Transmission paths
Jul 1, 1999 12:00 PM,
R. David Read
As everyday participants in the design, selection, purchase andinstallation of wire and cable for low-voltage usage, we need to becognizant of the technological factors associated with the means whereby wetransmit signals from sources to outputs. Wire and cable are notinterchangeable commodity items that are unworthy of serious consideration.The low-cost solution might not prove to be the most equitable choice.Today’s primary choices for the low-voltage contractor are traditionalcopper wire and multiple-conductor cable, coaxial cable and fiber opticcable. We will attempt to cover the tradeoffs in these transmission methodsand the various characteristics that affect the selection of the medium forthe appropriate communication path.
Traditionally, copper wire and cable have been the predominant choice wheninstalling such low-voltage, low-frequency analog networks as audio,security, fire alarm and telephone circuits. Not many years ago (circa1960), communication wiring networks were relatively simple. Connectionbetween occupants of a building and the outside world was largely afunction of the telephone and was controlled in most parts of the UnitedStates and Canada by AT&T and the local Bell Telephone Company. Fire alarmwiring was a 60 ma, 24 V system that connected pull-stations with alarmpanels and activated signaling devices. Internal, commercial andinstitutional intercoms were interconnected via large multiple-conductorcables. CCTV was in its infancy, and data transmission was the stuff ofscience-fiction magazines. Today, we specify, design, install and maintaincable plants that would have seemed unfathomable to a communications plantdesigner in 1960.
Coaxial cable has been around since the advent of video transmission. Inthe 1960s, however, parallel, 600 V twin lead between the off-the-airantenna and the receiver was the standard in the majority of residentialand commercial installations. Coax was a province of video andhigh-frequency transmission line installations. This took the form ofantenna transmission lines connecting radio frequency transmitters withtheir respective antennas.
Microwave and metallic tube wave-guide were largely confined to militaryand government installations. When I left the railroad communication sceneabout that time, it, along with some of the major oil companies and thetelephone companies, was just beginning to implement microwave to replaceits difficult-to-maintain, open-wire and cable networks. The collectiveknowledge of microwave technology was pretty sparse. When we implementedSanta Fe’s microwave link between Winslow, AZ, and Barstow, CA, we went togreat lengths to ensure that Morse code signals could be accommodated overthe microwave channel. This turn of events was a necessity because therewere several branch line stations south of Ashfork, AZ, whose sole contactwith the rest of the railroad was via Morse telegraphy. Fiber optic was aphysics lab curiosity at best. It would be another 30 years before thismedium could even be considered as a practical means for transmittingcommunication signals.
Cable characteristicsFor the purpose of this discussion, only those types of cable normallyencountered in the low-voltage sound and communications market will becovered. Hence, specialized mediums used for broadcasttransmission-hardline, Helix, microwave waveguide and other similarproducts-are omitted.
All transmission media have the properties of resistance-the property thatopposes the flow of DC (measured in ohms) as expressed by the R = E/I whereR is resistance; E is DC voltage, and I is DC current. Impedance-theproperty that opposes the flow of AC current as a combined function of theresistance, inductance and capacitance of the medium at some definedfrequency (measured in ohms) is expressed as Z = E/I where Z is impedance;E is AC sine wave voltage, and I is AC current.
Because all metallic wire and cable possess some degree of inductance andcapacitance, these two properties combine to form the reactance of thecircuit (XL inductive) and (Xc capacitive). These reactive characteristicsthen combine with the resistance of the circuit, which, in turn,establishes the impedance. Being frequency dependent, impedance must becalculated using frequency as a factor. Also, the values of resistance andreactances cannot be merely added together; they must be combinedvectorially.
