All About Crimping

MAKING CAT 3, 4, 5, 5E TWISTED-PAIR NETWORK CABLES Last updated:

INTRODUCTION. This article explains how to make cables used to network two or more computers together to form local area networks (LANs) for home or office. Patch cables for more complex networks are also covered. There are three common unshielded twisted-pair (UTP) network standards in current use. These are 10 Mhz 10BASE-T, 100Mhz 100BASE-TX, and 1000BASETX (1 GHz) Ethernet. Although the 100BASE-TX standard is presently the most common LAN standard for new installations, there are still many installations that rely on both 10Base2 (Coax) and 10BaseT (Twisted Pair-10 MHz). For new wiring, we suggest that CAT 5E wiring be installed. In general, the cost of the cable is only a small part of the cost of wiring for Ethernet and having 350 MHz capable CAT 5E wiring in place allows everything from phone lines to 10BaseT, 100TX or 1000BaseTX Ethernet. This article covers 10BaseT and 100TX wiring cable standards. CAT 3, 4, 5 AND 5E EXPLAINED The "CAT" is short for Category. The international standards committee rates the wiring of interest into three major catagories; CAT 3, CAT 4, CAT 5 and CAT5E. CAT 3 is the common phone wiring found in many homes and most businesses. It is capable of operating as LAN wiring for 10BaseT connectivity (10 MHz) but will not serve as wiring for 100 MHz 100 TX or 1000BaseTX Ethernet. CAT 4 is wire that was used for regular 10BaseT before 100TX became available. CAT 4, like CAT 3, is incapable of 100 TX operation at full 100 TX speeds. These Categories differ primarily in the number of twists per foot of length, the specification and quality of the wires and their coverings. CAT 3 is suitable for phone service and 10BaseT Ethernet, CAT 5 is suitable for all the above plus 100BASE-TX and CAT 5E serves all of the above plus connectivity at up to 350 MHz using 1000BaseT. PHONE WIRING Phone wiring uses the same 8 wires as Ethernet wiring. However, when used for phone connectivity, the center two wires are used for line 1. Depending upon which wiring standard is used, the next two wires adjacent to the center two form Line 2 or Line 3 and the pairs on either side are used for the 3rd and 4th or 2nd and 4th phone connections. Each of the lines is carried by one pair of wires, twisted together in a helix pattern. If one has a four line phone and its cord is plugged into the 8 conductor RJ45 outlet with four phone lines, that phone can access all four lines by pressing the appropriate line button 1-4.

1

Tip and Ring Or T & R are Telco Designations For The Two Wires Used For A Single Phone Connection HYBRID PHONE/LAN WIRING Many builders of modern homes advertise that they pre-wired the home with CAT 5 wiring and homeowners often attempt to use this pre-wiring to run their computer networks, only to run into problems. The reason is that the CAT 5 pre-wired builder's installation usually consists of running a single CAT 5 (8 conductor) cable from the central wiring closet to several outlets in a room or group of rooms in daisy chain fashion. There will be a number of runs from the wiring closet to various areas of the home but these are generally daisy chained to multiple outlets in a room or even groups of rooms. Connecting several phones on a single daisy-chained wiring run will work just fine... if two or more people pick up the same line together, they're simply on a party line and everyone can talk and listen on the single connection. Unfortunately, the same is not true for LAN wiring. Each client computer system must have its own set of transmit and receive pairs. If

