Technical Papers


Wire, Cable, and Fiber Optics
Steve Lampen
Technology Specialist, Multimedia Products
Belden

SBE Ennes Workshop
St Louis, Missouri
 June 17, 2000

 

Basic constructions
Wire and cable consists, for the most part, of four basic constructions:

 

Solid and Stranded
Conductors come in two variations, solid and stranded. Solid (Figure 1) offers slightly lower resistance. The key to solid conductors is better performance at high frequencies.

Figure 1

Stranded (Figure 2) offers greater flexibility, that is limpness, and greater "flex-life", or flexes until failure.

Figure 2

 

Multiconductor
As the name implies, multiconductor cables consist of many conductors. (Figure 3.) They are common in control applications but are rarely used for signal applications, and therefore, we will not be focusing on them for this paper.

Twisted Pairs
Twisted pairs (Figure 4) consist of two insulated wires twisted together. They are specifically intended for carrying signals and were invented in the 1880’s for wiring up the early telephone systems. Twisted pairs offer low noise pick-up and low noise emission from a cable because it is a balanced line and because a balanced line offers "common-mode noise rejection".

What is a Balanced Line?
A balanced line is a configuration where two wires are kept close together, usually by twisting them (Figure 5). Conductors need to be the same length, the same size, with a constant distance between them.

 

Figure 5

It should be noted that the signal, at any instant in time, is exactly the same but opposite polarity on the two wires. Another way of saying that is, if you note the signal voltage at any point of the cable, they should add up to zero.

 

When Noise Appears….
Noise is a fact of life. It is electromagnetic radiation and can come from many sources including fluorescent lighting, motors, car ignition systems, equipment such as hospital analyzers, transmission equipment from CB’s, truck, taxis, radio and television broadcasters, and natural sources such as the sun.

When noise appears, and hits the two wires in our twisted pairs (Figure 6), the electromagnetic radiation of the noise induces a voltage in both wires. However, the direction is the same or "common mode" in both wires.

Figure 6

When the two noise signals reach either end of the cable, there is either a passive balancing device (such as a transformer shown in Figure 6) or the equivalent active balanced input. As you can see, the two noise signals on the two wires cancel each other out. In this way, the noise cancels out and the signal (which is "differential mode") can continue through.

 

Coaxial Cable
Coaxial cable, or coax, for short, is the other most popular cable configuration. The signal on the cable is not the same on the two conductors since the shield carries both ground and signal. The two conductors are not the same size, same resistance. Therefore, coax is not a balanced line. It is an unbalanced line.

Coax does not have the inherent noise rejection of twisted pairs. But performance of coaxial cable can be far superior to twisted pairs. First, coax has extremely stable performance. The various parts of the cable are "locked together". (Figure 7) This, as we will discuss later, gives much better performance especially at high frequencies.

Figure 7

Shields
Shields are added to twisted pairs, or multiconductor cables, to help prevent in ingress (interference) or egress (radiation) of noise. Shields are an inherent part of coax cable. There are six basic shield configurations:

 

Unshielded
Unshielded cable is appropriate where no noise is present, such as no crosstalk from adjacent wires. Or it can be used if you don’t care if there is noise, that noise cannot substantially affect the signal on the cable.

Unshielded cables are especially prevalent in the data world where pairs have very tight twists, or may use conductors that are bonded together. Such high-performance pairs are good to excellent at not picking up or radiating noise.

 

Serve/Spiral Shields
Serve or spiral shields can be made to be ultra-flexible. However, serve shields can open up when flexed, which compromises shield effectiveness.

A spiral of wire obviously affects the inductance of the shield. Therefore spiral shields are rare in video and are usually audio only. There is a double spiral serve, also known as a "Reussen" shield. This configurations "shorts out" the inductive effect of a signal spiral, but the shield can still open up when flexed. This double serve is common on many European and Japanese audio snakes. The ultra-flexibility of the cables is a key. However, it should be realized that users of this cable type often trade flexibility for performance.

 

Braid Shields
Braid shields are formed by spinning wires or groups of wires around a core. This slow and labor-intensive process makes braiding the most expensive single step of cable manufacturing. Single braid coverage of up to 95% can be realized. Double braid coverage can be up to 98% coverage. Since braids always have "holes" where the wires cross, 100% coverage not possible with braid.

