After you have all your cable runs in place, you're ready to make the connec-tions so that the computers can communicate with each other through the hubs.Depending on the type of cable installation you've performed—internal or external—the connection process can be extremely simple or quite complex. In some cases, you must be familiar with the function of each wire inside the UTP cable, whereas in others you never have to see the wires at all.
The simplest possible LAN consists of two computers, with network interface adapters installed, connected by a single cable. If the two computers are located in the same room, the cable installation should be very simple. However, if the computers are far away from each other, and especially if they're located in different rooms or on different floors, the cable installation might require special attention.
Back in the days when an Ethernet network meant coaxial cable, it was possible to connect the NICs in two computers with a Thin Ethernet cable to set up a simple network. Today, however, the standard for Ethernet networking is UTP cable, and this generally requires the use of a hub.
In Lesson 3: Network Hubs, in Chapter 2, "Network Hardware," you learned how the hub on an Ethernet network provides a vital service by crossing over the signals between the transmit and receive wires. This enables the signals sent over the transmit wires by each computer to arrive at the receive connections at the other computers. When you connect two Ethernet network interface adaptersdirectly using a UTP cable, there is no hub and this crossover is absent. For these two computers to be able to communicate, you must supply a special cable called a crossover cable, which wires the transmit contacts in each connector to thereceive contacts in the other connector.
One limitation of a UTP Ethernet network without a hub is that the two computers can be no more than 100 meters apart. On a standard UTP network, the Ethernet hub functions as a repeater, which enables each cable connecting a computer to the hub to be 100 meters long, for a total span between computers of 200 meters when separated by a single hub.
If you are connecting two computers in the same room, you can purchase a prefabricated crossover cable and simply plug the ends into the network interface adapters in the two computers. Be aware, however, that you might have trouble finding a crossover cable in your local computer store. Virtually all computer stores these days stock basic networking equipment, such as NICs, hubs, and prefabricated UTP cables. Larger stores might have crossover cables as well, but you might find it easier to order one from an online or catalog dealer, particularly if you need a relatively long one.
If you want to connect two computers in different rooms or on different floors usinga crossover cable, you might have to perform an internal installation by running cable through the building's walls, ceilings, or floors. If this is the case, the cable that you use for a crossover connection is the same as that for a hub-based network, and the procedures for pulling the cable are the same as those detailed in Lesson 1: Pulling Cable. The difference between a crossover installation and a standard installation is in the attachment of the wires to the connectors at each end of the cable.
As explained in Chapter 2, "Network Hardware," a UTP cable contains eight separate wires, which are joined together in four twisted pairs. The RJ-45 connector at each end of the cable (whether it is male, as on a patch cable, or female, as part of a wall plate or patch panel) has eight conductive contacts, to which the eight wires areattached. When you plug a male connector into a female one, the corresponding contacts touch, creating electrical circuits. Figure 15.14 shows the functions of the eight contacts on a standard 10Base-T or 100Base-TX Ethernet network.
Figure 15.14 RJ-45 connector contact assignments for 10Base-T and 100Base-TX networks
Whereas 10Base-T and 100Base-TX networks use only four of the eight wires in a UTP cable, a 100Base-T4 network uses all eight. The four that aredesignated as unused in the figure can carry signals in either direction on a100Base-T4 network.
Standard network cable runs and prefabricated cables use straight-through connections. In a straight-through connection, each wire is attached to the same contact in both connectors, as shown in Figure 15.15. The transmit contacts at one end are connected to the transmit contacts at the other end, and the receive contacts are connected in the same way. This is possible because the crossover circuit is supplied in the hub, which makes the job much easier for the cable installer.
Figure 15.15 Straight-through connections use the eight wires in a UTP cable to connect the corresponding contacts in the connectors at each end
To create a crossover connection in the cable, you must connect the two transmit contacts to their corresponding receive contacts, as shown in Figure 15.16. The positive transmit data (TD+) contact at each end is connected to the positivereceive data (RD+) contact at the other end. Likewise, the two negative transmit data (TD–) contacts are connected to the two negative receive data (RD–) contacts. When you install a cable using a crossover connection like this, you cannot use the cable run with a hub, because the crossover circuit in the hub would cancel out the crossover circuit in the cable. In other words, the TD+ contact that is crossed to the RD+ contact in the cable would be crossed again, back to the TD+ contact, inside the hub. The only way you could use this connection with a hub (in the event that you had to expand the network, for example) would be to plug the cable into the hub's uplink port, which does not run through a crossover circuit.
