Windows 98 Installation & Configuration Handbook

Previous chapterNext chapterContents


- 23 -

Configuring Network Hardware


by Christopher Gagnon

Installing Network Interface Cards

Hundreds of network cards are available today. If your card manufacturer has done its homework, installation can be as simple as installing the card and booting the PC. Unfortunately, with the recent boom in multiplayer gaming, everybody and his brother has decided to jump into the marketplace. You will undoubtedly find lots of great deals on network cards, but these prices are usually subsidized by unstable drivers and a lack of support. Remember that you get what you pay for. If the choice is between an unknown company and a big company with a $20 difference in price, spend the extra buck for the well-known company. Fortunately, Microsoft has eased the installation process of nearly every network adapter by including every network card driver it could get its hands on with Windows 98. In the following sections, you'll learn about the installation and configuration of network adapters in your system.

Checking Network Hardware on Your System

Before you install any new network cards in your machine, checking for any existing or previously installed equipment is a good idea. Network cards eat up system resources such as IRQs and static memory ranges. If you have multiple cards in your machine, you should note these settings before you install. This way, you can avoid conflicts during the installation process itself.

The easiest way to check the status of network equipment is through the Device Manager (see Figure 23.1). You can access the Device Manager by selecting the System from the Control Panel. The Device Manager lists all the hardware installed on your system. You can view the network cards installed in your system by clicking the plus (+) sign next to network adapters.

FIG. 23.1 The Device Manager allows you change resource settings for every device in the PC.


NOTE:etwork adapters do not appear in the Device Manager if you do not have a network adapter installed.

The three areas in the Network Adapters section of the Device Manager are as follow:

FIG. 23.2 The General tab gives you basic information about your network card.

FIG. 23.3 The Driver tab allows you to perform driver-specific functions.

FIG. 23.4 The Resources tab shows you which system resources are in use by your adapter.


NOTE: Some network cards do not allow you to change certain settings. If you have a resource conflict but cannot change the settings on your network card, you have to change the resource on the conflicting device.

Installing a Network Interface Card

The network interface card is your PC's passport to the network. Most network adapters contain a port for whichever type of network you are installing and a link light to verify the connection status. The installation process is pretty straightforward and happens in one of two ways. In the following sections, you will examine both the Plug and Play and manual installation methods.

Plug and Play  Plug and Play technology, which was introduced in Windows 95, is one of the most useful features to carry over into Windows 98. This technology allows you to install a piece of supported hardware without much user input. To install a Plug and Play card, use the following process:

1. Install the card in a free expansion slot following the manufacturer's instructions.

2.
Boot the PC.

3.
Windows 98 then detects the card and informs you that it has detected new hardware.

4.
The installation process then completes, and the machine finishes booting.


NOTE: By default, Windows 98 assigns protocols and services to your network card. You need to verify that the protocols and services you need are configured correctly.

Depending on your network card, you may be asked to verify certain settings for the card. If you are unsure of the settings, just accept the defaults. In most cases, they work just fine. As you can see, the Plug and Play process is quick and easy. Sometimes, however, the card may not be detected properly, or you may just need to install manually. You'll examine that process next.

Manually Adding Network Cards  Although most network cards now support Plug and Play, some on the market still do not. If this is the case or if you just want to install the card yourself, you can add it manually. Be sure you have the following information if you are installing your network card manually:

This information will help you prevent resource conflicts before they happen.

The manual installation process can start from either the Add New Hardware interface or from the Network Properties dialog box (see Figure 23.5). Both are accessible from the Control Panel. Because the Network Properties dialog box allows you to alter all network settings from a unified interface, I cover that method here. The Add New Hardware Wizard is covered elsewhere in this book. If you choose to use it to manually add your network card, you should review that section of the book before proceeding. To install a card manually, follow this procedure:

FIG. 23.5 The Network Properties dialog box allows you to add components manually.

1. In the Network Properties dialog box, click the Add button. This action opens a selection window asking you what you want to add (see Figure 23.6).

