot keep
pace with server requests.
The Bytes Total/Sec measures the rate that bytes,
including data and framing characters, are sent and
received on the network interface. If the value is close
to or matching the network capacity, the network may
be saturated.
The packets outbound errors are the number of
outbound packets that could not be send because of
error. If you are seen an increasing amount of errors
then more sophisticated troubleshooting tools need to
be used to figure out the exact problem.
The packet received errors are the number of inbound
packets that had to be discarded because they
Troubleshooting the IA Platform
107
contained errors that prevented them from being
delivered to the proper destination protocol stack.
There are several views in System Monitor including chart and report. A real-time
activity chart displays the value of the counter over time in a graph. You want to use a
chart to investigate why a computer or application is slow or inefficient, to continuously
monitor the system to find intermittent performance problems and to discover why the
capacity of a subsystem needs to be increased. A report view allows you to display
constantly changing counter and instance values for selected objects.
Figure 27: Performance Monitor Chart
Figure 28: Performance Monitor Report
To add counters, right-click the chart or report pane (right-pane) and select the Add
Counters option. After you have identified the counters you want to monitor, you can
Intel Customer Support (ICS)
Troubleshooting the IA Platform
108
save the information and reuse it later. To save chart settings, right-click the chart or
report pane and select the Save as option.
The System Monitor supports three types of logs: counter logs, trace logs and alert logs.
Counter logs record data from local or remote computers about hardware usage and
system service activity. Trace logs are event-driven, recording monitored data such as
disk I/O or page faults. When a traced event occurs, it is recorded in the log. Alert logs
take trace logs one-step further. Alert logs are trace logs that monitor counters, wait for
them to exceed user-defined tolerances and log the event. You can even setup the alert
log to send a message or run an application when a particular value is exceeded.
To create a log, click on the Performance Logs and Alert in the left pane of System
Monitor, and open the folder for the type of log that you want. Right-click the empty
space in the Details window and choose an option from creating a new log from the
pop-up menu that appears. If you save the file as a binary file (*.BLG), you can use
System Monitor to open the log and view it later. If you save the file as a commadelimited file (*.CSV) or a tab-delimited file (*.TSV), you can open them with Excel to do
perform your own data analysis.
15 DIRECTX*
DirectX* is an Application Programming Interface (API) developed by Microsoft that
enables programs to write programs that access hardware features of a computer
without knowing exactly what hardware is installed on the machine. DirectX was original
designed to allow Windows 95 to become a gaming platform but has grown into a
multimedia platform. DirectX achieves this by creating an intermediate layer that
translates generic hardware commands into specific commands for particular pieces of
hardware.
A version of DirectX 5, included with
Windows 98, you can check for a proper
DirectX setup by doing the following:
1. Click on the start button and
selecting the run option.
2. In the Open field type DXINFO
and click on the OK button
The screen that appears is a split
window. See figure 29. On the left-hand
side, at the bottom, you will find a
section entitled DirectX Drivers. It lists
Figure 29: DirectX 5 DXINFO
the drivers being utilized by the DirectX
subsystem. At the end of each of the
lines it should show CERTIFIED. If one or more of the items does not say CERTIFIED,
this is most likely causing the problem you are having. Most often, the drivers that are
Intel Customer Support (ICS)
Troubleshooting the IA Platform
109
not certified are either the Primary Display Driver and/or the Audio driver. Most
manufacturers of video cards and sound cards are releasing updated drivers for
Windows to meet the DirectX standard. You will want to contact the manufacturer of any
component that is not supported and ask them about obtaining drivers that support
DirectX.
If you experience trouble finding CERTIFIED drivers for a device, please be sure to
download the newest End-User version of DirectX from Microsoft's website. Often the
newer versions have support for hardware that was previously unsupported by DirectX.
They should have a copy of DirectX at the following URL:
http://www.microsoft.com/directx
You may also want to check with Microsoft's Hardware Compatibility Labs to see if your
device is listed as having drivers that are compatible.
With DirectX 6 or greater, you can check for a proper DirectX setup by doing the
following:
1. Click on the start button and selecting the run option.
2. In the Open field type DXDIAG and click on the OK button. See figure 30.
Figure 30: DirectX 6 DXDiag
16 DATA
Data is the most important part of the computer. Your data usually represents hours of
work and data is sometimes irreplaceable. Since data is the most important part of the
computer, it is essential that you protect it. Data loss can be caused by many things.
including hardware failure, power fluctuations, software problems, viruses and user
error. The best method to protect data is to perform a backup. Of course, a backup is an
extra copy of data. Unfortunately, it is often ignored by most people and they do not
think about doing it until it is too late.
Intel Customer Support (ICS)
Troubleshooting the IA Platform
110
17 VIRUSES
A virus is a program designed to replicate and spread, generally without the knowledge
or permission of the user. Computer viruses spread by attaching themselves to other
programs or to the boot sector of a disk. When an infected file is executed or accessed
or the computer is started with an infected disk, the virus spreads onto the computer.
Depending on the virus, some viruses are cute, some are annoying and others are
disastrous. Some of the disastrous symptoms of a virus include:
1.
2.
3.
4.
5.
6.
7.
8.
9.
10.
11.
Computer fails to boot.
Disks have been formatted.
The partitions are deleted or the partition table is corrupt.
Can’t read a disk.
Data or entire files are corrupt or are disappearing
Programs don’t run anymore.
Files become larger.
System is slower than normal.
System has less available memory than it should.
System is running erratic or exhibiting strange behaviors
Intercept information being sent to and from a device.
To avoid viruses, you should do the following:
1. You should not use pirated software, since there is more of a chance for them
having a virus.
2. You should treat files downloaded from the Internet and Bulletin boards with
suspicion.
3. You should not boot from or access a floppy disk of unknown origin.
4. You should educate your fellow users.
5. You should use an updated anti-virus software package that constantly detects
viruses.
6. You should backup your files on a regular basis.
Intel Customer Support (ICS)
Troubleshooting the IA Platform
111
A.1 COMPUTER DOES NOT COMPLETE BOOT
Computer Does Not Boot - No Lights or Fans – Nothing Displayed on the monitor
• Determine if the system worked before. Determine if there have been any recent
changes. Often if a recent change is been made, the recent change is the cause of
the problem.
• Make sure the computer is plugged in and on.
• Make sure that the monitor is plugged in and on.
• Use a voltmeter or an AC wall outlet tester found in a hardware store to confirm that
there is adequate AC voltage at the wall outlet. See section 6.4.
• If your AC outlet is connected to an on/off outlet switch, make sure that it is on.
• Determine power supply voltage selector setting (230 volts vs. 115 volts) is correct.
• Ensure the selected motherboard is appropriate for the processor model, frequency,
and stepping you are planning to use.
• Verify that your chassis/case and power supply is appropriate for the processor
model and frequency and the motherboard you are planning to use.
• Verify that the power supply has the capacity to power all the devices used in your
system.
• Make sure the power cables inside the computer are attached correctly and secure.
You will need to attach the main power connector and a 2x2 power and/or 1x6
optional power connector.
• Make sure the drive ribbon cables inside the computer are attached correctly and
secure. Be sure to check the orientation of pin 1 of Hard Drive. If the ribbon cable is
connected backwards may cause the computer not to power up.
• Check for foreign objects such as screws that may ground the motherboard and
make sure the screws that hold the motherboard are not too tight.
• Check the cables that connect from the case to the motherboard. Be sure to include
the power switch (PWR SW) and power LED (PWR LED). Refer to the motherboard
manual for more information. See section 6.2.
• Use a voltmeter to verify that each output from the power supply is correct. If any
output is very low (especially the +5 volt output, replace the power supply).
• Use a voltmeter to verify the PowerGood signal is +5 volts. If the signal is below 1.0
volts, there may be a short or overload causing a constant reset. Consider replacing
the power supply.
• Check for shorts and overloads inside computer by removing nonessential items
such as extra controller cards and IDE/ATAPI devices and turning the computer on
to see if it starts to boot. Leave the motherboard, power supply, RAM or processor. If
the problem goes away, there was a short or overload with one of the components
that you just removed or one of those components is faulty. Replace each of those
one at a time until you isolate which is causing the problem. If the problem still
occurs after removing the nonessential components, the problem has to be with the
motherboard, power supply, RAM or processor. See section 6.3.
• Remove the processor and RAM and reinstall them to make sure that they are
installed correctly.
• Make sure that you have mounted the motherboard correctly with the spacers/stand-
Intel Customer Support (ICS)
Troubleshooting the IA Platform
112
offs. In addition, make sure that when you insert the screws to tighten the
motherboard into place, make sure not to tighten the screws too much.
• Determine if motherboard/system has any security features, which would disable
boot.
• If you are using RDRAM, make sure that all memory sockets of a channel are filled
with either a memory chip or a continuity module. In addition, if the motherboard has
multiple channels, make sure that you fill the first channel once and that you check
to see which memory sockets go with each channel. Lastly, the RDRAM often has to
be installed in pairs of the same type of memory chips. For example, you would have
to install two sticks of 64 MB of RDRAM running at 800 MHz. See section 5.1.3.
• If you are using SDRAM or DDR-SDRAM, some motherboards require you to
populate the memory sockets starting with the first socket. Refer to your
motherboard documentation for more information.
• If the problem still persists, swap the RAM with known good RAM. In addition, test
the suspected RAM in another known working system.
• If the problem still persists, swap the processor with a known good processor. In
addition, test the suspected processor in another known working system.
• If the problem still persists, swap the motherboard with a known good motherboard.
In addition, test the suspected motherboard in another known working system.
Computer Does Not Boot - Has lights and/or fans – Nothing Displayed on the
monitor
• Determine if the system worked before. Determine if there have been any recent
changes. Often if a recent change is been made, the recent change is the cause of
the problem.
• Make sure that the monitor is plugged in and that the monitor is on.
• Make sure that the video cable is connected properly at the monitor and the PC.
• Determine if there is power light on the monitor. Most monitors will have a green light
indicating it has sufficient power and is getting a video signal and an amber light if it
has power but it is not getting a video signal from the computer. If you have no
lights, it is most likely a problem with the monitor. Make sure that the monitor is
connected to a working AC outlet, the AC power cord is plugged into the wall outlet
and the monitor. If you have any on/off switches for the outlet, make sure that the
switches are on. If the problem still persists, try replacing the monitor and try the
monitor on another AC outlet and another system. If you have an amber light, it is
most likely a problem with the computer.
• Check the brightness and contrast controls on the monitor. The monitor might be
dimmed where you cannot see anything on the monitor.
• If you have a light on the monitor, use a voltmeter or an AC wall outlet tester found in
a hardware store to confirm that there is adequate AC voltage at the wall outlet for
the computer.
• If your AC outlet for your computer is connected to an on/off outlet switch, make sure
that it is on.
• Ensure the selected motherboard is appropriate for the processor model, frequency,
and stepping you are planning to use.
• Verify that your chassis/case and power supply is appropriate for the processor
Intel Customer Support (ICS)
Troubleshooting the IA Platform
113
•
•
•
•
•
•
•
•
•
•
•
•
•
•
•
model and frequency and the motherboard you are planning to use.
Verify that the power supply has the capacity to power all the devices used in your
system.
Make sure the power cables inside the computer are attached correctly and secure.
Some systems may require a 2x2 and/or 1x6 power connector.
Make sure the drive ribbon cables inside the computer are attached correctly and
secure. Be sure to check the orientation of pin 1 of Hard Drive. If the ribbon cable is
connected backwards may cause the computer not to power up.
Check for foreign objects such as screws that may ground the motherboard and
make sure the screws that hold the motherboard are not too tight.
Check the cables that connect from the case to the motherboard. Be sure to include
the power switch (PWR SW) and power LED (PWR LED). Refer to the motherboard
manual for more information.