Braid shields are most effective at frequencies from 1,000 Hz to 50 MHz. For these frequencies, the low resistance of a braid gives good coverage. Below 1,000 Hz there is no standard braid material which is effective. The wavelengths are so long, and the low frequency energy so pronounced, that the only effective shielding is solid steel conduit. And, at 60 Hz, even steel conduit gives 27 dB of noise reduction.

At frequencies above 50 MHz, braid becomes "wavelength dependant" where the holes look larger and larger as the wavelength gets smaller and smaller. The effective coverage of a braid gets worse and worse, especially compared to a foil shield, which has no holes.

 

French Braid Shields
French braid shields are a combination of serve and braid. A French braid consists of two serves braided along one axis. This gives cables excellent flexibility, rivaling those Reussen shield in European and Japanese cables. And yet, unlike Reussen shields, lab tests indicate that French braids have excellent RF performance. This may be partly because the braiding "shorts out" the inductive effect of serve shields and "shorts out" the RF noise too.

Because it takes the same number of steps as a regular braid, French braids are the same cost as regular braid. Maximum coverage of a French braid is 98%.

 

Foil Shields
Foil shields are the easiest and cheapest to apply. They can be applied as fast as the cable will run. Foil shields actually consist of two layers, a metal layer and a plastic substrate of polyester. This can be easily seen since the foil is silver on one side and colored (red, blue green or other colors) on the plastic side.

Since foil shields lack the mass and low resistance of a braid shields, the exhibit poor to average low-frequency performance. However, after 50 MHz, foil shields have excellent high frequency coverage. Since foil is a continuous sheet of metal, coverage can be 100%.

 

Combination Shields
Combination shields consist of foil and braid combined. Occasionally there can be more than one layer of each, such as "quad" cable television cable, so called because it has two layers of foil and two layers of braid. Because of this, combination shields are the most expensive of all. But they also give the best broadband coverage, since it contains a braid for low frequencies and a foil for high frequencies

The difference between broadcast coax cables, which often contain foil and braid in digital applications, and CATV/broadband cable is that CATV cables use low coverage braid (sometimes as low as 40). The reason is that these cables only operate above 50 MHz. At those frequencies, braid shields are ineffective. It is actually the foil shield that is doing all the noise reduction. The braid shield is there to give the F-connector something to grab onto. It’s a reliability issue, not a performance issue. CATV braids are aluminum belying their low cost and indicating that this braid is not included for performance.

Combination braids are required for digital video such as SDI or HD. The Broad frequency range of SDI (135 MHz) or HD (750 MHz) make a combination shield a requirement. That being said, it should be notes that double-braid cables (such as Belden 8281) can still operate at these high frequencies. It is simply that the effective distance they can run is severely reduced compared to cables with foil + braid (among other improvements). Most precision digital cables contain 95% braid + 100% foil

 

Cable Parameters
Cables are made up of metals and plastic. The choice of metals and the choice of plastic can have a significant effect on the performance of that cable. Effects such as capacitance, impedance, inductance and skin effect are derived from the choice of materials.

 

The Conductor
When comparing metals, each can be characterized by resistance as the next table shows. Silver, while the best conductor, is expensive and difficult to work with. Copper is the most common metal.

Metals Resistance

Circular mil-ohms per foot at 20şC

Silver 9.9
Copper 10.4
Gold 14.7
Aluminum 17
Nickel 47
Steel 74

One advantage of copper is its ability to be "annealed". After being drawn through dies from large rod to small wire, copper will get brittle. By placing it in an oven at around 700oF, the copper will become flexible again.

Gold is most commonly used on connectors because it will not oxidize. Aluminum is often used in low-cost cable constructions such as CATV/broadband shields, or in low-cost consumer audio interconnect cables.

 

Wire Gage
The size of each wire is describes as the gage size, and is measured in units of American Wire Gage (AWG). Below is a list of gages with a description of how small or large that size is:

40 AWG smaller than a hair
30 AWG sewing thread
20 AWG diameter of a pin
10 AWG knitting needle
1 AWG pencil
1/0 "1-aught" finger

 

Resistance
The choice of metal, the gage size of the wire, and the length of the wire can determine the resistance of any conductor. Charts are available, such as in the back pages of the Belden Master Catalog, which shows the resistance for stranded wire from 36 AWG to 10 AWG, and the resistance for solid 40 AWG to 10 AWG.

