Testing
Testing hardware is a major part of any organisation that refurbishes used computers. The
purpose of testing is to determine which equipment is still usable, and which cannot be
passed on to other users. The process of testing is essentially diagnostic – it aims to answer
the question „Can this computer part be used again?‟
In order for testing to be really effective, you can‟t just test the whole computer at once. Each
component part must be tested individually.
Luckily, it doesn‟t mean that the computer has to be taken apart entirely. To test each part of
the machine, you can use diagnostic software to probe and examine each piece of the
computer.
There are a number of commercial hardware testing programs available, and there are a few
free and open source programs available as well.
Unfortunately, most programs that are free are incomplete, and the ones that are give the
ease and function a centre needs but are expensive. However, given that testing is a core
function of any refurbishment centre, it‟s a worthwhile investment to pay for a good, reliable
product that will help to streamline and standardize the process of testing.
The industry standard product is American Megatrends‟ AmiDiag Suite. It costs US$259, but
comes with special kinds of hardware that allow all parts, including serial ports, which for
instance, external modems connect to, to be tested. (Visit http://www.amidiag.com/products/
for details.) AmiDiag is user friendly and well documented. It can be purchased online, but not
all African countries are eligible to use credit cards online. To get around this problem, it might
be possible to ask a supplier – one with whom you have an existing relationship - to buy the
product on the centre‟s behalf and include it in a shipment of computers. The software can be
paid for at the same time as the invoice for computers and other equipment is settled.
Many freeware utilities exist; many cheaper options exist as well.
There are some good free utilities that test some parts and not others. One very capable
memory testing tool is called MemTest86. It is widely used, well supported and easy to use. It
can be easily downloaded from the project website. http://www.memtest86.com/. Bootable
versions are available; they can be created on both DOS and Linux platforms.
There are many other testing utilities such as TuffTest (http://www.tufftest.com/index.htm;
US$29.95) and PC Check (distributed by Eurobyte). TuffTest-Lite is a free, limited-
functionality distribution of Tuff Test Pro. It‟s available at http://www.tufftest.com/tt01-lite-
dwnld.htm. One major drawback of the free utilities is that they don‟t include the whole suite of
tests offered in the commercial versions – TuffTest Lite can‟t, for instance, test a CD-ROM,
one of the most common parts to found to have failed in the process of testing. It also tests
only the first 8mB of RAM – even though today‟s computers may commonly have eight or
sixteen times that amount. It only tests the first 540mB of a hard disk‟s capacity. Still, TuffTest
Lite can be used, but only if you use it in conjunction with other tools in order to cover the
areas that the free utility leaves out.
This document is based on draft text taken from the bridges.org report
"Set-up and Operation of a Computer Production and Support Centre - A
How-To Guide", to be published July 2004.
Another issue that a refurbishing workshop needs to address involves hard-drives and the
data that comes on donated machines. The problem is really two-fold: not only does the
centre need to make sure the drive works but corporations and private donors often need
assurances that the data on its hard drives will be destroyed such that no one would have
access to private or confidential information that was once stored on a disk. By using a testing
procedure that reads, writes and destroys data, you can combine these convergent needs.
A Note on Data Destruction
In 1998, the UK government mandated that companies which collect or store personal or
confidential data are legally responsible for ensuring that all traces of personal data are
destroyed when a computer is decommissioned or resold. Fines amount to GBP 5,000. This
means that any hard drive that once contained correspondence, tax information, personnel
records, or virtually any other kind of document must be not just thoroughly erased, but
overwritten three or four times.
The U.S. military takes the issue so seriously it has issued guidelines about how to destroy its
data. Many companies will claim that their data destruction methods comply with the
Department of Defense standards, but no claim about any piece of data-destroying software
is true: the DoD mandates that its employees, in the last stage of a destruction process, use a
strong magnet called a DeGausser that not only destroys data, but also the hard-drive‟s ability
to read and write data.