Figure 15.16 A crossover cable connection attaches the TD contacts in one connector to the RD contacts in the other, eliminating the need for a hub
If you've installed prefabricated cables externally, making your final connections is simply a matter of plugging them in to the hub and the computers. Set up the hub in a central location, preferably where it is protected from traffic or vibrations that can pull on or loosen the cable connections, and connect it to a power source. Plug the connector for each of your cables into one of the hub's ports. Push it firmly into the socket until it clicks. Do not use the hub's uplink port for a computer connection unless the port has a switch that allows the crossover circuit to be disabled. Most hubs have light-emitting diodes (LEDs) that correspond to the ports; these will not be lit until you connect the other end of the cables to the computers and turn them on.
At the other end of each cable, you should have a computer that is set up and ready to go. Shut the computer down and plug the network cable into the jack provided by the computer's network interface adapter. Again, make sure that it clicks into place. If the jack does not fit in the socket, you're probably trying to plug the cable into a modem jack, which won't work.
Most Ethernet NICs have an LED next to the RJ-45 connector; most networkinterface adapters built into the motherboard do not show this LED. The LED lights up when the NIC is connected to an operating hub. When you turn on the computer, the NIC generates a signal called a link pulse and transmits it over the cable. When the hub receives the signal, it responds with a signal of its own. If either the NIC or the hub is a Fast Ethernet device, the devices use these link pulse signals to negotiate the fastest speed they have in common. For example, when you plug a dual-speed NIC into a Fast Ethernet hub, the link pulse signals enable the two devices to determine that they are both capable of operating at 100 Mbps, and they configure themselves to use that speed.
If you connect a dual-speed NIC to a standard Ethernet hub, the NIC determines that it must run at 10 Mbps to use the hub, and it adjusts itself accordingly. When this negotiation is complete, the LEDs on both the hub and the NIC should light up, even if you haven't yet installed the network interface adapter driver on the computer. Some dual-speed NICs have two LEDs, one of which specifies the speed at which the card is operating. If the LEDs don't light up, there might bea problem with your cable connection, or possibly with the NIC or hub. SeeChapter 17, "Network Troubleshooting Procedures," for more information about what to do next.
Assuming that the LEDs on both the NIC and the hub do light, your hardware installation is complete. If you haven't done so already, you must install the networking software components on your computers, after which your network should be operational.
If you have installed bulk cable internally, the process of making your finalconnections is more complicated. The essential steps for making each cableconnection are as follows:
When you install bulk cable, you must purchase the connectors you need and the tools for attaching the connectors separately. Most internal installations use wall plates for the computer end of each cable run and one or more patch panels for the hub end. A wall plate is a metal or plastic face plate that screws into a hole in a wall, much like an electrical outlet, except that the wall plate contains female RJ-45 connectors (jacks) instead of electrical outlets. A connector on the back of the wall plate jack contains the contacts to which you attach the wires inside the UTP cable. For each of your cable runs, you must connect the eight wires to a jack at each end of the cable. When the cable is connected and the wall plate installed, the cable is hidden in the wall, and the only part visible is the front of the wall plate. You can then plug a patch cable into the jack, just as you would a telephone cable.
As shown in Figure 15.17, some wall plates have integrated jacks, whereas others are modular. You can buy wall plates that hold one, two, four, or more jacks, and you can insert different types of jacks to support various cable connections. For example, in new construction, it's possible to install telephone and data network cables simultaneously and to use a single wall plate as the terminus for both networks. If you do this, be sure to label the jacks carefully so that users don't confuse them.
Figure 15.17 Wall plates and jacks
Although it's perfectly acceptable to use one wall plate for both telephone and data network connections, these two services must use separate cables. Some people assume that because two wire pairs are left unused in the typical UTP Ethernet network, it's safe to use those wires for voice traffic. This is most assuredly not the case, as the voice signals can cause crosstalk that interferes with the data signals on the other wires.