Use a voltmeter to verify that each output from the power supply is correct. If any
output is very low (especially the +5 volt output, replace the power supply).
Use a voltmeter to verify the PowerGood signal is +5 volts. If the signal is below 1.0
volts, there may be a short or overload causing a constant reset. Consider replacing
the power supply.
Check for shorts and overloads inside computer by removing nonessential items
such as extra controller cards and IDE/ATAPI devices and turning the computer on
to see if it starts to boot. Leave the motherboard, power supply, RAM or processor. If
the problem goes away, there was a short or overload with one of the components
that you just removed or one of those components is faulty. Replace each of those
one at a time until you isolate which is causing the problem. If the problem still
occurs after removing the nonessential components, the problem has to be with the
motherboard, power supply, RAM or processor.
Remove the processor and RAM and reinstall them to make sure that they are
installed correctly.
Make sure that you have mounted the motherboard correctly with the spacers/standoffs. In addition, make sure that when you insert the screws to tighten the
motherboard into place, make sure not to tighten the screws too much.
Determine if motherboard/system has any security features, which would disable
boot.
If you are using RDRAM, make sure that all memory sockets of a channel are filled
with either a memory chip or a continuity module. In addition, if the motherboard has
multiple channels, make sure that you fill the first channel once and that you check
to see which memory sockets go with each channel. Lastly, the RDRAM often has to
be installed in pairs of the same type of memory chips. For example, you would have
to install two sticks of 64 MB of RDRAM running at 800 MHz.
If you are using SDRAM or DDR-SDRAM, some motherboards require you to
populate the memory sockets starting with the first socket. Refer to your
motherboard documentation for more information.
If the problem still persists, swap the RAM with known good RAM. In addition, test
the suspected RAM in another known working system.
If the problem still persists, swap the processor with a known good processor. In
addition, test the suspected processor in another known working system.
Intel Customer Support (ICS)
Troubleshooting the IA Platform
114
If the problem still persists, swap the motherboard with a known good motherboard. In
addition, test the suspected motherboard in another known working system.
Operating System starts but does not finish boot
• Determine if it started to load the OS. Either it will load a Windows startup logo
(splash) screen or it will display one of the following messages: Starting MS-DOS,
Starting Windows 9X, Starting OS Loader. If it does not, the computer is having
trouble finding or loading the OS.
• Determine if the system worked before. Determine if there have been any recent
changes. Often if a recent change is been made, the recent change is the cause of
the problem.
• Determine if the machine always stops in the same place every time or it seems
random. If it seems to stop in different places, it may be a hardware problem.
• Update the system BIOS to the newest version. In addition, check to see if your
system has a firmware that could also be updated to the newest version. Refer to
your motherboard documentation.
• Check the BIOS using the BIOS/CMOS setup utility, particularly the boot order.
Make sure that your system is detecting all of the drives and make sure that the
drive you are trying to boot is listed. See section 4.2, 4.3 and 5.1.
• Check the BIOS using the BIOS/CMOS setup utility, particularly the memory
settings.
• Try booting the operating system in Safe mode/VGA mode. If it works fine in safe
mode, it is mostly likely a driver or software that is not loaded correctly during Safe
mode. See section 13.2 and 14.2.
• If the operating system supports, try a Step-by-Step boot. See section 13.2 an 14.2.
• Look at the boot log to find out what the last driver that it loaded. To access the
advanced boot menu, press F8 as soon as Windows starts to load. For Windows 9X,
the boot log is located at the C:\bootlog.txt. For Windows 2000 and Windows XP, the
boot log is NTBTLOG.TXT file located in the C:\WINNT directory. See section 13.2
and 14.2.
• If the processor can boot to safe mode, try a Selective Startup using MSCONFIG
(Windows 98, Windows Me or Windows XP only) program to isolate the offending
driver or program. See section 13.2 and 14.2.
• For Windows NT, Windows 2000 or Windows XP, start the installing program for
Windows from the installation CD and choose the R to start the Repair option. This
will check the master boot record and the boot files. Refer to the Troubleshooting for
the IA Platform PDF document for more information. See section 14.2.
• Suggest running ScanDisk to see if your hard drive has a lot of lost clusters and
other anomalies. If you don’t properly shutdown your computer using the Start button
shutdown option, you may see some lost clusters and possible bad clusters.
• Suggest running an updated virus checker to see if your system is affected by a
virus.
• Suggest reloading the OS, drivers or program.
• Suggest reformatting the hard drive and reinstalling everything
• Make sure that your system has enough ventilation around the fan outlet and that
the fan is operating correctly.
Intel Customer Support (ICS)
Troubleshooting the IA Platform
115
•
•
•
•
•
•
•
•
•
Make sure that the thermal solution for the processor is installed correctly and make
sure that processor fan is operating correctly.
Check for foreign objects such as screws that may ground the motherboard and
make sure the screws that hold the motherboard are not too tight.
Use a voltmeter to verify that each output from the power supply is correct. If any
output is very low (especially the +5 volt output, replace the power supply).
Use a voltmeter to verify the PowerGood signal is +5 volts. If the signal is below 1.0
volts, there may be a short or overload. Consider replacing the power supply.
Remove the processor and RAM and reinstall them to make sure that they are
installed correctly.
Make sure that you have mounted the motherboard correctly with the spacers/standoffs. In addition, make sure that when you insert the screws to tighten the
motherboard into place, make sure not to tighten the screws too much.
If the problem still persists, swap the RAM with known good RAM. In addition, test
the suspected RAM in another known working system.
If the problem still persists, swap the processor with a known good processor. In
addition, test the suspected processor in another known working system.
If the problem still persists, swap the motherboard with a known good motherboard.
In addition, test the suspected motherboard in another known working system.
•
You get a “The system did not load because of the computer disk hardware
configuration problem. Could not read from the selected boot disk.”
The location of your OS is not configured correctly in the BOOT.INI file.
• You need to boot using an Emergency Repair Disk or the Recovery Console. At the
command prompt, type attrib –s –r C:\boot.ini. Type edit boot.ini to read and change
the file. Check to make sure that the path of the default OS in the boot loader section
is identical to one listed in the “operating systems” section. If you recently added any
CD, DVD, or disk drives or created any primary partitions, then the path to your OS
probably has chanced. Refer to section 14.2.
After installing a new BIOS or replacing the CMOS battery, the system now asks
for a password
This is not the Windows password, but a BIOS password asked during POST.
• If you set a password using the BIOS Setup program, try the password. Be sure to
check your caps lock key. If the password still does not work, turn the CAPs lock on
and try it again. Also try a blank password where you just press enter.
• If you still cannot get in, clear the old configuration by looking for a jumper on the
motherboard to “Clear CMOS” or “Clear Password” or remove the CMOS battery.
The PC does not boot after flashing the BIOS
If you flashed the BIOS and it stopped for whatever reason or you flashed the wrong
BIOS version, there is nothing that you can do except to replace the BIOS IC.
• If the system has a boot block to restore the original BIOS, use it.
• If not, you will need to contact the system or motherboard manufacturer for a
replacement.
Intel Customer Support (ICS)
Troubleshooting the IA Platform
116
The system hangs up when the SCSI BIOS headers appear during POST
• Check SCSI Termination.
• Check the SCSI ID numbers.
• Check the system resources (BIOS address, I/O address, DMA and interrupts) of the
SCSI controller card.
Intel Customer Support (ICS)
Troubleshooting the IA Platform
117
A.2 BIOS BEEP CODES
AMI BIOS Beep Codes
You are getting beeps that indicate a memory failure:
1 Beep – Refresh Error (with nothing on the screen and it is not a video
problem)
2 Beep – Parity Error
3 Beep – Base 64 K memory failure
• Reseat the memory.
• Make sure that the contacts on the memory and the socket are clean.
• Try removing one bank of memory modules at a time. Note: Some systems might
need to have a memory module in Bank 0.
• Try using RAM chips from the same manufacturer with the same part number and
speed.
• Check for a faulty memory module by trying the memory in a known good system.
• Trying known good memory in the system.
• Check the power supply and check for power fluctuations.
• Swap the motherboard.
You are getting 4 beeps indicating a Timer not operational:
• Check for foreign objects such as screws that may ground the motherboard and
make sure that screws that hold the motherboard are not too tight.
• Swap the motherboard.
You are getting 5 beeps indicating a Processor error:
• Check for foreign objects such as screws that may ground the motherboard and
make sure the screws that hold the motherboard are not too tight.
• Make sure the processor is seated properly.
• Swap the processor.
• Remove the motherboard and reinstall it.
• If the problem still persists, swap the motherboard.
You are getting 6 beeps indicating 8042 – gate A20 failure:
• Check for foreign objects such as screws that may ground the motherboard and
make sure the screws that hold the motherboard are not too tight.
• Try reseating the keyboard controller chip (if possible).
• Swap the keyboard.
• Swap the processor.
You are getting 7 beeps indicating Processor exception interrupt error
• Make sure the processor is seated properly.
• Swap the processor.
• Swap the motherboard.
You are getting 8 beeps indicating Display memory read/write failure
• Make sure that you have a video card on your system.
• If possible, swap the memory on the video card.
• Swap the video card.
• Swap the motherboard.
Intel Customer Support (ICS)
Troubleshooting the IA Platform
118
AMI BIOS Beep Codes
You are getting 9 beeps indicating ROM checksum error
• If possible, try reseating the System ROM BIOS chip.
• If possible, try reflashing the system ROM BIOS.
• Swap the motherboard.
You are getting 10 beeps indicating CMOS shutdown register read/write error
• Swap the motherboard
You are getting 11 beeps indicating Cache memory bad
• Make sure the system is configured properly (CMOS Setup Program and jumpers on
the motherboard) for the RAM cache.
• If possible, replace the RAM cache.
• Swap the processor
• Swap the motherboard
Award BIOS Beep Codes
You are getting one long and two short beeps indicating a video problem
• Make sure that you have a video card on your system.
• If possible, swap the memory on the video card.
• Swap the video card.
• Swap the motherboard.
Any other beeps:
• Determine if the system worked before.
• Determine if there have been any recent changes
• Reseat the memory.
• Try a different motherboard.
• Make sure that the contacts on the memory and the socket are clean.
• Try removing one bank of memory modules at a time. Note: Some systems might
need to have a memory module in Bank 0.
• Try using RAM chips from the same manufacturer with the same part number and
speed.
• Check for a faulty memory module by trying the memory in a known good system.
• Swap RAM, motherboard and processor one at a time. If suspect one of these
components, try it in another PC.
Intel Customer Support (ICS)
Troubleshooting the IA Platform
119
Phoenix Beep Codes – Fatal System Board Errors
Beep Description
Recommended Solution
Code
None CPU register test in • Swap the processor.
progress
• Try reseating the motherboard
• Swap the motherboard.
CMOS write/read
1-1-3
• Try reseating the motherboard.
failure
• Swap the motherboard.
ROMBIOS checksum • Try reflashing the BIOS.
1-1-4
failure
• Swap the motherboard
Programmable
1-2-1
• Try reseating the motherboard.
interval timer failure • Swap the motherboard.
DMA initialization
1-2-2
• Try reseating the motherboard.
failure
• Swap the motherboard.
DMA page register
1-2-3
• Try reseating the motherboard.
write/read failure
• Swap the motherboard.
RAM refresh
1-3-1
• Reseat the memory.
verification failure
• Make sure that the contacts on the memory and the
None First 64K RAM test in sockets are clean.
progress
• Try removing one bank of memory modules at a
First 64 KB RAM chip time. Note: Some systems might need to have a
1-3-3
memory module in Bank 0.
or data line failure,
multibit
• Try using RAM chips from the same manufacturer
with the same part number and speed.
