Selected Topics EC Engineering I – Lecture IX – Fiber Optic Fundamentals Part I N Pillay
I. Fiber-Optic Fundamentals Part I
The first known implementation of the principles of fiber-optical systems have been dated
back to the 1800’s when telephone inventor Alexander Graham Bell used a beam of light in
air to transmit telephony messages. At that stage the technology was not at a practical level
for real-world use. In the 1950’s fibers were used in medical procedures, but in 1970’s the
cable loss was designed to be low enough for practical communications application.
Core
Cladding Jacket
Figure I1: Cross section of fiber-optic
An optical fiber, shown in Figure I1, is a waveguide that carries light waves from some point
to another desirable point. The cross section is shown above and at the simplest level
consists of the core, usually made of glass or plastic and the cladding, also made of glass or
plastic and the jacket usually made of plastic. The refractive index of the cladding needs to
be lower than that of the core in order to refract the light waves back into the core and thus
propagates along the fiber. Figure I2 shows the structure of a basic optical communication
system. A signal source generates an electrical signal that is converted into an optical signal.
The optical signal is then propagated along the cable and at the receiving front-end is
converted back into electrical and sent to the sink or further processing. The direct source
for the communication system can be a binary message switching a LED on and off. These
two voltages are then transmitted to the RX as light waves.
The fiber cable is used as a substitute for Cu cable and twisted pairs or a point-to-point
microwave radio link. Some of the advantages of using fiber over other copper would be
the high bandwidth, weight-cost ratio, and security. The high bandwidth provided by these
cables make it possible to transmit many signals over the line using multiplexing techniques.
Fiber can also be designed with lower loss than copper and thus longer lengths of cable can
be used between repeaters. The fact that optical signals are being used implies isolation.
Fiber is therefore used to control high voltage circuits in power stations and are often used
in medical equipment to isolate patients from possible electrical shorts. Geostationary
1|P ag e
Selected Topics EC Engineering I – Lecture IX – Fiber Optic Fundamentals Part I N Pillay
satellite links and point-to-point microwave links are also often replaced by fiber-optic
cables. The advantage of the former radio links is that there doesn’t have to be cables laid
out and the area between two links can have unaffected terrains. However, the round-trip
propagation delay of satellite links make fiber-optics much more attractive and the
bandwidth of fiber is much higher. Fiber also offers much higher privacy. Fiber solutions
are often used for long distances high bandwidth applications. Cu cables will be favoured
for short distance low bandwidth applications.
It is important to note that fiber is not used for the transmission of power but rather
information at relatively low powers.
Signal Optical
E/O
Source TX
Fiber-Optic cable
Signal Optical
O/E
Sink RX
Figure I2: Basic fiber-optic communication system
Figure I3: Fiber-optic cables
2|P ag e
Selected Topics EC Engineering I – Lecture IX – Fiber Optic Fundamentals Part I N Pillay
We have spoken earlier about total internal reflection. We know from second year
university physics that the refractive index is given by
= ………………………………… .1
Where , , are the refractive index of some medium, the speed of light in free space and
the velocity of light in a medium, respectively.
We can also recall that the dielectric constant or the relative permittivity can be expressed
as
ℇ = ………………………….. .2
From which it is clear that
= ℇ ………………….. .3
Also recall Snell’s law,
Ѳ = Ѳ …………………. .4
The angle is measured between the ray and the normal to the interface between the media.
For optical cables the first medium is the core and the second is the cladding. The refractive
index of the core must always be greater than the refractive index of the cladding.
Rearranging I.4,
Ѳ = Ѳ …………………… .4
There will be some value of Ѳ less than 90 degrees, for which Ѳ will be unity, i.e. Ѳ
will be 90 degrees and the refracted ray will lie along the interface between the media. For
a greater angle of Ѳ the ray will be reflected and Snell’s law ceases to operate. The value
of Ѳ for which Ѳ is 90 degrees is termed the critical angle.
This is defined as,
Ѳ = arcsin …………………………. .5
The numerical aperture is related to the critical angle and is often used in specifications for
optical fiber and its related components. The numerical aperture is defined as the sine of
the maximum angle a ray entering the fiber can have with the axis of the cable, and still able
to propagate by internal reflection.
We define numerical aperture as,
. = − …………………………….. .6
3|P ag e
Selected Topics EC Engineering I – Lecture IX – Fiber Optic Fundamentals Part I N Pillay
I.6 assumes that light enters the fiber from free space. The total angle of acceptance is
twice that of the N.A. Light entering from outside the cone of acceptance will be refracted
into the cladding – no propagation.