All wire has resistance. Resistance affects the signal by turning part of the signal into heat. This creates a voltage drop on the wire when one end is compared to the other. The voltage drop can be determined by one of the formulas of Ohm’s Law, E =I2R, where E is the voltage drop on the wire, I is the current in amps, and R is the resistance in ohms.

While a voltage drop in the presence of any resistance is unavoidable, picking a larger conductor with lower resistance can reduce the effect. Also, resistance is linear over frequency, meaning that resistance affects all frequencies equally. It is therefore often ignored since the effect may be a minor drop in overall level.

 

Insulation
Basic insulation prevents wires from touching each other and creating a short circuit or grounding portions of a circuit that should not be grounded.

When the insulation affects the signal being carried on the wire, it is called a "dielectric". Every non-conductor varies in its ability to insulate. Plastics, and other materials, can be compared by a number that describes their quality, called a "dielectric constant". Below is a list of materials and their dielectric constant. Note that vacuum is the standard by which all other materials are compared, and therefore, has a dielectric constant of one.

Vacuum = 1
Air = 1.0167
TeflonTM = 2.1
Polyethylene = 2.25
Polypropylene = 2.3
PVC = 3 to 5

Air is so close to "1" that it is most often used as "1", in formulas. As we will see, the dielectric constant of air makes it a highly prized commodity in dealing with cable construction. The simple question is how can you put air into a cable and yet keep everything inside from moving around?

 

Velocity of Propagation
Velocity of Propagation (Vp) compares the speed of a signal down a wire to the speed of light. The speed of light in a vacuum is the standard by which all other signals are measured. The reason there is such an effect is because the signal consists of an electromagnetic field around the wire. That field travels in the plastic or other insulation on the wire.

Velocity of Propagation is a cousin to dielectric constant and the relationship can be described in the formula shown in Figure 8.

Figure 8

Velocity of Propagation is what you most often see in a catalog, and comparing VP between cables gives you some idea of which cable performs better at high frequencies. A faster velocity means less high-frequency loss and flatter frequency response overall.

Here is our same list of dielectrics with their velocities.

 

Capacitance
A capacitor is a device that holds an electrical charge. It consists of two metal plates with insulation in between. Well, isn’t that exactly what a cable is? Two metal plates (or wires) with an insulator (dielectric) in between. Figure 9 shows the specific parts of a twisted pair and coax cable that are involved with capacitance.

While cables do have capacitance, it is very small, due mostly to the fact that the wires are also small. Capacitance in cables is almost always measured in picofarads-per-foot. (pF/ft.) A picofarad is a trillionth of a farad, the unit of capacitance. So why would we have any interest in an effect that small? Because you don’t use just one foot of cable. Most often you are using tens, hundreds, even thousands of feet. And this capacitive effect adds up. That is, a 1,000-ft. cable will have 1,000 times the capacitance as a one-foot piece. Then you can get up to some serious capacitance!

The real problem with capacitance is that it is affected by the frequency of the signal on the cable. The higher the frequency, the more the capacitance "stores" that signal as a charge. This "reaction" to frequency creates "capacitive reactance" also measured in Ohms, like resistance. But the effect changes with frequency, which resistance does not. Being "Frequency-dependant", capacitance is responsible for the "Frequency response curve" of any cable.

 

Inductance
The electrical signal down a wire also creates a magnetic field down that wire. This effect is called "inductance". However, on most cables, the inductive effect is so tiny, that it is never listed in a catalog. The effect, with a frequency running on the cable, is called "inductive reactance".

Because the inductance is tiny on most cables, the inductive reactance is also tiny. Inductance and capacitance are reverse effects. Therefore, they cancel each other out. But, in almost every cable, the capacitance and capacitive reactance and so much greater that they cancel out the inductance and inductive reactance/ But there is still capacitance, and capacitive reactance, left. This is why capacitance is a critical number in almost every cable type from analog audio to high-speed UTP, and inductance is essentially ignored.

Inductance is based mainly on the size of the wire (AWG) and can be most easily changed by changing the size of the wire.

