Tuesday, May 15, 2012

Coaxial Cable Connectors

Unless you have operated a 10Base-2 or 10Base-5 Ethernet network, you are probably familiar only with the coaxial connectors you have in your home for use with televisions and video equipment. Actually, a number of different types of coaxial connectors exist.

F-Series Coaxial Connectors

The coax connectors used with video equipment are referred to as F-series connectors (shown in Figure 1). The F-connector consists of a ferrule that fits over the outer jacket of the cable and is crimped in place. The center conductor is allowed to project from the connector and forms the business end of the plug. A threaded collar on the plug screws down on the jack, forming a solid connection. F-connectors are used primarily in residential installations for RG-58, RG-59, and RG-6 coaxial cables to provide CATV, security-camera, and other video services.

Figure 1: The F-type coaxial-cable connector
F-connectors are commonly available in one-piece and two-piece designs. In the two-piece design, the ferrule that fits over the cable jacket is a separate sleeve that you slide on before you insert the collar portion on the cable. Experience has shown us that the single-piece design is superior. Fewer parts usually means less fumbling, and the final crimped connection is both more aesthetically pleasing and more durable. However, the usability and aesthetics are largely a function of the design and brand of the two-piece product. Some two-piece designs are very well received by the CATV industry.
A cheaper F-type connector available at some retail outlets attaches to the cable by screwing the outer ferrule onto the jacket instead of crimping it in place. These are very unreliable and pull off easily. Their use in residences is not recommended, and they should never be used in commercial installations.

N-Series Coaxial Connectors

The N-connector is very similar to the F-connector but has the addition of a pin that fits over the center conductor; the N-connector is shown in Figure 2. The pin is suitable for insertion in the jack and must be used if the center conductor is stranded instead of solid. The assembly is attached to the cable by crimping it in place. A screw-on collar ensures a reliable connection with the jack. The N-type connector is used with RG-8, RJ-11U, and thicknet cables for data and video backbone applications.

Figure 2: The N-type coaxial connector

The BNC Connector

When coaxial cable distributes data in commercial environments, the BNC connector is often used. BNC stands for Bayonet Neill-Concelman, which describes both the method of securing the connection and its inventors. Many other expansions of this acronym exist, including British Naval Connector, Bayonet Nut Coupling, Bayonet Navy Connector, and so forth. Used with RG-6, RG-58A/U thinnet, RG-59, and RG-62 coax, the BNC utilizes a center pin, as in the N-connector, to accommodate the stranded center conductors usually found in data coax.
The BNC connector (shown in Figure 3) comes as a crimp-on or a design that screws onto the coax jacket. As with the F-connector, the screw-on type is not considered reliable and should not be used. The rigid pin that goes over the center conductor may require crimping or soldering in place. The rest of the connector assembly is applied much like an F-connector, using a crimping die made specifically for a BNC connector.
Figure 3: The BNC coaxial connector
To secure a connection to the jack, the BNC has a rotating collar with slots cut into it. These slots fit over combination guide and locking pins on the jack. Lining up the slots with the pins, you push as you turn the collar in the direction of the slots. The slots are shaped so that the plug is drawn into the jack, and locking notches at the end of the slot ensure positive contact with the jack. This method allows quick connection and disconnection while providing a secure match of plug and jack.
Be aware that you must buy BNC connectors that match the impedance of the coaxial cable to which they are applied. Most commonly, they are available in 75 ohm and 50 ohm types, with 93 ohm as a less-used option.
With all coaxial connectors, be sure to consider the dimensions of the cable you will be using. Coaxial cables come in a variety of diameters that are a function of their transmission properties, series rating, and number of shields and jackets. Buy connectors that fit your cable.

Friday, May 11, 2012

Shielded Twisted-Pair Connectors

In the United States, the most common connectors for cables that have individually shielded pairs in addition to an overall shield are based on a pre-1990 proprietary cabling system specified by IBM. Designed originally to support Token Ring applications using a two-pair cable (shielded twisted-pair, or STP), the connector is hermaphroditic. In other words, the plug looks just like the jack, but in mirror image. Each side of the connection has a connector and a receptacle to accommodate it. Two hermaphroditic connectors are shown in Figure 1. This connector is known by a number of other names, including the STP connector, the IBM data connector, and the universal data connector.