Mere erasure or formatting is not enough because some data recovery tools can restore data
that has been erased by a user. Today‟s security advisors demand not just deletion, but a
kind of data destruction: they ask that the substance that once held the data be overwritten
multiple times. Overwriting with different patterns three or four times makes sure that almost
every last trace of old data is unrecoverable. A Professor Gutmann in New Zealand has
devised a method that overwrites the data on a hard drive 35 times, and uses a different
secretly generated number at eight different points in the process. The method was so
rigorous it became known as Gutmann‟s protocol. This is probably a little too intensive a
method for most people – it also takes eight or more hours.
The problem of data destruction is so common that there are many suitable utilities for
performing a conscientious but not excessive data destruction routine on hard drives.
Here is a list of two: one is called Active@ KillDisk (free version is available at
www.killdisk.com/downloadfree.htm). Another uses a standard linux bootdisk called Tom‟s
Root Boot (see Boot Disks for download and creation instructions) and a conventional Linux
utility called badblocks. The advantage of using badblocks with Tom‟s Root Boot is that it
allows several disks to be blanked at once. When a single test can take an hour or more, it
can save a lot of time.
Sometimes you need to run diagnostics on hard drives too. In these cases, it‟s best to turn to
the hard drive companies themselves. Most hard drive manufacturers issue their own free
diagnostic tools for use testing their own products. It‟s a good idea to create a library of
manufacturer‟s tools to test hard drive performance. The site
http://www.motherboard.cz/diagtest/ provides links to tools created by the most common
manufacturers. You can also search Google for such things: try search terms „hard drive‟ and
„test‟ or „utility‟. Manufacturers themselves sometimes call their products „drive fitness test‟.
Lastly, monitors need to be tested for quality and clarity. A good free utility is available from E-
Leader. Its website is Monitortest.net. Monitor assessment is more qualitative than
quantitative; monitor tests display a series of patterns on a screen that allow a technician to
judge certain properties of the screen.
Summary
This document is based on draft text taken from the bridges.org report
"Set-up and Operation of a Computer Production and Support Centre - A
How-To Guide", to be published July 2004.
The main criteria a testing utility should assist a refurbishing workshop with is decision-
making: does the testing software give a clear verdict on what‟s wrong? Does it tell the user if
the parts work or not?
The most useful kind of pronouncement a testing or diagnostic utility should give is error
reporting detailed enough for the workshop to determine a computer‟s PASS or FAIL rating. If
the software doesn‟t give a pass or fail rating, or doesn‟t report errors so that a technician can
pass or fail a part based on the number or type of errors encountered, then it is not software
appropriate for a workshop.
For these reasons, AmiDiag may well be a worthwhile investment. TuffTest, which is a tenth
the price, may be a good option for smaller volume refurbishing workshops, and may be a low
cost option for new organisations. MemTest86 is free, robust, well supported and decisive –
it‟s a reliable choice.
This document is based on draft text taken from the bridges.org report
"Set-up and Operation of a Computer Production and Support Centre - A
How-To Guide", to be published July 2004.
Testing Software – Getting and Installing
This section explains how to find and install the following software:
MemTest86
AmiDiag
TuffTest Lite
KillDisk
Tom‟s RootBoot (see BootDisks)
VideoTest1.exe from E-Leader‟s Monitortest.net
MemTest86
MemTest86 runs from a bootable floppy drive. It runs under Linux, but you can use a DOS-
based program as well as Linux ones to create it.
To create MemTest86 bootable floppies with DOS, download the zipped contents from
http://www.memtest86.com/memt31a.zip. Decompress the files into a folder. Open a terminal
by going to clicking on Start > Run; type command to bring up a command line. Cd into the
folder where you decompressed memtest86. type install.bat. The prompts will guide you what
to do: insert a disk into the floppy drive. Identify the drive where the floppy is by its letter (a or
b, typically). Press enter.
To see how to use MemTest86 to test RAM, go to Testing Procedures.
AmiDiag
Protecting The Program: because the software is so expensive, the originals should be kept
in a safe place. Only copies of the originals should be allowed to be used in the workshop.
