In this article, and in two that will follow, we will look at the basics of data communications using copper cabling. Obviously, there is a lot of material to cover on this subject, so we will be covering only the basic information. We will attempt not to duplicate much of what we've covered in recent datacom articles. But before I get into the technical information, I think it is important to point out why now is a critical time to become involved with datacom.
During the 1990s, electrical contractors began migrating into data cabling. In the past two or three years, they have been making some real money. This work, which barely existed in 1990, was 28% of all work done by electrical contractors in 1995. Today, it is probably 35% to 40%, and continues to rise. Doing nothing takes you out of the datacom market. Right now that means that you are simply passing up one-third of the work available to you. In a few years that will mean taking a pass on half of the available work. Obviously, the basic reason for getting into datacom is that the mark-up that can be charged for low voltage work is frequently 50% to 100% higher than the mark-up that can be charged for power wiring.
But along with the promise of higher profit percentages, the timing for getting into datacom is ideal. Right now, you have an opportunity to catch up with the pioneers. This opportunity will likely vanish within a year or two. Here's why: For the past several years, datacom customers have been obsessed with Category (Cat) 5 cabling. Now, the entire business is changing over to new types of cabling, and forcing everyone to relearn the technologies.
Cat 6, Cat 7, and fiber is replacing Cat 5. Cat 5 cannot keep up with the bandwidth demands. When companies need an extension to their network (or a new network), they are asking, "Can you install Cat 6?" In other words, customers will no longer care about your experience (or lack thereof) with Cat 5.
But they will want to know about Cat 6 or 7 (or, possibly, fiber optics). Because Cat 6 and 7 are new, you'll have as much experience with it as the pioneers. This is probably the last chance you'll have to take advantage of this situation, because the datacom market is unlikely to ever again be obsessed with one type of cabling, as was the case with Cat 5.
Two circuit concerns
There are two separate concerns you have for data circuits:
There must be a clear path from one machine to the next. Here we are concerned with the signal's strength; it must arrive at the far end of the line with enough strength to be useful.
The signal must be of good quality. If we send a square-wave digital signal into one end of a cable, we want a good square-wave coming out of the far end. The main problem with power wiring is a loss of power (voltage drop). The same problem exists with data signals transmitting through conductors with too much resistance (usually due to distance). With data cabling, this is attenuation, and it is virtually the same as voltage drop-not enough power getting through.
The problem of signal quality is a completely different concern. With data signals, it is our job to make sure the signal quality is good. Getting enough signal from one end of the cable to another is one thing we must do. Therefore, we must also make sure that the signal does not distort.
In order for data signals to mean anything, they must arrive at the far end of their path resembling the input signal. The receivers in data systems require a clear signal. Since data signals are always digital (all zeros and ones), a signal that is too distorted means nothing at all.
Pulse spreading is the most common type of signal distortion. The digital signal sent into the fiber is a square wave. As the signal travels down the fiber, it distorts and spreads.
Pulse spreading is not a loss of signal; it is a distortion of the signal. If the pulses spread too much, they will be unintelligible to the receiver. The communication will be lost.
How networks work
The purpose of networking is to allow a number of computers to operate together, share information, and allow one computer operator to read and use programs in another computer. Here are the fundamental parts of networks.
Connection to the computer: If the network is going to allow other computers to access the information in one computer, the network must have some way of getting into it. Use a network interface card (NIC) to do this. The NIC is nothing more than a circuit card that fits into one of the expansion slots in the back of a personal computer. It interprets information between the computer and the network, and feeds information in and out of the computer. This is the most vital link between the processor and the network.
Communication means: Once you have a network interface in place, you need some method of getting the data signals from one computer on the network to another. The most common method uses unshielded twisted-pair copper (UTP) cables. Sending 60 cycle utility power through a wire rarely presents a difficulty, but sending a 100 million cycle signal can be a little more tricky. For this reason, the method of sending signals and the transporting materials are important.
Coordinate all parts of the system to send, carry, and receive the same types of signals. Usually, these details are not something that you have to consider, as long as all parts of your network come from (or at least are specified by) the same vendor. If you ever have trouble with your network, however, you will have to check the signals, and make sure that they are of the correct types. If there are problems with these signals, your network will not function properly.
Connection pattern: There are several methods for connecting all of the computers on the network together. Some of the most common methods are:
1. A star pattern.
2. A ring pattern.
3. A bus pattern.
4. A mesh pattern.
In addition to these connection patterns, there are others. Among these are the tree structure (a group of stars, connected to a bus) and a star-ring.
The brand of network that you plan to install determines which connection pattern you need. There can be advantages and disadvantages to all types. Other parts of a network are the processor hierarchy (defines which one of the processors-computers-in the network is the central "brain") and appropriate network software.