 

Impedance
Of all the effects of frequency on a cable, impedance is the hardest specification to understand. That is because it is the sum-total effect of resistance, capacitance, and inductance when a frequency or band of frequencies is applied to the cable. Since it describes the "total opposition to current flow" caused by these three factors, it too is measured in Ohms.

But impedance is more than that. It is also a number that describes the relationship of dimensions in the cable. In fact, if you can provide three numbers for any cable, there is a simple algebraic formula that can then tell you the impedance of the cable. Figure 10 shows you those "three numbers".

Figure 10

In the case of the twisted pairs on the left, just tell me (1) the size of the conductors, and we are assuming the two conductors are the same size, (2) the distance between them, and (3) the dielectric constant of the material in between.

In the case of the coax cable on the right, it is almost the same. Just tell me (1) the size of the center conductor, (2) the distance from the center conductor to the shield, and (3) the dielectric constant of the material in between.

If you are clever, you can already see the resistance, capacitance, and inductance factors all combined. The size of the conductor is obviously resistance. (Gage size determines resistance.). The size of the conductor also gives you the inductance. The distance between conductors (or center and shield), with the dielectric constant, gives you the capacitance. So you can see all three factors combined.

Every cable has impedance. From the lowliest twisted pair, to the fanciest coax. If it has two conductors (which is the definition of a "cable"), it has impedance. And like the capacitive and inductive reactances that it contains, impedance changes with frequency with one special difference. The impedance changes until you get to a certain frequency where you reach a "characteristic impedance". This occurs somewhere around 10 MHz. From that point on, the cable will be one impedance value.

There are three formulas for impedance based on capacitance and inductance. Figure 11 below shows all three.

Figure 11

In the first formula, on the left, works where resistance is a major factor. You will note that it starts at a very high number, (10,000 ohms at the 10 Hz), and descends as the frequencies get higher. The formula on the right is used where resistance is no longer a factor. In that case the inductance and capacitance are the key factors. Since they do not change, the impedance of the cable remains the same for all frequencies. The third formula, in the middle, is the transitional formula between the two other formulas.

You can see that, at low frequencies, there is no "characteristic" impedance. It is always changing. This is why it would be impossible to build an 8-ohm speaker cable, for instance, because it would only be 8 ohms at one particular frequency, in the analog audio spectrum of 20 Hz to 20 kHz. At any other frequency above or below the chosen frequency, it would be a different impedance.

Lest you think this is all scientific gobbledy-gook, Figure 12 shows the actual measured impedance of two cables.

Figure 12

The only difference between the cables is the gage size of the wires. Since this is a resistive different, the two traces are not in the same place. However, as they get closer to the characteristic impedance, you will note the get closer and closer together. If they have the same capacitance and inductance, the will eventually have the same value of impedance.

OK, so what is impedance and why is it important? To understand it’s importance, we have to understand one more specification, wavelength.

 

Wavelength

Wavelength describes the length of a wave. Figure 13 is a picture of a wave with an arrow indicating the wavelength. This picture could represent the movement of a string in a piano. If it is "middle A", it would make this shape 440 times a second. (OK, 436 times a second for you musical purists)

Figure 13

This picture could also show the air in the room being compressed or expanded 440 times per second by the piano string. It could also show the movement of your eardrum 440 times per second as you hear the sound. If we built a device that converted acoustical energy into electrical energy (a microphone), the wave would also represent the electrical flow, back and forth, down the cable from the microphone. It is this electrical flow in which we are most interested. So how long is the wavelength? It is different for every frequency. But we have a formula (Figure 14) that will tell us.

This says that the wavelength of any signal (in meters…we can convert to feet when we’re done) can be determined by dividing a big number (300 million) by the frequency we’re interested in.

In any system which uses a range of frequencies (such as audio, which is 20 Hz to 20 kHz) the number to put in the formula is the highest frequency. If your car will go 150 miles per hours, 65 miles per hour will be no big thing. In the same way, if your cable will handle 20 kHz, any frequency below that will be easy to carry.

OK, so the wavelength is nine miles. So what? Well, here’s a simple rule:

When any signal on any cable is at least one-quarter of a wavelength then the impedance of that cable is important.