Use diskcopy to make copies of the original. (From a DOS command line, type diskcopy a: a:
and follow the prompts). On a linux machine, create a directory to store the backup. (It‟s not a
bad idea to create a whole backup directory, and to back up software and data to
subdirectories on a regular basis).
mdir AMIbackup
This document is based on draft text taken from the bridges.org report
"Set-up and Operation of a Computer Production and Support Centre - A
How-To Guide", to be published July 2004.
and move into it.
cd AMIbackup
Use
mcopy a:[space]. (don‟t type [space], insert a space with the spacebar.)
to copy the DOS-formatted disk contents to your current directory.
mcopy . a:
To copy the files to a new floppy (it should be formatted first). (see Linux guides in the
appendix for more information).
Using the Program.
The full user manual appears in the appendix, and should also be included with an orginal
purchase. AmiDiag can run from floppies, but to use the whole suite of utlilities, it needs to be
installed.
Since the drive will be formatted after it‟s been tested, it‟s sensible to install the whole
program to the drive. Boot a computer using a bootable DOS disk (see Boot Disks for sources
and guidance).
Insert an AmiDiag disk and type install. Press any key to start installation, then press enter to
accept the default installation directory (c:/diag/amidiag). Follow the instructions.
Operation: AmiDiag is menu-driven and also uses the function keys. Move around menus with
arrows; make selections with the enter key. Its interface should be especially familiar to
anyone who has used a BIOS setup program. Function keys also work well.
TuffTest Lite – Getting and Installing
Download Tufftest Lite from http://www.tufftest.com/tt01-lite-dwnld.htm.
Find a floppy/stiffydiskette.
Double-click ttlsetup.exe to launch the installer. Follow the prompts.
Insert the diskette when prompted.
Follow the instructions that appear on screen. (You can choose to enter your data; you can
also fill in the fields with a random letter).
This document is based on draft text taken from the bridges.org report
"Set-up and Operation of a Computer Production and Support Centre - A
How-To Guide", to be published July 2004.
Read the How To Load directions that appear on the last screen.
Read the help document that opens automatically for any further guidance.
Tom’s RootBoot.
Tom‟s root boot is a bootable disk. How to get it is outlined in the section called Boot Disks.
Active@ Kill Disk:
Active@ Kill Disk is comes as a 600Kb utility that can be downloaded in a few minutes.
Download the software to your hard drive from http://www.killdisk.com/downloadfree.htm.
Download and run Bootable Floppy Disk Creator for Active@ KillDisk.
Use:
To create a floppy, double click the program called KillDiskFloppySetup.exe.
VideoTest1.exe
A website called monitortest.net provides a free utility that operates on a windows platform. It
can be downloaded from
http://www.monitortest.net/downloader.asp?ref=VideoTest1.exe
It runs from within a running Windows environment.
This document is based on draft text taken from the bridges.org report
"Set-up and Operation of a Computer Production and Support Centre - A
How-To Guide", to be published July 2004.
Testing: Preparing the Workshop
Testing should be run in an area of the workshop dedicated to finding out if the computers
coming into the workshop have any faults, and what, if any, parts are usable.
Testing is the first stage in refurbishing a computer. Its purpose is to remove faulty parts and
computers from the production stream, and to verify that parts that are destined for use by a
client of the supply centre do actually work.
Testing also gives an opportunity to take an inventory. Most testing programs also identify the
hardware present; technicians can test items and record the specifications of each computer
that enters the workshop.
Workshop testing should focus on four things: the integrity of the memory, the function of the
motherboard and other devices, and the quality of the hard drive and the quality of the
monitors.
Central to the idea of testing is the importance of controlling the number of variables in the
testing environment. This means standardizing both how something is tested – by using the
same software testing programs) and in what computer it is tested – by using the same
equipment to test the same pieces.
This means that some specific, dedicated testing systems need to be built.