It is beyond the scope of this article to explain how to calculateimpedances. It should be kept in mind, however, that in the transmission ofAC signals, the reactive elements will cause some degree of phase shift. Asfrequencies increase and data speed rises, severe distortion in datasignals can occur. This becomes a significant performance factor whenhigh-speed digital data is transmitted. When a manufacturer of data-basedtransmission equipment specifies that the cable used for the installationshould not exceed a given capacitance per unit of length, it is notsomething you can ignore. If this parameter is violated, you can expect anincrease in the number of data error rates. Capacitive values of cable aregenerally stated by cable manufacturers in terms of pf (picofarads) perfoot (or meter).
Fiber optic cable avoids this problem. The transmission medium is opticalrather than metallic. This is not to say there are not transmission lossesin fiber, but this has to do with properties unrelated to electrical signaltransfer.
The National Electrical Code (NEC)One of the first steps when specifying or selecting cable for a givenproject is to consider the ramifications of the NEC. Considering that theNEC manual runs about 1,069 pages, and the attendant handbook, whichattempts to demystify the code, takes 1,016 pages, it is no wonder there isa fair amount of confusion when it comes to the proper selection andapplication of wire and cable.
The NEC is a voluntary guideline publication, but if the local governingboard adopts the provisions of the NEC, then the NEC is the law within thatjurisdiction. Also, in certain sectors of the country, code authorities mayhave adopted regulations that are more stringent than those contained inthe NEC. You would be well advised to check out your local buildinginspection department, the fire marshal’s office and any other agencieshaving jurisdiction before you start pulling wire.
The NEC does not address the performance of cable, nor does it passjudgement or test on matters that are not directly associated withflammability and emission of toxic fumes when wire and cable are subjectedto fire or how they might react to corresponding safety issues. Virtuallyall communication cables fall within five NEC ratings. Unrated cables arenot tested to NEC specifications, and unless installed in conduit, thesewill not pass code. Residential rated cables are for use in residentialbuildings only-romex and telephone quad cable, for example. Commercialrated cables can pass through a wall without conduit but not verticallybetween floors or in drop ceilings that are a part of any air recyclingsystem. Riser rated cable is acceptable for any of the above applicationsand for installation in vertical shafts between floors in a commercialbuilding. Plenum rated Cable is designated for use in any of the above, andit is applicable for installation in plenum areas, any hidden area that isused as a return-air handling space for the building’s air-conditioningsystem. Beware, many inspectors may consider all drop ceilings or raisedfloors as plenum spaces, even though they may not handle return air. Youcan fight such interpretations, but you stand a good chance of losing.
When the NEC first established the conditions for plenum cable, the onlycable jacketing material that could meet the flammability requirement wasFEP Teflon. Because Teflon is expensive and difficult to work with, variouscable manufacturers have developed other proprietary alternatives, butthere is still a tendency on the part of designers and inspectors to equateTeflon with plenum. Actually, any cable marked as CMP, CL2P, CL3P, FPLP orCATVP is an acceptable substitute for Teflon, regardless of theconstruction material. Teflon is plenum-rated, but not all plenum-ratedcable needs to be Teflon. In cases where the cable must continue tofunction at high temperature, it may be necessary to use Teflon. Teflon hasa meltdown of 200degreesC (392degrees F); other acceptable plenum cableusing PVC formulations fail at 75degreesC (167degreesF).
To meet code, all installed cables must be clearly identified as to theirelectrical and flammability characteristic class. Normally, thisinformation is printed directly on the jacket of the cable. Manufacturersof Teflon cable print the identification on the inner jacket and thenenclose it in a layer of Teflon. Recently, some cable suppliers have begunproviding additional information printed on the jacket. This might includethe number of feet (meters) from the start of the run, special color codingand identification of the applicable use for the particular cable at hand.
The NEC manual is divided into a number of articles. These roughlycorrespond to chapters in a book. The articles that have particular contentthat bears consideration by low-voltage, communication system contractorsare summarized in Table 1. Attention should also be given to Articles 640and 645. Article 640 covers sound recording, sound reinforcement andsimilar audio systems. Article 645 addresses the installation ofinformation technology equipment, which was adopted by the NEC 1996 codeand covers data equipment, computer networks and control systems.