First 64 KB RAM
1-3-4
odd/even logic failure • Check for a faulty memory module by trying the
memory in a known good system.
Address line failure
1-4-1
• Trying known good memory in the system.
first 64 KB RAM
Parity failure first 64 • Check the power supply and check for power
1-4-2
fluctuations.
KB
• Swap the motherboard.
First 64 KB RAM
2-x-x
Failure
Slave DMA register
3-1-1
• Try reseating the motherboard.
failure
• Swap the motherboard.
Master DMA register
3-1-2
failure
Master interrupt
3-1-3
mask register failure
Slave interrupt mask
3-1-4
register failure
Keyboard controller
Check keyboard XT/AT switch, keyboard cable,
3-2-4
test failure
keyboard or motherboard.
Intel Customer Support (ICS)
Troubleshooting the IA Platform
120
Phoenix Beep Codes – Fatal System Board Errors
Screen initialization
3-3-4
• Make sure that you have a video card on your
failure
system.
Screen retrace failure • If possible, swap the memory on the video card.
3-4-1
• Swap the video card.
Search for video
3-4-2
ROM in progress
• Swap the motherboard.
Phoenix BIOS Beep Codes – Nonfatal System Board Errors
Beep
Description
Recommended Solution
Code
4-2-1
Timer tick interrupt
• Try reseating the motherboard.
test in progress or
• Swap the motherboard.
failure
4-2-2
Shutdown test in
First check the keyboard for problems. If nothing, you
progress or failure
have a bad motherboard.
4-2-3
Gate A20 failure
• Check for foreign objects such as screws that may
ground the motherboard.
• Try reseating the keyboard controller chip.
• Swap the keyboard.
• Swap the processor.
• Try removing and reinstalling the motherboard.
• If the problem still persists, swap the motherboard.
4-2-4
Unexpected interrupt • Check for a bad expansion card or bad motherboard.
in protected mode
4-3-1
RAM test in progress • Replace the motherboard.
or address failure
>FFFh
4-3-3
Interval timer
• Try reseating the motherboard.
Channel 2 test or
• Swap the motherboard.
failure
4-3-4
Time-of-day clock
• Try running the setup program that comes with the
test or failure
computer. Check the date and time. If that doesn't
work, replace the battery or motherboard.
4-4-1
Serial port test or
• Reseat, or replace, the I/O card. If the serial port is on
failure
the motherboard, disable
4-4-2
Parallel port test or
• Reseat, or replace, the I/O card. If the serial port is on
failure
the motherboard, disable
4-4-3
Math coprocessor
• Run a test program to double-check it. If the problem
test or failure
does exist, replace the math coprocessor (or CPU).
Low
System board select • Replace the motherboard.
1-1-2
failure
Low
Extended CMOS
• Replace the motherboard.
1-1-3
RAM failure
Intel Customer Support (ICS)
Troubleshooting the IA Platform
121
A.3: Operating Problems
The date and time seem to drift or you get one of the following messages
indicating problems with the CMOS:
- CMOS battery failed
- CMOS BATTERY HAS FAILED
- CMOS Battery State Low
- CMOS CHECKSUM ERROR
• Replace the CMOS battery.
• If the problem still persists, try replacing the motherboard. Note: You may also
consider using a battery backup. Refer to your motherboard manual.
The system boots with no problem, but crashes or freezes after several minutes
of operation.
• Determine if the system worked before. Determine if there have been any recent
changes. Often if a recent change is been made, the recent change is the cause of
the problem.
• Check to see if your system feels warm and check the processor for overheating.
NOTE: Use extreme caution when checking for heat, since you can be easily
burned.
• Make sure that the power supply fan is running properly and any other external case
fans are running properly. Make sure that the air intakes for the external fans are
unobstructed and have at least several inches away from walls and other items.
• Make sure the fans on the processor are connected properly and running.
• Make sure that the thermal interface material or the thermal grease is applied
properly.
• Disable thermal measurements and power saving features in the BIOS setup
program.
• Update the system BIOS to the newest version. In addition, check to see if your
system has a firmware that could also be updated to the newest version. Refer to
your motherboard documentation.
• Check the BIOS using the BIOS/CMOS setup utility, particularly the boot order.
Make sure that your system is detecting all of the drives and make sure that the
drive you are trying to boot is listed.
• Do not overclock your system. Your system can become unreliable, may shorten the
life of your PC components, may damage your PC components and may void your
warranty.
• Check the BIOS using the BIOS/CMOS setup utility, especially the RAM settings.
See section 5.2.
• Suggest running ScanDisk to see if your hard drive has a lot of lost clusters and
other anomalies. If you don’t properly shutdown your computer using the Start button
shutdown option, you may see some lost clusters and possible bad clusters.
• Suggest running an updated virus checker to see if your system is affected by a
virus.
• Suggest reloading the OS, drivers or program.
• Suggest reformatting the hard drive and reinstalling everything
Intel Customer Support (ICS)
Troubleshooting the IA Platform
122
•
Verify that your chassis/case and power supply is appropriate for the processor
model and frequency and the motherboard you are planning to use.
• Verify that the power supply has the capacity to power all the devices used in your
system.
• Check for foreign objects such as screws that may ground the motherboard and
make sure the screws that hold the motherboard are not too tight.
• Make sure the power cables inside the computer are attached correctly and secure.
You will need to attach the main power connector and a 2x2 power and/or 1x6
optional power connector.
• Use a voltmeter to verify that each output from the power supply is correct. If any
output is very low (especially the +5 volt output, replace the power supply).
• Use a voltmeter to verify the Power Good signal is +5 volts. If the signal is below 1.0
volts, there may be a short or overload causing a constant reset. Consider replacing
the power supply.
• Check for shorts and overloads inside computer by removing nonessential items
such as extra controller cards and IDE/ATAPI devices and turning the computer on
to see if it starts to boot. Leave the motherboard, power supply, RAM or processor. If
the problem goes away, there was a short or overload with one of the components
that you just removed or one of those components is faulty. Replace each of those
one at a time until you isolate which is causing the problem. If the problem still
occurs after removing the nonessential components, the problem has to be with the
motherboard, power supply, RAM or processor.
• Remove the processor and RAM and reinstall them to make sure that they are
installed correctly.
• Make sure that you have mounted the motherboard correctly with the spacers/standoffs. In addition, make sure that when you insert the screws to tighten the
motherboard into place, make sure not to tighten the screws too much.
• If the problem still persists, swap the RAM with known good RAM. In addition, test
the suspected RAM in another known working system.
• If the problem still persists, swap the processor with a known good processor. In
addition, test the suspected processor in another known working system.
• If the problem still persists, swap the motherboard with a known good motherboard.
In addition, test the suspected motherboard in another known working system.
Error loading GDI.EXE. You must re-install Windows.
•
The Fonts folder is moved, renamed or deleted. Recreate the Fonts folder in the
C:\WINDOWS directory and extract he files form the Windows installation CD-ROM.
•
Reinstall Windows.
You get an “Error loading Kernel. You must reinstall Windows” error message.
• The Kernel32.dll file, which contains the Windows kernel is corrupt or missing.
• Extract the Kernel32.dll file form your Windows setup disk and copy it to the
appropriate folder on your hard drive (typically \windows\system).
• Reinstall Windows.
Intel Customer Support (ICS)
Troubleshooting the IA Platform
123
A.4 Processor Problems
The processor is not identified correctly during POST or in the BIOS setup
program.
• Check the CMOS parameters or jumpers settings on the motherboard for the
processor.
• Check to make sure the motherboard supports the processor.
• Update to the newest BIOS version and use the Processor Update Utility. See
section 3.8 and 4.2.
• Run the Frequency ID Utility to verify the processor. See section 3.7.
The processor is not identified correctly by the operating system or an
application.
• Check the CMOS parameters or jumpers settings on the motherboard for the
processor.
• Check to make sure the motherboard supports the processor.
• Update to the newest BIOS version and use the Processor Update Utility. See
section 3.8 and 4.2.
• Run the Frequency ID Utility to verify the processor. See section 3.7.
• Since many programs detect the processor so that they can better utilize their
features, the software may have been written before the processor existed.
Therefore, check with the software manufacturer for a patch or update. Note: If an
application cannot identify a processor, the software will not run with the processor
or it will assume it is an older processor without many of the features that the
processor offers resulting in slower performance.
You are experiencing frequent processor failures.
• Make sure the motherboard supports the specific processor.
• Check the jumpers or BIOS setup program for the voltage settings of the processor.
• Check the jumpers or BIOS setup program for the operating frequency settings for
the processor.
• Run the Frequency ID Utility. See section 3.7.
• Checking the thermal solution (heat-sink and fan). See section 3.2.
• Check for power fluctuations. Make sure that you have a reliable surge protector.
See section 6.5.
• Use a voltmeter to verify that each output from the power supply is correct. If any
output is very low (especially the +5 volt output), replace the power supply. See
section 6.1.
Processor is running hot
• Determine if the system worked before. Determine if there have been any recent
changes. Often if a recent change is been made, the recent change is the cause of
the problem.
• Do not overclock your system. Your system can become unreliable, may shorten the
life of your PC components, may damage your PC components and may void your
warranty.
• Verify that your chassis/case and power supply is appropriate for the processor
model and frequency and the motherboard you are planning to use.
Intel Customer Support (ICS)
Troubleshooting the IA Platform
124
•
•
•
•
•
•
•
•
•
If applicable, verify the voltage settings for the processor.
Verify that your thermal solution for the Pentium 4 processor is adequate for the
processor and frequency of the processor.
Check to see if your system feels warm and check the processor for overheating.
NOTE: Use extreme caution when checking for heat, since you can be easily
burned. Many people may be running a software package or using their BIOS setup
program, which is telling them the processor is running warmer than normal. Many
times, these programs and/or BIOS setup programs may be wrong. Therefore, if you
not feel a large amount of heat coming from the processor and heat sink, then the
processor is probably not overheating.
Make sure that the power supply fan is running properly and any other external case
fans are running properly. Make sure that the air intakes for the external fans are
unobstructed and have at least several inches away from walls and other items.
Make sure that you have power cable for the fan connected to the correct fan header
specifically for the processor. Refer to your motherboard documentation for more
information.
Make sure the fans on the processor are connected properly and running.
Make sure that the thermal interface material or the thermal grease is applied
properly.
Disable thermal measurements and power saving features in the BIOS setup
program.
Update the system BIOS to the newest version. This will often correct problems with
how the system measures the processor temperature.
Intel Customer Support (ICS)
Troubleshooting the IA Platform
125
A.5 Memory Problems
You get a Parity Error
• Reseat the memory.
• Make sure that the contacts on the memory and the socket are clean.
• Check the memory parameters in the BIOS setup program including the number of
wait states. If the memory is too slow, you need to increase the number of wait
states. See section 5.2.
• Make sure that you have the right memory module for your system.
• Try removing one bank of memory modules at a time. Note: Some systems might
need to have a memory module in Bank 0.
• Try using RAM chips from the same manufacturer with the same part number and
speed.
• Check for a faulty memory module by trying the memory in a known good system.
• Trying known good memory in the system.
• Check the power supply and check for power fluctuations. See section 6.1.
You get a ROM Error displayed on the monitor during POST
To guarantee the integrity of system ROM, a checksum error test is performed as part
of the POST.
• Try reflashing the System ROM BIOS (if possible).
• Some motherboards have a dual system ROM BIOS so that the second can be used
to restore the first one when the first becomes corrupt.
• Replace the System ROM BIOS chip or the motherboard.
You see a general RAM error with fault addresses listed. For example:
Memory address line failure at <XXXX>, read <YYYY>, expecting <ZZZZ>
Memory read/write failure at <XXXX>, read <YYYY>, expecting <ZZZZ>
• Reseat the memory.
