1. Introduction

2. Interfaces
  2.1 Connectors
  2.2 Signaling Types
  2.3 Every Day Problems With Interfaces
    2.3.1 Identity
    2.3.2 Gender
    2.3.3 Distance
    2.3.4 Interface Conversion
    2.3.5 Media Conversion
  2.4 Wiring Adapters
  2.5 Factors Influencing The Choice Of An Interface Converter
  2.6 Interface Conversion With Work Stations

3. Wiring Alternatives - Media
  3.1 Data Cable
  3.2 Twisted Pair
  3.3 Coaxial Cable
  3.4 Fiber Optic Cable

4. Extending Distance - The Need For Short Haul Modems
  4.1 Application
  4.2 Environment
  4.3 Selection
  4.4 PC Extenders

5. Resource Sharing
  5.1 Multiplexing
  5.2 Port Sharing

6. Lightning And Surge Protection
  6.1 Why Worry About Lightning?
  6.2 Where Should You Worry About Lightning?
  6.3 When Should You Worry About Lightning?
  6.4 Equipment Protection

7. Test Equipment
  7.1 Low Cost Data Line Analyzer
  7.2 Data Analysis
  7.3 Protocol Analyzers
  7.4 Future Trends

8. Local Area Networks
  8.1 Why Use LANS
  8.2 LAN Architectures
    8.2.1 Ethernet Bus Architecture
    8.2.2 Token Ring Architecture
    8.2.3 Comparing Ethernet And Token Ring
  8.3 Ethernet Implementation

References
Glossary

INTRODUCTION

Welcome to the subject of premises data communications. In this book we will explore its many facets, expose the various problems encountered in this area of interest and illustrate the way that these problems can be solved. The approach is very "results oriented." The view taken is not that of the academic network architect who has the opportunity and freedom of designing and specifying everything in advance by "starting with a blank page." Rather, for the most part, it is from the perspective of the person who has a premises data communications capability and for some reason or other is unhappy with it. It has problems which need to be dealt with. Either something needs to be fixed, expanded or protected. This book is a "Physician’s Desk Reference" not a "Gray’s Anatomy." Furthermore, we assume that our audience may or may not have a technical background in data communications. This book is written for the "hands on" user of data communication equipment, not the experienced design engineer.

It is best to begin by describing the setting for the communications problems of interest. Premises data communications deals with data communications in the office building, factory, manufacturing plant site or campus. An example of this environment, the small office building, is shown in Figure 1. The distances involved may vary from several inches to several thousands of feet. This is the world of what is commonly called short haul, limited distance or local area data communication. It is necessary to emphasize that this is not the world of long haul communication or Wide Area Networks-the world usually associated with long distance telephone dial-up circuits, leased lines, ISDN, packet switching, ATM or frame relay. Both data communication environments are important, but, the economics, technical problems and solutions differ.

Within the premises data communications setting there are data producing and using devices which need to be connected in order to support the modern automated office, factory or campus. These devices, these data sources and sinks, may be computers, computer terminals, badge readers, bar code readers, printers, plotters or other peripherals. These may even be the communication interfaces to the outside world, devices like high speed dial-up modems or packet switches.

The most commonly encountered situation in the premises data communications environment is that of the multi-user computer placed at one location e.g. the computer room, with terminals scattered throughout the rest of the building.



Figure 1: Typical Premises Data Communications Environment

In a sense from a user perspective it is a much "cozier" environment than that of the wide area network. The user can often "see," "touch" and play with all elements of the data communication system. The user has a closer connection to the data communication devices supporting the applications.

The subject of premises data communications deal with the methods and equipment needed to accomplish and support the connections of the data devices in this setting. While we said that our view is not "architectural" we must mention that there are three(3) architectural alternatives to solving this connection problem.

The first is the extreme of not really needing data communications at all due to the proliferation of self-contained work stations in the limited environment. All of the equipment needs of such a work station (both data processing and data communications) are contained within it and supplied by the same manufacturer or integrator. Here the problem of needing connection across the office building, factory or campus for the most part presents few challenges. We will touch upon these from time-to-time, but only in passing.

The second alternative is that of accomplishing the connectivity through the use of a Local Area Network (LAN). Here all data equipment resources "tap onto" the LAN. True distributed computing, sharing of all data resources, is then possible. This is the approach of choice when one has the freedom and the money to design a system starting essentially with a "blank page." We will deal with this in some limited depth towards the end of this book in Chapter 8.

The third alternative is the central focus of our attention. Here premises data communications is accomplished across limited distances using an empirical approach. There is no "grand architectural" design. Rather, the needed capabilities evolve in a haphazard manner driven by the demand of an operating business to have something "up and running." Devices are connected in the most straightforward manner as they are purchased, dealing with problems as they arise, on a case by case basis. The premises data communications capability in this environment usually grows like "Topsy" in "Uncle Tom’s Cabin." It is driven by the need to service user connections of the type illustrated in Figure 2. Here we show terminals which need to be connected to a multi-user minicomputer or now a 286/386/486 PC running UNIX/XENIX with many users or some equivalent system.

The problems that are encountered come under a number of different headings: Equipment Interfaces, Wiring Alternatives, Extending Long Distance, Protection, Resource Sharing and Testing.

In the following chapters we will deal with these subjects. The most commonly encountered situations will be discussed. We will illustrate how the problems can be dealt with and when appropriate indicate which Telebyte products can be used as solutions.



Figure 2: Typical Multi-User PC Environment

Pin	Signal	Pin	Signal
A	Chassis Ground	B	Signal Ground
C	Request to Send	D	Clear to Send
E	Data Set Ready	F	Receive Line Signal Detect
H	Data Terminal Ready	J	Ring Indicator
P	Transmitted Data (Signal A)	R	Recieved Data (Signal A)
S	Transmitted Data (Signal B)	T	Received Data (Signal B)
U	Terminal Timing	V	Receive Timing A
W	Terminal Timing	X	Receive Timing
Y	Transmit Timing	AA	Transmit Timing

Pin	EIA CKT	Description	From DCE	From DTE
1	AA	Protective Ground
2	BA	Transmitted Data		*D
3	BB	Received Data	*D
4	CA	Request to Send		*C
5	CB	Clear to Send	*C
6	CC	Data Set Ready	*C
7	AB	Signal Gnd/Common Return
8	CF	Rcvd. Line Signal Detector	*C
12	SCF	Secondary Rcvd. Line Sig. Detector	*C
13	SCB	Secondary Clear to Send	*C
14	SBA	Secondary Transmitted Data		*D
15	DB	Transmitter Sig. Element Timing	*T
16	SBB	Secondary Received Data	*D
17	DD	Receiver Sig. Element Timing	*T
19	SCA	Secondary Request to Send		*C
20	CD	Data Terminal Ready		*C
21	CG	Sig. Quality Detector	*C
22	CE	Ring Indicator	*C
23	CI	Data Sig. Rate Selector (DCE)	*C
23	DA	Transmitter Sig. Element Timing		*T
Signal Type: D=Data, C=Control, T=Timing
Note: On the DB25 connector that is commonly used for RS232:       Pins 9 and 10 are reserved for Data Set Testing.       Pins 11, 18 and 25 are undefined.       Pin 23 may be defined as CI or CH.