Physical properties of cable constructionSingle-conductor wire and multi-conductor cables are constructed of someconductive metal, the most common being copper. Copper is not the mostconductive material available, but it is the best compromise among cost,strength and flexibility. Conductors are classified in accordance withtheir cross-sectional diameter in terms of wire gauge. In North America,the standard is AWG (American Wire Gauge).
Conductors also are available in two versions-solid wire (one solidconductor) or stranded (a collection of smaller wires bundled into a largerconductor). For example, an AWG#20 stranded wire may be constructed bycombining seven strands of number AWG#28 or 19 strands of number AWG#32.Either method will produce a corresponding AWG#20 conductor. For anincreased price, the higher strand method will produce a cable with ahigher flexibility and flex-life.
With fixed installations, where the cable will not be subject to frequenthandling or movement (mic cable or outside drop-wire installation, forexample), flex-life is probably inconsequential. If your installationinvolves the use of mic cables that will be subject to frequent handling,flex-life and flexibility are a matter of primary concern. In cableterminology, the two terms-flex-life and flexibility-are notinterchangeable. Flex-life refers to the number of times the cable may bebent before the conductors will fail; flexibility refers to the somewhatsubjective properties describing the ease with which the cable can behandled. In particular, certain mic cables may be considered flexible anduser-friendly; others may be classified as stiff and hard to handle. Thosecables with more strands and with tighter twists per unit of length willdisplay better flexibility and flex-life. This is also a function of theshielding used and the jacket composition, which will be discussed later.
The conductivity of the conductor material and the gauge of the wire willdetermine the DC resistance of the medium. Because most communication cableis constructed of copper, which has a mil-ohm per foot value of 10.4 V, theprincipal contributing factor of resistance is the gauge of the conductor.As the wire diameter increases, the resistance of the conductor decreases.This can produce a dramatic effect in such circuits as loudspeaker linesconnected to an amp. For example, you can expect to see a 3 dB(50%) signalloss when driving a 4 V loudspeaker load over 85 feet (26 m) with atransmission line of AWG#24 cable; this 50% loss factor will not beencountered until a similar 1,150 feet (350.5 m) of number AWG#12 cable isput in place. The cross-sectional area of the conductor (the gauge) has adirect relationship to the DC resistance of the cable and the subsequentsignal losses. These adverse effects can be mitigated by usingimpedance-matching transformers.
ShieldingCommunication cables designed for audio, video and data are often providedwith a shield to offset the adverse effects of EMI (electromagnetic) and/orRFI (radio frequency) interference. Cable shields are constructed usingcopper, tinned wire or aluminum foil surrounding the signal carryingconductors. Cable shielding is provided in a variety of methods andcorresponding prices. The application of the cable and the degree ofshielding required will dictate the selection and the subsequent cost ofachieving the desired results. When specifying or purchasing the type ofshielded cable required, consider the characteristics shown on Table 2.
In Table 2, it is assumed that the shielding characteristics are applicableto two single conductors in a balanced transmission circuit. Circuits thatuse the shield as a signal carrying conductor (unbalanced or high-impedancetransmission lines) compromise the shielding effect of EMI and RFIinterference. The effective length of such an unbalanced audio circuit is,at best, 25 feet (7.6 m) before adverse interference. Such unbalancedcircuits also give rise to grounding problems and the possibility ofcoupling sometimes lethal voltages to source generators. There have beenmany reported incidents where performers using unbalanced mic circuits havereceived serious electrical shocks from ungrounded (unbalanced) micapparatus.