• Make sure that the contacts on the memory and the socket are clean.
• Check the memory parameters in the BIOS setup program including the number of
wait states. If the memory is too slow, you need to increase the number of wait
states.
• Try removing one bank of memory modules at a time. Note: Some systems might
need to have a memory module in Bank 0.
• Try using RAM chips from the same manufacturer with the same part number and
speed.
• Check for a faulty memory module by trying the memory in a known good system.
• Trying known good memory in the system.
• Check the power supply and check for power fluctuations.
Window 95, 98 or Me get a “Windows Protection Error” during boot.
• Reseat the memory.
• Make sure that the contacts on the memory and the socket are clean.
• Check the memory parameters in the BIOS setup program including the number of
wait states. If the memory is too slow, you need to increase the number of wait
states. See section 5.2.
• Try removing one bank of memory modules at a time. Note: Some systems might
Intel Customer Support (ICS)
Troubleshooting the IA Platform
126
need to have a memory module in Bank 0.
• Try using RAM chips from the same manufacturer with the same part number and
speed.
• Check for a faulty memory module by trying the memory in a known good system.
• Trying known good memory in the system.
• Check the power supply and check for power fluctuations. See section 6.5.
Invalid VxD Dynamic Call to Device 3 Service B
• Reseat the memory.
• Make sure that the contacts on the memory and the socket are clean.
• Check the memory parameters in the BIOS setup program including the number of
wait states. If the memory is too slow, you need to increase the number of wait
states. See section 5.2.
• Try removing one bank of memory modules at a time. Note: Some systems might
need to have a memory module in Bank 0.
• Try using RAM chips from the same manufacturer with the same part number and
speed.
• Check for a faulty memory module by trying the memory in a known good system.
• Trying known good memory in the system. See section 6.5.
• Check the power supply and check for power fluctuations.
HIMEM.SYS had Detected Unreliable XMS Memory at <address>
• Reseat the memory.
• Make sure that the contacts on the memory and the socket are clean.
• Check the memory parameters in the BIOS setup program including the number of
wait states. If the memory is too slow, you need to increase the number of wait
states. See section 5.2.
• Try removing one bank of memory modules at a time. Note: Some systems might
need to have a memory module in Bank 0.
• Try using RAM chips from the same manufacturer with the same part number and
speed.
• Check for a faulty memory module by trying the memory in a known good system.
• Trying known good memory in the system.
• Check the power supply and check for power fluctuations. See section 6.5.
• Replace the motherboard.
You encounter random “fatal exception” errors under Windows 95, 98 or Me.
• Check for buggy or corrupted drivers by trying safe mode or by using the MSConfig
to systematically disable software and drivers in order to isolate the offending
software.
• Reseat the memory.
• Make sure that the contacts on the memory and the socket are clean.
• Check the memory parameters in the BIOS setup program including the number of
wait states. If the memory is too slow, you need to increase the number of wait
states. See section 5.2.
• Try removing one bank of memory modules at a time. Note: Some systems might
need to have a memory module in Bank 0.
• Try using RAM chips from the same manufacturer with the same part number and
Intel Customer Support (ICS)
Troubleshooting the IA Platform
127
speed.
• Check for a faulty memory module by trying the memory in a known good system.
• Trying known good memory in the system.
• Check the power supply and check for power fluctuations. See section 6.5.
An error indicates that there is not enough memory to start Windows 95, 98 or ME
or to start an application.
There is not enough real and virtual memory to start the Windows shell or a particular
program.
• Make sure you have lots of free disk space on your hard drive. Try to boot from
floppy and free some disk space.
• Try to Defrag the drive.
• Trying adding RAM.
• If the problem still persist, it could be caused by corrupt software. Try to reinstall the
software.
• Reformat the system and reinstall the software.
You get the “Limited Virtual Memory. Your system is running without a properly
sized paging file. Please use the virtual memory option of the System applet in
the Control Panel to create a paging file, or to increase the initial size of your
paging file.”
The size of the paging file is too small.
• Make sure that you have plenty of free disk space on your hard drive.
• Use the System applet in the Control Panel, access the virtual memory tool to set
the paging file size to 0. Reboot the system and access the virtual memory tool
again. Look for the recommended minimum and maximum sizes and set the values
accordingly. Reboot the system.
Intel Customer Support (ICS)
Troubleshooting the IA Platform
128
A.6 Disk Problems
You see an “Invalid Media” error message when formatting.
• Make sure that the hard drive power connector is connected.
• Make sure that the ribbon is connected properly.
• Check for viruses.
• Enter the BIOS setup program and make sure that the correct CMOS settings are
chosen for the drive and make sure that the hard drive and floppy controllers are
enabled. Also check the boot order. Note: on must systems, you can use the
autodetect feature to determine the correct CMOS parameters and to see if it can
see the drive.
• Check the drive’s power including the +5 and +12 volt connectors.
• If it an IDE drive, check the jumper settings of all of the IDE devices, especially the
master, slave or stand-alone jumpers.
• If it is a SCSI drive, check the jumper settings of all SCSI devices, especially the
termination and SCSI ID numbers. To make sure that the primary hard disk is set to
SCSI ID 0.
• Check to see if all of the device drivers have been installed correctly.
• For hard drives, try a low-level format. Then partition, format and add the system
files to the hard drive. For floppy disks, try another floppy disk.
• Swap the controller card, cable and hard drive, one at a time.
• Swap the motherboard.
Hard Drives
The hard drive is completely dead.
The drive does not spin up and the drive light doesn’t illuminate during power-up.
• Make sure that the hard drive power connector is connected. See section 8.
• Make sure that the ribbon cable is connected properly. See section 8.
• Check the drive’s power including the +5 and +12 volt connectors.
• Swap the hard drive.
You see drive activity, but the computer will not boot from the hard drive.
• Make sure that the hard drive power connector is connected.
• Make sure that the ribbon cable is connected properly.
• Run the BIOS setup program and make sure that drive and the controller is enabled.
In addition make sure that the drive has the correct parameters; most likely you can
try the autodetect feature.
• If it an IDE drive, check the jumper settings of all of the IDE devices, especially the
master, slave or stand-alone jumpers. See section 8.2.
• If it is a SCSI drive, check the jumper settings of all SCSI devices, especially the
termination and SCSI ID numbers. To make sure that the primary hard disk is set to
SCSI ID 0. See section 8.4.
• If it is a SCSI drive, check during POST to see if the SCSI BIOS is loading and
finding any SCSI devices. If not, check the host adapter installation including IRQ,
DMA and I/O settings. See section 8.4.
Intel Customer Support (ICS)
Troubleshooting the IA Platform
129
• Check to see if all of the device drivers have been installed correctly.
• Check the drive’s power including the +5 and +12 volt connectors.
• Swap the controller card, cable and hard drive, one at a time.
• Swap the motherboard.
The following error messages indicate that the computer cannot find an operating
system to boot:
- Non-System disk or disk error, Replace and strike any key when ready
- No ROM Basic
- Disk Boot Error, Replace and Strike Key to Retry
- DISK BOOT FAILURE, INSERT SYSTEM DISK AND PRESS ENTER
- BOOT: Couldn’t find NTLDR, Please insert another disk
• If you are trying to boot from the hard drive, make sure that there is no disk in drive
A.
• If you are trying to boot from the floppy disk drive, try to boot from the hard drive and
try to access the floppy disk. Then try to make sure that the disk has the necessary
boot files and that the necessary boot files are not corrupt.
• If you are trying to boot from the floppy disk drive, try to boot from another bootable
floppy disk.
• Make sure that the hard drive power connector is connected.
• Make sure that the ribbon cable is connected properly.
• If using a removable/mobile IDE rack for the hard drive, verify that drive tray has
been locked.
• Try reinstalling the necessary boot files.
• If you cannot access the drive, enter the BIOS setup program and make sure that
the correct CMOS settings are chosen for the drive and make sure that the hard
drive and floppy controllers are enabled. Also check the boot order. Note: on must
systems, you can use the autodetect feature to determine the correct CMOS
parameters and to see if it can see the drive.
• For hard drives, verify that your primary partition is active.
• For DOS, Windows 95, Windows 98 and Windows Me, try to reinstall the necessary
boot files.
• For Windows NT, 2000 and XP, start the Windows installation program and choose
R to repair. You will then be asked for an emergency repair disk. If you don’t have
one, don’t repair the Registry when asked to do so. When you are in the Repair
utility, the repair utility will detect the system file and reinstall it in the correct place.
• Try to reformat the drive and make the drive bootable again. See section 14.2.
• Try a low-level format. Then partition, format and add the system files to the hard
drive.
• Check the drive’s power including the +5 and +12 volt connectors.
• If it an IDE drive, check the jumper settings of all of the IDE devices, especially the
master, slave or stand-alone jumpers.
• If it is a SCSI drive, check the jumper settings of all SCSI devices, especially the
termination and SCSI ID numbers. To make sure that the primary hard disk is set to
SCSI ID 0.
• If it is a SCSI drive, check during POST to see if the SCSI BIOS is loading and
Intel Customer Support (ICS)
Troubleshooting the IA Platform
130
•
•
•
finding any SCSI devices. If not, check the host adapter installation including IRQ,
DMA and I/O settings.
Check to see if all of the device drivers have been installed correctly.
Swap the controller card, cable and hard drive, one at a time.
Swap the motherboard.
You get a “No Fixed Disk Present” or “Error reading fixed disk” error message on
the monitor.
• If you are trying to boot from the hard drive, make sure that there is no disk in drive
A.
• If you are trying to boot from the floppy disk drive, try to boot from the hard drive and
try to access the floppy disk. Then try to make sure that the disk has the necessary
boot files and that the necessary boot files are not corrupt.
• If you are trying to boot from the floppy disk drive, try to boot from another bootable
floppy disk.
• Make sure that the hard drive power connector is connected.
• Make sure that the ribbon cable is connected properly.
• If you cannot access the drive, enter the BIOS setup program and make sure that
the correct CMOS settings are chosen for the drive and make sure that the hard
drive and floppy controllers are enabled. Also check the boot order. Note: on must
systems, you can use the autodetect feature to determine the correct CMOS
parameters and to see if it can see the drive.
• For DOS, Windows 95, Windows 98 and Windows Me, try to run the FDISK /MBR
command.
• For Windows NT, 2000 and XP, start the Windows installation program and choose
R to repair. You will then be asked for an emergency repair disk. If you don’t have
one, don’t repair the Registry when asked to do so. When you are in the Repair
utility, the repair utility will detect the system file and reinstall it in the correct place.
See section 14.2.
• Try a low-level format. Then partition, format and add the system files to the hard
drive.
• Check the drive’s power including the +5 and +12 volt connectors.
• If it an IDE drive, check the jumper settings of all of the IDE devices, especially the
master, slave or stand-alone jumpers.
• If it is a SCSI drive, check the jumper settings of all SCSI devices, especially the
termination and SCSI ID numbers. To make sure that the primary hard disk is set to
SCSI ID 0.
• If it is a SCSI drive, check during POST to see if the SCSI BIOS is loading and
finding any SCSI devices. If not, check the host adapter installation including IRQ,
DMA and I/O settings.
• Check to see if all of the device drivers have been installed correctly.
• If you have two drives on the same IDE cable, try swapping the master/slave.
• Swap the controller card, cable and hard drive, one at a time.
• Swap the motherboard.
Intel Customer Support (ICS)
Troubleshooting the IA Platform
131
Your drive spins up, but the system fails to recognize the drive.
• If you are trying to boot from the hard drive, make sure that there is no disk in drive
A.
• If you are trying to boot from the hard drive, try to boot from a bootable floppy disk
and try to access the hard drive. Then try to make sure that the disk has the
necessary boot files and that the necessary boot files are not corrupt.