FIG. 23.6 Select Adapter from the dialog box.

2. Choose Adapter and click Add. The next screen presents you with a list of all the manufacturers that provided drivers to Microsoft that are included on the CD.

3.
Choose the manufacturer of your network card on the left side of the screen. A list appears on the right showing every model of network card that is included with Windows 98 for this manufacturer (see Figure 23.7).

FIG. 23.7 This dialog box lists almost every network card manufacturer on the market.

4. Scroll through the list, select your card (as shown in Figure 23.8), and click OK.

FIG. 23.8 Select your card from the list of models on the right side of the window.


TIP: You can get to Network Properties more quickly. Just right-click the Network Neighborhood icon and choose Properties from the list.

After you click OK, you are prompted to insert your Windows 98 distribution CD. After the files are copied, you are returned to the Network Properties dialog box. You then need to reboot for the changes to take effect.


CAUTION: Because you are installing the adapter manually, Windows 98 assigns only Client for Microsoft Windows and TCP/IP to your card (see Figure 23.9). This may not be the logical configuration you need. Be sure you understand the concepts of network services and protocols before you attempt a manual installation. This process is covered in the next chapter.

FIG. 23.9 After you add a network card manually, the default network bindings are TCP/IP and Client for Microsoft Networks.

Configuring Your Network Interface Card

Sometimes network cards require additional configuration because they require several resources that may be in use by other hardware. The three most important network card settings are Memory Range, IO Address, and Interrupt Request. These settings must be unique to the network card and cannot be in use by any other equipment on the PC (see Figure 23.10). Depending on the equipment in your PC, you may have to make some changes to these settings. These changes can be made from the Device Manager.

You can access the Device Manager by selecting the System icon from the Control Panel. As described earlier, the Device Manager allows you to change resource settings for every piece of hardware in the PC. It also allows you to track down resource conflicts between devices.

FIG. 23.10 This figure shows a resource conflict between the network card and other devices.


TIP: You also can use a shortcut to get to the Device Manager. From the desktop, simply right-click My Computer and choose Properties from the list.

To change the resource settings for your adapter, go into the Device Manager, select your network card from the Network Adapters list, and select the Resources tab. Deselect the Use Automatic Settings check box. Now you can change any of the settings that are relevant to the card. For example, to change the IRQ, you double-click the IRQ listing and choose a new setting from the list (see Figure 23.11). On the Resources tab, the Conflicting Device List lets you know if the new resource is in use by another device. Keep changing the resource value until you find a setting that has no conflicts.

FIG. 23.11 To change the conflicting IRQ, just scroll through the list until you find a setting with no conflicts.

If you try to change a setting and receive a message stating that the desired resource cannot be changed, your network card is forced to use that setting. If you have a conflict, you need to change the resource setting on the conflicting device (see Figure 23.12).


NOTE: Some network cards install a new icon in the Control Panel to allow you to configure the card's resources. Although handy, these interfaces do not include the Conflicting Device List, which lets you know in real time if your new settings conflict with other devices.

FIG. 23.12 The Conflicting Device List shows no conflicts after the IRQ setting is changed.

Configuring the Identification Tab

The Identification tab in the Network Properties dialog box allows you to identify your computer to the network (see Figure 23.13). The three settings on this tab are as follow:

Configuring the Access Control Tab

The Access Control tab allows you to choose the type of security you want on your shares. The two types of security available in Windows 98 are both described here.


NOTE: The Access Control tab may not be available if you do not have File and Print Sharing installed. This topic is covered in the next chapter.

FIG. 23.13 The Identification tab controls how your PC is identified on the network.

Share-Level Security  Share-level security allows you to assign permissions on a per-share basis (see Figure 23.14). This means that each share you create on your machine has its own set of permissions. The disadvantage to this strategy is that you have to manage passwords for each share you create. If you create a large number of shares, keeping up with all the passwords can be difficult.