Four kinds of testing systems need to be built:
A RAM testing area should be able to allow a number of computers without peripherals to be
put on a bench, plugged into a monitor and keyboard that are known to work, and be able to
be left unattended for the duration of the test. It‟s best to test RAM in the machine it first
arrived with; not all motherboards are compatible with all speeds and types of RAM. For more
on this issue see Crucial.com‟s interactive RAM diagnostic tool at http://www.crucial.com
A hard-drive testing machine should have all the parts of a computer and all peripherals,
but no dedicated hard drive. All parts – including the IDE cable – should have been tested
beforehand. Each hard drive to be tested should be installed on each available connector on
an IDE channel (see PC Assembly for a description). This allows up to four drives to be tested
at once. Since the tests take several hours, at times it may be necessary to build more hard
drive testing stations if a shipment has been particularly large.
A components testing area should be built that allows multiple machines to be easily
swapped in and out from a table fitted with a set of peripherals. The picture below, taken from
an assembly centre in Johannesburg, shows two full sets of peripherals set up around a work
area. The open table allows two technicians easy access to the insides of the machines;
putting the monitor on a shelf above increases the work area. The standard set of peripherals
allows technicians to isolate and attribute any errors they find to the machines themselves,
because the peripherals are known to function.
This document is based on draft text taken from the bridges.org report
"Set-up and Operation of a Computer Production and Support Centre - A
How-To Guide", to be published July 2004.
A monitor testing area should be built to test several monitors at once. It requires its own
fusebox, some shelving, a video splitter, and one fully functional, dedicated testing machine.
Fusebox: Monitors carry a lot of voltage, and if they‟re faulty, they can cause a short circuit.
It‟s best to protect the rest of the work area from the monitor testing area by making sure the
test area is on its own fused circuit. (These details are be covered in the Workshop
Modifications: Electricity section of the Workshop module).
Shelving: Monitors need to be tested over a long period of time, so it‟s best to create a system
where many can be tested in parallel. Build a set of shelves that house a minimum of eight
monitors, with plenty of space on both sides of the shelves to make it easy to move bulky
monitors in and out.
Video splitter: A video splitter is a small device that shares the video signal from one PC to
several monitors at once. Typical splitters have two, four or eight output connectors; linking
two eight-way splitters allows a technician to test 15 monitors at once. Splitters cost about
US$100 online. As with most video technology, different brands likely have the same
components inside, so there may not be any disadvantages to buying the cheapest unit. Use
the term „vga splitter‟ or „video splitter‟ to search online vendors and auction sites such as
froogle and eBay for pricing. See if an overseas supplier can‟t ship one along with a container
of computers.
Video/VGA Splitter: notice the nine ports. One takes a feed in from the central computer; its
signal is replicated through eight output ports.
The construction of the burn-in testing area is described in the Multiple Installations section.
This document is based on draft text taken from the bridges.org report
"Set-up and Operation of a Computer Production and Support Centre - A
How-To Guide", to be published July 2004.
Testing Procedures
This section outlines procedures for testing memory, hard drives, components and monitors.
To see how to install or use the testing software, see Testing Software: Getting and Installing.
Memory Testing
Using Memtest86 to Test Memory
(adapted from FreeGeek.org‟s Memory testing procedure –
http://www.freegeek.org/howto/testing/memory/index.html).
Insert the Memtest86 disk into the floppy drive. Turn on the computer.
The tests will start automatically.
If the beeps wildly, there is either a memory or a keyboard problem. Double check the
keyboard. Otherwise the memory is bad.
Watch the POST happen. The memory test should indicate the amount of RAM in the
system. If you see an error here, the RAM is bad.
Boot from the MEMTEST86 floppy.
The process can take a while. Keep an eye on the system, and look for errors:
The screen is divided into three parts (upper left, upper right, and lower)
Several tests will be run on the memory. This can take several minutes to run.
In the upper right part, you will see two status bars made up of "####" marks, indicating the
percentage complete. The top of these status bars indicates the overall status. The lower of
these indicates the current test.
Below the status bars somewhere an error count is displayed. If this error count goes above
zero, the memory is bad.
As errors are encountered, they are listed in the lower portion of the screen.
If you do not encounter errors, mark the RAM as good, mark its size, and file it in the correct
storage box.