• If you are trying to boot from the floppy disk drive, try to boot from the hard drive and
try to access the floppy disk. Then try to make sure that the disk has the necessary
boot files and that the necessary boot files are not corrupt.
• If you are trying to boot from the floppy disk drive, try to boot from another bootable
floppy disk.
• Make sure that the hard drive power connector is connected.
• Make sure that the ribbon cable is connected properly.
• Try reinstalling the necessary boot files.
• If you cannot access the drive, enter the BIOS setup program and make sure that
the correct CMOS settings are chosen for the drive and make sure that the hard
drive and floppy controllers are enabled. Also check the boot order. Note: on must
systems, you can use the autodetect feature to determine the correct CMOS
parameters and to see if it can see the drive.
• For hard drives, verify that your primary partition is active.
• For DOS, Windows 95, Windows 98 and Windows Me, try to reinstall the necessary
boot files.
• For DOS, Windows 95, Windows 98 and Windows Me, try using the FDISK /MBR
command.
• For Windows NT, 2000 and XP, start the Windows installation program and choose
R to repair. You will then be asked for an emergency repair disk. If you don’t have
one, don’t repair the Registry when asked to do so. When you are in the Repair
utility, the repair utility will detect the system file and reinstall it in the correct place.
See section 14.2.
• Try to reformat the drive and make the drive bootable again.
• Try a low-level format. Then partition, format and add the system files to the hard
drive.
• Check the drive’s power including the +5 and +12 volt connectors.
• If it is an IDE drive, check the jumper settings of all of the IDE devices, especially the
master, slave or stand-alone jumpers.
• If it is a SCSI drive, check the jumper settings of all SCSI devices, especially the
termination and SCSI ID numbers. To make sure that the primary hard disk is set to
SCSI ID 0.
• If it is a SCSI drive, check during POST to see if the SCSI BIOS is loading and
finding any SCSI devices. If not, check the host adapter installation including IRQ,
DMA and I/O settings.
• Check to see if all of the device drivers have been installed correctly.
• If you have two drives on the same IDE cable, try swapping the master/slave.
• Swap the controller card, cable and hard drive, one at a time.
• Swap the motherboard.
Intel Customer Support (ICS)
Troubleshooting the IA Platform
132
The following error messages indicate that the computer cannot communicate
with the hard drive:
- Hard Disk Failure
- Hard Disk Controller Failure
- HDD Controller Failure
• Make sure that the hard drive power connector is connected.
• Make sure that the ribbon cable is connected properly.
• If you cannot access the drive, enter the BIOS setup program and make sure that
the correct CMOS settings are chosen for the drive and make sure that the hard
drive and floppy controllers are enabled. Also check the boot order. Note: on must
systems, you can use the autodetect feature to determine the correct CMOS
parameters and to see if it can see the drive.
• If you have no IDE drives, select no hard drives installed in the BIOS setup program.
• Check the drive’s power including the +5 and +12 volt connectors.
• If it is an IDE drive, check the jumper settings of all of the IDE devices, especially the
master, slave or stand-alone jumpers.
• If it is a SCSI drive, check the jumper settings of all SCSI devices, especially the
termination and SCSI ID numbers.
• Swap the controller card, cable and hard drive, one at a time.
• Swap the motherboard.
You get an “Error reading drive C:” error message.
• Make sure that the ribbon cable is connected properly.
• Enter the BIOS setup program and make sure that the correct CMOS settings are
chosen for the drive. Note: on must systems, you can use the autodetect feature to
determine the correct CMOS parameters and to see if it can see the drive.
• Check for viruses.
• Run ScanDisk or some other similar utility.
• For DOS, Windows 95, Windows 98 and Windows Me, try running the FDISK /MBR
command.
• For Windows NT, 2000 and XP, start the Windows installation program and choose
R to repair. You will then be asked for an emergency repair disk. If you don’t have
one, don’t repair the Registry when asked to do so. When you are in the Repair
utility, the repair utility will detect the system file and reinstall it in the correct place.
See section 14.2.
• Check the drive’s power including the +5 and +12 volt connectors.
• Swap the controller card, cable and hard drive, one at a time.
• Swap the motherboard.
• Try to reformat the drive and make the drive bootable again.
• Try a low-level format. Then partition, format and add the system files to the hard
drive.
Intel Customer Support (ICS)
Troubleshooting the IA Platform
133
You see a “No SCSI Controller Present” error message
• Check the host adapter installation including IRQ, DMA and I/O settings.
• Check to see if all of the device drivers have been installed correctly.
• Try the controller in a different system. If the control fails in a different PC, the
control is probably faulty.
• Try upgrading the System ROM BIOS and the SCSI ROM BIOS (if you are able).
You see a “No SCSI device found” or similar error
• Make sure SCSI devices have power and ribbon cables connected properly.
• Check SCSI Termination
• Check the SCSI ID numbers.
• Swap the SCSI device.
• Swap the SCSI Controller.
You see an error such as “Unknown SCSI Device” or “Waiting for SCSI Device”
• Check that the primary hard disk is set to SCSI ID 0.
• Make sure that the drive is partitioned and formatted at the primary active drive.
Make sure the partition is active.
You get a “Sector not found” error message.
• Use a utility such as ScanDisk to examine the drive for bad sectors. Note: if a file is
using a failed sector, the file will be corrupt and should be restored from backup.
• Check for viruses.
• If the problem continues, try a low-level format.
• Try swapping the cable, card and drive, one at a time.
The system reports random data, seek or format errors.
• Use a utility such as ScanDisk to examine the drive for bad sectors. Note: if a file is
using a failed sector, the file will be corrupt and be restored from backup.
• Check for viruses.
• If the problem continues, try a low-level format.
• Try swapping the cable, card and drive, one at a time.
The drive will work as a primary drive, but not as a secondary (or vise-versa)
• The drive does not spin up and the drive light doesn’t illuminate during power-up.
• Make sure that the hard drive power cable is connected.
• Make sure that the ribbon cable is connected properly.
• If you cannot access the drive, enter the BIOS setup program and make sure that
the correct CMOS settings are chosen for the drive and make sure that the hard
drive and floppy controllers are enabled. Also check the boot order. Note: on must
systems, you can use the autodetect feature to determine the correct CMOS
parameters and to see if it can see the drive.
• For hard drives, verify that your primary partition is active.
• Check the drive’s power including the +5 and +12 volt connectors.
• If it an IDE drive, check the jumper settings of all of the IDE devices, especially the
master, slave or stand-alone jumpers.
• Swap the controller card, cable and hard drive, one at a time.
• Swap the motherboard.
Intel Customer Support (ICS)
Troubleshooting the IA Platform
134
The FDISK procedure hangs up to fails to create or save a partition record for the
drive(s).
• Make sure that the hard drive power connector is connected.
• Make sure that the ribbon cable is connected properly.
• Check for viruses.
• If you cannot access the drive, enter the BIOS setup program and make sure that
the correct CMOS settings are chosen for the drive and make sure that the hard
drive and floppy controllers are enabled. Also check the boot order. Note: on must
systems, you can use the autodetect feature to determine the correct CMOS
parameters and to see if it can see the drive.
• For hard drives, verify that your primary partition is active.
• Try to recreate the master boot record by using the FDISK /MBR command.
• For Windows NT, 2000 and XP, start the Windows installation program and choose
R to repair. You will then be asked for an emergency repair disk. If you don’t have
one, don’t repair the Registry when asked to do so. When you are in the Repair
utility, the repair utility will detect the system file and reinstall it in the correct place.
See section 14.2.
• Try a low-level format. Then partition, format and add the system files to the hard
drive.
• Check the drive’s power including the +5 and +12 volt connectors.
• If it is an IDE drive, check the jumper settings of all of the IDE devices, especially the
master, slave or stand-alone jumpers.
• If it is a SCSI drive, check the jumper settings of all SCSI devices, especially the
termination and SCSI ID numbers. To make sure that the primary hard disk is set to
SCSI ID 0.
• If it is a SCSI drive, check during POST to see if the SCSI BIOS is loading and
finding any SCSI devices. If not, check the host adapter installation including IRQ,
DMA and I/O settings.
• Check to see if all of the device drivers have been installed correctly.
• Swap the controller card, cable and hard drive, one at a time.
• Swap the motherboard.
You get a “Track 0 bad, disk unusable” error message.
• Try to boot from the floppy disk drive and access the hard drive.
• If you can, make sure that the bootable files are there. Try reinstalling the necessary
boot files.
• From the floppy disk drive, run FDISK to check the partitions on your hard drive.
Make sure there is at least one DOS partition and that is active.
• If you cannot access the drive, enter the BIOS setup program and make sure that
the correct CMOS settings are chosen for the drive and make sure that the hard
drive and floppy controllers are enabled. Also check the boot order. Note: on must
systems, you can use the autodetect feature to determine the correct CMOS
parameters and to see if it can see the drive.
• Make sure that the ribbon cable and power connector is connected properly.
• Replace the ribbon cable.
Intel Customer Support (ICS)
Troubleshooting the IA Platform
135
• Check the drive’s power including the +5 and +12 volt connectors.
• Swap the controller card (or motherboard), cable and hard drive, one at a time.
• Swap the motherboard.
My hard drive seems to be running slow.
• Check for viruses.
• Defrag the hard drive.
• Check to see if you have a slow device on the cable.
• For SCSI drives, check termination.
• For IDE drives, enter the BIOS setup program and check to see if it is set to Ultra
DMA or PIO mode. For SCSI drives, enter the SCSI BIOS and check the settings.
• Check to see if any one did a low-level format on the drive. If they don’t use the
proper low-level format program, essential tracking information could have been
erased.
• Try a clean boot/install
You are experiencing frequent drive failures
• Make sure that all fans are working properly in the case.
• If you have several drives running in the system, check to see how warm the case is
and how warm the exhaust coming from the fan.
• Make sure that you are using the proper screws in mounting the drives in the case.
• Check for low-level vibrations.
• Check for power fluctuations. Make sure that you have a reliable surge protector.
• Use a voltmeter to verify that each output from the power supply is correct. If any
output is very low (especially the +5 volt output), replace the power supply.
• Check other environment factors such as smoke, heavy dust, high humidity and
anything that may generate a strong electromagnetic field such as heavy machinery
and motors.
Floppy Disk Drives
Your system generates a “FDD Controller Failure” or similar error message
• Make sure that the floppy disk drive power connector is connected.
• Make sure that the ribbon cable is connected properly.
• If you cannot access the drive, enter the BIOS setup program and make sure that
the correct floppy disk is connected and that the floppy disk drive controller is
enabled.
• Check the drive’s power including the +5 and +12 volt connectors.
• Swap the ribbon cable.
• Swap the floppy disk drive.
• Swap the controller card (if it is not built into the motherboard).
• Swap the motherboard.
When a new disk is inserted in the drive, a directory from a previous disk
appears.
This is called a phantom directory caused by not receiving the change disk signals (Pin
34).
• Make sure the ribbon cable is attached properly.
Intel Customer Support (ICS)
Troubleshooting the IA Platform
136
• Swap the floppy drive cable.
• Swap the floppy disk drive.
• Swap the floppy disk drive controller/motherboard.
The system does not recognize the floppy disk drive.
• Make sure that the floppy disk drive power connector is connected properly.
• Make sure that the ribbon cable is connected properly.
• If you cannot access the drive, enter the BIOS setup program and make sure that
the correct floppy disk is selected and that the floppy disk drive controller is enabled.
• Check the drive’s power including the +5 and +12 volt connectors.
• Swap the ribbon cable.
• Swap the floppy disk drive.
• Swap the controller card (if it is not built into the motherboard).
• Swap the motherboard.
You cannot read from or write to a floppy disk.
• Try the floppy disk in another system to make sure that the disk is good.