Figure 1: Hermaphroditic data connectors
The original Token Ring had a maximum throughput of 4Mbps (and later 16Mbps) and was designed to run over STP cabling. The 16Mbps Token Ring used a 16MHz spectrum to achieve its throughput. Cables and connectors rated to 20MHz were required to allow the system to operate reliably, and the original STP hermaphroditic connectors were limited to a 20MHz bandwidth. Enhancements to these connectors increased the bandwidth limit to 300MHz. These higher-rated connectors (and cable) are designated as STP-A.
STP connectors are the Jeeps of the connector world. They are large, rugged, and versatile. Both the cable and connector are enormous compared to four-pair UTP and RJ-type modular plugs. They also have to be assembled and have more pieces than an Erector set. Cabling contractors used to love the STP connectors because of the premium they could charge based on the labor required to assemble and terminate them.
Darwinian theory prevailed, however, and now the STP and STP-A connectors are all but extinct—they've been crowded out by the smaller, less expensive, and easier-to-use modular jack and plug.

Tuesday, May 8, 2012

Crossover Cables | Modular Jacks and Plugs

One of the most frequently asked questions on wiring newsgroups and forums is "How do I make a crossover cable?" Computers that are equipped with 10Base-T or 100Base-TX network adapters can be connected "back-to-back"; this means they do not require a hub to be networked together. Back-to-back connections via crossover cables are really handy in a small or home office. Crossover cables are also used to link together two pieces of network equipment (e.g., hubs, switches, and routers) if the equipment does not have an uplink or crossover port built-in.
A crossover cable is just a patch cord that is wired to a T568A pinout scheme on one end and a T568B pinout scheme on the other end. To make a crossover cable, you will need a crimping tool, a couple of eight-position modular plugs (a.k.a. RJ-45 plugs), and the desired length of cable. Cut and crimp one side of the cable as you would normally, following whichever wiring pattern you desire, T568A or T568B. When you crimp the other end, just use the other wiring pattern.
As mentioned several times elsewhere in this book, we recommend that you buy your patch cords, either straight through or crossover, instead of making them yourself. Field-terminated patch cords can be time-consuming (i.e., expensive) to make and may result in poor system performance.
Table 1 shows the pairs that cross over. The other two pairs wire straight through.
Table 1: Crossover Pairs 
Side-One Pins
Wire Colors
Side-Two Pins
1 (Transmit +)
3 (Receive +)
2 (Transmit –)
6 (Receive –)
3 (Receive +)
1 (Transmit +)
6 (Receive –)
2 (Receive –)

Saturday, May 5, 2012

Using a Single Horizontal Cable Run for Two 10Base-T Connections

Let's face it, you will sometimes fail to run enough cable to a certain room. You will need an extra workstation in an area, and you won't have enough connections. Knowing that you have a perfectly good four-pair UTP cable in the wall, and that only two of those pairs are in use, makes your mood even worse. Modular Y-adapters can come to your rescue.
Several companies make Y-adapters that function as splitters. They take the four pairs of wire that are wired to the jack and split them off into two separate connections. The Siemon Company makes a variety of modular Y-adapters (see Figure 1) for splitting 10Base-T, Token Ring, and voice applications. This splitter will split the four-pair cable so that it will support two separate applications, provided that each application requires only two of the pairs. You must specify the type of splitter you need (voice, 10Base-T, Token Ring, etc.). Don't forget, for each horizontal cable run you will be splitting, you will need two of these adapters: one for the patch-panel side and one for the wall plate.

Figure 1: A modular Y-adapter for splitting a single four-pair cable into a cable that will support two separate applications
Many cabling professionals are reluctant to use Y-adapters because the high-speed applications such as 10Base-T Ethernet and Token Ring may interfere with one another if they are operating inside the same sheath. Certainly you should not use Y-adapters for applications such as 100Base-TX. Furthermore, Y-adapters eliminate any chance of migrating to a faster LAN system that may utilize all four pairs.