A patch panel, sometimes called a punchdown block, is similar in function to a wall plate, except that it supports many more ports. A patch panel is essentially a face plate or box with a number of RJ-45 jacks mounted in it. It provides a row of ports on its front, as shown in Figure 15.18. A patch panel is not a hub; it is nothing more than a nexus that is a convenient place to terminate the hub end of all your cable runs. You plug patch cables into the patch panel's ports to connect them to hub ports, thus completing the connection at that end. Patch panels are available in a variety of sizes and configurations, and are either mounted on a wall or integrated into a rack-mounted system.
Figure 15.18 A patch panel
Make sure that all the jacks you use on your network conform to the same rating as your cables. If you are installing Category 5 cable, you must use jacks that are rated for Category 5 as well.
The process of connecting the ends of your bulk cable runs to the jacks in your wall plates and patch panels is called punching down. Each jack contains eight sets of contacts that correspond to the eight wires in the cable. Punching down a cable consists of the following steps:
Remember that you must repeat this process at both ends for each of your internal cable runs. This can be a lot of work but, fortunately, there are tools that simplify the process. A punchdown block tool, shown in Figure 15.19, is a handheld device that you use to insert each wire between its set of contacts. The tool strips the insulation off the wire, presses it into place between the contacts, and cuts off the excess wire. This is an essential tool for an internal UTP cable installation. Without it, the process of stripping, installing, and cutting each wire is very laborious.
Figure 15.19 A punchdown block tool
The punchdown block tool you purchase must be the same type as your jacks. The types usually refer to the configuration of the blade that cuts off the wire ends. The jacks (or blocks) most often used today are called 110-style. You can purchase a tool designed specifically for this type of block, or a modular one with interchangeable parts that supports multiple block types.
The most important part of the punchdown process is matching up the wires with the correct contacts. The wires inside the UTP cable are color-coded orange, green, blue, and brown. The positive wire in each pair is solid-colored, and the negative wire has a white stripe. You can buy jacks that have corresponding colors on the contacts, so that you simply have to match up the wires with the same-colored contacts when punching down.
There are two standards used today that specify which color wire in a UTP connection should be associated with each contact on the connector. These standards are called 568A and 568B. They are actually the configurations found in two versions of the Electronics Industry Association/Telecommunications Industry Association (EIA/TIA) 568 "Commercial Building Telecommunication Cabling Standard" document. The colors and the corresponding contacts for each of the standards are shown in the following illustration.
Which standard you choose when wiring your connectors is not important. The two are functionally identical, and you could even make up your own color code if you want to. What is important (indeed, crucial) is that you select a wiring configuration and use it consistently throughout the entire cabling project. A cable wired per the 568A standard on one end and the 568B standard on the other will not function. If there are several people working together to install the cable, make sure that all are using the same wiring standard. You need not be concerned about the wiring standard used to construct prefabricated cables. As long as each contact in one connector is wired to the corresponding contact in the other connector, the cable will function properly.
To punch down a cable, begin by stripping about 2 inches of sheathing off the end, and then untwist each of the four wire pairs. You then lay down the cable in the center of the jack and spread out the wires so that they lay out between the appropriate sets of contacts, as shown in Figure 15.20. To protect the wires, the beginning of the cable sheath should be no more than one-eighth of an inch from the jack. You should also be careful to untwist each wire pair only as much as is necessary for the wire to fit between the contacts. The wire pairs are not twisted simply for organizational purposes. The twists provide an essential function by preventing the signals on the various wire pairs from interfering with each other. Each pair uses a different number of twists per foot, and you want to preserve this configuration as much as possible.
Figure 15.20 Lay out the wires between the appropriate contacts and use the tool to punch them down
When you have the wires laid out on their respective contacts, take the punchdown tool and place it over the first set of contacts with the blade on the outside of the jack and the handle of the tool tilted slightly outward. Press down firmly on the tool. This presses the wire into place, stripping off the insulation as it goes, and cutting off the loose wire end. Repeat this process for the remaining seven wires and be sure to remove the wire ends that are cut off. This process takes a bit of getting used to, so it's a good idea to buy some extra jacks for practice before you start working with your actual cables. This is also another good reason to allow some extra slack in your cable runs. If you make a mistake, you can simply cut off the end of the cable and start again with a new jack.