• If you are getting an access denied or similar message, make sure that the tab on
the disk is not set to lock.
• Make sure that the floppy disk drive power connector is connected properly.
• Make sure that the ribbon cable is connected properly.
• If you cannot access the drive, enter the BIOS setup program and make sure that
the correct floppy disk is connected and that the floppy disk drive controller is
enabled.
• Check the drive’s power including the +5 and +12 volt connectors.
• Swap the ribbon cable.
• Swap the floppy disk drive.
• Swap the controller card (if it is not built into the motherboard).
• Swap the motherboard.
Compact Disk Drives
You get a “Buffer under-run error XX” or similar error message when data is
getting transferred from one storage device to another such as a CD-R or CD-RW
drive.
• Close all applications before running the CD recording software.
• Run Disk defragmenter and ScanDisk.
• Make sure you are using updated device drivers.
• Try using a different brand of storage media.
• If possible, try to reduce the recording speed.
You get a read error when reading a CD.
• Try the compact disk in another drive. If the CD is having problems with both drives,
then try to clean it.
• Try another compact disk in the drive. If the drive cannot read either disk, try to clean
the drive heads with a special CD cleaning disk.
• Check the CMOS setup program.
Intel Customer Support (ICS)
Troubleshooting the IA Platform
137
•
•
•
Check the power and ribbon connections.
Replace the CD drive.
If the problem still exists, the problem is either the motherboard or related
components. It is suggested that you contact your distributor or notebook vendor to
setup additional troubleshooting or replacement.
Intel Customer Support (ICS)
Troubleshooting the IA Platform
138
A.7 General Expansion Card Problems
You get a “Optional XXXX Optional ROM Bad, Checksum=YYYY” error message.
A ROM chip on an expansion card is having problems.
• Remove one card at a time until you can isolate which card is causing the conflict.
• Check to see if you have a memory address conflict with another card, especially if
you just installed another device recently.
• Replace the expansion card with the faulty ROM chip.
You get no dial tone for your modem
• If it is a modem built into the motherboard, check the CMOS setup program to
ensure that the modem is enabled.
• In addition, ensure that the appropriate drivers are loaded and working correctly.
• If it is an external modem, make sure the modem has power and is connected
properly to the PC.
• Make sure that the phone line is active.
• Replace the phone cable.
• Replace the modem.
• Replace the motherboard
Intel Customer Support (ICS)
Troubleshooting the IA Platform
139
A.8 Video Problems
Windows seems to start normally, but the monitor display is unreadable, has bad
static, or is otherwise screwed up.
• Boot your system to VGA mode to correct your video setting. Check the video driver,
resolution, number of colors and refresh rate.
The image is saturated with red blue or green or appears greenish-blue (cyan),
bluish-red (magenta) or yellowish.
• If there are any color controls on the monitor, make sure those controls have not
been accidentally adjusted.
• Replace the cable that connects the monitor and video card.
• Check the connection between the monitor and video card.
• Check to see if any of the pins on the monitor plug to see if they are bent or
mangled.
• Try swapping the monitor.
• Replace the video card.
The display appears wavy or the image appears to shake.
• Make sure the power cable and the connection between the monitor and video card
is connected properly.
• If your monitor has the option, try to degauss the monitor.
• Check the AC power and look for power irregularities. You may also try a different
AC outlet. See section 6.3 and 6.4.
• Replace the cable that connects the monitor and video card.
• Replace the monitor. It is usually indicates the power supply in the monitor is going
bad.
• Replace the video card.
• Check for environment factors that may generate strong magnetic fields or power
irregularities such as heavy machinery or motors.
On your monitor, you appear to have the colors bleed or smear.
• Make sure the power cable and the connection between the monitor and video card
is connected properly.
• Check the AC power and look for power irregularities. You may also try a different
AC outlet. See section 6.3 and 6.4.
• Replace the video card.
• Replace the monitor. It is usually indicates the power supply in the monitor is going
bad.
You cannot select a higher display resolution in the Windows Display Properties
dialog.
• Make sure that you have the correct monitor driver installed.
• Make sure that you have the correct video card driver installed.
• Make sure that you have enough video memory installed to support the higher
display resolution.
Intel Customer Support (ICS)
Troubleshooting the IA Platform
140
A.9 Sound Problems
Speakers do not produce any sound
• Check the volume control in Windows and the volume control in the application to
see if the sound has been muted or has been adjusted low.
• Check the volume control on speakers.
• Make sure the speakers are turned on and have power.
• Make sure the speaker cable is installed into the correct port on the back of the
sound card.
• Make sure the sound drivers are installed and functioning properly.
• Swap the speaker cables and power cable one at a time.
There is no audio when playing an Audio CD
• Check the volume control (specifically Master Volume and CD Player) in Windows
and the volume control in the application to see if the sound has been muted or low.
• Check the volume control on speakers.
• Make sure the speakers are turned on and have power.
• Check to see if the four-wire CD audio cable is connected between the CD drive and
the sound card.
When trying to play a high-quality audio file through a USB speaker system, you
get an error message saying “Out of bandwidth”
• Disconnect extraneous USB devices such as joysticks, scanners, and others.
• Purchase an additional USB controller and attach the speakers to it so that they can
have sole access to all of its bandwidth.
• Play the audio file at a lower playback quality or use a lower-quality version of the
file.
Intel Customer Support (ICS)
Troubleshooting the IA Platform
141
A.10 Problems with Input Devices
Keyboards
Your keyboard fails to be recognized during boot up or the keyboard is
completely dead and no keys appear to function.
• Swap the keyboard. Note: you should reboot the system when a keyboard is
replaced.
• Use a multimeter to check the +5 V supply at the keyboard connector.
• Check to see if the motherboard has a fuse to protect the +5 V supply feeding the
keyboard connector.
• Replace the motherboard.
Joysticks
The Joystick does not respond
• Make sure the joystick is plugged into the game port correctly.
• When the game port has more than one connector, be sure that the joystick is
plugged into the correct connector (joystick 1 or joystick 2).
• If the game port is running through a sound card, make sure the sound card is
configured to use the part as a game port instead of a MIDI port.
• Refer to the application and see that it is configured to run from the joystick. Note:
since many new joysticks appear with supplemental functions (hat switches, throttle
control, etc), make sure that the application is written to take advantage of the
particular joystick.
• Using the Windows Device Manager, make sure that the game port is set for the
proper I/O address and that the proper driver is loaded.
• Swap the joystick.
Joystick performance is erratic or choppy and the rest of the system is running
fine.
• Make sure the joystick is plugged into the game port correctly.
• Make sure that you have the driver loaded correctly.
• Swap joysticks.
The joystick is sending incorrect information to the system or the joystick
appears to be drifting.
• Make sure the correct driver is loaded.
• Make sure that the joystick is calibrated correctly. Refer to the documentation that
came with the joystick and refer to the documentation of the application.
• Check for I/O conflicts.
• Try another joystick.
• Try another game port.
Intel Customer Support (ICS)
Troubleshooting the IA Platform
142
A.11 Serial Ports, Parallel Ports and Modems
Serial Port
The serial port on the motherboard doesn’t work.
• Check the BIOS setup program and make sure that the serial port is enabled.
• Check for resource conflicts such as I/O addresses and IRQs.
• Replace the motherboard.
Parallel Port
The parallel port on the motherboard doesn’t work.
• Check the BIOS setup program and make sure that the parallel port is enabled.
• Make sure that the correct transfer mode (ECP, EPP and etc...) has been selected
for the parallel port that relates to the device currently connected
• Check for resource conflicts such as I/O addresses and IRQs.
• Replace the motherboard.
Data is randomly lost or garbled
• Make sure the parallel cable is connected properly.
• Make sure that the cable is less then 2 meters or 6 feet long.
• Replace the cable.
• Try to connect to a different peripheral.
• Try to connect to the peripheral using a different computer.
• Try swapping the motherboard.
• Check for environmental factors such as devices that generate electromagnetic
fields such as heavy machinery and motors.
Modem
When you try to use your modem, you may receive an error message such as
“Could not open port.”
• Check for resource conflicts such as I/O addresses and IRQs.
• A program is loading in the Windows Startup folder that opens a COM port for some
other use other than the modem.
• Try rebooting to see if the port was opened by another program. Rebooting may
correct the problem.
The PC or communication software refuses to recognize the modem.
• If it is an external modem, make sure the modem has power and it is turned on. In
addition, make sure the cables are connected to the serial port.
• If it is an internal modem, make sure the modem is seated properly in the slot.
• Check in the BIOS setup program that the modem and/or serial port is enabled.
• Check for resource conflicts such as I/O addresses and IRQs.
• Try the modem in another system.
• Replace the modem. Be sure to try the modem in another system.
Intel Customer Support (ICS)
Troubleshooting the IA Platform
143
A.12 Problems Shutting Down
The system will not turn off when you press the power button.
• For ATX power supplies, the power button is designed to turn off the system only
when you press and hold the power button for more than 5 seconds. Therefore, if
you to reconfigure you’re the power button for “instant off”, use the BIOS setup
program.
• If you cannot reconfigure the power button for “instant off” you may need a BIOS
upgrade.
Problems Shutting Down Windows 95/98/Me
When Windows does not shut down properly, it may appear to stop responding for
several minutes while the “Please wait while your computer shuts down” is displayed on
the screen or you may see only a blank screen and a blinking cursor, or your computer
may restart instead of shutting down.
Windows 98 Second Edition tries to perform many functions during the shutdown
process, including completion of all disk write functions, flushing the disk cache, running
the Close Window code to close all currently running programs and transitioning all
Protected-mode drivers to Real mode.
Shutdown problems can be caused by any of the following issues:
• The Fast Shutdown registry key is enabled.
• There is a damaged Exit Windows sound file.
• A program or terminate-and-stay-resident (TSR) program may not close correctly.
• An incompatible, damaged, or conflicting device driver is loaded.
• Determine if shutdown problems are being caused by a virtual device driver (.vxd
file)
• There is an incompatible Advanced Power Management (APM) or Advanced
Configuration and Power Interface (ACPI) setting.
• There is an incompatible Basic Input/Output System (BIOS) configuration setting.
• The computer contains incorrectly configured or damaged hardware.
• There is a video adapter that is not assigned an IRQ in real mode.
The Fast Shutdown Registry Key Is Enabled
WARNING: Using Registry Editor incorrectly can cause serious problems that may
require you to reinstall your operating system. Intel cannot guarantee that problems
resulting from the incorrect use of Registry Editor can be solved. Use Registry Editor at
your own risk. Note that you should back up the registry before you edit it. If you are
running Windows NT or Windows 2000, you should also update your Emergency Repair
Disk (ERD).
Intel Customer Support (ICS)
Troubleshooting the IA Platform
144
Problems Shutting Down Windows 95/98/Me (Continued)
The Microsoft System Configuration utility includes an option to disable Fast Shutdown.
If this option is not set in Windows, your computer may reboot instead of shutting down.
To resolve this issue, change the FastReboot data value from 1 to 0 in the following
registry key:
HKEY_LOCAL_MACHINE\System\CurrentControlSet\Control\Shutdown
Note: After you apply the Windows 98 Second Edition Shutdown Supplement (available
from Microsoft), the Disable Fast Shutdown option is no longer listed on the Advanced
tab in Msconfig.exe.
There is a damaged Exit Windows sound file
• Click Start, point to Settings, click Control Panel, and then double-click Sounds and
Multimedia.
• Record the name and location of the Exit Windows sound, click Exit Windows in the
Sound Events box, and then click None in the Name box.
• Click OK, and then restart your computer.
If the issue is resolved, your Exit Windows sound file is damaged and must be replaced.
To replace the sound file, restore the file from a backup, or reinstall the program that
provided the sound file.