2

the Local Area Network (LAN) cards of two computers are bussed together (wired in parallel), neither will work. When the homeowner with a pre-wired CAT 5 home attempts to set up their home network using these cables, they discover that they can have one Ethernet connection on a wiring run or up to four different phone lines. But, in order to use the one cable for both phone and Ethernet service, one must assure that the phone and Ethernet connections are not shorted together. If only a single line phone and two wires are used for the phone service, the other wires can serve the Ethernet application. Be sure to construct the patch cable from the wall to the client computer so that it skips the center two connections. The reason is that one cannot be sure of the termination the Network Card, Hub or Switch will make for these two wires. If they are shorted to ground, goodbye phone service. People often think that the center two wires are required for full duplex 100 TX operation but this is untrue, Both 10BaseT and 100 TX require only four conductions for full operation and both standards skip the center pair of wires (the phone's line number 1). One must also assure that the phone connections for shared Phone line 1 and Ethernet use of the CAT 5 cable uses only the center wire pair. Connecting the other pairs to the phone can result in poor or no Ethernet connectivity. It's best to construct a phone connection cable using only the two center wires. LANS SIMPLIFIED. The simplest LAN consists of two computers, each having a network interface card (NIC), or network adapter and running network software. These NIC cards must be connected together so that the transmit pair from one NIC card goes to the receive pair of the other. When only two computers are networked, a special cable can be used for this application that has the transmit and receive pairs "crossed over" and is referred to as a crossover cable. When more than two computers must be connected, the crossover trick won't work and devices known as a Hubs and/or Switches must be used. In this situation, each of the computers is plugged into a port on the Hub or Switch using a straight-thru cable. The hub or switch has its ports (female RJ45 jacks) wired so that its Transmit lines will go to the Receive connections of the RJ45 jack on a NIC card using a straight through cable. Switches and Hubs often have one RJ45 wired with

3

the same transmit/receive pair pattern as a NIC card. This is used to "Uplink" one HUB or SWITCH to another HUB or SWITCH. Sometimes a switch is provided, marked "Uplink", that swaps the Transmit and Receive lines for one of the RJ45 Jacks so that it can either serve as another client LAN card connection or to uplink to another Hub or Switch. Another common practice is to put in two RJ45 receptacles for one of the ports, with one wired for conventional connection to a NIC card and the other wired with the transmit/receive lines wired to uplink to another hub or switch. This 9th port is usually labeled "uplink." If the hub or switch has nine ports and one is marked "uplink", don't think that there are nine ports. There are only 8 and if the uplink port is used, the normal port 8 cannot be used. [ Top | Contents | Next - Cable and Connectors ]

NETWORK CABLE AND CONNECTORS Last updated:

There are several classifications of cable used for twisted-pair networks. Don't try to save a few cents by using a low grade cable. Install only Category 5E ( CAT 5E) cable for all new installations. There are several fire code classifications for the outer insulation of CAT 3, 4, 5 and 5E cable. The cost varies based on the quality of manufacturer as well as the insulation and sheathing used. It is not unusual to find a price differential of three hundred percent (300%) between the lowest grade and the highest grade insulations. Most wiring jobs can use the less expensive CMR cable ( "riser" cable). However, if the wiring is in an office complex and runs through suspended ceilings, Local, State or National codes may require the use of more expensive CMP or so-called plenum cable. Plenum is the name of the ducts used to distribute air in a building. Local, state or national codes, or the landlord's lease, may require the use of the more expensive plenum-jacketed cable so if the wiring is in a business environment, be sure to check the wiring code and the landlord's rules. If in doubt, use plenum. CMR cable is generally acceptable for all applications not requiring plenum cable.

4

CAT 3, 5 and 5E wire is available in reel-in-a-box packaging and on spools. Reel-in-a-box is very handy for pulling the wire without putting twists in it. For spooled cable, a cable reel stand or special reel handling tools are required. With spooled wire, pulling wire is usually a two-person job. Before reel-in-box packaging became available, broom handles were sometimes used as an axel to hold wire reels. One person would hold the broom handle that ran through the center of the wiring spool and the other would pull and measure the cable. Otherwise, the wire usually ended up in a tangled mess. Stranded wire patch cables are generally specified for cable segments running from a wall jack to a PC and for patch panels. They are more flexible than solid core wire and are well suited for patch cables. The rational for using stranded cable is that the constant flexing of patch cables may wear-out solid core cable and break it. This shouldn't be a real concern in the average small network. A downside to stranded cable is its susceptibility to moisture degradation. in general, stranded cables should not be used for runs of more than about 10 or 20 feet. Solid core cable is quite suitable for connecting computers directly to other computers or hubs. They serve this purpose for many home and small business networks. They are also quite acceptable for use as patch cables. For notebook computers that are constantly moving around, consider stranded wire patch cables. CAT 3, 4, 5 and 5E cables have four twisted pairs of wire for a total of eight individually insulated wires. Each pair is color coded with one wire having a solid color (blue, orange, green, or brown) twisted around a second wire with a white background and a stripe of the same color. The solid colors may have a white stripe in some cables. Cable colors are commonly described using the background color followed by the color of the stripe; e.g., white-orange is a cable with a white background and an orange stripe. CONNECTORS. The straight through and cross-over patch cables discussed in this article are terminated with CAT 3, 4, 5 or 5E RJ-45 modular plugs (RJ means "Registered Jack"). RJ-45 plugs are similar to those you'll see on the end of telephone cables except they have eight versus four or six contacts on the end of the phone cords, so they are about twice as wide as 4 wire phone connectors. Make sure the RJ45 connectors are rated for the class of wiring being installed. CAT5E connectors will work fine for CAT 5, 4 and 3 applications but a lower rated connector will not work for higher rated wiring. Also, there are RJ-45 plugs designed to handle either solid core wire and stranded wire. Others are designed specifically for one kind of wire or the other. Be sure to buy connectors appropriate for the wire being used.