Figure I4: Illustration of numerical aperture
As the frequency increases the losses in conductors and dielectrics increase. At microwave
frequencies, waveguides provide an alternative. A waveguide can be thought of as a pipe,
through which waves are allowed to travel by reflecting of the walls. These operate like
high-pass filters, and are not used for frequencies below the GHz range. The losses in
waveguides are also limited due to the large surface area of the wall, radiation losses and
dielectric (usually air) losses are zero or very small.
Electrical energy can propagate in a number of ‘modes’ in waveguides (see Figure I5). For
high order modes rays strike the wall at greater angles. For low-order modes this angle is
reduced. High-order modes cause slower propagation. Low-order modes creates fast
propagation. Consider a brief pulse of microwave energy applied to the entrance of the
waveguide. The pulse will arrive at the other end at several different times – according to
the modes. This causes the brief pulse to be spread over time and another pulse following
may cause interference with this. We refer to this effect as dispersion. Because of
dispersion it is not desirable to have more than one mode propagating. A fiber-optic cable
is in fact a waveguide for light.
4|P ag e
Selected Topics EC Engineering I – Lecture IX – Fiber Optic Fundamentals Part I N Pillay
Figure I5: Modes
Single mode fiber has much lower dispersion but the manufacturing process is more difficult
compared to multimode fibers.
The maximum allowable diameter for a single-mode fiber is,
0.383
= ……………………… .7
and varies with the wavelength of light.
Both single-mode and multimode fibers are otherwise referred to as step-index fibers. This
is because the refraction index changes distinctly between the core and cladding. Graded-
index fiber is different in this sense. It is actually designed so that the refractive index
decreases gradually moving away from the centre. Thus light will travel faster closer to the
cladding. Optical fibres are normally made of high quality glass to reduce losses. Some low-
cost multimode fiber is made of acrylic plastic and used in vehicles, etc. Single mode fibers
are almost free from dispersion effects. Multimode fiber on the other hand is not. Graded
index multimode fibers tend to reduce the amount of dispersion. This is possible since,
signals propagating in high-order modes spend more time close to the outside of the fiber.
Since the index is smaller in these regions these signals are made to propagate much faster.
The little dispersion exhibited by single-mode fibers is referred to as intramodal dispersion
or chromatic dispersion because it results from the presence of different wavelengths of
light.
Dispersion increases with the bandwidth or linewidth of the light source. Thus lasing diodes
are preferred to normal LEDs as light sources. Also an increase in the length of the fiber
causes higher dispersion.
5|P ag e
Selected Topics EC Engineering I – Lecture IX – Fiber Optic Fundamentals Part I N Pillay
Multimode fibers are rated according to bandwidth-distance product.
We define the dispersion per kilometre as,
= ∆ …………. .8
Where ∆ is the linewidth of the source.
is the chromatic dispersion in picoseconds per nanometer of linewidth per kilometre of
fiber (ps/nm/km). And is expressed in ps/km.
The total dispersion is then,
∆ = . ……… .9
∆ is total dispersion in picoseconds.
In most cases the manufacturer will provide a means for calculating the dispersion. The
SMF-28 single mode fiber has the following specification.
( )= − … … … … … … … … . . . 10
4
, are the zero-dispersion wavelength and zero-dispersion slope specified in the
accompanying data.
We will speak about and bandwidth and losses in fiber-optic systems in the next lecture.
Optical fibers are extremely thin and need protection from external forces. Damage from
corrosive conditions must also be accounted for. More often than not several fibers are run
in a single cable. These cables will also include protective jackets and strength members to
increase robustness, for installation and lifetime. There are two basic types of fiber – loose-
tube fiber and tight-buffer fiber. Figure I6 shows the cross section for each of these.
Core
Cladding
Fiber Buffer
Air or Gel
Tube
Figure I6: Loose tube construction
6|P ag e
Selected Topics EC Engineering I – Lecture IX – Fiber Optic Fundamentals Part I N Pillay
Core
Cladding
Fiber Buffer
Plastic Buffer
Kevlar
Plastic Jacket
Figure I7: Tight buffer construction
Popular strength members are Kevlar or steel. Kevlar is usually used in small cables for
flexibility. There are various types like flat cables, armoured or waterproofed cables and
fiber cables including copper pairs.
Figure I8: Tapping into a fiber-optic cable
Copper cable has been tapped easily for years, by exploiting the fields around the cable. It
turns out that it is also possible to tap into a fiber-optic communication system. The cable if
bent acutely will begin to leak light, a photo detector can then be used to tap in. An optical-
electrical converter can be used in conjunction with a sniffer.