Pin	Signal	Pin	Signal
1	Data Strobe	19	(R) Data Strobe
2	Data Bit 1	20	(R) Data Bit 1
3	Data Bit 2	21	(R) Data Bit 2
4	Data Bit 3	22	(R) Data Bit 3
5	Data Bit 4	23	(R) Data Bit 4
6	Data Bit 5	24	(R) Data Bit 5
7	Data Bit 6	25	(R) Data Bit 6
8	Data Bit 7	26	(R) Data Bit 7
9	Data Bit 8	27	(R) Data Bit 8
10	Acknowledge	28	(R) Acknowledge
11	Busy	29	(R) Busy
12	Paper End	30	(R) Paper End
13	Select	31	Input Prime
14	Supply Ground	32	Fault
15	OSCXT	33	Undefined
16	Logic Ground	34	Undefined
17	Chasis Ground	35	Undefined
18	+5V	36	Undefined

Pin		EIA CKT	Description	From DCE	To DCE
A	B
1 2		SI	Shield Signaling Rate Indicator	*C
4 5	22 23	SD ST	Send Data Send Timing	*T	*D
6 7	24 25	RD RS	Receive Data Request to Send	*D	*C
8 9	26 27	RT CS	Receive Timing Clear to Send	*T	*C
10 11	29	LL DM	Local Loopback Data Mode	*C	*C
12 13	30 31	TR RR	Terminal Ready Receiver Ready	*C	*C
14 15		RL IC	Remote Loopback Incoming Call	*C	*C
16 17	35	SR TT	Signaling Rate Selector Incoming Call		*C *T
18 19		TM SG	Test Mode Signal Ground	*C
20 28		RC IS	Receive Common Terminal in Service		*C
32 33		SS SQ	Select Standby Signal Quality	*C	*C
34 36 37		NS SB SC	New Signal Standby Indicator Send Common	*C	*C
Signal Type: D = Data, C = Control, T = Timing Note: On the DB37 connector that is commonly used for RS449; Pins 3 and 21 are undfined B = Return

Figure 3d: Interface Specification for RS-449

Cables are used to connect the interface of one data equipment unit to another. But even the simple problem of cable connection presents situations of which you should be aware. For example, consider the need to connect a computer to a printer with the computer having a female interface and the printer having a female interface. In order to accomplish the connection you would need a cable with male connectors at both ends and consistent with both the computer and printer interfaces.

2.3 EVERY DAY PROBLEMS WITH INTERFACES

The typical premises data communication user is always encountering operational problems related to interfaces.

These problems are grouped as: 1) identity, 2) gender, 3) distance, 4) interface conversion and 5) media conversion.

2.3.1 IDENTITY

To discuss the identity problem it will be worthwhile to introduce some nomenclature. Data processing equipment is usually referred to as data terminal equipment and abbreviated DTE. These equipments such as terminals, computers and printers are considered as a class separate from equipment used to transport or communicate data such as modems, line drivers and multiplexers. Such communication equipment is referred to as data communications equipment and abbreviated DCE. Whether or not an equipment unit is a DTE or a DCE has significance with regards to the interface. Interfaces are usually defined assuming that the equipment of which the interface is the debarkation point is in fact a DTE.

A typical misconception is that the pin of the interface labeled Transmit Data (TD) implicitly assumes that the equipment of which it is part is a DTE. Along with this the interface is defined assuming that the data coming out of the TD pin is supposed to go into a data communication equipment unit, a DCE. The interfaces of a DTE and DCE must mate consistent with this assumption. The TD pin of a DTE interface must mate with the TD pin of the DCE interface.

However, what happens if you want to connect not a DTE and a DCE but two DTE’s? If you think this is a rare possibility then think again. Actually, it is quite a common situation. It occurs when you want to connect a computer and a printer. Now show me the facility where someone does not need a computer and a printer linked. The computer outputs data on the TD pin, Pin 2, of its interface. However, if a one-to-one cable is used a printer’s receive data pin, RD, Pin3, can not be mated properly to capture this data. No need to worry. This problem can be solved quite easily. Just use a passive modem eliminator in order, effectively, to change the wiring between the pins of the interface so that the computer’s TD pin mates with the printer’s RD pin. The device is called a modem eliminator because, historically, people used a "Rube Goldberg" approach to deal with this problem. They inserted DCE types of devices between the two DTE devices needing to be connected in order to get the consistent mating of their plugs. That is, in the early days of premises data communications users set up a system where the connector went from the computer to a modem which was connected back-to-back to another modem and then to a printer. This approach is illustrated in Figure 4.



Figure 4: Early Solution to Unscramble Interfaces

This haphazard approach did accomplish the proper mating of the pins. Of course, with this you were using the modem in a very uneconomic, costwise ineffective, manner. It was being used just to solve an interface problem and not being used for what it was designed, for extending the distance - one of the other subjects that we are going to be discussing. Much of the internal electronics of the modem was just being wasted The modem eliminator shown in Figure 5 is a device which effects the pin conversion but without the complexity and cost of the modem’s signaling capacity.



Figure 5: Present Modem Eliminator

2.3.2 GENDER

Gender mending sounds exotic but is really quite easy to describe. We have talked about the interfaces, possibly being female or male. Yet, what happens when one wants to connect equipment units directly together where both units only have male connectors?

To deal with this problem you can either use a connecting cable with a female interface connector at each end or you can use a device called a gender mender which is sometimes called a sex changer. This small device plugs into the interface on one of the equipment units and changes gender, either from male to female or female to male, to be consistent with the sex of the interface on the other equipment unit. You can use two connecting cables with the same gender connector on both ends. A low profile, gender changer is shown in Figure 6.



Figure 6: Low Profile Gender Changer

2.3.3 DISTANCE

There is always concern with extending the distance between two data equipment units that you are trying to connect by joining their interfaces together with a cable. The various interface standards shown in Figure 3, as well as any other interface standard, are usually defined for signalling at a known rate at a specific distance between the two data equipment units to be connected. That is, the voltage levels on the various pins are guaranteed to interchange data and control signals reliably, provided that the cable interconnecting the two interfaces together is no more than a certain length. As mentioned previously, for the EIA-232C interface, this distance can be at most 50 feet at 9600 Baud. The distance specifications for common interfaces are provided in Table 1.

Notes:
*CLOOP is Current Loop
**EIA-530 defines connector signal assignment, not voltage levels
Table 2
2.3.5 MEDIA CONVERSION

As we note in the next section, the common situation encountered will be where the premises data communications system is implemented with a single media type, usually Unshielded Twisted Pair (UTP) cable. However, you may come upon implementation problems where the system uses mixed media, some combination of UTP, coaxial cable and fiber optic cable. Mixed media usually comes about when the system evolves and specific segments of it may have to operate in a harsh environment such as a factory floor. The UTP with which the premises data communications system began its growth may not provide sufficient protection in such an environment from electromagnetic interference. The better shielding available with either Shielded Twisted Pair (STP), coaxial cable or fiber optic cable may be required.

In implementing such systems you may be faced with converting the signal modes propagating on one media type so that they are appropriate for another. Adapters for such conversions are readily available.

2.4 WIRING ADAPTERS

A variant on the problem of interface conversion is that of the plug-a-verter. Consider two data equipment units, again a computer and a printer, both of which have an EIA-232C interface. In this situation there is really no need to do interface conversion. Most people do not realize that data processing systems such as this often use the X-ON/X-OFF characters to control the flow of data. These systems only need three wires to transfer data between the two data devices. They do not need all of the control signals; that is, all of the pins provided by the EIA-232C interface and its bulky and expensive cable. The data can be transferred from the EIA-232C interface using a modular wiring adapter or Plug-A-Verter and skinny modular cable to connect the three wires as shown in Figure 7.