5

[ Top | Contents | Previous | Next - Tools ]

NETWORK CABLE TOOLS Last updated:

Modular Plug Crimp Tool. Crimping requires a modular crimp tool. This one works for all kinds of telephone cable work and it works well for Ethernet cables. Buy a quality tool, not one of the cheap plastic and tin models. The bad taste of unreliable connections will soon replace the initial sweet smell of a low price. You don't need a lot of bells and whistles, just a tool which will securely and precisely crimp RJ-45 connectors. The tool should have precision machined metal dies that crimp the RJ-45 connectors. Better tools will also have a ratchet assembly built in that prevents the crimping action from reversing itself. The ratchet requires that the full crimp motion be executed before the handles can be returned to the open position. The tools we stock are of the highest quality, yet are reasonably priced. We believe they are the best tools made for the money. Even though the crimper has cutters that can be used to cut the cable and individual wires, and possibly even stripping the outer jacket, separate stripping and cutting tools will do a better job.

Universal UTP Stripping Tool. This model works well and does an excellent job of cutting and trimming cables. For a professional result, one of these will serve you well.

6

UTP Cable Sheath Stripper. This is one of the neatest tools we've seen for cutting away the outer sheath of CAT 5 cable; without so much as nicking the insulation on the inner wire pairs. Just push the wire into the hole, spin it around once and off comes the cable sheath. This can also be done using the Universal UTP stripping tool, but this little gadget will do it quicker and easier.

Diagonal Cutters ("4 to 6"). If a technician could only take one tool to a job, this is the one that 99% would choose. With a little practice, this becomes a wire stripper as well as a wire cutter. By knowing how hard to squeeze, this tool will cleanly strip the insulation off any size wire from 12 gage up. It's also much easier to use diagonal cutters (sometimes called "diags" or "dikes") to cut the cable off at the reel and to fine tune the cable ends during assembly. The cable sheath can also be stripped using the diagonal cutter and a small knife (X-acto, utility, etc.). To remove the sheath, carefully slice the outer jacket down the wire sheath and then use the diagonal cutter to trim it off around the circumference.

Punch Tools. This tool is not used for making patch cables but is a must for professional, between room, wiring using RJ45 Jacks and Patch Panels. These devices have V-Grove cutting/holding jaws that are designed to cut into the insulation on CAT 3, 5 or 5E wires and make contact between the wire and the RJ45 Jack's connector pins, either on the Patch Panels or individual Jacks. The punch down tool spreads the V-Groove jaws apart and pushes the wire into the groove. Interchangeable punch blades provide a variety of functions such as trimming excess wire next to the V-Grooves. Our punch tools are of the highest quality and a large selection of interchangeable punch tool cutter/dies are available. [ Top | Contents | Previous | Next - A Little Theory ]

A LITTLE THEORY Last updated:

10BASE-T and 100BASE-TX Ethernets consist of two sets of transmission lines. Each transmission line set is a pair of twisted wires. One pair receives data signals and the other pair transmits data signals. A balanced

7

line driver or transmitter is at one end of one of these lines and a line receiver is at the other end. A (much) simplified schematic for one of these lines and its transmitter and receiver is shown below:

Data pulses travel down the transmission line but what began as a square wave of energy is attenuated as it travels along the transmission line. The line itself has resistance as well as capacitance and inductance. These cause the corners of the square wave to become rounded and attenuated. When the length of the cable exceeds design limits for a particular protocol and frequency, communication becomes erratic and unreliable. See http://www.einsite.net/ednmag/index.asp?layout=article&articleid=CA149133&industryid =2282 for a more complete discussion. Using quality CAT 5 Cabling, good wiring practices and no EMI problems, the maximum recommended transmission line length between nodes for 10BaseT and 100TX is 100 Meters (about 329 feet). For Gigabit wiring using 1000BaseT operating at 350MHz, the limitation is 82 feet.. The primary characteristic used for communication is is the voltage potential between the transmission pair and current flowing near the surface of the wires. This energy also resides in the magnetic field which surrounds the wires and the electric field between the wires. In other words, the data pulse forms an electromagnetic wave which is guided by, and travels along the wires. The characteristics of the wire are determined by the quality of the manufacturing process and the materials used, so there is little the installer can do to improve the wire itself once a particular brand of wire is selected. However, there are many things the installer can do to maximize the quality of transmission/ reception. A major concern is the transient magnetic fields that surround the wires and the magnetic fields generated externally by the other transmission lines in the cable, other network cables, electric motors, fluorescent lights, telephone and electric lines, lightning, etc. This is known as EMI (ElectroMagnetic Interference) noise. Strong external magnetic fields couple to

8

and induce their own pulses in a transmission line and can literally bury the Ethernet data pulses. This can and often does prevent reliable data reception at the other end. Twisted-pair Ethernet uses two primary methods to combat EMI noise: 1) The first is the use of balanced transmitters and receivers. A differential amplifier is used so the signal pulse actually consists of two simultaneous pulses going in opposite directions relative to ground; a negative pulse on one line and a positive pulse on the other. The differential receiver at the other end detects the total difference between these two pulses. Since pulses of EMI noise (shown in red in the positive direction and blue in the negative direction in the diagram) usually produces pulses of the same polarity on both lines, one pulse essentially cancels out the other at the receiver. In addition, the magnetic field surrounding one wire from a signal pulse mirrors that of the magnetic field surrounding the other wire. At very short distance from the two wires, the magnetic fields are oppose and theoretically cancel each other out. This reduces, but does not eliminate, the line pair's impact on other pairs of wires in the same cable or in adjacent cables. 2) The second is a means of reducing interference from close-by lines, sometimes referred to as cross-talk ...the term cross-talk is a Telco term stemming from the fact that cross-talk between adjacent phone lines literally allows one user hear conversations on other phone lines. The twisting of the line pairs in a double helix configuration is designed to reduce cross-talk between adjacent line pairs. The double helix configuration theoretically produces symmetrical (identical) noise signals in each wire. In theory, their difference, as detected at the receiver, is zero. In practice, it is much reduced, but by no means eliminated. [ Top | Contents | Previous | Next - Color Code Standards ]

COLOR-CODE STANDARDS Last updated:

These are the pin-out diagrams for straight through and crossover UTP Ethernet cables.

9

The TX (transmitter) pins are connected to corresponding RX (receiver) pins, with plus to plus and minus to minus. A coss-over cable must be used to connect units with identical interfaces. When straight-through cables are used to connect Ethernet devices, one of the two units must, in effect, perform the cross-over function. This is the reason that straight through cables work directly between hubs or switches and NIC cards.... the Hub or Switch is designed so that their RJ45 Jacks are pre-wired with the transmit and receive pairs already reversed. There are two color-code standards in common use: EIA/TIA 568A and EIA/TIA 568B. These standards derive from TELCO usage and the pairs shown correspond to four phone lines, each with its own line pair. This same wiring was adopted for LAN standard Ethernet RJ45 wiring as well. RJ45 receptacle wiring for both standards are shown below:

Note: Only pairs 2 and 3 are used for Standard Ethernet wiring. Pairs 1 and 4 can be used for other purposes such as telephones or even a second separate, complete Ethernet connection. Straight-Through Wiring Using The 586A Standard

10

The flat wiring diagram, above, shows the 568A color code standard as the wiring for the PC side of the cable and the same 568A standard for the Hub, Switch or Router side of things (assuming that the Hubs, Switches or Routers are wired internally to perform the cross-over function). The illustration depicts the wiring arrangement before insertion into an RJ45 connector prior to crimping. Cross-Over Wiring Using The 568A to 586B Standards