The shielding methods described in Table 2 will also provide varyingresults depending on the frequency of the signal the cable will betransmitting. In general, foil will provide better shielding at highfrequencies and braid or spiral shielding will provide better results atlower frequencies. For the best of both worlds, cables that employ a braidshield over a foil coating will provide maximum broadband shielding. Beprepared to pay a premium price for cable that has that degree ofshielding. Also, be aware that the degree and type of shielding can produceunpleasant side-effects in the amount of time and labor that might berequired to strip and terminate connections. Proper field connection ofbraid+shield cable can double the time required per connection as opposedto the time required for drain wire type connections.
Frequency responseAs the frequency of the signal increases, the signal tends to flow closerto the surface of the wire, which is called the skin-effect. At lowfrequency, analog and DC transmission rates, this effect is virtuallyimperceptible. As signal frequencies reach or exceed 50 MHz, this becomes aserious consideration. Consequently, if the application necessitates thetransmission of low- and high-frequency signals over a common carrier, theconductor must be all copper or, for best results, silver-coated copper.This cable construction will ensure that the low frequencies will betransmitted in the copper core while the skin-effect high frequencies willbe carried by the silver coating. This is not an inexpensive solution, andit should be used only in cases where broadband signal transmission isexpected.
Jacket considerationsJacket material definitely affects cable performance. Almost all cable inuse today is coated with an extruded plastic for insulation and thencovered with an additional material to form an overall jacket. Both ofthese materials can be considered to be the dielectric of the cable. Allcables exhibit some degree of capacitance, whose value will depend on thespacing, the twist rate and the dielectric material’s constant.
One way to reduce capacitance is to increase the spacing of the twoconductors in a cable pair. This can be accomplished by increasing thethickness of the dielectric. This will, however, increase the mass of thecable with a corresponding difficulty involved in stripping and terminatingthe material. Frequently, cables like these are best terminated using IDCs(insulation displacement connectors). Another manufacturing solution to thecapacitance problem has been the introduction of better materials that havea lower dielectric constant. This will be accompanied with an attendantincrease in cost.
In recent years, great strides have been made in cable characteristics.Unshielded and shielded twisted pair (UTP/STP) can now be used for signaltransmission once considered impossible.
In pre-transistor days when signal circuits were terminated in highimpedance tube grid circuits, the subject of characteristic impedance was amajor consideration. Telephone and similar audio circuits were intended toterminate at 600 V. It quickly became apparent that twisted pair cablelacked the stability to exhibit the consistent impedance required for thetransmission of video, high-speed data and digital signals. To realize theconstant impedance characteristics required, coaxial cable was developed.
The two wires in a cable are like the two plates of a capacitor. Tomaintain a consistency of capacitance, these two wires must be preciselyspaced over the entire length of the cable run. To maintain thesetolerances, the conductors and their associated dielectrics must beuniformly maintained.
In the analog world of audio, DC security and fire alarm, the more esotericprovisions of a cable’s characteristic impedance are not too stringent. Onthe other hand, as frequencies increase, data speeds become faster andsignal complexity increases, it becomes increasingly critical to maintain aconstant-impedance match between the transmitting device and its intendedreceiving device. If the cable impedance varies, it will introduce standingwaves (SWR) or, to use another phrase for the same phenomena, structuralreturn losses (SRL). The power transfer between the sending device and thereceiver will not be efficient, and some of the energy is reflected backtowards the source.
An example of the effect of SRL is the case of an RF transmitter connectedto its associated antenna by a length of transmission cable. Both thetransmitter and the antenna have a defined value of matching impedancesthat must be maintained if maximum transfer of RF energy can be sustained.If the transmission line’s impedance varies from this set standard,standing waves (SRL) will be introduced, and losses will occur. If thisimpedance mismatch is severe, it could reach a point where almost notransfer of energy will occur.
The impedance of a coaxial cable is a function of the diameter of thecenter conductor, the diameter of the shield, the relative distance(spacing) between the two conductors and the dielectric constant of theseparation material. To maintain a fixed impedance, these factors must beconsistent.
