Table of Contents
This laboratory is a practical and must be completed within the time allocated. See the class schedule for the specific due date and time.
This practical comprises two elements. First, you will learn how to build a simple ethernet cable by installing an RJ-45 plug. Then, you will take a virtual tour of a structured cabling plant, and record details particular to a certain machine, as if you’d just installed it into the network.
You will take a little time to learn about cabling tools and concepts such as cable types and wiring standards. If you are already familiar with them, you can skip over to Section 1.7, “Building the Cable” to follow the steps for building an Ethernet cable.
We also provide information regarding how to install an Ethernet Network Interface Card (NIC) in Section 2, “Install an Ethernet Network Interface Card (NIC)” and Ethernet autonegotiation and flow control in Section 3, “Autonegotiation and Flow Control”. Much of the time you will not need to know much about them, but occassionally they can be useful for your understanding and future career.
Finally, Section 4, “Structured Cabling” provides you information regarding the assessment of structured cabling after watching the video.
You must complete the two parts of the laboratory to attain the marks for this lab.
Your teacher will give a demonstration of the procedure for installing an RJ-45 plug onto a UTP cable.
You will need the following tools and materials, shown in Figure 2, “Ethernet Cabling Equipment”, which will be provided for you. For your reference, I have provided the approximate cost of them for you.
UTP Cable Stripper. (optional, ~ $15)
RJ45 plug. (~ $0.30 each in bulk)
Serrated Snips. (optional)
Cable Crimper. (~ $50, buy quality)
Ethernet Cable Tester. Very Basic (~ $15+)
(Not shown) Cat5e or better UTP cable. (~ $1/m wholesale)
If you look at five different videos of how to do this procedure, you will find about five different methods that people use, but the basics are the same: strip the cable, sort the pairs, crimp the cable.
What method you use will depend on the tools you have available, and the amount of money you are willing to spend on your tools. Professional cable engineers have this as part of their ‘plant’, which means from a taxation point, they receive depreciation on their tools.
Professional grade tools can be expensive, but they are worth it if you are doing this regularly. This is particularly good advice when buying a crimper; buy one with a metal body, as a sloppy hip joint will give mostly bad crimps.
There are a few possibilities as far as Ethernet cables are concerned. The most common is Cat6, Cat5E, Cat5, or Cat3 in slightly older (10Base-T) networks. All of these are UTP. New structured cabling installations will have at least Cat5e.
In very old Ethernet networks, you also see Thin-net and Thick-net, examples will be shown in the lab. We will not be concerning ourselves with such antiquated hardware.
Nor will we be concerning ourselves with Optical Fibre, as it takes specialised tools and training to deal with, and you’re not likely to need it in a small network.
There are three types of cables you might see in Ethernet networks. The first is often called Silver Satin, and is a flat cable suitable only for 10Base-T networks. It’s not used in modern networks, as it’s more susceptible to noise, because the pairs are not twisted.
Cat3 and above cables come in twisted pair format. They are generally unshielded, although you can get Shielded Twisted Pair (STP) for environments where noise may be an issue, such as factory environments. They have 4 pairs of wires. Each pair is twisted, and the twists should be maintained as far as possible to the plug for Fast Ethernet and above. The number of twists/foot is what helps it to be less susceptible to interference, and is one of the things that is checked when a new installation is certified.
The core of each wire can be either solid or stranded. Solid core is used for permanent installations, as it has better signal characteristics. Stranded core is often used for patch cords, or anywhere where the cable may be moved a lot. Stranded cores are less susceptible to breaking when they are bent, and are quite whippy, whereas solid core cable tends to keep it’s shape when you bend it.
Stranded cable requires a slightly different RJ45 plug than solid core. The differences are very subtle, and affect how the pins on the plug bite into the cable.
Cables have certain performance specifications, and are graded by the baud (signalling) rate, and other signal characteristics, such as susceptibility to noise, attenuation, etc. For 10Base-T networks, Cat3 cable is used, it can carry 10Mbps. Cat4 cables can carry 16Mbps, and is used in older Token Ring. Cat5 is common today, and can be used with 100Base-Tx networks. Cat5e and Cat6 are higher grade cables that are suitable for Gigabit applications and for ‘future applications’. All of these cables have four pairs of cores. However, only Gigabit Ethernet uses all four pairs.
Naturally, you can use a higher grade cable than you need to and everything will still work fine, but the reverse is not true. If you were to use Cat3 in a 100Mbps application, you would get a lot of errors, and the network would be even slower.
There are two standards for wiring twisted pair ethernet cable: T568-A (Preferred) and T568-B (Optional). Both can be used together, but for optimal performance, the same should be used throughout the network. In fact, in Cat5 installations and above, the standards say not to mix the two.
We have listed the Preferred and Optional wiring standards in Figure 3, “TIE-EIA T568 Ethernet Wiring Standards”, as laid out by the TIE-EIA standards body. The optional scheme comes about from the prominence of AT&T wiring schemes, and is still very common.
Note the the words Preferred and Optional may not reflect reality on your site.