Figure 7: Plug-A-Verter

A Plug-A-Verter is simply a DB-25 to RJ-11 adapter. This is not as expensive and takes up much less space. If the two data equipment units require a more complicated handshake or control signal, for example a handshake in one direction, then you can go to the next level of complexity using four wires with a Plug-A-Verter. Using the example of the printer, the printer usually has a busy line which would be the fourth wire.

Another level of complexity is the situation where the data processing system needs a bi-directional handshake. This can be accomplished with a Plug-A-Verter with six wires. All of these Plug-A-Verter schemes are more attractive than using the bulky EIA-232C cable with the full interface. More complex systems can use eight wires. As a final point, note that Plug-A-Verters with the associated modular interfaces have all kinds of modular support equipment available on the market. These include wall plugs, wiring closets and multiple port harmonicas.

2.5 FACTORS INFLUENCING THE CHOICE OF AN INTERFACE CONVERTER

Given that you have to carry out an interface conversion of a certain type there are a number of factors which may influence your choice of a particular converter model. Below are some of these factors although the discussion is not all inclusive.

Chief among these is whether you want the conversion done by an external box, a card in a rack or a card which fits into a slot on the backplane of a computer. The most common converters are external boxes. But, cards are also available for certain situations. A card fitting into a slot on the backplane has the advantage of being able to use the computer’s electrical power source as its own.

You may also be concerned about the effects caused by ground loops and look for a converter which has optical isolation.

Finally, you may need a converter with surge suppression built into it or which has Tempest protection for a military application where low emissions are needed for security.

2.6 INTERFACE CONVERSION WITH WORK STATIONS

We have mentioned that the self contained work station usually has within it all of its data communication equipment needs. Nonetheless, there are cases where even these self contained work stations need interface converters.

One situation in which this often occurs is when some military agency procures, commercially available, off the shelf, work stations with EIA-232C interfaces. An unexpected need arises to output data from these units to some equipment using one of the military standard interfaces such as MIL-STD-188-144 Type 1. An interface converter then is obviously required.

Another case develops when a work station with only a parallel output (i.e. Centronics) needs to connect to a serial input device (e.g. a plotter) or when one with a serial output needs to connect to a parallel input printer. A bi-directional parallel/serial converter-buffer is then needed. This is an interface converter.

Table 4

With twisted pair there are four major sources of performance deterioration: noise, distortion, attenuation and crosstalk [Ref. 1]

Excess noise comes from two sources radio frequency interference (radio and television transmitters) and electromagnetic interference (fluorescent lights, arc welders, motors).

Distortion is caused by capacitance and increases with the cable length over which a signal is transmitted.

Excess noise coupled with significant attenuation is a serious problem with twisted pair cable. Long cable runs can realize a low gain antenna and "pick up noise." This coupled with excess attenuation can seriously degrade performance.

An interfering signal in a twisted pair that originates in another pair is termed "crosstalk." The number of cable twists per foot and the dielectric have a significant effect on the amount of crosstalk. The tighter the twists the lower the crosstalk and the more limited its degrading effect.

3.3 COAXIAL CABLE

Coaxial cable consists of two cylindrical conductors, one placed concentric within the other, but separated by an insulator. Figure 8 illustrates the anatomy of coaxial cable. As compared to twisted pair, coaxial cable has a much higher signal bandwidth. It can, as a consequence, support much higher data rates. When it is used in a single ended system, that is, when the outer conductor is grounded, there is an inherent shielding capability which unshielded twisted pair does not have. This shielding capability makes transmission immune to Electro-Magnetic Interference (EMI), Radio Frequency Interference (RFI) and other forms of interference. It provides more reliable communication but is generally more costly than twisted pair. On average it is four times more expensive. Also connecting devices to it (the termination problem) is much more difficult.


Figure 8: Detailed Illustration of Segment of Coaxial Cable

An excellent detailed discussion of the performance characteristics of coaxial cable is provided in [Ref.2]. Particular elements are summarized below.

As illustrated in Figure 8 coaxial cable is a class of cable that is best characterized as having several layers of material surrounding a common axis. A center conductor (solid or stranded) is surrounded by dielectric, or nonconductive, material and then shielded. This shield is often a wire braid or foil jacket, and is covered by an abrasion-resistant jacket. For outdoor applications, the spaces between the dielectric shield, and outer covering can be filled with an inert, waterproof gel for extra protection. This would also be done for buried cable. The dielectric material and outer jacket may be made of plenium or non-plenium material (often Teflon or polyvinyl chloride, respectively).

The cable thickness as well as the composition of the dielectric material, and construction techniques determine the signal properties of a coaxial cable. As with twisted pair, four parameters determine signal quality: characteristic impedance, mutual capacitance, attenuation and velocity of propagation. DC resistance is sometimes mentioned as a fifth parameter. Tables 5 and 6 [Ref.2] provide the electrical and physical characteristics of various types of coaxial cable.

Electrical Characteristics of Various Types of Coaxial Cable Obtained from Well Known Commercial Sources

Table 6

To get a feel for the meaning of these parameters observe Figure 9. It illustrates the combined effect of attenuation and capacitance on the transmitted signal at some distance from the source.


Figure 9: Illustration Showing Mutual Capacitance and Attenuation Significantly Corrupting a Transmitted Signal on Coaxial Cable

Determining the correct coaxial cable type for your application can be confusing. Several dozen types are available. The coaxial cable types most often encountered include RG-58, RG-58A, RG-58A/U, RG-58C/U, 802.3 thick and thin ETHERNET cable, RG-8, RG-8/U, RG-59,RG-62,RG-62A and RG-62A/U, all in polyvinyl chloride and plenum equivalents.

Many cable types, such as RG-58C/U, also meet military specifications, ("mil-specs") and may cost more due to the required certification testing.

Proper reference is made to [Ref. 2] as the original source of the figures and tables of this section.

3.4 FIBER OPTIC CABLE

Fiber optic cable consists of a strand of glass or plastic. A gradient in the refractive index across the diameter of this strand allows the cable to act as a waveguide for light. This provides the essential communication connectivity. Figure 10 [Ref.3] shows the details of the structure of fiber optic cable.

Fiber optic cable is a cylindrical structure made of a dielectric material transparent in the visible and near-infrared region of the spectrum. This structure consists of an inner region called the "core," where light essentially propagates. It is surrounded by a region called "cladding." This outer region has a refractive index smaller than that of the core. It is this constraining condition which allows light to confined in the core region.

The cladding also serves another function. It provides mechanical protection for the core. However, additional outer protection is usually needed for practical handling. Such additional protection typically consists of plastic material.

The basic material for the cladding is silica (SiO2). For increasing the core refractive index, germania (GeO2) is usually added to silica, while fluorine is typically added to decrease the silica refractive index, when needed.

There are many different fiber optic cable types. Fortunately, they can be conveniently grouped into three broad categories, step-index multimode fibers, graded-index multimode fibers and single mode fibers.


Figure 10: General Structure of Fiber Optic Cable

Step-index multimode fibers are characterized by a homogeneous core region with a constant refractive index. As to be expected they are surrounded by a cladding of lower refractive index. Step-index fibers were the first exceptionally low loss fibers to be developed. They have sizes of around 80 um for the core diameter, and 110 to 150 um for the cladding diameter. To date they have not found application in telecommunications outside the laboratory and have not been standardized.

