Vous êtes sur la page 1sur 69

I

TABLE OF CONTENTS Page


List of Figures ........................................................................................................................... III
Table of Abbreviations ............................................................................................................. VI
Abstract .................................................................................................................................. VII
.................................................................................................................................. VIII
Chapter 1 Introduction...............................................................................................................1
1.1 Free Space Optics To The Rescue ..............................................................................1
1.2 The Local Area Network Extension Problem .............................................................1
1.3 Project Description ....................................................................................................2
1.4 Project Block Diagram .............................................................................................3
1.5 Report Layout ............................................................................................................5
Chapter 2 FSO as A Digital Communication System .................................................................6
2.1 Modulation Scheme in FSO ..............................................................................6
2.1.1 Unipolar Bi-Phase Characteristics ..................................................................7
2.2 FSO Channel Modeling .............................................................................................9
2.2.1 Attenuation ....................................................................................................9
2.2.2 Frequency Response ..................................................................................... 15
2.2.3 Induced Noise .............................................................................................. 15
2.2.4 Link Budget Equation .................................................................................. 16
Chapter 3 The Optical Transmitter ............................................................................................ 18
3.1 Light Characteristics ................................................................................................ 18
3.2 Eye Safety ............................................................................................................... 19
3.3 Driving and Conditioning Circuit ............................................................................. 20
3.4 Light-Emitting-Diodes ............................................................................................. 20
3.4.1 LED Electrical Characteristics ...................................................................... 20
3.4.2 LED Optical Characteristics ......................................................................... 23
3.5 The Designed Transmitter ........................................................................................ 25
3.5.1 Transmitter operation and simulation ........................................................... 25
II

3.5.2 PCB manufacturing and implementation ...................................................... 26
3.5.3 Practical Results ........................................................................................... 28
Chapter 4 The Optical Receiver................................................................................................. 30
4.1: Introduction ............................................................................................................ 30
4.2 Photodiodes ............................................................................................................. 30
4.2.1 Introduction.................................................................................................. 30
4.2.2 Positive-Intrinsic-Negative (PIN) Photodiode ............................................... 31
4.2.3 PIN characteristics ....................................................................................... 31
4.2.4 Responsivity ................................................................................................ 33
4.2.5 PIN and the datasheet ................................................................................... 34
4.3 Receiver Stages ....................................................................................................... 35
4.3.1 Transimpedance Amplifier ........................................................................... 35
4.3.2 Post Amplifier .............................................................................................. 37
4.3.3 Limiting Amplifier ....................................................................................... 38
4.3.4 Output Stage ................................................................................................ 38
4.3.5 Received Signal Strength Indicator (RSSI) ................................................... 38
4.4 System design .......................................................................................................... 39
4.4.1 Design Revision 0 problems ......................................................................... 39
4.4.2 Design Revision 1 ........................................................................................ 39
4.4.3 The PCB ...................................................................................................... 41
4.5 Practical results ........................................................................................................ 42
Chapter 5 The Interfacing Circuit ............................................................................................ 45
5.1 Ethernet Signaling ................................................................................................... 45
5.2 Interfacing Ethernet to optical transceivers ............................................................... 46
Chapter 6: The power supply ..................................................................................................... 49
Chapter 7 Conclusion and Problems .......................................................................................... 53
REFERENCES ......................................................................................................................... 54
Appendix (A) The Link Budget .............................................................................................. A.A
Appendix (B) Datasheets ...B.1
III

List of Figures
Figure 1. 1: LAN extension using FSO ....................................................................................................................... 2
Figure 1. 2: Transmitter Block Diagram ................................................................................................................... 3
Figure 1. 3:Receiver Block Diagram ......................................................................................................................... 4

Figure 2. 1: Manchester and unipolar bi-phase modulation ............................................ ! .
Figure 2. 2: PSD of unipolar bi-phase signaling ............................................................... ! .
Figure 2. 3:Divergence reduction using collimating lens .................................................. ! .
Figure 2. 4: Geometric attenuation ................................................................................ ! .
Figure 2. 5: Angular alignment ....................................................................................... ! .