The processor hierarchy:The processor hierarchy defines which one of the processors in the network is the central "brain." Or, the hierarchy may indicate that all computers on the network are equal. This is important to define because many types of networks use one computer (usually the most powerful) as the heart of the network, and the other computers on the network as satellites. The center computer is generally called the server, or file server, and the satellite computers are commonly called nodes or clients.
Appropriate software: Because networks send various routing commands through their system, software that is not written for networks (where it will be exposed to these strange commands), will often not work properly. Because of this, the software used on networks is often specifically written for networks. Many of the more popular types of software come in a "network version." It is very important to verify if your software is compatible to your network before you install it.
Here are a few of the devices that are commonly used in networks:
A repeater receives and then immediately retransmits each bit. It has no memory and does not depend on any particular protocol. It duplicates everything, including the collisions between data streams.
A bridge receives the entire message into memory. If the message was damaged by a collision or noise, it is discarded. If the bridge knows that the message was being sent between two stations on the same cable, it discards it. Otherwise, the message is queued up and will be retransmitted on another network cable. The bridge has no address. Its actions are transparent to the client and server workstations.
A router acts as an agent to receive and forward messages. The router has an address and is known to the client or server machines. Typically, machines send messages directly to each other when they are on the same cable, and they send the router messages addressed to another zone, department, or sub-network.
Why networks fail Networks fail in three common ways: 1. A nail or other object can break one of the conductors.
2. A screw or other object can touch one or more of the conductors and short them to an external grounded metal shield, conduit, or other grounded metal.
3. A station on the network can break down and start to generate a continuous stream of electronic noise, thus blocking legitimate transmissions.
A time domain reflectometer (TDR) is used to find problems in a network. It plugs into any attachment point in the cable, and sends out its own voltage pulse. The effect is similar to a sonar ping. If the cable is broken, then there is no proper terminating resistor. The pulse will hit the loose end of the broken cable and will bounce back. The test device sensesthe echo, computes how long the round trip took, and then reports how far away the break is in the cable.
If a network cable is shorted out, a simple voltmeter determines that the proper resistance is missing from the signal and shield wires. Again, by sending out a pulse and timing the return, the test device can determine the distance to the problem.
Newer generations of "smart" hubs can perform part of the error detection and reporting function. For example, they could isolate a problem in the connection to a particular desktop workstation and automatically isolate that unit from the rest of the network.
Baseband network. A baseband network is one that provides a single channel for communications across the physical medium (cable), so that only one device can transmit at a time. Devices on a baseband network, such as Ethernet, are permitted to use all the available bandwidths for transmission, and the signals they transmit do not need to multiplexed onto a carrier frequency. An analogy is a single phone line: Only one person can talk at a time-if more than one person wants to talk everyone has to take turns.
Broadband network. A broadband network is in many ways the opposite of a baseband network. With broadband, the physical cabling is virtually divided into several different channels, each with its own unique carrier frequency, using a technique called frequency division modulation.
These different frequencies are multiplexed onto the network cabling in such a way to allow multiple simultaneous "conversations" to take place. The effect is similar to having several virtual networks traversing a single piece of wire. Network devices tuned to one frequency can't hear the signal on other frequencies, and vice-versa. Cable-TV is the best example of a broadband network, with multiple conversations (channels) transmitted simultaneously over a single cable; you pick which one you want to see by selecting the frequencies.
OSI Model. The Open Systems Interconnect (OSI) reference model is the ISO (International Standards Organization) structure for network architecture. This Model outlines seven areas, or layers, for the network. These layers are (from highest to lowest):
7. Applications: Where the user applications software lies. Such issues as file access and transfer, virtual terminal emulation, interprocess communication and the like are handled here.
6. Presentation: Differences in data representation are dealt with at this level. For example, UNIX-style line endings (CR only) might be converted to MS-DOS style (CRLF), or EBCIDIC to ASCII character sets.
5. Session: Communications between applications across a network is controlled at the session layer. Testing for out-of-sequence packets and handling two-way communication are handled here.
4. Transport: Makes sure the lower three layers are doing their job correctly, and provides a transparent, logical data stream between the end user and the network service. This is the lower layer that provides local user services.
3. Network: This layer makes certain that a packet sent from one device to another actually gets there in a reasonable period of time. Routing and flow control are performed here. This is the lowest layer of the OSI model that can remain ignorant of the physical network.
2. Data Link: This layer deals with getting data packets on and off the wire, error detection and correction, and retransmission. This layer is generally broken into two sub-layers: The LLC (Logical Link Control) on the upper half, which does the error checking, and the MAC (Medium Access Control) on the lower half, which gets the data on and off the wire.