The flat wiring illustration, above, shows cross-over cable wiring using the 568A color code standard as the wiring for the PC side of things and the

11

568B standard for wiring to the other PC. Note that in both cases, all eight wires are shown but only four are actually needed. Pins 4, 5, 7, and 8 and the blue and brown pairs are not used in either standard. Contrary to common tech-lore and what you may have read elsewhere, these pins and wires are not used or required to implement 100BASE-TX duplexing. In fact, they can be used for other purposes such as a single line phones or even operating two separate Ethernet channels, provided care is taken to assure that these wire pairs are isolated from the other wires. In practice, making actual RJ45 Patch cables is not physically that simple. The connections of the pairs to the pins in the RJ45 jack isn't wire pair by wire pair. Instead, the orange pair of wires are not adjacent and the blue pair is upside-down. If fact, flattening out the cables in the correct order for insertion into the RJ45 jack before crimping is by far the most complex part of the job of making twisted pair Ethernet patch cables. One cannot use flat-untwisted telephone cable for a network cable that runs any appreciable distance. One must use a pair of twisted wires to connect a set of transmitter pins to their corresponding receiver pins. One cannot use a wire from one pair and another wire from a different pair... See the theory page for the reasons why. Blue Max designed a RJ45 wiring card so new technicians won't have trouble remembering how to wire straight-through and cross-over cables. These are business card size and are printed on both sides of business card stock. One side shows the 568A standard at both ends (a regular straight-through cable). There's a fold mark at about 1/3 of the way across the card. To see the wiring for a cross-over cable, the user simply folds the card to see the wiring of a cross-over cable with the 568A standard on one end and the 568B standard on the other. A sample of the card is shown below, along with a down-loadable .PDF File of the front and rear of the card along with the front and back of the card in Microsoft Word. These are designed for printing on Avery Stock 8371.

12

Front of Card

Back of Card

If you'd like a free card, fill in the order form by clicking below and we'll mail you one. Or, if you'd like to download the Word file and print your own, click the download button

[ Top | Contents | Previous | Next - Make It Simple ]

KISS Last updated:

Note that there are only two unique cable ends in all of the preceding diagrams. They correspond to the 568A and 568B RJ45 jacks. A sample of the A standard is shown to the right. Wires with colored backgrounds may have white stripes and may be called out that way in diagrams seen elsewhere. For example, the green wire may be labeled Green-White. However, in this article, the background color is always specified first. All one needs to remember the proper configuration for cables are the diagrams for the two cable ends and the following rules: • •

A straight-thru cable has identical ends. A crossover cable has different ends.

It makes no functional difference which standard is used for a straight-thru cable. One can start a crossover cable with either standard as long as the other end uses the other standard. There is no practical or functional difference which end is which. A 568A patch cable will work in a network with 568B wiring and 568B patch cable will work in a 568A network. The

13

connectivity is what counts and from a connectivity standpoint, both are identical. Most professionals prefer to use the 568A standard for straight-thru cables and to start crossover cables with a 568A end. That way, all one needs to remember is the diagram for the 568A end, that a straight-thru cable has two of them, and that the green and orange pairs (pairs 2 and 3) are swapped at the other end of a crossover cable. [ Top | Contents | Previous | Next - Make Some Cables ]

MAKING CAT 3, 5, 5E RJ45 ETHERNET CABLES Last updated:

An easy way to measure cable is to place marks at 5 and 10 feet from the cable box or reel. Then, pull the cable off the reel or out of the box to the desired lengths and cut. For cable lengths between the marks, estimate the length as a percent of the distance between known marks. One can get a rough estimate by stepping off the distance, assuming that a normal human foot with shoes on is about one foot long. If numerous cable of the same length are needed, make the mark from the box or reel at the exact distance for the cables being made. For cables longer than 10 feet, pull the cable out to the ten foot mark and then go back to the box and repeat this procedure until until the desired cable length (rounded down to the nearest 10') has been pulled out. The remainder is then measured out by estimating or "walking" it out using the "foot" measurement method. Routing cable is best done before terminating the ends with RJ45 connectors. This way, one doesn't have to worry about snagging the RJ45 connector's locking tabs when running them through walls, floors, holes, cracks and other small spaces. This is especially true when a cable must be pulled backwards, since the locking tabs will invariable snag, either breaking the tab and requiring re-termination or pulling the connector off entirely. Wiring runs should always follow the wiring rules listed at the end of this article. It's much easier to do the job right the first time than having to go back and figure out why the connectivity is poor; after the wire has been run. The total length of twisted pair wire segments between a PC and a hub or between two PC's should not exceed 100 Meters (329 feet or about the length of a football field) for 100BASE-TX or 10BASE-T cable runs. The maximum length for 1000BaseTX twisted pair is presently 82 feet. The Tables, below show the minimum and maximums for 10Base2, 10BaseT, 100TX and 1000BaseTX