This means that, in our audio example, one-quarter of nine miles is 2 1/4 miles. That means the cable carrying that 20 kHz (and below) has to be at least two miles long before we even care what the impedance is. So what is the impedance of a speaker cable? It doesn’t matter. What’s the impedance of that analog microphone cable? It doesn’t matter, as long as they are less than two miles long.

One other thing affects wavelength, and that is the dielectric. So, in our example above, you would actually multiply the distance by the Vp. So, assume our cable was low-grade PVC with a Vp of 50%. Then the critical distance will be 2 1/4 miles x 50%, or 1 1/8th miles, still a very long way.

What about other frequencies? Below is a table that shows the critical distance for various signals. Appropriate velocities of plastic are included.

Frequency

Wavelength (meters)

Wavelength (ft.)

Quarter-
wavelength
(ft.)

Common cable velocity

Distance where impedance is important

20 kHz

15,000

49,200

12,300

50%

6,150 ft.

100 kHz

3,000

9,840

2,460

50%

1,230 ft.

1 MHz

300

984

246

66%

162 ft.

4 MHz

75

246

61.5

66%

40.6 ft.

25 MHz

12

39.36

9.84

78%

6.89 feet

135 MHz

2.2

7.29

1.82

83%

22 inches

750 MHz

0.4

1.3

0.328

83%

3.27 inches

Table 1

You can see that, as the frequency gets higher, the wavelength gets shorter and the more critical the impedance becomes. And, when the frequencies are very high, everything the signal runs on is critical in its impedance, including cable, connectors, patch cords, and patch panels. Even traces on printed circuit boards are critical at those high frequencies.

The frequencies above were not chosen at random. They are the top critical frequencies of many broadcast standards.

Frequency

Standard

Application

20 kHz

none

Analog audio

4 MHz

NTSC

Analog video

25 MHz

AES/EBU

Digital audio (192 kHz sampling)

135 MHz

SMPTE 259M

SDI serial digital video

750 MHz

SMPTE 292M

HD video uncompressed

Table 2

So you can see that cable impedance can important, even critical, with certain high-frequency applications. And what happens if your cable (or connector) does not have the correct value of impedance? You get "return loss"

 

Return Loss
Broadcast engineers are familiar with return loss. They know it as Voltage-Standing-Wave-Ratio (VSWR), sometimes called SWR. In high frequency systems, when a signal goes down a cable which the wrong impedance, the signal will reflect and "return" to the source. This is sometimes mistaken for natural attenuation, or even resistive loss in the cable. However, it looks like a very long cable is attached, when the cable is not long at all. In digital systems, it can increase bit errors, even to the point of signal failure, if the impedance problem is severe enough.

 

Skin Effect
As frequencies get higher, the signal tends to migrate to the "skin" of the conductor. This is why resistance becomes less and less of a factor in impedance. (See "Impedance" above.) The formula for the skin depth of copper is quite simple. Here it is in Figure 14.

 

Figure 15

This shows the depth in inches (Din) determined by the frequency (F in Hertz). Below is a table that shows the skin depth of various conductors at various frequencies. The percentage of conductor used is then determined. Be aware that, since skin effect is a gradient from the inside of the conductor to the outside, the "percentages" shown indicate the majority of wire conduction

Frequency

Skin Depth

AWG

Diameter

Percentage of conductor used

20 kHz

.0184 in.

24

.024 in.

100%

20 kHz

.0184 in.

22

.031 in.

100%

20 kHz

.0184 in.

12

.093 in.

75%

20 kHz

.0184 in.

10

.115 in

68%

4.2 MHz

.0127 in.

20

 

68.6%

25 MHz

.00527 in.

24

.024 in.

4.39%

135 MHz

.00225 in.

20

 

12.5%

750 MHz

.000953 in.

20

 

5.29%

Table 3

 

Fire Ratings
The National Fire Protection Agency (NFPA) is a voluntary non-profit organization that puts out the National Electrical Code (NEC). This book sets suggested standards for safe construction of buildings. These standards include flammability testing of wire and cable.

The NEC code is voluntary. This means that a state, county, or city may or may not adopt the code. Some cities, such as Las Vegas and Chicago, have stricter codes. If you are planning an installation, you should check with your architect, contractor, Building Inspector, Fire Marshal, Planning Board or other authority as to what your community uses as a standard. The majority of the states and communities subscribe to the NEC, but you can’t know for sure unless you ask.