After you have punched down all eight wires, you can insert the jack into the wall plate or patch panel (if necessary). You can then mount the wall plate into the hole you cut previously, pushing all of the excess cable inside the wall. Mount the patch panel on the wall or rack after you've punched down all of your cables.
A patch cable is simply a shorter length of cable with standard male RJ-45 connectors on both ends that you use to connect a wall plate to a computer's network interface adapter or to connect a patch panel port to a hub port. You can purchase prefabricated cables for this purpose, or you can build them yourself. Making the final connections is no different than the process for an external cable installation, as described earlier in this chapter. When you have an unbroken connection between a NIC and a hub, and both devices are switched on, the link pulse LEDs at both ends should light up, indicating that communication is possible. If the LEDs don't light up, the troubleshooting process is a bit more involved than that for an external cable installation because there are more components to check for problems. See Chapter 17, "Network Troubleshooting Procedures," for more information.
Although wall plates and patch panels make for the neatest installation, you don't have to go this route if you don't want to. You can also attach male RJ-45 connectors to the ends of your cables and plug them directly into your hubs and computers, just as you would with prefabricated cables. You can also attach these connectors to shorter lengths of cable if you want to build your own patch cables.
Male RJ-45 connectors for UTP come in three configurations; ensure that your RJ-45 connectors are compatible with the selected cable, as follows:
Silver satin cables are designed for telephone network connections and should not be used for data networking.
Attaching male RJ-45 connectors to UTP cable requires another special tool, called a crimper, which is shown in Figure 15.21. A crimper is a jawed device that looks like a pair of pliers. It has a set of dies in it that enables you to squeeze the two halves of an RJ-45 connector together with the wires inside. As with the punchdown process, you strip some of the sheath off a cable and lay the wires out in the bottom half of the connector, making sure you use the same wiring standard at both ends. You then lay the other half of the connector on top of the wires and squeeze the handles of the crimper to lock the two halves together. This process is trickier than using a punchdown tool, because you have to get all eight wires in place at the same time. Some practice is necessary to get the hang of it. When you consider the price of the crimper and the dies (about $50), plus the bulk cable and the connectors you'll ruin while learning how to crimp, not to mention your valuable time, buying prefabricated patch cables might be a more economical alternative.
Figure 15.21 A crimper
Testing is an essential part of every cable installation. You test your cable runs by simply connecting up your computers and hubs to see if they work, but professional cable installers use a special cable-testing device to check for problematic conditions that might not be immediately apparent in a real-world test. For more information about testing cable runs, see Lesson 3: Network Testing and Monitoring Tools, in Chapter 18, "Network Troubleshooting Tools."
Fiber optic cables are different from copper cables in most ways, and attaching connectors is no exception. Unlike the connectors used on copper cables, which completely contain the end of the cable and provide their own conductors, the connectors used on fiber optic cables, called straight tip (ST) and subscriber connectors (SC), are really just sleeves that fit around the end of the cable and let the central core protrude out the end. The only function of the connector is to lock the signal-carrying core in place when it's plugged into the jack.
The process of attaching a connector to a multimode fiber optic cable basically consists of stripping off the outer sheath from the end of the cable, gluing the connector in place using an epoxy adhesive, allowing the adhesive to cure, and then polishing the protruding core so that the pulses of light carried by the cable reach their terminus in the best possible condition. Single mode cables are terminated by permanently splicing a "pigtail" to them, which is a short length of cable with a connector already attached to it. This is necessary because the tolerances for singlemode fiber are much tighter than those of multimode.
Professional fiber optic cable installers typically use a tiny electric oven to cure the epoxy; otherwise, the adhesive must be left to cure overnight. There are several products on the market that are designed to speed up or simplify this process, such as quick-setting adhesives and connectors that crimp on with no adhesive at all. Some professionals swear by these, but others prefer to stick with the traditional method.
Match the tools in the left column with the proper functions in the right column.
| || |