FIG. 23.14 Share-level security allows you to assign permissions to the share itself.

When share-level security is enabled, you have three security options for each share (see Figure 23.15), as listed here:

FIG. 23.15 Share-level security gives you two security levels: Full and Read Only.

User-Level Security  User-level security is much more robust than share-level security because you assign permissions to users instead of particular shares. This way, you don't have to keep up with passwords for each share. When you are setting up user-level security on a share, you choose the users who you want to access the share along with the access level. The disadvantage to this type of security is that you must have a Windows NT or a Netware Server on the network (see Figure 23.16) because Windows 98 does not contain a system for managing and tracking individual users across the network. To implement user-level security, Windows 98 uses a method called passthrough security. Windows NT and Netware servers are capable of managing thousands of users across the network. When a Windows 98 machine needs to validate a particular user, it queries the Windows NT or Netware server for the user's list of access permissions (see Figure 23.17). If the user has been granted the appropriate permissions, the Windows NT or Netware server sends the okay to the Windows 98 computer, and the user is given access to the share. Because Windows NT and Netware are expensive and are not required for small workgroup networks, this option is less popular than share-level security.

FIG. 23.16 When selecting user-level security, you are required to provide the name of the Windows NT server that validates users.

FIG. 23.17 User-level security allows you to assign share access by adding valid users to a list.

Installing Network Cables

Computer networks have seen a burst of evolution in just a few decades. This rapid growth has led to several types of networking strategies, equipment, and cables. In the following sections, I'll briefly review the most common cable types while pointing out the benefits and shortcomings of each type. I'll also explain the correct installation procedures for each type.

Unshielded Twisted Pair

Unshielded twisted pair (UTP) is the most common type of network cable in use today. It is fast, reliable, and relatively inexpensive. Category 5 UTP, which is used for networking, consists of four pairs of thin copper wire. Each pair is twisted together--hence, the name twisted pair. Twisting the pairs together helps to prevent a phenomenon known as crosstalk. Crosstalk occurs when the signal from one wire is picked up by a nearby wire. This creates a false signal that can cause problems on the network. The four pairs are encased in a plastic jacket. Because the wires are not shielded, they are more susceptible to electromagnetic noise; thus any application of this cable would need to contain some type of error checking. Category 5 UTP is used primarily in Ethernet networks. The connector used in this configuration is called an RJ-45 connector. It looks like a big phone plug, but it has eight leads (to connect the four twisted pairs) instead of the six leads present in the RJ-11 connectors used for telephones. To install the cable, all you need to do is plug one end of the connector into the port on the network card and the other end into the hub.

Shielded Twisted Pair

Shielded twisted pair (STP) is similar to UTP, but it consists of only two wires that are twisted together in a single pair. The pair is then wrapped in a shielding material before the outesjacket is applied. The shield is usually made of either a wire mesh screen or a metallic foil. This shield protects the twisted pair from electromagnetic interference, providing a cleaner, more reliable connection. The connector used with STP is called a D-shell connector. It is a large plastic connector that is shaped like the letter D. The most common use for STP is in token-ring networks. Because token ring is not generally used in workgroup situations, I will not cover it in much detail.

Coaxial Cable

Coaxial cable is constructed from a single conductive strand surrounded by insulation and a second conductive sheath. A plastic jacket is then wrapped around the cable. Coaxial cable can be used in a variety of applications, including cable television and Ethernet networks. Because Ethernet over coaxial cable is installed in a bus topology, a central connection point like a hub or a switch is not needed. The connector used with coaxial cable is called a BNC connector. Unlike cable television connectors that either push or screw into place, the BNC connector has a quarter-turn locking system, which holds it to the device it is connected to. When coax is connected to a PC, a component called a T-adapter is plugged into the PC. The coax is then plugged into one side of the T-adapter. The next PC in the chain is connected to the other side of the T-adapter. This process is used to chain the PCs together to form the network. The PC at each end of the chain is then affixed with a terminator that reflects the signal back onto the network when it reaches the end of the bus. The disadvantage of this type of network is, if a machine is unplugged from the network, the network is effectively cut in two. Because the two network halves are both unterminated at one end, network reliability drops to nil.