Graded-index multimode fibers exhibit a refractive index profile n(r) (r is the radial coordinate with origin at the core axis). This features a gradual increase from a minimum value n(a) at the core-cladding boundary (generally coincident with the cladding value) to a maximum value of n(0), generally found at the core center. The main advantage of graded-index over step-index fibers is the greatly enhanced bandwidth. This is due to the fact that the speed of the different modes can be almost equalized by suitably shaping the index profile.

Single-mode fibers correspond to the core diameter of the optical fiber taking values comparable to the wavelength of the radiation to be propagated. Reaching this condition allows only the fundamental mode to be guided along the fiber. The most spectacular advantage of single-mode fibers is the increase of bandwidth due to the absence of modal dispersion. The bandwidth limiting factor is essentially constituted by chromatic dispersion which in ordinary step-index profiles has zero value in the region around 1.3 um. As a consequence, with suitably narrow line width sources, the total bandwidth might be two or three orders of magnitude greater than in multimode fibers. Single-mode fibers have other advantages: lower loss, great upgrade capability in view of future ultra-high bandwidth transmission, easy and accurate system design, and compatibility with integrated-optics devices. For these reasons at longer distances than those used in premises data communications single-mode fibers are the only fiber type used. Of course, multi-mode fibers are perfectly acceptable for premises distances.

Essentially, fiber optic cable has a much higher bandwidth than both coaxial cable and twisted pair. Therefore, it has the ability to support much higher data rates. As the demand for bandwidth increases and the cost comes down this will become the media of choice.

The nature of fiber optic cable, the glass or the plastic material from which it is constructed, essentially allows it to have complete immunity to EMI, RFI and other forms of extraneous signals. It is a highly appropriate transmission media for the heavy industrial environment where such forms of interference are a problem. Fiber optic cable is immune to signal leakage and is secure against eaves dropping. It has a much higher reliability than either coaxial or twisted pair. Since fiber optic cable is nonconductive its use eliminates ground loop problems. It also does not need protection from electrical surges or lightning.

Against these advantages it is 25 times more expensive (on average) than twisted pair. Terminating devices to its ends is much more difficult and requires a high degree of skill by the field technician. But, there is some light at the end of this tunnel (No pun intended). Increased demand and utilization in expectation for bandwidth needs is causing cost to come down. Technician handling costs and more convenient termination techniques are likewise arising.

Most fiber optic cable in use today is optimized for operation at either wavelengths of 820 nm or 1300 nm. Table 7 [Ref.4] indicates the attenuation losses of cables at these wavelengths in various situations.


Fiber Optic Cable Attenuation Losses at Premises Communications Distances

RESOURCE SHARING

Within the premises data communication environment economies can be gained by sharing resources. Such resource sharing may take a number of different forms. The communication capacity, that is the actual cable, can be shared by a number of different data sources or their applications. Communication and peripheral equipment can also be shared.

5.1 MULTIPLEXING

Let’s consider the problem of capacity sharing first. If you were to take a careful look at the actual cable connecting a single pair of communicating data equipments you would find that its signalling capacity, the rate at which it can transfer data, probably far exceeds the needs of the particular data processing application being run. To the point, of course the cable can be used to connect a single data equipment type to another isolated data equipment type. However, in general, the data transferred in a single application is far less than the actual capability of the wire. You can then conclude that the single cable need not be dedicated just to a single pair of communicating data units. Rather, it can be shared among a number of different users.

You may ask how may this sharing occur and the resultant cable cost reduction be achieved? It is really quite simple. Consider the top illustration, Figure 12. Here is a group of different terminals. Also shown is a multi-user computer to which the terminals must be connected. the computer may be somewhere else on the premises perhaps in the basement of the office building.

When the terminals are scattered throughout the building the most straightforward approach to connecting them to the computer is also probably (close to) the most economical. This is shown in the middle illustration of Figure 12. Here, each terminal has its own dedicated pair of wires connecting it to the computer. Short haul modems are used for signal boosting if required.

When the group of terminals is not scattered but clustered in the same office or floor there is a more economical alternative. A single cable pair has the capability to effect the transfer of the data from all the terminals. Consequently, the cluster of terminals can share the single cable bringing about a savings in cabling and equipment costs. A device called a short haul multiplexer is required to bring about the efficient and reliable sharing of the single cable. The bottom illustration of Figure 12 shows this. The cabling cost here is one-sixth that of the the situation shown in Figure 12B with the terminals at the same locations. Of course, one has to compare the cost of the short haul multiplexer with its significant intelligence with that of the distinct short haul modem pairs to determine the actual cost savings.


Figure 12A:Terminal Cluster Isolated from Multi-user Computer


Note:All Terminals Use Short Haul Modems to
Support Cable Lengths in Excess of 50 Feet
Figure 12B:Terminals in cluster each connected by dedicated cables to multi-user computer


Figure 12C:Terminals sharing a single cable to multi-user computer by multiplexing

Different short haul multiplexers manufactured by Telebyte are shown in Figure 13.

LIGHTNING AND .... SURGE PROTECTION

I have alluded to the possible problems caused by lightning and also by other electrical surges in the premises data communications environment. It is now the place to discuss these at more length.

Lightning has long fascinated the technical community. Ben Franklin studied lightning’s electrical nature over two centuries ago and Charles P. Steinmetz generated artificial lightning in his General Electric laboratory in the 1920’s. As someone concerned with premises data communications you need to worry about lightning. Here I will elaborate on why, where and when you should worry about lightning. I’ll then discuss how to get protection from it.

6.1 WHY WORRY ABOUT LIGHTNING?

It is unfortunate, but a fact of life, that computers, computer-related products and process control equipment found in premises data communications environments can be damaged by high-voltage surges and spikes. Such power surges and spikes are most often caused by lightning strikes. However, there are occasions when the surges and spikes result from any one of a variety of other causes. These causes may include direct contact with power/lightning circuits, static buildup on cables and components, high energy transients coupled into equipment from cables in close proximity, potential differences between grounds to which different equipments are connected, miswired systems and even human equipment users who have accumulated large static electricity charge build-ups on their clothing. In fact, electrostatic discharges from a person can produce peak Voltages up to 15 kV with currents of tens of Amperes in less than 10 microseconds.

A manufacturing environment is particularly susceptible to such surges because of the presence of motors and other high voltage equipment. The essential point to remember is, the effects of surges due to these other sources are no different than those due to lightning. Hence, protection from one will also protect from the other.

When a lightning-induced power surge is coupled into your computer equipment any one of a number of harmful events may occur.

Semiconductors are prevalent in such equipment. A lightning-induced surge will almost always surpass the voltage rating of these devices causing them to fail. Specifically, lightning-induced surges usually alter the electrical characteristics of semiconductor devices so that they no longer function effectively. In a few cases, a surge may destroy the semiconductor device. These are called "hard failures." Computer equipment having a hard failure will no longer function at all. It must be repaired with the resulting expense of "downtime" or the expense of a standby unit to take its place.

In several instances, a lightning-derived surge may destroy the printed traces in the printed circuit boards of the computer equipment also resulting in hard failures.

Along with the voltage source, lightning can cause a current surge and a resultant induced magnetic field. If the computer contains a magnetic disk then this interfering magnetic field might overwrite and destroy data stored in the disk. Furthermore, the aberrant magnetic field may energize the disk head when it should be quiescent. To you, the user, such behavior will be viewed as the "disk crashing."