Attenuation is also a consideration. This is a function of the gauge of thecenter conductor although the type of dielectric and the capacitancecreated during construction will have a bearing on the attenuation factor.Regardless of the attenuation factor, the impedance can be maintained byadhering to the spacing ratio of the conductors, which explains how you canhave RG6/U, RG8/U and RG12/U with reduced attenuation values while havingthe same impedance value. Attenuation of coaxial cable is expressed indecibels (dB) per unit length at a specific frequency. For example, a givencable might have a specification of 6 dB loss per 100 feet (30.5 m) at 10MHz. These specifications bear serious consideration when choosing a cablefor any given application.
How quick is your cable?The dielectric constant is a major factor in determining how well the cablewill transmit high frequency and data-based signals. The higher thevelocity of propagation (Vp), the better the cable will perform at highfrequencies. The velocity of propagation in a vacuum is defined as being100%.Air is close to this value, but as various forms of plastics areintroduced as dielectric materials, Vp will fall to 66% or less. Ideally,the coax dielectric would be a vacuum, air or an inert gas. This is not apractical solution because the center conductor must be positionedprecisely in the middle of the structure to maintain a consistentimpedance. Some high-power, broadcast transmission cable providesessentially air (or inert gas) filled cable that uses Teflon spacers tokeep the center conductor stabilized.
Cable delay characteristics are also associated with the dielectricconstant. Time delay (Dn) is measured in nanoseconds per meter and isdirectly proportional to the velocity of propagation. This conditionmanifests itself when signals having different frequencies are transmittedover the same cable. Considering that Vp varies as a function of frequency,the higher frequency signal will arrive later than the low-frequencycontent. Below 4.2 MHz or in the case of composite video, this willprobably not create a problem, but if the transmission is RGB video orhigh-speed data, this group delay factor needs consideration.
Foam cableBecause the velocity of propagation in air is close to 100%, the higheramount of air (or other inert gases) injected into the plastic, the higherthe percentage of Vp that will be achieved. Various cable manufacturers usedifferent methods to introduce foaming into their product in an effort toimprove Vp. Some of these methods tend to be more consistent than others.The illustrations of Figures 5 and 6 are representations of microscopiccross-sectional views of two foamed cables. Note in Figure 5 theinconsistency of the size of the air bubble and the random spacing of theelements of the dielectric. Under conditions like this, Vp will varyconsiderably. Also, in this example of a coax with poor foaming, the centerconductor will likely migrate in a vertical, horizontal or diagonalposition, relative to the exterior shield conductor. This will adverselyeffect the cable’s impedance. Hence, bear in mind that not all cables arecreated equal.
Figure 6 is a representation of a foamed cable that exhibits the desiredconsistent size and placement. Air bubbles are more uniform, and spacing ismore consistent. This type of cable construction will display a moreconstant Vp and have relatively little center conductor migration.
Similar to other types of cable, considerations for coax with respect toflex-life, flexibility, jacket physical properties and NEC ratings areimportant when evaluating the suitable cable for the intended application.
When evaluating coaxial cable for a particular application, attentionshould be given to NEC/NFPA ratings, attenuation per foot (meter) shieldmaterials-percentage of shielding required, nominal impedance, velocity ofpropagation, sweep frequency attenuation measurements, jacket/dielectricflexibility and conductor flex-life
Fiber Optic InstallationsOver the past few years, fiber-optic cable has become a cost-effectivemedium for wiring installations. The costs of the materials have droppedsignificantly; connectivity problems have been correspondingly reduced, andlabor involved has sharply declined. Hence, fiber has becomecost-competitive with other forms of cable.
You should consider using fiber-optic cable if the equipment being used isfiber-optic compatible, if the service provider has provided fiber to thepremise and extension to the network is contemplated, if the customerdemands fiber, if long runs are under consideration, runs from thousands offeet (meters) or miles (kilometers) or in cases where noisy EMI or RFIinterference pollution may be encountered, exposed environmental conditionsmay be conducive to oxidation or corrosion, weight and size are majorconsiderations and high bandwidths or high data-rate signal transfersignals must be accommodated.