The NEC book lists many different cable ratings. The most common for audio and video are:

  1. Unrated
  2. CM
  3. CMR
  4. CMP

Unrated cables are those which will not be installed, and which will be visible when in operation such as microphone cables. In the 1999 NEC, they now state that any cable installed must carry a rating. If this is how your inspector interprets the new rules, then unrated cable cannot even be installed in a conduit.

CM or "commercial" grade cables can go through a wall without being in a conduit. CM is the most common rating

CMR is "riser" version of CM. This cable type can be run vertically between floors without being in conduit.

CMP is the plenum version of CM. This can be placed in the most fire-critical areas such as drop ceiling, or raised floor. These areas are often where air conditioners have an air return. Anything burning and creating smoke will have the smoke fed into the air conditioning thus causing a hazard to all other areas of the building. So plenum cable is intensely tested to avoid being fuel for a fire and to limit the smoke produced.

You need to be aware that the NEC ratings have nothing whatsoever to do with cable performance. You can get a plenum rated speaker cable, or a plenum rated high-definition video cable and all they have in common is their fire rating.

 

What Should I Buy?
With all the discussion we have had about cable types, constructions, and performance, the question still remains, how can you select the appropriate cable for any application? So below, is a list of each type of cable we have discussed with a list of the important or critical parameters to be considered when choosing cable.

 

What Cable is Required for Each Application?

Analog audio cable requires…

  • Low capacitance
    1. 50 pF/ft. for short runs under 100 ft.
    2. 30 pF/ft. for longer runs up to 500 ft.
    3. 25 pF/ft for longer runs up to 750 ft.
    4. 13 pF/ft. for runs over 1000 ft. or where ultimate performance is desired
  • Low resistance
    1. 26 AWG for short runs under 100 ft. where ruggedness is not an issue
    2. 24 AWG for runs up to 1000 ft.
    3. 22 AWG for runs over 1000 ft. or where ruggedness is essential
  • Low crosstalk between pairs
  1. Foil shields for RF protection
  2. Braid shields for low self-noise and protection 1 kHz to 50 MHz

 

AES/EBU Digital audio twisted-pair cable requires…

  • Specific impedance 110
  1. May vary between 88 and 132
  • Low capacitance
  1. Below 20 pF/ft.
  2. Most AES pairs are 13 pF/ft.
  • Low resistance
  1. Gage size a factor due to skin effect
  2. 24 AWG is most common
  3. 22 AWG can go about 10% farther
  • Moderate crosstalk
  1. Only 30dB required
  2. Even UTP can meet this

 

AES3id or S/PDIF Digital audio coax cable requires…

  • Coax cable
  1. Distance about 2-3 times that of twisted pairs
  2. Lacks common mode noise rejection of twisted pairs
  • Specific impedance
  1. 75 required
  • Low capacitance
  1. <20 pF/ft.
  2. Capacitance controlled by impedance dimensions
  3. Precision video cables suggested
  • Low resistance
  1. Large gage coax such as 20 AWG or 18 AWG suggested

 

Analog video coax cable requires…

  • Specific impedance
  1. 75 required
  • Low capacitance
  1. ~20 pF/ft.
  2. Capacitance controlled by impedance dimensions
  3. Precision video cables suggested
  • Low resistance
  1. Large gage coax such as 20 AWG or 18 AWG suggested

 

SDI Digital video coax cable requires…

  • Impedance critical
  1. 75 +/- 3
  • Low capacitance
  1. <20 pF/ft.
  2. Capacitance controlled by impedance dimensions
  3. Precision video cables swept to 400 MHz minimum
  • Low resistance
  1. Large gage coax such as 20 AWG or 18 AWG suggested
  • Return loss
  1. To be considered

 

HD-SDI video cable requires…

  • Impedance critical
  1. 75 +/- 3
  • Low capacitance
  1. <20 pF/ft.
  2. Capacitance controlled by impedance dimensions
  3. Precision video cables swept to 2.25 GHz minimum
  • Low resistance
  1. Large gage coax such as 20 AWG or 18 AWG suggested
  • Return loss
  1. >15 dB SMPTE standard to 2.25 GHz
  2. >20 dB suggested to 2.25 GHz