Some computer equipment may have magnetic relays. The same aberrant magnetic fields which cause disk crashes may activate relays when they shouldn’t be activated, causing unpredictable, unacceptable performance.

Finally, there is the effect of lightning on program logic controllers (PLCs) which are found in the manufacturing environment. Many of these PLCs use programs stored in ROMs. A lightning-induced surge can alter the contents of the ROM causing aberrant operation by the PLC.

So these are some of the unhappy things which happen when a computer experiences lightning. But you may say, "Come on, equipment hit by lightning, that’s like winning the lottery. It has never happened and I doubt that it ever will." This is a typical reaction and unfortunately it is based on ignorance. True, people may never, or rarely, experience, direct lightning strikes on exposed, in-building cable feeding into their equipment. However, it is not uncommon to find computer equipment being fed by buried cable. In this environment, a lightning strike, even several miles away, can induce voltage/current surges which travel through the ground and induce surges along the cable, ultimately causing equipment failure. The equipment user is undoubtedly aware of these failures but usually does not relate them to the occurrence of lightning during thunderstorm activity since the user does not experience a direct strike.

In a way, such induced surges are analogous to chronic high blood pressure in a person; they are "silent killers." In the manufacturing environment, long cable runs are often found connecting sensors, PLCs and computers. These cables are particularly vulnerable to induced surges.

6.2 WHY WORRY ABOUT LIGHTNING?

This question primarily relates to the geographical location of computer equipment end-users. When other interfering phenomena which can cause a deterioration of performance is considered, it matters little where the equipment is geographically located.

Unfortunately, this is not the case with lightning. Life is unfair for computer users in certain regions of the United States. Take a look at the map shown in Figure 14.

Produced by the Electric Power Research Institute, the map denotes the mean annual days of thunderstorm activity for the continental United States. Upon examination, the map shows the high point of the thunderstorm activity as being in western Florida. It also shows a ripple effect out from this focus of lightning activity. In addition, it shows several areas outside of the southeast where intense thunderstorm activity perturbs the "dying ripple." Notice, for example, the intense areas of Colorado, New Mexico and Arizona. Also observe that the high level of the southeast extends all the way west to Texas.

There are a number of other maps of thunderstorm activity which you may find of interest. These include: 1) National Weather Service map showing average number of annual thunderstorm days by regions, 2) IEEE Working Group Report indicating Thunderstorm-Hour Data in various regions and 3) Annual Lightning flash contours. These maps are available from the author on request.

Examining the map shown in Figure 14 and the other referenced maps pretty much answers the original question. If you’re in Florida, lightning protection absolutely must be a concern. In fact, it should be throughout the whole southeast United States extending west to Texas and northwest to Missouri. The map shows where other intense pockets of activity occur. Apparently, only the Pacific coast can be "light hearted" about lightning as a problem.

For outside the continental United States the principal source for lightning statistics is a "world thunderday map" produced by the World Meteorological Organization of Geneva Switzerland. The map is very detailed but I have still produced it here in Figure 16. Yes, it is difficult to read without a magnifying glass. Call me if you want an enlarged copy. But, I will make a few observations.

Notice that outside the continental United States the places to worry about lightning are Mexico, Central and South America, Southern Africa and the southern Pacific Rim. These are all locations where manufacturing activities are on the rise and consequently, lightning will pose more of a problem than in the past.


Figure 15: Mean Annual Days of Thunderstorm Activity in the Continental United States


Figure 16: World Thunderday Map Produced By the World Meteorological Organization of Geneva Switzerland

One passing note, these maps show contours in terms of number of thunderdays. A more accurate characterization of lightning in a geographical region would be by the lightning flash density, the rate at which flashes occur. There may be many flashes on a given thunderday. It is flash density which is the real parameter of interest in choosing lightning protection. Unfortunately, for outside the continental United States there is even less quantitative data about lightning flashes than about thunderdays. It is possible to relate flash density to the number of thunderdays. In fact, there are a number of such relations. Most are of the form:

Ng = a Tb per km per km per year.

Here Ng is the ground flash density, T the annual thunderdays and a and b empirical constants. Typically, a = 0.1 -to- 0.2 and b=1.

6.3 WHEN SHOULD YOU WORRY ABOUT LIGHTNING?

This question really deals with two separate issues: 1) What is the seasonal variation of lightning and 2) What is the variation during a thunderstorm.

At present, the seasonal variation question can only be answered in detail for the continental United States. It is only here that significant measurements have been made and analyzed. The lightning season in the continental United States extends from April until October. Figure 16 illustrates this seasonal variation. The figure is based upon over 13.4 million cloud-to-ground lightning flashes as recorded by the National Lightning Detection Network. Notice the slow rise in the lightning rate in the Spring and the relatively fast decrease in the Fall during 1989.


Figure 17: Seasonal Variations of Lightning in 1989

Elsewhere in the world it is much more difficult to characterize the daily and seasonal behavior of lightning. Simply put, there have not been intensive efforts to date to collect the data. However, the situation is not completely vacuous. The availability of near Earth orbiting satellites has made it possible to map, in a limited way, worldwide lightning activity by detecting the electromagnetic (light and radio frequency) emissions of a discharge. From experiments reported and carried out by a number of distinguished scientists the following points have been made (under the caveat that they are still based on very limited data):

1) There is 1.4 times more lightning flashes during the summer in the Northern Hemisphere than in the Southern Hemisphere.

2) 37% of the global lightning activity originates over the ocean at dawn and 15% at dusk.

3) The global lightning flash rate is estimated at 64 per second for the Northern Hemisphere Spring, 55 per second for the Northern Hemisphere Summer, 80 per second for the Northern Hemisphere Fall and 55 per second for the Northern Hemisphere Winter.

When do you have to worry during a thunderstorm? Typically, thunderstorms are characterized as intense individual rain cells or showers embedded in a broad area of light rain. These intense cells are only over a fixed location for a few minutes. They are on average, several miles in each direction. In the continental United States thunderstorm cells move from west to east along a squall line as shown in Figure 17. This squall line is about 12-30 miles in width and up to 1,250 miles long. The speed at which the thunderstorm cell moves is generally 30 knots (approximately 34.4 statute miles per hour).

Remember when I raised the question about the possible effects of induced voltage/current surges from lightning striking several miles from computer equipment. Taking this into account, you may believe that a thunderstorm has passed by and the vulnerability to damage has ceased. Actually, a moving thunderstorm having passed, but striking several miles away, may still cause damage. The time interval of vulnerability may be tens of minutes depending upon whether the cell is moving directly along the squall line or obliquely to it.


Figure 18: Thunderstorm Cell Movement

6.4 WHEN SHOULD YOU WORRY ABOUT LIGHTNING?

Coming right down to it, a lot can be done as far as protection is concerned. However, it is best to begin by describing the magnitude of the threat from which you need protection.

The first stroke of lightning during a thunderstorm can produce peak currents ranging from 1,000 to 100,000 Amperes with rise times of 1 microsecond. It is hard to conceive of, let alone protect against, such enormous magnitudes. Fortunately, such threats only apply to direct hits on overhead lines. Hopefully, this is a rare phenomenon.

More common is the induced surge on a buried cable. In one test, lightning-induced voltages caused by strokes in ground flashes at distances of about 5 km were measured at both ends of a 448 meter long, unenergized power distribution line. Typical test results are illustrated in Figure 18. Notice that the maximum-induced surge exceeds 80 Volts peak-to-peak. This is more than enough to destroy semiconductor devices and computer related equipment. Yet, 80 Volts is well within the range of affordable protection.