A program or terminate-and-stay-resident (TSR) program may not close correctly
Determine if the shutdown problem is caused by a command or driver that is loaded
automatically from the Win.ini or System.ini files:
1. Click Start, click Run, type msconfig, and then click OK.
2. Click Selective Startup, and then click to clear the following check boxes:
a. Process Win.ini file
b. Process System.ini file
NOTE: If any of these items is unavailable, that particular file does not contain any
items to load.
3. Click OK, and then click Yes to restart the computer. Test to determine if the issue is
resolved by restarting your computer.
4. If the problem still exist, re-enable the processing of the Win.ini and System.ini files.
Then to determine if the problem is with a TSR or drivers loaded in the
CONFIG.SYS and AUTOEXEC.BAT file.
5. Click Start, click Run, type msconfig, and then click OK.
6. Click Selective Startup, and then click to clear the following check boxes:
a. Process Config.sys
b. Process Autoexec.bat
7. Click OK, and then click Yes to restart the computer. Test to determine if the issue is
resolved by restarting your computer.
Intel Customer Support (ICS)
Troubleshooting the IA Platform
145
Problems Shutting Down Windows 95/98/Me (Continued)
A program in the StartUp folder may not close correctly StartUp folder
Determine whether a program in the StartUp folder conflicts with the startup process:
1. Click Start, click Run, type msconfig in the Open box, and then click OK.
2. Click Selective Startup, click to clear the Load Startup Group Items check box, click
Apply, click OK, and then click Yes to restart your computer.
Test to determine if the issue is resolved by restarting your computer. If your computer
does not shut down correctly If your computer shuts down correctly, one of your startup
programs is probably causing this issue. To determine which startup program is causing
this issue:
1. Click Start, click Run, type msconfig in the Open box, and then click OK.
2. On the Startup tab, click to select a check box, click OK, and then click Yes to restart
your computer. Test to determine if the startup program you disabled caused the
issue. If not, repeat these steps but enable a different program. When your computer
does not start correctly, you have determined that the last startup program you
enabled is the cause of the issue.
If disabling startup programs does not resolve the problem, re-enable the startup
programs.
Determine if the problem is listed in the startup log file. You can look in your startup log
file (Bootlog.prv) to find information about problems that occur during shutdown. The
Bootlog.prv file is a hidden file in the root folder of drive C. If Windows is configured to
hide system files, use these steps to display hidden files:
1. On the Windows desktop, double-click My Computer.
2. On the Tools menu, click Folder Options.
3. Click the View tab.
4. Click to clear the Hide protected operating system files check box, and then click
Yes when you are prompted to confirm this action.
5. Click OK, and then look for the Bootlog.prv file again.
6. Click Start, click Shut Down, click Shut Down, and then click OK.
If your computer does not automatically shut off, press the power button on your
computer to turn it off. Leave your computer off for approximately 15 seconds, and then
turn your computer back on. While your computer is starting, press and hold the CTRL
key. When the Startup menu is displayed, use the UP ARROW and DOWN ARROW
keys to choose Logged (\BOOTLOG.TXT), and then press ENTER.
After Windows starts, click Start, click Shut Down, click Restart, and then click OK.
To review your startup log file:
1. Double-click My Computer, double-click drive C, and then look for the Bootlog.prv
file.
2. Double-click the Bootlog.prv file. If you are prompted to select a program to use to
open the file, click WordPad, Notepad, or any text editor in the list.
3. Look for "Terminate=" entries. These entries, located at the bottom of the file, might
provide clues about the cause of the problem. If a process that is started by a
"Terminate=" entry is completed successfully, the file contains a matching
"EndTerminate=" entry.
Intel Customer Support (ICS)
Troubleshooting the IA Platform
146
Problems Shutting Down Windows 95/98/Me (Continued)
4. Check the last line in the Bootlog.prv file to see if it is one of these items:
- Terminate=Query Drivers: This is a Memory Manager problem. Your computer
might have defective memory chips or damaged files. You might need to reinstall
Windows.
- Terminate=Reset Display: Try installing an updated video adapter driver.
- Terminate=RIT: There may be timer-related problems with the sound card or an
older mouse driver. Try installing updated drivers for these devices.
- Terminate=Win32: A program is not shutting down properly. Try quitting all
programs before you shut down Windows.
An incompatible, damaged, or conflicting device driver is loaded
Some of the most common device drivers that may cause problems are sound card
drivers, video adapter drivers, and network adapter drivers. To disable device drivers:
1. Right-click My Computer, click Properties, and then click the Device Manager tab.
2. Double-click a device type, for example, Sound, Video, and Game Controllers, to
expand it.
3. Click one of the listed devices, and then click Properties.
4. Click to select the Disable In This Hardware Profile check box.
IMPORTANT: Record all of the devices you disable in this manner so that you can
enable them again later.
5. Click OK, click OK, and then restart your computer.
Test to determine if the issue is resolved by restarting your computer. If your computer
does not shut down correctly, repeat the steps but with a different component.
To test your video adapter, click Start, click Run, type msconfig, click OK, click
Advanced, click to select the VGA 640x480x16 check box, and then click OK. If you
disable your sound card, video adapter, and network adapter, but the issue is still not
resolved, re-enable these devices by clicking to clear the Disable In This Hardware
Profile check box.
If you are unable to change the video adapter by using the preceding steps, change
your video adapter driver to the Standard VGA driver by using these steps:
1. Click Start, point to Settings, click Control Panel, and then double-click Display.
2. Click the Settings tab, click Advanced, click the Adapter tab, and then click Change.
3. Click Next, click Display a list of all the drivers in a specific location, so you can
select the driver you want, and then click Next.
4. Click Show All Hardware, click Standard Display Types in the Manufacturer box,
click Standard Display Adapter (VGA) in the Models box, and then click Next.
5. Click Yes, click Next, and then click Finish.
6. Click Close, click Close, and then click Yes when you are prompted to restart your
computer.
7. If changing your video adapter to the Standard VGA driver resolves the issue,
contact your video adapter manufacturer to inquire about the availability of an
updated Windows Me video adapter driver.
Intel Customer Support (ICS)
Troubleshooting the IA Platform
147
Problems Shutting Down Windows 95/98/Me (Continued)
Determine if shutdown problems are being caused by a virtual device driver (.vxd
file):
1. Click Start, click Run, type msconfig, and then click OK.
2. Click Selective Startup, and then click to clear the Load Static VxDs check box.
3. Click OK, and then click Yes to restart the computer.
Test to determine if the issue is resolved by restarting your computer. If your computer
does not shut down correctly, click to select the Load Static VxDs check box, and then
continue with the next step. If the problem is fixed, suspect one of the following files:
This is a partial list of Microsoft static .vxd files:
Name of static .vxd file Provides support for
Vnetsup.vxd Microsoft networking
Ndis.vxd Microsoft networking
Ndis2sup.vxd Microsoft networking
Javasup.vxd Java support
Vrtwd.386 Clock
Vfixd.vxd Video Phone helper
Vnetbios.vxd Microsoft networking
Vserver.vxd Microsoft networking
Vredir.vxd Microsoft networking
Dfs.vxd Microsoft networking
Ndiswan.vxd Microsoft networking
Msmouse.vxd Microsoft Mouse
Check the properties of the incompatible static .vxd file to determine which program it is
associated with. Uninstall and then reinstall the program to replace the .vxd file with a
new copy. To check the properties of a file, click Start, click Search, type the name of
the file in the For Files or Folders box, and then click Search Now. After the file you are
looking for is displayed in the list, right-click the file, and then click Properties.
There is an incompatible Advanced Power Management (APM) or Advanced
Configuration and Power Interface (ACPI) setting
Determine if Advanced Power Management (APM) is causing the shutdown problem by
disabling it (Assuming that your system has APM):
1. Click Start, point to Settings, click Control Panel, double-click System, and then click
the Device Manager tab.
2. Double-click the System Devices branch to expand it.
3. Double-click Advanced Power Management in the device list, click the Settings tab,
and then click to select the Force APM 1.0 Mode and Disable Power Status Polling
check boxes.
4. Click OK until you return to Control Panel, and then restart your computer.
5. Test to determine if the issue is resolved by restarting your computer. If your
computer does not shut down correctly, repeat the steps, but click to clear the Force
APM 1.0 Mode and Disable Power Status Polling check boxes.
Intel Customer Support (ICS)
Troubleshooting the IA Platform
148
Problems Shutting Down Windows 95/98/Me (Continued)
There is an incompatible Basic Input/Output System (BIOS) configuration setting
- IRQ Steering
This option allows several PCI devices to share the same interrupt request line (IRQ). If
the BIOS is not fully compliant, this option may cause your computer not to shut down
properly, even if two or more devices are not sharing an IRQ. To disable PCI bus IRQ
Steering, follow these steps:
1. Click Start, point to Settings, click Control Panel, and then double-click System.
2. On the Device Manager tab, click System Devices.
3. Double-click PCI Bus, and then click to clear the Use IRQ Steering check box on
the IRQ Steering tab.
4. Click OK, click OK, and then restart your computer.
5. After you restart the computer, attempt to shut down your computer again.
If your computer shuts down successfully, you may need to change the BIOS
configuration or you may need a BIOS update. For information about how to do so,
contact your BIOS manufacturer.
There is an incompatible Basic Input/Output System (BIOS) configuration setting
- Resume on Ring and LAN
Some shutdown-related issues may be solved by disabling the "Resume on Ring and
LAN" feature in the computer's BIOS. For information about how to do so, contact your
computer or BIOS manufacturer.
There is an incompatible Basic Input/Output System (BIOS) configuration setting
- NVRAM/ESCD
There are specific settings for how the BIOS and Windows interact during the startup
and shutdown processes. To check this, disable the NVRAM/ESCD updates feature to
determine if doing so resolves the shutdown problem. To do this, follow these steps:
1. Click Start, point to Settings, click Control Panel, and then double-click System.
2. On the Device Manager tab, double-click System Devices.
3. Click Plug and Play BIOS, and then click to select the Disable NVRAM / ESCD
updates check box on the Settings tab.
4. Click OK, click OK, and then restart your computer.
5. After you restart the computer, attempt to shut down your computer again.
There is an incompatible Basic Input/Output System (BIOS) configuration setting
- Plug and Play BIOS
In some cases, the BIOS and Windows may not be communicating properly with the
computer hardware during the shutdown process. You can configure Windows to ignore
the presence of a Plug and Play BIOS and communicate directly with the hardware.
NOTE: Use this method for testing purposes only. Leaving the Plug and Play BIOS
disabled may cause some hardware to stop working.
1. To configure Windows not to use the Plug and Play BIOS:
2. Restart your computer, and press and hold CTRL until you see the Windows Start
menu.
Intel Customer Support (ICS)
Troubleshooting the IA Platform
149
Problems Shutting Down Windows 95/98/Me (Continued)
If the problem still exist, do the following:
1. Type the following line at the command prompt:
cd \windows\system
2. Rename the Bios.vxd file to Bios.old.
3. Restart your computer.
4. After the computer restarts, attempt to shut down Windows.
If the computer shuts down correctly, the system BIOS is likely to be contributing to the
shutdown problems. Contact the motherboard or BIOS manufacturer for a possible
update.
Antivirus Program
If you have an antivirus program that is configured to scan your floppy disk drive when
you shut down your computer, your computer may stop responding.
Intel Customer Support (ICS)
Troubleshooting the IA Platform
150
A10 Windows NT 4.0, Windows 2000 and Windows XP Stop Errors
STOP 0x0000000A (0xA) IRQL_NOT_LESS_OR_EQUAL
Parameter:
1
2
3
4
Memory referenced
IRQL
Value: 0 = read operation, 1 = write operation
Address that referenced memory
An attempt was made to touch pageable memory at a process internal request level
(IRQL) that is too high. In other words, a driver is using improper memory addresses. It
is usually caused by a bad device driver, or using a Windows NT 4.0 driver on a
Windows NT 3.51.