 

Fiber Optics
There are three basic kinds of fiber optic cables:

  1. Plastic fiber, of which the most common is for high-end audio and is called "Toslink". This fiber has a diameter of 900 µm (micrometers) or almost 1 mm. We will see that this is "huge" in the fiber world, and huge is not good when considering performance.
  2. Then we have "Multimode" fiber. This is made of glass and can come in various diameters. The most popular is 62.5 µm, although recently 50 µm has been making a comeback in the data world.
  3. And then we have "Single Mode" the king of fiber, with a diameter of 8 µm. This is so small that you cannot see it without a microscope, and is therefore very hard and expensive to connectorize, at least compared to the two fiber optics cables above. It is also the king of performance.

 

Plastic Fiber
Plastic Fiber uses visible light as the signal-carrying medium. Because of the long wavelengths and the relatively huge size of the fiber, the light bounces around while passing down the fiber (called "dispersion"). The effect is that the signal can only go a few feet, maybe 20 or 30 feet.

On the other hand, plastic fiber is amazingly easy to connectorize. It is also limited to low bandwidth of a few megahertz. It is most often used for high-end consumer interconnection of devices and is especially popular carrying digital audio.

 

Multimode Fiber
Multimode fiber uses much shorter wavelengths than the plastic fiber. In multimode fiber there are two places in the spectrum which give the lowest loss and are called "windows" Instead of describing them by "megahertz", windows are describes by the wavelength of the light in nanometers. The characteristics of the fiber, based on its diameter, determine the performance or bandwidth in each window. The bandwidth is specified for 1 km (1 kilometer or approximately 3280 ft.) Shorter runs would have less dispersion and therefore be higher bandwidths. Longer runs would have greater dispersion and therefore have less bandwidth.

For 50 µm multimode fiber the two windows are 850 nm (with a bandwidth of 500 MHz) and 1300 nm (which also has a bandwidth of 500 MHz). So the total bandwidth of 50µm fiber is 1 GHz. How much data you can put down such a cable is determined by the compression, bit reduction, and other techniques. Therefore, telling someone the bit rate running down the fiber is almost meaningless, since you have no idea what techniques were used to get that much data on the fiber. The bandwidth is the only way to compare fiber optic cables.

On the more popular multimode fiber, 62.5 µm the windows are the same but the bandwidth of each is different. The first is 850 nm (with a bandwidth of 160 MHz). The second window is1300 nm (with a bandwidth of 500 MHz). So 62.5µm fiber has a total bandwidth of 660 MHz. One can see why 50 µm is making a comeback: greater bandwidth. There are also some 62.5 µm fibers that are selected for higher first-window bandwidth, some above 200 MHz, giving a total above 700 MHz.

Considering that the hardware, especially connectors, is most common for 62.5 µm, one should consider seriously a choice between 62.5 µm and 50 µm. The latter died, and is coming back from the dead, so not all manufacturers may support it. Or the lead-times for their products may be unacceptable

 

Single Mode
Then we come to the king of fiber optics, Single Mode. It also has two windows, but they are at different wavelengths, 1300 nm and 1550 nm. What is the bandwidth in these windows? Truthfully, nobody knows. In fact, the bandwidth seems to be limited only by the performance of the devices to which it is connected.

Single mode has bandwidth well into the GHz with most common equipment. Top-of-the-line equipment will get you into a total bandwidth in the 40 GHz range. Experiments have been done with 3000-mile long samples where they successfully passed 100 GHz of data. Such a bandwidth would be enough to cover almost every signal requirement imaginable, and far into the future.

 

So Which Do I Use?
Most professional audio and video equipment still has copper connections. When this changes, and when you have fiber optic connections on the back of equipment, that is the most likely time you will start using fiber.

The other time you will use fiber is when the bandwidth required, or the distance required, cannot be satisfied by a copper cable. For instance, the farthest one can go on a copper coax running HD-SDI (750 MHz/1.5 GHz) is somewhere around 900 ft. If you want to go a thousand feet, fiber might be your next best option.

However, at each end of the fiber you will need to convert from electrical to optical and back again. This adds to the cost and complexity, and reduces reliability. You can’t solder or crimp a fiber, and it is recommended that you take a class in connectorizing fiber.