1. Physical: The nuts and bolts layer. Here is where the cable, connector and signaling specifications are defined. 10Base5, 10BaseT, 10Base2 and 10Broad36 are some of the IEEE names for the different physical types of Ethernet. The "10" stands for signaling speed: 10MHz. "Base"means Baseband, "broad" means broadband. Initially, the last section was intended to indicate the maximum length of an unrepeated cable segment in hundreds of meters. This convention was modified with the introduction of 10BaseT, where the T means twisted pair, and 10BaseF where the "F" means fiber.
In actual practice:
10Base2 Is 10MHz Ethernet running over thin, 50 Ohm baseband coaxial cable. 10Base2 is also commonly referred to as thin-Ethernet.
10Base5 Is 10MHz Ethernet running over standard (thick) 50 Ohm baseband coaxial cabling.
10BaseF Is 10MHz Ethernet running over fiber-optic cabling.
10BaseT Is 10MHz Ethernet running over unshielded, twisted-pair cabling.
10Broad36 Is 10MHz Ethernet running through a broadband cable. Driver. The software that allows a NIC card in a computer to decode packets and send them to the operating system and encode data from the operating system for transmission through the network. By handling the hardware interface chores, it provides a device-independent interface to the upper layer protocols, thereby making them more universal and easier to develop and use.
Attenuation is the loss of signal power. An attenuating signal is a weakening signal. At a given frequency, measure attenuation as the number of decibels per 100 feet. For LAN (Local Area Network) work, attenuation is measured in decibels over the link, or decibels per 100 meters.
Signals sent over copper wires deteriorate differently at different frequencies-the higher the frequency, the greater the reactance, and the greater the attenuation. These losses come primarily from the capacitance of the cable, the inductance of the cable, or from copper losses (simple resistance, resulting in heat). Attenuation is a problem, since only a sensitive receiver can pick-up a weakened signal.
Signal attenuation also depends on the construction of the cables, particularly the dielectric characteristics of the cable insulation. For example, you could have two 100 ohm, 24-gauge cables with one of the cables having a lower impedance than the other, because of the construction characteristics of the cables. Use a cable with lower impedance over longer distances to achieve better results.
Cable impedance. Impedance, which is the total opposition to current flow, is an important consideration. Cables used in computer networks are rated by their impedance characteristics. For example, the cable previously mentioned was a 100-ohm cable. That means that the cable has a characteristic impedance of 100 ohms.
Characteristic impedance is an old radio term, and refers to the internal signal-transmission characteristics of the cable. You cannot measure the characteristic impedance of a cable with an ohmmeter. The number does not refer to a number of ohms per foot or per hundred feet. The important thing with the characteristic impedance of cables is that you never switch to a cable with a differing impedance. If you are running 100-ohm cable, you should not insert a piece of 150-ohm cable in the network. If you were to do this, the signal would tend to reflect off the 100-to-150-ohm junction, ruining the transmission. In short, the link will not pass the signal if you mix cables of different impedance.
Characteristic impedance determines the amount of power transferred between devices. Whenever two electrical devices are connected, equal impedance means that the maximum power transfers from one to the other. (The output impedance of the driving device being equal to the input impedance of the driven device). Any amount of impedance mismatch will cause some of the power to reflect back into the source device. Use impedance-matching transformers instead of directly connecting cables of differing impedance.
Capacitance. Capacitance is the most common element of attenuation. The capacitance usually comes in the form of mutual capacitance, which is capacitance between the conductors within the cable. Capacitance is a big factor: data transmissions are square wave signals. When the capacitance is too great, it has a tendency to roundoff the data signals.
When data signals have a clear square wave, they are easily intelligible to the receiver. If, however, the signals become rounded, they can be confusing to the receiver, causing malfunctions.
Crosstalk. Crosstalk is the amount of signal picked up by a quiet conductor (a conductor with no signal transmitted over it) from other conductors that are conducting data. This phenomenon happens because of electromagnetic induction, the same principal by which transformers operate. Crosstalk contaminates adjacent lines and can cause interference, overloaded circuits, and other problems.
Cable shielding. The most common method to prevent crosstalk and electromagnetic interference is to place a shield around the conductors. The three principal types of shields are:
1. Longitudinally-applied metallic tape.
2. Braided conductors (commonly used in coaxial cables).
3. Foil laminated to plastic sheets. Shielding stops interference and cross- talk by absorbing magnetic fields since the shielding is conductive. It does induce a current into the shield. The flat surface of the shield spreads the current. This diffuses the current rendering it of no consequence to the transmission of data.
Foil shields will usually reduce electromagnetic interference by 35 decibels. Wire braid shields will generally reduce this type of interference by 55 decibels. A combination of the two types of shields will reduce interference by more than 100 decibels. It is possible to use filters to reduce interference, but this requires changing the filter every time the data transmission rate of the network alters.
Structured cabling systems typically do not use shielded cables. However, Cat 7 cabling will probably require shielded cables.