14

10Base2/10BaseT Maximum Cable Lengths

Thin-net(Coax): 600 Feet Twisted Pair: 329 Feet Fiber 3,000: Feet

Minimum Cable Lengths

Thin-Net (Coax): 1.5 Feet

Maximum Number of Stations per Cable

Thin-net: 30 Twisted Pair: 2 Fiber: 2

Maximum Number of Stations per Logical Network

1024

Maximum Segments

5 Segments only three of which are populated

Maximum Overall Length of Logical Network

3,000 Feet

100TX Maximum Cable Lengths

Twisted Pair: 329 Feet Fiber :650 Feet

Minimum Cable Lengths

None

Maximum Number of Stations per Cable

Twisted Pair: 2 Fiber: 2

15

Maximum Number of Stations per Logical Network

1024

Maximum Segments

2 Segments

Maximum Overall Length of Logical Network

650 Feet

1000BaseTX Maximum Cable Lengths

Twisted Pair: 82 Feet but 329 feet with CAT 5E operating at 350 MHz Fiber: 1640 Feet

Minimum Cable Lengths

None

Maximum Number of Stations per Cable

Twisted Pair: 2 Fiber: 2

Maximum Number of Stations per Logical Network

1024

Maximum Segments

1 or 2 Segments

Maximum Overall Length of Logical Network

Media Dependent

Although this page describes making cables that may be hundreds of feet long, we do not suggest this method for cable runs between rooms. The professional way to run Ethernet twisted pair wiring is to use a patch panel and RJ45 wall jacks between rooms and patch cables from the jacks to the

16

PC's or hubs/switches/routers. The cables described herein are patch cables used to connect to the wall jacks or patch panels that terminate the between room wiring. Patch panels and RJ45 wall jacks not only provide a finished, professional look but they provide easy methods of rerouting connectivity, ease of trouble-shooting problems and a far easier method of terminating long cable runs. Be sure to use patch panels and wall jacks that have at least the same, or higher, ratings than the cable being used. One cannot use CAT 5 rated cable with a CAT 3 rated RJ45 jack or patch panel and expect to have anything better than CAT 3 performance. See our section on punch down tools and techniques and RJ45 wall jack and patch panel standards for punch down RJ45 jacks and panel wiring. After cutting and pulling the cable, strip one end of the cable with a specialized stripper tool or a knife and diagonal cutters as detailed elsewhere. Depending on the stripper used, the cable is placed in the groove of the blade (left) side of the stripper and the end of the cable is aligned with the right side of the stripper. The goal is to end up with the sheath stripped, leaving about 1/2" of flattened evenly cut wires extending past the cut-back sheath. The stripper should be turned about 1 1/4 turns and the cable pulled out, leaving behind at least enough wire to end up with 1/2 inch of evenly-cut, flattened cable extending past the cut-back sheath. If the tool is rotated more than 1 1/4 turns, the wires can be nicked and require re-cutting. This means leaving 3/4 or 1 inch of exposed wire so that it can be trimmed evenly after the wires are flattened and arranged in the right order. When using a knife and diagonal cutters, be careful to slit only the sheath and not the wire insulation around the twisted pairs. Carefully slit the cable by an inch or so and neatly trim around the circumference of the cable with diagonal cutters to remove the jacket. Nicks are a No-No, so carefully inspect the wires for nicks. If any nicks are found, cut the end off and start over. Be sure the cutting blade has been properly adjusted using the screw at the front of the stripper. Cable diameters and jacket thicknesses vary. Prepare the pairs for insertion by spreading and arranging them, roughly in the order of the desired cable end.