To connect from a hub or switch to a computer (non-like devices), you use a straight-through cable. To connect a like devices, such as hub/switch to a hub/switch (without using an uplink port on the hub or switch), or a computer to another computer, you would use a crossover cable. Note that a router interface would be wired the same as the computer.
If you don’t get a link light when you plug in the cable, you’re probably using the wrong cable for the port you’re plugging into. Crossover cables are sometimes marked with an X on the ends to note that it is a crossover cable. Get used to looking at the both ends of the cables to judge reliably if you have a straight-through or crossover cable. A lot of switches and laptops being sold on the consumer market today are autosensing, which means it will detect if you’re using a straight-through or crossover cable.
A crossover cable for 10Base-T and 100Base-T (i.e., Fast Ethernet) is constructed simply by wiring one end in Preferred, and the other in Optional, as shown in Figure 3, “TIE-EIA T568 Ethernet Wiring Standards”. However, for Gigabit Ethernet and 10Gigabit Ethernet, you have to follow the standards to make crossover cables. Fortunately, such network cards generally have have an autosensing feature to detect which type of cable is plugged in and whether a crossover needs to be added internally or not. Therefore, there is no particular need to use crossover cables for Gigabit and 10Gigabit Ethernet.
Straight-through (normal) cables have both ends the same, either both Preferred, or both Optional. Short cables that are wired this way are generally referred to as patch cables, as they are used for wiring patch panels and for the computer-WAO (Work Area Outlet) link.
The demonstator will show you in detail the steps for putting a RJ-45 plug onto the end of the cable. Here are the steps in brief so you can prepare.
In case you want to use a protective boot for the cable, thread this onto the cable first, ensuring you put it on the right way. Boots are useful, because they prevent the snap on the plug from breaking off when the cable is pulled backwards, often from a tangle of other cords. However, we won’t be using boots in this lab.
Either use a suitable cable stripper to remove the cladding. You don’t need to cut more than 1 inch of cladding away. The tool should ideally cut almost through the cladding. There is often an adjustment mechanism (usually a screw) for tuning this,
or use a sharp knife, such as a craft knife or Stanley knife. Score a ring around the cable, about 25mm (1 inch) from the end. If done well, the cable casing should snap off when pulled or twisted at the score line, leaving a tell-tail light-blue ring around the inner-wall of the cladding where the blade did not penetrate. This provides a handy guarantee that you didn’t cut into the actual cable pairs.
Cut off the ripping cord (a thin string), if present. This is present so that cable installers can open up a large section of the cable if needed.
Untwist and sort out the pairs and cores, according to whatever wiring scheme you are creating on the cable. Do not untwist further than the edge of the casing; in fact, try to stop untwisting a little bit before it. Practice will tell you when to stop. See Figure 3, “TIE-EIA T568 Ethernet Wiring Standards” for the wiring sequence. Looking at the natural lay of the pairs will help to create a tidier cable. Remember, if it is a normal (straight-through) cable, both ends must be the same. For crossover they must be different.
Straighten and flatten the cores so they lie side by side, and parallel to each other. Massaging them between finger and thumb while swaying from side to side (the cores, not you) is a technique I find effective. You want to get them nice and straight close to the bottom also. Rotating the cable cladding to and fro while holding the cores between finger and thumb can be useful here.
Using your cable stripping tool, or serrated snips, cut off the cores 12mm from the edge of the cladding, being careful that the cut is square. If its not square, some of the cores may pull free during its working life. I don’t recommend using diagonal pliers (sidecutters), because it’s harder to make a level cut. Cable crimpers, and also strippers often have a part of the tool for this, which can be useful for getting the correct length.
Another thing you can do here, if you cut them about 1mm or so longer first, is to insert the cores into the plug, so that they straighten, then take the plug off. Trim the cable to the correct length and straight across.
Holding the cable so the green (or orange depending on the wiring scheme) core is on the left, place the plug into the cable, with the snap of the plug away from you. Push the plug in, gently but firmly if required. Make sure the cores go all the way into the plug. You should be able to see the copper cores when you look at the end of plug.
You want to push the cladding up into the cable. When the plug is crimped, part of it will pinch the cladding, making sure the cladding takes the strain and not the cores.
Double-check the pinning. This is your last chance.
Place the cable and plug into the crimper, making sure it is in the correct way (there should be a notch for the snap to fit into). Squeeze firmly on the crimper. With some plugs you may hear a snapping noise. Don’t be alarmed.
Take out the cable, and check that all of the brass pins at the top are well pushed in. If the plug fails to go into a socket, or is a tight fit, it’s likely that it needs recrimping.
Once the cable is complete, put both ends in a tester and ensure that the wiring is correct.
If you have the type of tester that has all lights on at the same time, firmly hold the plugs into the sockets, and give each end of the cable a wiggle. Look for any flickering of the lights, indicating a poor crimp which will soon fail.
If a plug is found to be faulty, cut it off and start again.
Once you have completed a cable, use it to assemble a very small network comprising a couple of switches and stations; there should be a physical-layer network map shown in the classroom, which you will need to replicate using the hardware provided.
 Although you can get flat cable that has twisted pairs, such as Flat Cat 6, which is useful for laptops.