Installing a Workgroup Hub

Hubs are centralized connection points for networks with a star topology. The main purpose of a hub is to redirect all network traffic to each machine on the network. In this section, I'll discuss the purpose of hubs and walk you though the installation process.

Hub Description

A hub is a small box with RJ-45 ports in it. Hubs can contain anywhere from 5 to 24 ports with a link light for each port. All hubs are designed differently, but usually the link light is either green if a connection is made or off if a problem occurs with the connection. Check your documentation for your hub's link light configuration. When you plug your PC into the network, you physically plug it into a hub. Because hubs are passive and are not directly attached to the PC, you don't need to install a driver. In fact, Windows 98 is not even aware of the hub. The network card handles all communication with the hub.

To truly understand the purpose of a hub, you need to be aware of the network topology behind it. The network in this case is laid out in a star topology. Every machine on the network is plugged into a central point, or hub. The benefit of this strategy is that you can unplug a machine from the network without interrupting service to any of the other machines. Another benefit is that you can monitor the network load because all traffic flows through a central location.

Installing the Hub

Hub installation is quick and easy. Usually, the hub is a small plastic box with a power supply and several RJ-45 ports. When placing the hub, you should find a location that allows you to plug in each PC without creating a big mess of cables. After you select a location, all you need to do is plug it into a power source and attach the network cables from the PCs. Because hubs are passive devices, they need no further configuration.

Troubleshooting Network Hardware

Setting up a network requires so many steps that you are bound to forget something along the way. Here's a list of simple things to check that will solve 50 percent of your network problems:

You might not get a successful link for several reasons. They could range from malfunctioning equipment to using the wrong cables. Use the following steps to troubleshoot your connection.

1. The first step is to check your cables. Cables are much more susceptible to malfunctions than network cards and hubs. The easiest way to test a cable is to try to establish a link with a known good cable. Also, you can try the suspect cable on another machine that has an established connection. Make sure that you are using the correct kind of cable for your network configuration. Several types of devices that use RJ-45 connectors may look like network cables. One way to check your cable type is to look at the pin-outs. UTP has four pairs of wires, for a total of eight wires. Standard Ethernet cable uses a straight-through pin-out. This means that if you look at the RJ-45 connectors side by side, the wire colors match from left to right. For example, if one end of the cable has a pin-out of red-orange-yellow-green-blue-white-brown-black from left to right, the other end of the cable should have the same configuration. Also, make sure that you are using network-grade cable. Standard Ethernet cable uses Category 5 cabling. Sometimes you can find Category 3 cabling with RJ-45 connectors in the correct pin-out configuration. This cable is rated only for telephone applications and is not reliable enough for data transmission.

2.
If the cable is working, you need to check your network card. If Windows 98 reports that the card is working with no conflicts, you should try to install the card in another machine. If you have an identical card handy, replace the suspect card with it. Most network cards come with diagnostic disks. The easiest way to check a network card is to run the manufacturer's diagnostics. If the diagnostic program reports that the card is not installed in the system, you probably have a bad card.

3.
You need to check the hub. One of the ports in the hub may be malfunctioning. This malfunction causes the problem to appear to come from the card because the other machines in the network work fine. The easiest way to check for a bad port in a hub is to plug a working machine into it. Also, be sure that the port is not set up as a crossover port. Some hubs have a port with the cable pin-outs reversed so that you can attach a second hub. If you plug a computer into a crossover port, it does not work. If the port is not a crossover port, and you still cannot get a link, the port is probably bad.

4.
Be sure that your network cards are all set to the same speed. Many cards are now supporting 100Mbps as well as 10Mbps.


Previous chapterNext chapterContents


© Copyright, Macmillan Computer Publishing. All rights reserved.