Figure 3. 1 The Transmitter Block Diagram ............................................................................................................ 18
Figure 3. 2 The electromagnetic spectrum with respect to the wavelength ............................................................ 19
Figure 3. 3 Characteristics curves for different types of material along with a set of gap energy ............................ 21
Figure 3. 4: BAND gap energy dependency on temperature for varous materials ................................................... 22
Figure 3. 5 The definition of rise time and fall time ................................................................................................ 25
Figure 3. 6 The Designed Optical Transmitter ........................................................................................................ 26
Figure 3. 7 Simulation results ................................................................................................................................ 27
Figure 3. 8 The Top layer of the Transmitter .......................................................................................................... 28
Figure 3. 9 The Bottom layer of the Transmitter .................................................................................................... 28
Figure 3. 10 The figure shows a 10Mbps transmitted signal and the received signal, they are almost identical ....... 29

Figure 4. 1: Receiver (Rx) block diagram ................................................................................................................ 30
Figure 4. 2: Effect of intrinsic layer on the depletion region in the PIN [10] ............................................................. 31
Figure 4. 3 PIN mode...........33
Figure 4. 4 PIN Photodiode.................................................................................................................................... 32
Figure 4. 5: Current-voltage characteristics ........................................................................................................... 33
Figure 4. 6: Spectrum sensitivity ............................................................................................................................ 34
Figure 4. 7 Dark current relative to reverse voltage ............................................................................................... 34
Figure 4. 9 Responsivity ratio relative to light falling angle (Directional curve) ....................................................... 35
Figure 4. 8 Reverse voltage vs. capacitance ........................................................................................................... 35
Figure 4. 10 Transimpedance Amplifier ................................................................................................................. 36
Figure 4. 11 The Receiver (Revision 0) ................................................................................................................... 39
Figure 4. 12 The Receiver (Revision 1) ................................................................................................................... 40
IV

Figure 4. 13 Receiver (Revision 1) Simulation ......................................................................................................... 41
Figure 4. 14 The top layer of the receiver .............................................................................................................. 42
Figure 4. 15 The bottom layer of the receiver ........................................................................................................ 42
Figure 4. 16 The Receiver ...................................................................................................................................... 43
Figure 4. 17 Rx and Tx signals at 1MHz ................................................................................................................. 44
Figure 4. 18 Rx and Tx signals at various signal shape ........................................................................................... 44

Figure5. 1 The interfacing circuit ........................................................................................................................... 45
Figure5. 2 10Base-T Ethernet frame ...................................................................................................................... 46
Figure5. 3 10Base-T Ethernet signaling ................................................................................................................. 46

Figure 6. 1 Block diagram of the power supply ..................................................................................................... 49
Figure 6. 2 Protection circuit of the power supply .................................................................................................. 50
Figure 6. 3 Inverting regulator .............................................................................................................................. 50
Figure 6. 4 High current +12V to +5V regulator ..................................................................................................... 51
Figure 6. 5 -12V to -5V regulator ........................................................................................................................... 51
Figure 6. 6 The top layer of the power supply ........................................................................................................ 52
Figure 6. 7 The bottom layer of the power supply .................................................................................................. 52


V

List of Tables
Table 2. 1: The q coefficient versus visibility .................................................................... ! .

Table 4. 1: Photodiodes materials and their characteristics [12] ............................................................................ 33


VI

Table of Abbreviations
AELs Allowable Exposure Limits
AGC Automatic Gain Control
APD Avalanche Photo-Diode
AWGN Additive White Gaussian Noise
BER Bit Error Rate
BPF Band Pass Filter
DD Direct Detection
FOV Field Of View
FSO Free Space Optics
HPF High Pass Filter
IEC International Electrotechnical
Commission
IM Intensity Modulation
IR Infra Red
ISI Inter Symbol Interference
ISP Internet Service Provider
LA Limiting Amplifier
LANs Local Area Networks
LD Laser Diode
LED Light Emitting Diode
LNA Low Noise Amplifier
LOS Line Of Sight
LPF Low Pass Filter
MC Media Converter
NEP Noise Equivalent Power
NIC Network Interface Card
NLP Normal Link Pulses
PA Post Amplifier
PCB Printed Circuit Board
PD Photo Detector
PSD Power Spectral Density
QoS Quality of Service
RSSI Received Signal Strength Indicator
SNR Signal to Noise Ratio
STP Shielded Twisted Pair
TIA Transimpedance Amplifier
WANs Wide Area Networks