7|P ag e
Selected Topics EC Engineering I – Lecture IX – Fiber Optic Fundamentals Part I N Pillay
Splices and connectors are related but differ. Splices refer to connections between two
pieces of fiber cable that is meant to serve as permanent. Connectors on the other hand are
temporary and can be removable.
Figure I9a: Coupling losses
Figure I9b: Coupled to smaller N.A
These splices and connectors are complete with cable losses. Figure I9a shows different
scenarios for coupling loss. A misalignment of the fibers by the connector causes some light
to be lost. Air gaps and rough ends are also common causes. Light sources normally come
with a pigtail (a short length of fibre for joining). If the new fiber has a smaller numerical
aperture or cone of acceptance then light will again be lost. Refer to Figure I9b.
Usually fiber spools are available at around 10 metres. Thus if repeaters are placed more
than 10m apart then splices will need to be used. Splices have typical losses of 0.02 dBs.
While connectors have around 0.2dB losses. This is normally due to the small airgap
included to avoid damage to fiber ends when disconnecting and reconnecting. The next few
pages give a guide on connectors.
8|P ag e
Selected Topics EC Engineering I – Lecture IX – Fiber Optic Fundamentals Part I N Pillay
9|P ag e
Selected Topics EC Engineering I – Lecture IX – Fiber Optic Fundamentals Part I N Pillay
10 | P a g e
Selected Topics EC Engineering I – Lecture IX – Fiber Optic Fundamentals Part I N Pillay
11 | P a g e
Selected Topics EC Engineering I – Lecture IX – Fiber Optic Fundamentals Part I N Pillay
It is also possible to have couplers and switches for fiber-optic cabling and systems.
Couplers can be directional or bi-directional depending if light can travel in both directions
or not. A directional coupler is shown in Figure I10. A transmissive directional coupler is
shown in Figure I11 and a reflective bidirectional coupler is shown in Figure I12.
Figure I10: Directional
Figure I11: Transmissive directional
Figure I12: Reflective bi-directional
Passive optical couplers divide the input power among two or more outputs. There will also
be some additional loss involved in this case. These losses are normally less than a decibel.
In some cases switches might need to be used to switch optical signals from one fiber to
another. This can be done in several ways. The signal can be converted to electrical then
switched but it would be cheaper and more convenient to use light switches instead. One
possible approach would be to use a solenoid in conjunction with a prism. See Figure I13.
However this mechanical relay would be rather slow and switching speed will be seriously
limited. A more elegant approach use crystals whose refractive index changes with applied
12 | P a g e
Selected Topics EC Engineering I – Lecture IX – Fiber Optic Fundamentals Part I N Pillay
voltage. As the voltage is changed the refractive index changes and can be used to focus
beams in respective points. These types of switches are high speed.
Figure I13: Prism switch
LEDs can be used as light sources for fiber-optic communications. The output is normally
coupled to a fiber pigtail as shown in the the Burrus etch-well cross section in Figure I14.
Figure I14: Cross section of Burrus etch-well LED
LASER diodes are more commonly used. This is due to the narrower linewidth that is
available from the lasing diode. The cross section is shown in Figure I15. These diodes are
almost ideal for narrow cone cables. LASER diodes are however more difficult to operate
than LEDs. The current needs to be controlled carefully and the current also varies with
temperature. Very high currents destroy the device but currents must be above the
threshold (Figure I16) for lasing to occur. Normally thermoelectric coolers and
photodetectors are used in conjunction to limit temperature and to allow current to be
adjusted automatically.
13 | P a g e
Selected Topics EC Engineering I – Lecture IX – Fiber Optic Fundamentals Part I N Pillay
Figure I15: Cross section of lasing diode
Figure I16: Light-current relation for laser diode
At the receiving end of the fiber-optic cable a detector is employed. A optical detector like a
PIN diode or an avalanche diode is normally used. These are both operated in reverse bias
and are able to detect the light signals from the transmitter.
These devices have been covered in second and third year electronics so it will not be
repeated here.
What is dark current??
Some datasheets have been included.
14 | P a g e
Selected Topics EC Engineering I – Lecture IX – Fiber Optic Fundamentals Part I N Pillay
15 | P a g e
Selected Topics EC Engineering I – Lecture IX – Fiber Optic Fundamentals Part I N Pillay
16 | P a g e
Selected Topics EC Engineering I – Lecture IX – Fiber Optic Fundamentals Part I N Pillay
17 | P a g e
Selected Topics EC Engineering I – Lecture IX – Fiber Optic Fundamentals Part I N Pillay
18 | P a g e