If you encounter errors, the RAM may be bad. If the RAM appears bad follow this
criteria:
If it is a DIMM, hold onto it. It may test without errors in a different machine.
If it is a SIMM, only one of the two may be bad. If there are more than two SIMMs that
work together, you can systematically swap out RAM until you isolate the bad stick. Mark that
one as bad, but keep the other.
At the end of a testing session take all the DIMMs that tested bad during MemTest86 and sort
them according to speed and memory type using an interactive diagnostic tool available at
http://www.crucial.com. Source a machine with a compatible motherboard and chipset and
retest.
If RAM tests bad on a machine whose motherboard is known to be compatible, try at least
one more compatible machine before dispatching those DIMMs to the recycler. RAM is
expensive and worth trying to save.
This document is based on draft text taken from the bridges.org report
"Set-up and Operation of a Computer Production and Support Centre - A
How-To Guide", to be published July 2004.
Hard Drive Formatting and Testing
How to test, label, and sort IDE hard disk drives.
(adapted from FreeGeek‟s Hard Drive Testing How To
http://www.freegeek.org/howto/testing/harddrive/oldharddrive.html)
Overview:
Attach up to four IDE drives to the cables on the testing station. Turn it on. Run the disk
formatting and checking utilities. Check for errors. Detach harddrives.
Detailed Instructions
Introduction
The IDE Hard Drive testing Station has no operatring system installed that boots from a hard
drive. This allows all 4 IDE connections to be used for testing simultaneously. Use the "Tom's
Root Boot" floppy to boot to a workable command line interface. This document assumes you
have some basic familiarity with what IDE is, what the difference between master and slave,
etc.
Setup
First, up to 4 IDE drives will need to be connected. Next, determine the jumper settings for
master and slave. BE SURE THE POWER IS OFF ON THE TESTING STATION and connect
the drives. Once the drives are physically connected to the testing station, power on and boot
to the "Tom's Root Boot" floppy. Enter "root" for a login and "xxxx" for a password.
Hard Drive designations in Linux
These are different from the logical drive letters specified in DOS/Windows:
/dev/hda - Primary Controller, Master
/dev/hdb - Primary Controller, Slave
/dev/hdc - Secondary Controller, Master
/dev/hdd - Secondary Controller, Slave
Using fdisk to create a single partition on the drives
This is a different interface than the MS-DOS version of FDISK. Type in:
fdisk /dev/hda
This will bring you into the fdisk command for hda (Entering hdb, hdc, hdd will examine/modify
the rest.) Unlike MS-DOS FDISK, you must exit and re-enter the command to modify a
different physical drive's partitions, though a reboot is not required after modifying the
partitions.
Once you‟re at the fdisk command prompt, type ? to bring up a list of commands.
First, use the d command to delete any existing partitions, then create a primary partition with
the n command and make it active. You must enter P for primary type and the first and last
cylinder of the drive. Although not neccessary, we usually make the partition bootable with the
a command. Once created, use the w command to write the changes to the partition table and
exit the fdisk program. Repeat the above steps for hdb, hdc, and hdd.
Using mke2fs to create a file system (format) the drives
For each drive, enter:
This document is based on draft text taken from the bridges.org report
"Set-up and Operation of a Computer Production and Support Centre - A
How-To Guide", to be published July 2004.
mke2fs /dev/hdn1
... where n is the letter of the drive (a, b, c, d).
example: mke2fs /dev/hda1
Be sure to add the 1 - designating the 1st (and only) partition. This process goes rather
quickly compared to the DOS FORMAT command.
Next.check to be sure the file system was created with the e2fsck command:
e2fsck -f /dev/hdn1
This proccess again goes quickly - it's an empty file system.
Then check for bad blocks with the badblocks command
Here's where the real work starts. One of the nice aspects of Linux is that unlike DOS, it is a
multi-user, multi-tasking operating system. You switch between virtual consoles using the left
ALT key plus f1, f2, f3, f4, etc. For 4 drives, we'll use these four. Since we‟re already at the
first (ALT-f1) console, and it's nice to do things in order, enter:
badblocks -b 512 -svw /dev/hda1
This will destroy any existing data. While the Console #1 works away, use ALT-f2 and login
again as "root", using "xxxx" for the password. At the shell prompt, enter the badblocks
command just like above, only using hdb1 instead of hda1. Then ALT-f3 and repeat the
proccess for hdc1, and ALT-f4 and repeat for hdd1.