Conceptually, lightning protection devices are switches to ground. Once a threatening surge is detected, a lightning protection device grounds the incoming signal connection point of the equipment being protected. Thus, redirecting the threatening surge on a path-of-least resistance (impedance) to ground where it is absorbed.

Any lightning protection device must be composed of two "subsystems," a switch which is essentially some type of switching circuitry and a good ground connection-to allow dissipation of the surge energy. The switch, of course, dominates the design and the cost. Yet, the need for a good ground connection can not be emphasized enough. Computer equipment has been damaged by lightning, not because of the absence of a protection device, but because inadequate attention was paid to grounding the device properly.

The basic elements used as protective switches are: gas tubes, metal oxide varistors and silicon avalanche diodes (transorbs). Each has certain advantages and disadvantages.

Because they can withstand many kilovolts and hundreds of Amperes, gas tubes have traditionally been used to suppress lightning surges on telecommunications lines. This is just what is needed to protect against a direct strike. Because gas tubes have a relatively slow response time, this slowness lets enough energy to pass to destroy typical solid state circuits.


Figure 19: Measured Induced Voltage

Metal oxide varistors (MOVs) provide an improvement over the response time problem of gas tubes. But, operational life is a drawback. MOVs protection characteristic decays and fails completely when subjected to prolonged over voltages.

Silicon avalanche diodes have proven to be the most effective means of protecting computer equipment against over voltage transients. Silicon avalanche diodes are able to withstand thousands of high voltage, high current and transient surges without failure. While they can not deal with the surge peaks that gas tubes can, silicon avalanche diodes do provide the fastest response time.

Thus, depending upon the principal threat being protected against, devices can be found employing gas tubes, MOVs, or silicon avalanche diodes. This may be awkward, since the threat is never really known in advance. Ideally, the protection device selected should be robust, using all three basic circuit breaker elements. The architecture of such as device is illustrated in Figure 19. This indicates triple stage protection and incorporates gas tubes, MOVs and silicon avalanche diodes as well as various coupling components and a good ground.

With the architecture shown in Figure 19 a lightning strike surge will travel, along the line until it reaches a gas tube. The gas tube dumps extremely high amounts of surge energy directly to earth ground. However, the surge rises very rapidly and the gas tube needs several microseconds to fire.

As a consequence, a delay element is used to slow the propagation of the leading edge wavefront, thereby maximizing the effect of the gas tube. For a 90 Volt gas tube, the rapid rise of the surge will result in its firing at about 650 Volts. The delayed surge pulse, now of reduced amplitude, is impressed on the avalanche diode which responds in about one nanosecond or less and can dissipate 1,500 Watts while limiting the voltage to 18 Volts for EIA-232 circuits. This 18 Volt level is then resistively coupled to the MOV which clamps to 27 Volts. The MOV is additional protection if the avalanche diode capability is exceeded.

The robust structure shown in Figure 19 is embodied in Telebyte’s Model 22. This is shown in Figure 20. The Model 22 is designed to protect 4-wire lines as used in short haul modems operating at speeds up to 38.4 kBaud. Table 9 indicates other Telebyte lightning protection products.

As previously mentioned, the connection to earth ground can not be over emphasized. The best earth ground is undoubtedly a cold water pipe. However, other pipes and building power grounds can also be used. While cold water pipes are good candidates you should even be careful here. A plumber may replace sections of corroden metal pipe with plastic. This would render the pipe useless as a ground.


Figure 20: Architecture of a Robust Protection Device


Figure 21: The Telebyte Model 22 Lightning Protection Device

Lightning and Surge Protector Selection Charts

TEST EQUIPMENT

Test equipment that the user requires in the premises data communications environment varies from very simple devices to extremely complicated devices.

7.1 LOW COST DATA LINE ANALYZER

At the low end are simple line monitors referred to by a variety of different names. Telebyte calls its data line analyzer the Model 301 which is shown in Figure 21. This typical low end unit is made up of two components. It includes a data line monitor, the Model 43 Micropeeper, that allows the user to determine the actual status of the seven key signals of the RS-232 data path. All of the 25 signal leads are usually passed through with the seven key signals TD, RD, RTS, CTS, DSR, DCD and DTR being monitored and their status indicated on a set of dual colored LED’s. The LED’s show whether or not the EIA-232 pin is high, low or has no signal. The Model 301 also includes the Model 151 Mini-Patch Box to allow altering the connections in the data path.

7.2 DATA ANALYSIS - PC

A step up from such low end units are add-on cards for the PC, which allow the personal computer to perform as a full featured line monitor. These cards plug into a PC and convert it to a test instrument. They typically have the capability of viewing the bi-directional data and the controls signals of any EIA-232 communication link. The PC display allows visual presentation of the data on a serial line. It allows the operator to understand, quickly, the relationships between transmit data, receive data and the various control leads such as Request-to-Send and Clear-to-Send.

An example of such a test unit is Telebyte’s Model 903 PC Comscope shown in Figure 22. The Model 903 consists of a plug in card and a T cable, which allows the PC to tap into and capture the data flowing between data equipment devices. The PC Comscope allows the protocol that is used in the data transmission to be verified. It enables maintenance personnel easily to identify open data leads, missing clock signals, parity errors, incomplete handshakes, etc. It also allows you to emulate protocols such as BISYNC, X.25 and HDLC. Bit Error Rate Test (BERT) capability is also provided by the unit. This test device effectively lets you design your own protocol analyzer. The Model 903 operates in PC AT/386/486 machines or compatibles including lap tops.

While test instruments like the Model 903 are convenient they do have a significant limitation. They depend upon a spare slot being available in the PC platform which is their home.


Figure 22: Low Cost Data Line Analyzer


Figure 23: Model 903 PC Comscope

With the proliferation of add-on boards a slot for the Model 903 may not always be available. This is especially a problem if one wants to use a notebook PC. Telebyte addressed this issue by developing the Model 904 PC Notebook Comscope Protocol Analyzer. It is a small portable unit which is external to the PC but works with it via its serial port. The Model 904 is illustrated in Figure 23.


Figure 24: Model 904 PC Notebook Comscope Protocol Analyzer

7.3 PROTOCOL ANALYZER

Beyond this are datascopes which both monitor and emulate data sources. These allow quality assurance of communication circuits by doing a variety of tests. Datascopes are not usually put in as part of the PC but stand-alone with their own keyboards and screens. Depending on what the user is willing to pay he can have increasing speed and more and more features. Telebyte’s Model 901 Netscope was an example of such a tester. A photograph of the Model 901 is shown in Figure 24. However, such test units have lost ground to the PC based equipment.

7.4 FUTURE TRENDS

It is always dangerous to predict the future. Nonetheless, at present, we appear to be in the midst of a revolution with respect to datacom test equipment and certain trends are evident. The PC has become ubiquitous. As a result, field service technicians can generally count on a PC being present and available for their use no matter where they are sent. The technician need only put a card like the Model 903 in his attache case when going into the field and then commandeer an available PC. This we see as the "new portability."

Where does this leave the stand alone test units like the Model 901 Netscope? While the Model 901 was retired other stand alone types will not disappear. Rather, they will increase in capabilities especially in the areas of display and archival memory. The increase in capabilities will be matched by a high price. Their use will thus be limited to the laboratory rather than as a portable tool for the traveling technician.