VII

Abstract
The intended scope of this project is to design, analyze, and implement a solution for
extended 10Mbps Local Area Network topology, employing a Free Space Optical data link. This
communication system is based on point to point optical transceivers having a line of sight
alignment. An interfacing circuit mediates between an Ethernet port of a Local Area Network
segment and an optical transmitter at one end, which in turn converts Ethernet frames into a
modulated optical signal, namely an Infra Red beam, via a Light Emitting Diode, then it sends
the optical signal through the atmosphere to the other end, where an optical receiver detects the
optical signal by means of a low noise photodiode amplifier. The role of the interfacing circuit at
the receiving end is to retrieve the Ethernet frames from the optical receiver, and transmit them
to another segment of the Local Area Network, therefore interconnecting the distant segments
into one extended network. A complete design and analysis of a free space optical, full duplex
10Mbps Ethernet link is presented. The design consists of three main building blocks, namely an
optical transmitter employing a driving circuit, and an infrared Light Emitting Diode, an optical
receiver employing a photodiode, a low noise transimpedance amplifier, a post amplifier, and a
limiting amplifier. An interfacing circuit based on logic gates and shift registers.










VIII

10
.
. Ethernet
Ethernet
.
.
Ethernet
.
10 .
.
.
.


1


Chapter 1 Introduction
1.1 Free Space Optics To The Rescue
In the current state of telecommunication technology, bandwidth hunger seems to be the
most dominant aspect among others. The rapid development of the over IP services, and the
exponential growth of data rates demand dictate the scene, and make it inevitable for
telecommunication engineers to address the issue, and explore new frontiers.
In the recent years, attention was called to a new technology that could participate in the
ever increasing pace of todays telecommunication world, that is Free Space Optics (FSO). FSO
refers to a line of sight digital telecommunication technology of broadband data transmission
through the atmosphere by means of modulated optical beams. A cutting edge technology that
could revolutionize the way we think about data streaming!
Among the most notable attributes of FSO are high speed ratings, reliability, a relatively
good coverage for short and medium range applications, inherently high security, high megabit
per dollar figure of merit, protocol independence, and license free requirements.
There are countless applications where FSO could be handy in a way or another. The last
mile loop problem, disaster recovery, metropolitan area networks, backup redundant data links,
indoor solution for short range applications, and the Local Area Connection (LAN) extension
problem, are a handful of the most important examples on such applications.
1.2 The Local Area Network Extension Problem
Local Area Networks (LANs) are known to be implemented to cover small geographical
areas with a range of tens of meters where a shared computer environment is required. A larger
coverage of broad geographical areas is implemented with Wide Area Networks (WANs), where
leased telecommunication lines are provided by an Internet Service Provider (ISP). The speed
ratings of WANs are much less than those of LANs, not to mention that WANs require access to
third party equipment, such as that of an ISP, which jeopardizes security. Therefore extended
LANs present a solution for applications where high speed data transfer, secure data links, and
relatively wide geographical areas are needed.
2

LAN extension has been used to be implemented by means of bridged wireless links,
twisted pair or coaxial cables, and fiber optics. Ethernet over Free Space Optics (FSO) presents a
viable competitor to the latter techniques. The use of FSO avoids some technical and economical
limitations of other techniques. Some advantages FSO Ethernet links over other techniques are
ease of installation, cost effectiveness, immunity against interference, and scalability. The figure
below illustrates an extended LAN topology using an FSO link.

Figure 1. 1: LAN extension using FSO
1.3 Project Description
A typical FSO system consists of an optical transmitter, an optical receiver, and an
interfacing device which mediates between the data source and the optical transceivers. An
optical transmitter consists of a driving circuit that feeds a Light Emitting Diode (LED) with an
electrical modulating signal, which in turn is converted by the optical source into a modulated
optical beam, namely an Infra Red (IR) beam, which is then collimated by a lens before being
transmitted through the atmosphere.
At the receiving end, the optical beam is focused by a receiving lens on a high sensitivity
P-I-N photo-detector. The photo-detector converts a certain level of power intensity into a
corresponding level of electrical current. The photo-detector is connected to a Low Noise
Amplifier (LNA) for the sake of an excellent Bit Error Rate (BER) performance. The LNA in
turn converts the photo-current to an appropriate voltage level before passing it to subsequent
amplification stages, needed for a proper data retrieval.
3

In Ethernet over FSO, interfacing a LAN segment via an Ethernet interface to the FSO
system is accomplished by means of a Media Converter (MC). The role of the MC is to encode
some Ethernet special pulses before passing the Ethernet frames to the optical transmitter, and
decode those pulses at the receiver end before passing the received signal back to an Ethernet
interface of another LAN segment.
FSO requires a clear Line Of Sight (LOS) between the transmitter and the receiver so as
for the collimated optical beam to propagate through the atmosphere without being intercepted.
A proper alignment practice is required as well. Misalignment would result in a complete
dropout or degrading of the link in terms of the effective distance.
The fundamental limitations of FSO communications are imposed by atmospheric
phenomena, which tend to attenuate the optical power in the transmitted beam, hence decreasing
the link distance. Such phenomena are scattering and scintillation. Although insignificantly
affected by rain and snow; FSO communication systems Quality of Service (QoS) can be
extremely degraded by fog and atmospheric turbulence. These phenomena should be accounted
for when designing an FSO system.