Fiber-optic cable is an optical transmission medium in lieu of thetraditional copper electrical transfer mechanisms. In fiber-optictransmission, electrically generated signals are converted into opticallymodulated data streams and transmitted in optical fashion (modulated light)to the intended receiver, whereupon the signals are demodulated intoelectrical form. If the source and receiver equipment is outfitted toconnect to fiber optic, then a degree of compatibility exists. On the otherhand, if the transmitting and receiving devices are non-optical compatible,then some modulation/demodulation converters will be required. Although theprice of modulation converters has been steadily dropping, it is still acost not inherent in copper-to-copper connectivity.
Fiber-optic cable constructionThere are three general types of fiber-optic cable. One type is constructedof plastic while the other two are made of glass. Plastic construction hasthe advantage of being relatively inexpensive; on the other hand, plasticis larger in diameter, and its transmission losses limit its use to rathershort runs (20 feet to 30 feet or 6 m to 9 m).
Glass fiber-optic cable is classified as being either multi-mode orsingle-mode. Multi-mode refers to the characteristic that allows it toprovide multiple paths over a single-fiber conductor. Single-mode indicatesthat the conductor will be used for the transmission of a dedicated signal.Multi-mode has the advantage of being less expensive than single-mode andmore easily connected in the field. On the down-side, multi-mode has aninherently higher transmission loss than single-mode and is used primarilywhere cable runs are on the order of 1,000 feet (300 m).
Single-mode, which may be bundled in multiple conductor arrangements fromtwo to 240 conductors, is widely used to provide signal transmission overthousands of miles. This medium is used extensively by telephone companiesand other data service providers. The tradeoff is that single-mode is small(on the order of 5 mm to 10 mm). Hence, connections for this medium areexpensive and require microscopic connection procedures.
Fiber-optic cable (without metallic fillers) is considered nonconductive;however, it still comes under the scrutiny of the NEC with respect toflammability and toxic fume emission. In this respect, fiber-optic cablesare similar to code conditions for traditional wire and cable. Refer toarticle 770 of the NEC for ratings and installation methods for opticalfiber cable.
The Future in FiberIn Stephen H. Lamphen’s Wire, Cable and Fiber Optics for Video and AudioEngineeers the author provides a dissertation on the characteristics of anddirection for fiber-optic. This is repeated with permission of the author:
The bandwidth (and therefore the potential data rate) of multi-mode fiberis well-known; this is not true of single-mode. The theoretical limit ofsingle-mode was thought to be due to chromatic dispersion. Chromaticdispersion is the optical equivalent of group delay in copper cable. Thatis, fiber is nonlinear or, to be more precise, its index of refraction isnonlinear. This seemed to indicate that there was a maximum bandwidthbeyond which the parts of the signal could not be recombined.
One solution has been to create dispersion-limited fiber. Using this typeof technology, a 70 km cable run has been successfully installed near Tokyo.
There is also a more elegant solution-don’t change the fiber; change thelight. This solution is called solitron technology. Using this method, theoptical spectrum is divided into tiny channels, as small as 0.2nm. Whenapplied to signal transmission systems, this is referred to aswave-division multiplexing, or WDM. Solitron technology also pre-distortsthe pulses so that they stay in shape temporally (time-wise) and spectrally(light-wise) while trans-versing the fiber network.
In 1995, Neal Bergano and Carl Davidson of Lucent Technologies demonstrateda WDM system of light pulses that could send eight 5 Gbps channels over9,000 km (equivalent to crossing the Pacific Ocean). That was followedquickly with a demonstration of transmitting a 0.1 terahertz (twenty 5 GHzchannels) over 6,300 km (equivalent to the distance across the continentalUnited States).”
Bandwidth and data speed transmission characteristics of fiber are stillunder investigation, and there appears to be no end in sight for the futurepossibilities of this medium.