The process can easily take several hours, depending on the speed and capacity of the
drives. It will do four writes per disk, and four read/comparisons. Bad blocks will be displayed.
Use the ALT-f1 through ALT-f4 to switch between consoles and read the results. Once all 4
drives are finished, you can power down the system. Tom's Root Boot does not have to be
shut down like a normal Linux installation.
Mark the drives' test results. If by the time the utility has scanned and read and written onto
the surface four times in succession, and bad blocks (the smallest units on a drive) are still
appearing, that drive is should be checked with a manufacturer‟s utility.
The reappearance of bad blocks after four passes is a sign of ill health on the drive. Modern
hard disks have extra sectors and dynamically remap bad blocks to the spare blocks with no
user or software intervention. When a hard drive detects an error, it silently fixes itself. When
errors persist after four passes, it is a sign that the drive's spares are all filled and there are
still more problematic areas. The hard drive has many, many more bad sectors than are
visible. It should be disposed of.
This document is based on draft text taken from the bridges.org report
"Set-up and Operation of a Computer Production and Support Centre - A
How-To Guide", to be published July 2004.
Components Testing
Using TuffTest Lite
Put the TuffTest Lite floppy diskette into the drive. Turn on the computer.
While the computer is booting, check the following physical characteristics:
Does the power supply fan work? Put your hand behind the grille at the rear of the machine.
Does air blow out?
Find the CPU. Is the heatsink in place. Is there a fan? Does it spin?
Check the lights on the case: is there a power light? Disk activity light? Does the floppy drive
activity light turn on when the floppy is accessed?
Are all cables clear of the fans? Are they secure on the motherboard?
If all physical characteristics are okay, proceed to the diagnostics test.
Diagnostics
Follow the prompts. TuffTest Lite will start automatically. All components will be detected and
tested. Watch for failures. If any components fail, remove them (if possible) and start the test
again.
If the on-board video fails, remove the motherboard from the case. Put the motherboard into
the recycling area for Printed Circuit Boards.
If the CPU fails, remove the motherboard from the case. Put the motherboard into the storage
area. Mark the CPU for recycling.
If any slot devices fail, remove them. Reboot and test again.
If no items fail, mark the machine as good.
This document is based on draft text taken from the bridges.org report
"Set-up and Operation of a Computer Production and Support Centre - A
How-To Guide", to be published July 2004.
Components Testing
Using AmiDiag
Boot using a Windows 98 floppy.
While the computer is booting, check the following physical characteristics:
Does the power supply fan work? Put your hand behind the grille at the rear of the machine.
Does air blow out?
Find the CPU. Is the heatsink in place. Is there a fan? Does it spin?
Check the lights on the case: is there a power light? Disk activity light? Does the floppy drive
activity light turn on when the floppy is accessed?
Are all cables clear of the fans? Are they secure on the motherboard?
If all physical characteristics are okay, proceed to the diagnostics test.
Diagnostics
At the a:> prompt, insert disk one of the AmiDiag testsuite. Type install. Follow the prompts.
Once in the AmiDiag program, use the arrows to navigate to the Options Tab, all the way on
the right. Hit enter. Scroll down and choose “Toggle All Tests” and then further down to “Run
Selected Tests”. When the tests are complete, make a note of any errors found:
If the on-board video fails, remove the motherboard from the case. Put the motherboard into
the recycling area for Printed Circuit Boards.
If the CPU fails, remove the motherboard from the case. Put the motherboard into the
recycling area for Printed Circuit Boards.
If any slot devices fail, remove them. Reboot and test again.
If no items have failed, go to Options. Select System Info. Make a record of the System
Information on the product data sheet.