1.4 Project Block Diagram
Figure 1.2 shows a block diagram of the transmitter circuitry, and Figure 1.3 shows a
block diagram of the receiver circuitry.

Figure 1. 2: Transmitter Block Diagram

4


Figure 1. 3:Receiver Block Diagram
The transmitter interfacing circuit receives Ethernet frames from a 10Mbps Ethernet
LAN port at one end of the system, and provides the LED driver with the an appropriate signal
for transmission. The LED driving circuit amplifies the signal fed by the transmitter interface,
and provides the LED with the current level needed for a proper modulation.
At the receiving end, the intercepted optical beam generates a corresponding photo-
current in the PIN photodiode. The Trans Impedance Amplifier (TIA) converts the photo-current
into a corresponding voltage level. The TIA possesses a Low Pass Filter (LPF) characteristic,
which tend to integrate the received signal, and change its shape, therefore the post amplifier
compensates for the integrating effect of the TIA, and amplifies the signal further more. The
Limiting Amplifier (LA) ensures that the amplified signal swings between two definite voltage
levels acceptable by the receiver interfacing circuit. The Received Signal Strength Indicator
(RSSI) indicates the strength of the optical beam intercepted, and provides a way to measure the
misalignment losses, hence helping with a proper alignment practice. The receiver interfacing
circuit retrieves the Ethernet frames, and transmits them to a 10Mbps Ethernet LAN port at the
other end of the system.

5

1.5 Report Layout
This report is divided into six chapters. A brief description of each chapter is shown below:
In chapter 2 FSO system is analyzed as a digital communication system in terms of modulation
scheme, channel modeling, and link specifications.
In chapter 3 a complete design and analysis of the optical transmitter is presented, where LED
operation is illustrated, and a driving circuit is developed.
In chapter 4 a complete design and analysis of the optical receiver is presented, where
photodiode operation is analyzed, and a structured design of a TIA, a post amplifier, an LA is
developed.
In chapter 5 Ethernet operations is presented in terms of IEEE standards concerning 10base-TX
Ethernet. An interfacing circuit in compliance with 10base-TX and 10base-FL Ethernet standards
is analyzed.
In chapter 6 the power supply is introduced.
In chapter 7 conclusions and results of the design and analysis are listed.




6

Chapter 2 FSO as A Digital Communication System
2.1 Modulation Scheme in FSO
The selection of the right modulation scheme is of a great importance when designing an
FSO system. The modulation technique employed defines important parameters such as
bandwidth, bandwidth efficiency, and Bit Error Rate. However, there are a handful of
modulation techniques suitable for FSO transmission to be chosen among. One of the most
popular forms of modulation in FSO systems is Intensity Modulation (IM), where the waveform
of the information signal is used to modulate the instantaneous power intensity of the transmitted
energy. Power intensity is defined as the optical power emitted per solid angle. At the receiver
end; the information is recovered using a technique called Direct Detection (DD) in the sense
that the output photo-current of the photo detector is approximately proportional to the received
irradiance. Irradiance is defined as the incident power at a certain surface per unit area. [1]
In IM/DD links, the electrical data signal is used as the input of the transmitter driver,
where the instantaneous optical power intensity is modulated in accordance to the input electrical
current signal. The information sent on the channel is not contained in the amplitude, phase or
frequency of the transmitted optical waveform as it is in the heterodyning systems, but rather in
the intensity of the optical beam. The electro-optical conversion process is accomplished by a
Light Emitting Diode or a Laser Diode (LD). The opto-electrical conversion is usually performed
by a P-I-N photodiode, or an Avalanche Photo-Diode (APD).
This transmission-detection technique defines three fundamental characteristics of the
modulation scheme to be used in FSO systems:
1- In an IM/DD channel, the modulating current must be unipolar positive only, as the LED
or LD cannot be excited by a negative current.
2- The input power at the receiver is unipolar positive only, as photodiodes cannot be
excited by negative irradiance.
3- The LED modulating signal is preferably a baseband signal, in the sense that the signal
must have a zero or positive levels, thus having a frequency spectrum which is
necessarily measured from zero up to a certain frequency.
7