17

Untwist the pairs and then arrange the wires in the desired pattern for insertion and crimping to the RJ45 connector. The ends can be flattened by using the thumb and forefinger. They can also be flattened by placing them on a flat surface and using a smooth tool or finger or thumb. Once the desired pattern of flattened wires is in place, trim the ends of the wires so they are even with each other and about 1/2" long from the shortened sheath to the ends of the wires. It is very important that the un-stripped (un-twisted), evenly-cut wires be slightly less than 1/2" long. If longer than 1/2" the ends will "bottom out" at the end of the RJ45 wire channels and force the jacketed portion of the cable away from the cable clamping plug. When this happens, the jacket cinching plug will miss the jacket and the cable will not be securely connected to the RJ45 connector. If the flattened wires are too short, they will miss the cutting blades that connect them to the RJ45 connectors. Flatten the cable again after trimming. There should be little or no space between the wires. For final pre-crimp inspection, the RJ-45 plug should be held with the locking clip facing down or away. Push the wire firmly into the plug. Carefully inspect the assembly before crimping. Once crimped, problems can be resolved only by cutting off the end and starting over! Look through the bottom of the plug and make sure the wire on the far left side has a white background. The wires should alternate light/dark from left to right. The furthest right wire is always brown. All wires should end evenly at the front of the plug. The jacket should end just about as shown in the diagram -- exactly on the line. [ Top | Contents | Previous | Next - How to Crimp ]

HOW TO CRIMP CAT 3, 5 and 5E RJ45 CONNECTORS Last updated:

The keys to good crimps are to start with good tools and taking care that the ends of the wires are trimmed evenly. Then, make only one crimping motion. The last requirement is usually met by the first since good crimping tools incorporate a ratchet assembly that forces the user to complete a full crimping motion before the crimp jaws can be returned to their detent position. There are actually two simultaneous crimping actions performed by the crimping tool. One is to force a set of eight teeth through the insulation of the eight wires and through the wires themselves to form eight separate electrical connections. The secondary crimping forces a plastic wedge or plunger against the outer sheath that covers the four wire pairs. If the sheath is cut too far back from the ends of the wires being connected, the 18

plastic tab of the RJ45 will simply push against the individual covering of the the wire pairs and will not firmly hold the wires in place. Be sure to trim the outer sheath of the CAT 3, 4, 5 or 5E wire just enough to allow the individual wires to extend to the end of the RJ45, while leaving the outer cable sheath to extend past the plastic tab used to hold the wiring assembly in place. See the diagram below:

Trim the outer sheath of the cable as shown. The sheath should be cut back far enough to allow the wires (with insulation) to be trimmed to even lengths and extend exactly to the end of the wire channels of the RJ45. Not too short, but just long enough to press against the end of the channels as shown. If the sheath and wires don't go into the RJ45 as shown in the side view, cut the wire back and start over. There is nothing any more frustrating than a finicky, flaky intermittent wiring connection. Once the trimming is satisfactory and the individual wires are arranged in the proper order and flattened out for insertion into the RJ45, push the wire into the RJ45 and make sure the assembly looks like the side view drawing shown above (except for the fact that the plastic plunger and connecting blades will not be pressed down yet). Once satisfied that the assemble is ready for crimping, insert the RJ45 (with wire in place) into the jaws of the crimping tool (it should only go in one way). Next, firmly press down on the handles of the crimping tool, keeping the wire firmly pressed into the RJ45. If the correct crimping tool is used, one must continue to crimp until the process is complete. A one way ratchet device assures this. CRIMP ONLY ONCE If a quality tool is used, the ratchet assembly assures that the crimp was complete. Attempting to make a second crimp is more likely to bend or flatten pins or crack the plastic housing than it is to improve the crimp. If it was done correctly with a good tool, once is enough.