This document is based on draft text taken from the bridges.org report
"Set-up and Operation of a Computer Production and Support Centre - A
How-To Guide", to be published July 2004.
Monitor Testing
Patch the monitor into the video splitter.
Monitor testing requires one low-end computer running a Windows operating system for
testing. It should be installed with VideoTest1.exe. Double click VideoTest1.exe to launch the
program.
Common problems on old monitors include poor focus, a persistent flicker, a hum, or a
distorted screen. These tests will try to produce those effects.
First Stage: Screen Quality
Look at each screen carefully during each test. Take time to check all 15 screens.
Click on Colour > White. Is the colour even along the screen face? Any dark spots or rings?
Click on Focus and bring up one of the grid tests. Look at each monitor. Can you see each
dot in the middle of the square? Are the lines straight? Do they appear to move?
Left click to cancel the test and see the menu again. Choose Colour > Grayscale 1. Can you
see all the different shades of grey on each screen?
Left click to cancel the test and see the menu again. Choose Colour > Grays 3. Is each
square distinct? Does one edge bleed into another?
Left click to cancel the test and see the menu again. Choose Colour > Colour Scale. Does
each colour look the right shade? Adjust brightness and contrast if necessary.
Left click to cancel the test and see the menu again. Choose Concentric > Crosses 1. Check
for straightness and crispness of the lines.
Left click to cancel the test and see the menu again. Click Dynamic > Voltage regulation. Do
black and white alternate on screen? Does the monitor make any cracking noises when it
switches between tones?
Stage 2: Tolerance.
Stop the videotest by choosing exit. Start a screensaver (right click on the Desktop and
choose Properties to adjust the screensaver)
Leave the monitors to run for about two hours, then run stage 1 again. See if there are any
differences in the quality of the display.
Check each monitor for heat. Excessively hot vents, screens, power cords and video cables
mean that there is too much electronic resistance in the unit. Label it for recycling.
If colours are uneven or blotchy, the monitor may have to be degaussed. Degaussing is a
consequence of the high voltages and coils that a monitor produces.
Degaussing: from whatis.com. “Degaussing a computer display removes any undesirable
magnetic field that may have built up within the display, causing distortion of the image or
color. Display monitors with cathode ray tube (CRT) technology are subject to the buildup of
magnetic fields. Many display manufacturers include an internal coil that will degauss the
display when it is turned on. Some displays also include an external control that a user can
operate to degauss the display at any time.”
(http:www.whatis.com/definition/0,,sid9_gci213666,00.html).
This document is based on draft text taken from the bridges.org report
"Set-up and Operation of a Computer Production and Support Centre - A
How-To Guide", to be published July 2004.
Check on the monitor controls to see if there is a manual degaussing option. If there isn‟t
one, label it for recycling. (or see Advanced Monitor testing).
If the picture has deteriorated over two hours, label the monitor for recycling.
Stage 3: Sizing
Restart VideoTest1.exe. Choose Resolution > 600 x 800.
If the screen can‟t display 800x600, label it for recycling.
Measure the size of the display. Is it at least 15 inches in diameter?
If the screen is smaller than 15 inches in diameter, label it for recycling.
Keyboards and Mice
Keyboards and mice should be sorted on arrival according to their connection type; anything
that doesn‟t have a PS/2 connector (small, round, 7-pin connector with a notch in the top)
should be decommissioned.
Keyboards and mice can be tested as they‟re attached to a computer in the final stage of
configuration. The only barometer for testing is actual performance: to test a keyboard, check
for sticking and loose keys, and reject any where buttons don‟t work.
The sentence “Quick brown foxes jumped over the lazy dog” tests every letter of the alphabet.
Mice trackballs and rollers should be cleaned of any dirt and debris by opening the cage on
the underside. Technicians should check each button by double-clicking, right clicking and
single clicking while the computer is running.
Mice and keyboards are generally not valuable enough to spend time fixing. If they don‟t
work, recycle them.
This document is based on draft text taken from the bridges.org report
"Set-up and Operation of a Computer Production and Support Centre - A
How-To Guide", to be published July 2004.