Figure 25: Model 901 Netscope

LOCAL AREA NETWORKS

The subject of this book is premises data communications. I noted in Chapter 3 that there were three architectural alternatives to solving the connection problem posed in premises data communications. To this point our focus has been on only one of these approaches - the empirical techniques use to connect terminals to a 286/386 PC running UNIX/XENIX or some equivalent system. Another alternative is to use a Local Area Network (LAN). The employment of LANs will now be discussed in more detail.

8.1 WHY USE LANS?

Previously I made the following statement with regards to a LAN. "This alternative is the approach of choice when one has the freedom and the money to design their premises data communication system starting essentially with a `blank page’." This is a rather glib statement. The question arises as to why it is true?

Go back and take a look at the typical multi-user PC environment with its empirical connection solutions which was shown in Figure 2. Now take a look at the general pictorial of a LAN which is illustrated in Figure 25. The "cloud" represents the connecting media and all supporting protocols. There are many advantages which a LAN has over the data communications approach of Figure 2. I describe several below.

To begin with in Figure 2 one is only dealing with communications between the terminals and the multi-user computer. If there are printers or plotters or other equipment present they are essentially part of another data communications system with the multi-user PC at its center. In contrast as shown in Figure 25 a LAN allows full connectivity of all the data equipments connected to it. The terminals communicate with the computer through it. The computer can communicate with the printer or plotter through it. The terminals can even communicate with each other and with the printer or plotter through it. There is much more flexibility.

This advantage of a LAN in having full connectivity is often described in the context of it allowing peer-to-peer resource sharing.

The typical (UNIX-based) environment shown in Figure 2 does not have peer-to-peer capabilities. Rather, they designate the computers either as "clients" or "servers." A "client" is usually a computer at which a (human) end-user works. A "server" is a computer whose resources (e.g. printers, mass-storage devices, data files, program files, applications software) are available to all (human) end-users. Servers are often more powerful computers than the client computers. In the Figure 2 environment without peer-to-peer resource sharing the computer resources of a client are only available to the (human) end-user stationed at the client.


Figure 26: Data Equipments Connected By a LAN

However, a LAN has peer-to-peer resource sharing. All networked computers perform both client and server functions. This is the true meaning of full connectivity. It yields one of the major benefits which is file sharing among all the computers.

The multi-user computer system of Figure 2 will generally only make use of a data link level protocol to effect error control. LANs make use of packetization along with the higher levels of the OSI protocol stack. This leads to better error control as viewed from the user interfacing the application.

LANs are inherently faster. Data communications can be carried out at megabit per second (and above) speeds. They can support those applications which are becoming more and more common which need such speeds. Nowhere is this more evident than in the use if computer application packages using high resolution graphics. The multi-user computer system of Figure 2 with its 10’s of KBPS connections to terminals is a "time sharing system." Historically, it handled graphics by using "bit mapping." This was a relatively slow, cumbersome procedure for the UNIX/XENIX software and as a result graphics screens were painted "very slowly." Essentially, it was inappropriate for graphics. Recently, it has been improved by the development of the X-Terminals protocol. This replaces "bit mapping" with the transmission of "drawing descriptives." This makes the multi-user computer somewhat more appropriate for high resolution graphics. Nonetheless, the higher speed LANS have a significant advantage here.

8.2 LAN ARCHITECTURES

Work on the development of LANs began in the 1960’s. Since that time many different LAN architectures have been proposed, studied and implemented. However, at the present time only two architectures have survived to be of general interest, the ETHERNET BUS ARCHITECTURE and the TOKEN RING ARCHITECTURE. Neither of these has a clear advantage over the other. It is for that reason that both have been adopted as standards by the IEEE 802 committee. Other architectures may be of interest in special situations but these two are the only ones worth discussing of general interest. We do this below. Details are taken from [Ref.5] and [Ref. 6].

8.2.1 ETHERNET BUS ARCHITECTURE

This architecture had its origins in work done at XEROX’s Palo Alto Research Center (PARC) by Robert Metcalf in the early 1970’s. XEROX was later joined by DEC and INTEL in promoting ETHERNET as the coming LAN standard. Basically, Metcalf built upon work done by ARPA (now DARPA) and the University of Hawaii with a multiple access technique called ALOHA.

We can explain the operation of ETHERNET briefly with the aid of Figure 26. Here we see the data equipments which need to communicate all tapping onto transmission media (a cable) which we have labeled "Broadcast Channel-The ETHERNET BUS." The Bus Interface Units (BIUs) provide the essential interfacing between the data equipments and this channel-that is the transmit/receive capability and all needed intelligence. It is an essential feature of ETHERNET that by using the Broadcast Channel any data equipment can transmit to any other data equipment and any data equipment can listen to all transmissions on the channel whether intended for it or for some other data equipment user.

Now how does ETHERNET operate? It makes use of CSMA/CD - this is Carrier Sense Multiple Access/Collision Detection. The ETHERNET BUS - the connecting cable - is passive and can be used for broadcast type transmissions. Consider a specific data equipment unit, a terminal wanting to communicate with the computer. The terminal’s BIU before attempting to transmit a data packet onto the ETHERNET BUS first "listens" to determine if the BUS is idle-that is if there are no other packets from other data equipment units on the BUS. It senses the presence of a carrier on the BUS. An active BIU transmits its packet onto the BUS only if the BUS has been sensed idle. If the BUS is sensed busy then the BIU defers its transmission until the BUS becomes idle again. Due to propagation delays and carrier detection time, a collision may occur when a BIU senses an idle BUS and begins to transmit its packet while another data equipment’s BIU has already started transmitting a packet that has not yet propagated to this BIU. All BIU’s connected to the BUS have some means for "Collision Detection." When a collision occurs, all parties involved cease transmission and wait a random amount of time before initiating retransmission. If collision occurs again this random time wait is repeated but increased and increases at an exponential rate until the collision event disappears. This approach is called " exponential back off."


Figure 27: Ethernet Bus Architecture

8.2.2 TOKEN RING ARCHITECTURE

This architecture had its origins in work done in Great Britain. Its first known use in the United States was by Prime Computer in 1977. Since that time IBM has adapted this as its preferred LAN architecture.

The operation of the Token Ring Architecture can be explained with the help of Figure 27. In contrast to the ETHERNET BUS the Token Ring structure is a concatenation of point-to-point communications links arranged in a closed loop. Each link is terminated with an active repeater that detects a data packet on the in-bound link and re-transmits it on the outbound link. The detection, regeneration and all intelligence are carried out by the Ring Interface Unit (RIU) which interfaces the data equipment unit to the Ring. Basically, the ring transmission uses a token-passing access scheme. A special packet structure is called the "idle token." This circulates around the ring. When a connected data equipment user wants to transmit a data packet to some other user it may grab the "idle token," change it to a "busy state" and append its data packet to the busy token. At the end of the packet transmission, the data equipment unit issues another idle token. The Token Ring Architecture essentially behaves as a polling system. The most significant factor affecting the ring performance is the ring propagation time, the processing time for token recognition and regeneration at each RIU.