19

WHAT ACTUALLY TOOK PLACE The crimping dies pushes two plungers down on the RJ-45 connector. One forces a plastic plug/wedge onto the cable jacket, firmly clinching it and preventing it from pulling out of the RJ45 connector. The other seats the "pins," (each with two teeth at its end) through the insulation and into the conductors of their respective wires. TEST THE CRIMP If properly done using quality components and tools, the connector will stay firmly in place even if pulled apart with up to 25 pounds of force. A firm pull with an average person's bare hands should not separate the cable and connector. Far better to find out that the wiring assembly won't hold up before it's put into service than to spend hours tracing a poor connection once it's in place. Don't pull too hard when testing the crimp. If the cable is stretched, its characteristics will change. Just look at the side of the plug and see if it looks like the diagram and give it a fairly firm tug to make sure it is crimped well. Advantages of twisted-pair cables over the older thin-wire, coaxial cables include low cost, ease of termination and reliability and permanence of the connection. RJ-45 and its ease of installation are one of the reasons coaxial cable is no longer widely used for small Ethernets. However, the larger reason is that there is no coax 100BaseTX or 1000BaseTX. The world has moved beyond 10Base2's capabilities, so coax is out and CAT 5, 5E are in. One will still find coax in older installations and this presents a problem. The problem is that there is no RJ45 equivalent to the daisy chaining of systems that was possible with coax. After the connectors are attached, if both ends of the cable are within reach, hold them next to each other with the RJ-45 locking tabs facing away. Look through the bottom of the RJ45 connector. If the RJ45s are wired correctly, they are identical in a straight-thru cable and different in a crossover cable. Inexpensive test tools are available that claim to test cables. However, most of these only test the cable in a static, DC voltage environment, not at 10 or 100 MHz. If an operational network is available, test the cable as part of the network and copy some large files over the cable and check the results.

20

If the cable doesn't work, inspect the ends again and make sure you have the right cable and that it is plugged into the correct units for the type of cable. When many straight-through and cross-over cables are used in a system, it is a good idea to label or otherwise identify cross-over cables. One way is to use different color cable when making cross-over cables. Another is to use wiring boots such as those shown at the right. If all else fails, use a magic marker to place an "X" on both ends of cross-over cables. Hours can be wasted testing a system only to discover that the problem was that one of the cables was a crossover cable when a straight through cable was intended. [ Top | Contents | Previous | Next - Cat 5 Cabling Rules ]

CABLING RULES Last updated:

1. Use quality components and tools to construct cables. As the saying goes, "Buy Quality, Only Cry Once." 2. Under no circumstances should cable bends be less than four times the diameter of the cable. The Cat 5E standard is no bend radius less than 5 inches. 3. When bundling groups of cables together with cable ties (zip ties), keep the ties snug but not excessively so. Do not over-cinch them. Keep them snug but don't tighten them so much that any of the cables deform. 4. Keep cables away from devices that can introduce EMI noise. Among others, these include: copy machines, computer monitors, power supplies, UPS units, electric heaters, speakers, printers, TV sets, fluorescent lights, AC power cables, RF antennas or transmission lines, copiers, welding machines, radio transmitters, X-Ray Machines, un-shielded transformers, refrigerator compressor motors, dishwashing machine motors, microwave ovens, telephones, fans, electric garage door openers, elevator motors, electric ovens, dryers, washing machines, and shop equipment. 5. Power cables and Ethernet twisted pair cables don't co-exist well. Do not run Ethernet cables parallel to power cables. Yes, we know that this is a repeat of number 4 but it is worth repeating... DO NOT RUN ETHERNET CABLES PARALLEL TO POWER CABLES! 6. Do not stretch UTP cables when pulling cable. The maximum force on a cable should be 25 LBS or less. 21

7. Do not use metal staples or insulated metal U shaped cable clips to secure UTP cables. Use telephone wire hangers, preferably ones with plastic hangers for the wire. 8. Never, Never run UTP cable outside a building. It presents a very attractive lightning rod and will prove dangerous to you and your network's health! REFERENCES Charles Spurgeon's Ethernet Web Site Chris Brenton, Multi-protocol Network Design and Troubleshooting, Sybex Network Press IBSN: 0-7821-2082-2 ConnectWorld Cable FAQ David Hess and John Gold; A Practical Guide to Cable Selection, National Semiconductor Application Note 916, 0ctober 1993 Hubbell Premise Wiring Keneth M. True; Data Transmission Lines and Their Characteristics,, National Semiconductor Application Note 806, April 1992 Robert Grover Brown, et al; Lines, Waves, and Antennas, The Transmission of Electrical Energy, The Ronald Press Company, New York, 1973 [ Top | Contents | Previous | Next - Back to The Beginning ]

22