To isolate the cause of the STOP error, do the following:
•
•
•
Make sure that you have the most current BIOS.
Make sure that you have the newest Windows Service Pack.
Turn your PC off and take out all cards except the video card and SCSI card if
you are booting from the SCSI drive. After the OS is installed, shut down the
computer and plug in one card at a time, boot the OS and installed the drivers for
that card. If you get the error message again, try to update the driver.
STOP 0x0000001E (0x1E) KMODE_EXCEPTION_NOT_HANDLED
Parameters:
1
2
3
4
The exception code that was not handled
The address at which the exception occurred
Parameters 0 of the exception
Parameter 1 of the exception
While this is typically not a processor stop error, it is one of the most common bug code.
Usually the exception address pinpoints the driver or function that caused the problem.
Other causes include bad processor, bad memory, bad cache or bad BIOS or outdated
BIOS. Always note this address as well as the link date of the driver or image that
contains this address. A common error is exception code 0x80000003. This error
means a hard-code breakpoint or assertion was hit, but the system was booted with
the/NODEBUG switch. This problem should not occur very often. If it occurs repeatedly,
make sure a debugger is connected and the system is booted with the /DEBUG switch.
To isolate the cause of the STOP error, do the following:
•
Make sure that you have adequate disk space, especially for new installations.
Intel Customer Support (ICS)
Troubleshooting the IA Platform
151
•
•
•
•
•
•
•
•
Make sure that you have the most current BIOS.
Make sure that you have the newest Windows Service Pack.
Update the offending driver or reinstall the offending driver/software package.
You may need to use Last Known Good Configuration or Safe mode.
If the motherboard and hard drive support Ultra DMA-66 or faster, enter the BIOS
CMOS program and disable the Ultra DMA and set the PIO to mode 4 for that
channel. Then after Windows is installed, try the Ultra DMA to see if Windows is
stable. Be sure to use the 40-pin, 80 conductor cable.
Replace the memory
Replace the cache (if on motherboard)
Replace the motherboard
Replace the processor
STOP 0x00000023 and 0x00000024 FAT_FILE_SYSTEM or NTFS_FILE SYSTEM
This is usually caused by heavily fragmented drives, heavy file I/O, some types of drivemirroring software or some types of antivirus software.
To isolate stop error:
•
•
•
•
Temporarily disable any antivirus programs, backup programs and
defragmentation utilities.
Check for hard drive corruption by running ScanDisk.
Try Safe mode and Last Known Good Configuration
Delete all partitions, then repartition and reformat your system.
STOP 0x0000002E (0x2E) DATA_BUS_ERROR
Parameters:
1
2
3
4
Virtual address that caused the fault
Physical address that caused the fault
Processor status register (PSR)
Faulting instruction register (FIR)
This bug code is normally caused by a parity error in the system memory. This error can
also be caused by a driver accessing a 0x8XXXXXXX address that does not exist.
To isolate the cause of the STOP error, do the following:
•
•
•
•
Make sure that you have the most current BIOS.
Make sure that you have the newest Windows Service Pack.
Check for viruses.
If you added a new device driver, start the computer in Safe mode, uninstall the
driver and restart the computer.
Intel Customer Support (ICS)
Troubleshooting the IA Platform
152
•
•
Update the offending driver
Run hardware system diagnostic program. Run a general test for the entire
system and an extensive test for the memory.
•
Replace the memory
•
Replace the Video memory (or video card)
•
Replace the cache
•
Replace the hard driver
•
Replace the processor
•
Replace the motherboard
STOP 0x0000003E (0x3E)
MULTIPROCESSOR_CONFIGURATION_NOT_SUPPORTED
The system has multiple processors, but they are asymmetric in relation to one another.
To be symmetric, all processors must be the same type and level (same type, cache
size and frequency with compatible stepping). In addition, make sure the processor
supports multiprocessing. This information is stated in the Processor Specification
Update guides.
STOP 0x0000003F NO_MORE_SYSTEM PTES
A driver is not cleaning up properly
To isolate the cause of the STOP error, do the following:
•
Remove any recently installed software, including backup utilities or disk-intensive
applications such as defragmenting, virus protection and backup utilities.
STOP 0x00000058
The failure of a primary drive in a fault-tolerance set.
•
•
•
•
Using a Windows 2000 Startup disk, start your computer from the mirrored
(secondary) system drive.
Try Last Known Good Configuration.
Delete all partitions, then repartition and reformat your system.
If you are using an older machine and the drive is over 8,057 GB, you may need
to load some special software so that the system can recognize over 8 GB or update
the BIOS.
STOP 0x00000077 (0x77) KERNEL_STACK_INPAGE_ERROR
Parameters:
1 0
2 0
Intel Customer Support (ICS)
Troubleshooting the IA Platform
153
3 PTE value at time of error
4 Address of signature on kernel stack
or
1
2
3
4
Status code
I/O status code
Page file number
Offset into page file
The requested page of kernel data could not be read. This error is caused by bad block
in a paging file or a disk controller error and in extremely rare cases, it is caused by
running out of resources.
If the first and second arguments are 0, the stack signature in the kernel stack was not
found. This error is caused by bad hardware.
If the status code is 0xC000009A, it’s probably the L2 cache. If 0xC000009C and
0xC000016A, you have a failing hard drive. If the status is 0xC0000185
(STATUS_IO_DEVICE_ERROR) and the paging file is on a SCSI disk device, the
cabling and termination should be checked. Other items to check would include
defective L2 cache, defective memory, defective processor, defective motherboard and
defective hard drive.
STOP ERROR 0x00000079 MISMATCHED HAL
•
•
•
•
The Hardware Abstraction Layer is incorrect for the machine. This usually occurs
when one machine mixes single and dual processor configuration files. In Windows
NT, restart the install and when the message “Windows NT is examining the
hardware” appears, press F5. Choose your computer type from the list that appears
and continue with the installation. In Windows 2000, you should begin the installation
process again and install a clean copy of OS. See
Repartition and reformat your hard drive from scratch and install Windows.
Replace the motherboard.
Replace the processors.
Note: To change the HAL for Windows 2000 without reinstalling Windows 2000, see
section 14.4.
STOP 0x00000007B INACCESSIBLE_BOOT_DEVICE
A problem occurred during the initialization of the I/O system (usually the startup device
or the file system.
•
•
•
Make sure that you have the most current BIOS.
Make sure that you have the newest Windows Service Pack.
Check for viruses on your computer.
Intel Customer Support (ICS)
Troubleshooting the IA Platform
154
•
•
Remove any newly added hard drives or controllers.
If you are using a SCSI adapter, obtain the latest Windows driver form the
adapter vendor. In addition, try to disable sync negotiation, check termination and
check the SCSI Ids of the devices.
•
If you are using IDE devices, define the on-bard IDE part to Primary only. Check
the Master/Slave/Only settings for the IDE devices. Remove all Ide devices except
for the hard disk.
•
Run CHKDSK. The file system could be corrupt. If Windows cannot run
CHKDSK, you might have to move the drive to another computer running Windows
2000 or Windows XP and run the CHKDSK command on that drive.
•
Restart your computer. At the startup screen, press F8 for the Advanced Startup
option, and select Last Known Good Configuration.
STOP 0x0000007F UNEXPECTED_KERNEL_MODE_TRAP
This error means a trap (exception) occurred in kernel mode, either a kind of trap that
the kernel is not allowed to have or catch (a bound trap) or a kind of trap that is always
instant death (double fault). This can be caused by a hardware or software problem, but
is most commonly a hardware failure. The first number in the bug code parenthesis is
the number of the trap (8 = double fault). To learn more about what these traps are,
consult an Intel® Software Developer Manual, Volume 3: System Programming Guide,
Chapter 5. Check for mismatched RAM, bad RAM, bad processor or outdated BIOS.
To isolate the cause of the STOP error, do the following:
•
•
•
•
•
•
•
•
•
•
•
Make sure that you have the most current BIOS.
Make sure that you have the newest Windows Service Pack.
Run diagnostic software and test the RAM in the computer.
Disable memory caching.
Turn your PC off and take out all cards except the video card and SCSI card if
you are booting from the SCSI drive. After the OS is installed, shut down the
computer and plug in one card at a time, boot the OS and installed the drivers for
that card. If you get the error message again, try to update the driver.
Try removing or swapping out controllers, cards or other peripherals.
Check to see if the RAM modules are the same speed and that they are
compatible with each other.
Make sure that you have sufficient RAM to run Windows.
Try different RAM
Try a different motherboard.
Try a different processor.
STOP 0x00000080 NMI_HARDWARE_FAILURE
The HAL is supposed to report whatever specific data it has and to tell the user to call
his or her hardware vendor for support.
Intel Customer Support (ICS)
Troubleshooting the IA Platform
155
To isolate the cause of the STOP error, do the following:
•
•
•
•
•
•
Make sure that you have the most current BIOS.
Make sure that you have the newest Windows Service Pack.
Check to make sure the system is not using non-parity memory in a parity based
system or mixing of parity/non-parity memory
Try different RAM
Try different motherboard
Try different processor
STOP 0x0000009C Machine_Check_Exception
A fatal Machine Check Exception has occurred.
For Pentium processors, the parameters are:
1
2
3
4
Low 32 bits of P5_MC_TYPE MSR
Not Applicable
High 32 bit of P5_MC_ADDR MSR
Low 32 bits of P5_MC_ADDR MSR
Note: Machine-Check Exceptions are listed in the Intel Architecture Software Developer’s
Manual – Volume 3: System Programming Guide, chapter 12.
For a processor with a P6 architecture, the parameters are:
1
2
3
4
Bank number
Address of field of MCi_ADDR MSR for the MCA bank that had the error
High 32 bits of MCi_STATUS MSR for the MCA bank that had the error
Low 32 bits of MCi_STATUS MSR for the MCA bank that had the error
Note: Machine-Check Exceptions are listed in the Intel® Architecture Software
Developer’s Manual – Volume 3: System Programming Guide, chapter 12.
To isolate the cause of the STOP error, do the following:
•
•
•
•
•
Make sure that you have the newest Windows Service Pack.
Make sure that you have the most current BIOS update.
Replace the RAM
Try a different motherboard
Try a different processor or processors
STOP 0x000000B4 VIDEO_DRIVER_INIT
The video driver is bad, corrupted, missing, or disabled.
Intel Customer Support (ICS)
Troubleshooting the IA Platform
156
To isolate the cause of the STOP error, do the following:
•
Try starting your computer in safe mode or Enable VGA mode and trying removing
the driver and reinstalling the driver.
• Make sure that you have the newest Windows Service Pack.
• Make sure that you have the most current BIOS update.
STOP 0XC0000185 STATUS_IO_DEVICE_ERROR:
• Make sure the SCSI controller and SCSI cable is connected correctly.
• Check for Resource Conflicts for your SCSI card.
STOP 0x00000221 STATUS_IMAGE_CHECKSUM_MISMATCH
The installer program found a corrupt driver or DLL file for some hardware device.
•
•
•
Make sure that you have the most current BIOS.
Make sure that you have the newest Windows Service Pack.
Turn your PC off and take out all cards except the video card and SCSI card if
you are booting from the SCSI drive. After the OS is installed, shut down the
computer and plug in one card at a time, boot the OS and installed the drivers for
that card. If you get the error message again, try to update the driver.
Intel Customer Support (ICS)
Troubleshooting the IA Platform
157
Was this manual useful for you? yes no
Thank you for your participation!

* Your assessment is very important for improving the work of artificial intelligence, which forms the content of this project

Download PDF

advertising