Figure 28: Token Ring Architecture

Ring structure is a concatenation of point-to-point communications links arranged in a closed loop. Each link is terminated with an active repeater that detects a data packet on the in-bound link and re-transmits it on the outbound link. The detection, regeneration and all intelligence are carried out by the Ring Interface Unit (RIU) which interfaces the data equipment unit to the Ring. Basically, the ring transmission uses a token-passing access scheme. A special packet structure is called the "idle token." This circulates around the ring. When a connected data equipment user wants to transmit a data packet to some other user it may grab the "idle token," change it to a "busy state" and append its data packet to the busy token. At the end of the packet transmission, the data equipment unit issues another idle token. The Token Ring Architecture essentially behaves as a polling system. The most significant factor affecting the ring performance is the ring propagation time, the processing time for token recognition and regeneration at each RIU.

8.2.3 COMPARING ETHERNET AND TOKEN RING

I have said that neither of these architectures has a clear advantage over the other. With ETHERNET the data equipment units do not have to be synchronized, they do not have to control their transmissions. This is a big advantage in implementation. It comes at the cost of collisions. But in lightly loaded networks collisions do not occur that often. Luckily, this is the situation most of the time. In heavily loaded situations, of course, long delays may occur and ETHERNET may degrade significantly. But these heavily loaded situations are almost never seen in the commercial environment.

With Token Ring transmissions are controlled. What is more, the maximum value of delay can be controlled by limiting the amount of time that a data equipment unit can take the idle token out of circulation. One never needs to suffer the infrequent long delays due to collisions that ETHERNET may experience.

Thus, in comparing the two architectures one needs to know where such infrequent long delays are tolerable and where they are not. They most likely are quite tolerable in the ordinary office environment where no catastrophe will result from the delay-just annoyance. Here, we would expect ETHERNET to be quite popular. On the otherhand if the LAN were being used in a factory environment to support process control or some type of automation an occasional long delay may well be catastrophic. Here, we would expect Token Ring to be the preferred architecture.

8.3 ETHERNET IMPLEMENTATION

I have noted that both ETHERNET and Token Ring are the LAN architectures which are mostly employed. Yet, to the present date ETHERNET appears to dominate as measured strictly by the amount of ETHERNET equipment in use. This dominance is expected to continue. It will be worth while to describe the evolution of ETHERNET implementations.

The first implementations of ETHERNET employed thick coaxial cable. Termed Thicknet it was defined by the 10 Base 5 standard. Figure 28 provides a pictorial illustration of this type of ETHERNET implementation. By comparing it to Figure 26 you can make the correspondence between the broadcast channel-the ETHERNET BUS and the two coaxial cable segments which are shown as vertical "pipes." Short segments of these pipes are joined by barrel connectors. Long jumps between these pipes required repeaters. The BIU function was carried out by a transceiver which was also referred to as the Media Access Unit (MAU). The Attachment Unit Interface (AUI) connected the data equipment to the MAU.

Thicknet delivered data across the network with a worst case bit error rate of 1 in a 100 million. A LAN implemented by Thicknet was also quite extensive. The maximum network cable length could exceed 8,000 feet. Implementation with Thicknet did require an external transceiver for signaling.


Figure 29: Thick Ethernet Network

Unfortunately, the thick coaxial cable was difficult to work with. As a result the second wave implementations of ETHERNET employed thin coaxial cable. Termed Thinnet it was defined by the 10 Base 2 standard. Figure 29 gives a pictorial illustration of this type of ETHERNET implementation. Thinnet had a bit error rate degraded somewhat relative to Thicknet. It also was not as extensive. The maximum network cable length was reduced to be of the order of 3,000 feet. Implementation with Thinnet did not require an external transceiver for signaling. Notice the absence of these in Figure 29.

The Thinnet realization of ETHERNET has recently given way to the use of Unshielded Twisted Pair cable (UTP) under the 10 Base T standard. The use of UTP has given ETHERNET deployments a new burst of growth. The cable used for the connecting medium is the same as that generally (although not always) used as the telephone cable in office buildings for the last 20 years. Thus, ETHERNET under 10 Base T can make use of an existing wiring plant and the task and cost of pulling new cable can be avoided.

ETHERNET under 10 Base T delivers data at the same worst case bit error rate of Thicknet, 1 in 100 million. However, a LAN implemented under 10 Base T has significant distance constraints.The UTP cable length must be of the order of 300 feet at its maximum. Implementation with 10 Base T does not require an external transceiver for signaling unless the adapter card is not 10 Base T compatible.


Figure 30: Thin Ethernet Network

The topology of an ETHERNET LAN under 10 Base T is also quite different from that of Thicknet and Thinnet. This topology is illustrated in Figure 30 taken from [Ref. 7]. The basic unit of a 10 Base T implementation is called the "Work Group." Data equipments are tied separately by UTP to a Multipoint Repeater or hub. This is a star topology, quite a bit different than the topologies for Thicknet or Thinnet. In fact, it is similar to the Multi-user PC Environment topology of Figure 2. Nonetheless, from a logical point of view this Work Group star topology has the required "Broadcast - ETHERNET BUS" property need for operation. Any data equipment unit can transmit to any other and also "listen" to all transmissions.

While ETHERNET under 10 Base T has the convenience of using UTP it also has the "annoyance" of having to employ 2 wire pairs to tie each data equipment unit to the hub. One pair is needed for transmission and another pair for reception.

The severe distance limitations on the Work Group distance also pose a problem. However, this can be solved by tying Work Groups together. This is indicated at the top of Figure 30. Depending upon the distance, Work Groups can be connected by either 10 Base T, 10 Base 2, 10 Base 5 or even a fiber optic cable.


Figure 31: Ethernet Operating As A 10 Base T Work Group

Figure 31 illustrates the components needed to connect to an ETHERNET LAN from the point of view of a data equipment unit "user." At the bottom of the figure is the applications software residing in the data equipment unit. This generates and is the ultimate recipient of the data to be communicated across the LAN. "On top of this" is the LAN operating system. Novell is probably the most popular example of this at present. The operating system carries out the packetization of data, complete with addressing. It establishes sessions between communicating data equipments. When collisions are sensed it handles the generation of exponential backoff procedures. "Above" the operating system is the adapter card. This carries out the CSMA/CD procedures. If it detects collisions it passes this information down to the operating system. With Thicknet only an external transceiver or MAU is also required as shown.


Figure 32: Connecting To An Ethernet LAN

References

---

[REF 1]
Leeds, Frank and Chorey,Jim;"Twisted-Pair Wiring Made Simple," LAN Technology, April 1991, pp 49-61.

[REF 2]
Leeds, Frank and Chorey, Jim; "Cutting Cable Confusion: The Facts About Coax," LAN Technology, March 1991, pp 31-40.

[REF 3]
Costa,B.,"Fiber Optic Communications Handbook,Second Edition," Chapter 1 pp 2-31, TAB Books, Summit,Penna,1990.

[REF 4]
Head, Joe,"Fiber Optics in the ‘90s’: Fact and Fiction," Data Communications,Magazine, September 21,1990, pp 55-57.

[REF 5]
Tsao,C.David, "Performance Comparisons of LANs," in "Proceedings of the 1985 Symposium New Developments in Local Area Networks," IEEE Long Island Section, October 4, 1985.

[REF 6]
Hawe,Bill and Lauck, Anthony,"Datalink and Network Layer Interconnection of LANs," in "Proceedings of the 1985 Symposium New Developments in Local Area Networks,"IEEE Long Island Section, October 4,1985.

[REF 7]
Anderson,Rick and Woods, Kevin."10 Base-T Ethernet: The Second Wave," Data Communications,pp 49-64, November 21, 1990.

---