According to the latter characteristics, the modulation scheme chosen for this project is
unipolar bi-phase modulation. This scheme is a unipolar version of the Manchester line
encoding, which is used in 10 Mbps Ethernet LANs, therefore choosing a close version to
Manchester encoding as a modulation scheme is a fair justification, in order to simplify the
design of the optical transceivers.
2.1.1 Unipolar Bi-Phase Characteristics
As mentioned before, unipolar bi-phase modulation is a unipolar version of the
Manchester line encoding. In Manchester encoding used in 10 Mbps Ethernet; a bit 1 is
represented by a low to high transition of the signal, while a bit 0 is represented by a high to low
transition. The low level is a negative voltage level, and the high level is a positive voltage level
equal in magnitude to that of the negative voltage. In unipolar bi-phase modulated signal; a bit 1
is represented by a transition from zero voltage to a certain positive level, while a bit 0 is
represented by a transition from that positive level to zero voltage. The figure below illustrates
the Manchester and unipolar bi-phase modulation.

Figure 2. 1: Manchester and unipolar bi-phase modulation
According to [2], The Power Spectral Density (PSD) of unipolar bi-phase signal with bit
duration T, and pulse amplitude A can be expressed by:

2

2

2
1 +

=
. (2.1)

8

The figure below is a sketch of the PSD of unipolar bi-phase signaling:

Figure 2. 2: PSD of unipolar bi-phase signaling
An investigation of the PSD reveals that the first null-to-null bandwidth of a unipolar bi-
phase signal is given by:
=
2

. (2.2)
That means, the bandwidth of a unipolar bi-phase signal is twice of the bit rate.
Therefore, 10Mbps operation requires a system bandwidth of 20 MHz. Therefore, this
modulation technique is considered bandwidth wasteful, where the bandwidth efficiency is 0.5
bit per second/hertz.
It should be noted that the impulses located at

, for |n|=1,3,5could be used in


conjunction with a Phase-Locked Loop (PLL) to recover the data clock, and therefore could be
used to synchronize the transmitter and the receiver on the bit level[2]. Nevertheless in Ethernet
technology synchronization is accomplished on the frame level by means of a special segment of
the Ethernet frame called the preamble segment.
The Bit Error Rate (BER) of bipolar bi-phase modulation is given by [4]:
=
1
2

2
. . . (2.3)
9

erfc: is the complementary error function
SNR: is the Signal to Noise power Ratio at the receiver front end.
As a measure of system performance; it is customary in FSO systems to accept a
maximum BER that ranges between 10
12
and 10
9
. According to the approximation:
1
2

2
2

2
. . (2.4)
For BER = 10
9
, SNR should be equal to 36 , which is equivalent to 15.6dB
[4]
.

2.2 FSO Channel Modeling
A proper design of an FSO communication system requires a detailed understanding of
the channel in terms of power attenuation, frequency response, and induced noise. In this section
the channel model for IM/DD links is considered.
2.2.1 Attenuation
The free space optical link suffers from two kinds of attenuation: geometric attenuation
and atmospheric attenuation.
1: Geometric attenuation: This attenuation is due to the geometric arrangement of the
transmitter and receiver along with the optical beam model. In FSO systems which use LEDs as
beam sources; the lambertian distribution is used to model the beam profile of the LED, that is:
the power intensity distribution of the beam follows the lambertian cosine law which is given
by[5]:
(, ) =
+1
2
2

(cos)
m

2
. . . . . (2.5)
(, ): the power intensity measured at a distance R and angle

: The total transmitted power


R: The distance at which the intensity is being measured
: The angle between the emitters normal and the point of measurement .
m : is the mode number of the radiation lobe , and is given by :
=
2
(1
2
)

10

Where 1
2
is the half power angle or divergence angle, which is the angle that creates an
illumination cone containing half the transmitted power . 1
2
is usually given by the manufacturer
datasheet .
On the transmitter side, the beam source i.e the LED transmits an optical beam with a
certain initial divergence. As the beam propagates through the atmosphere; diffraction takes
place, and the beam suffers from further divergence. Divergence is a measure of how fast the
beam expands as it travels away from the emitting source. This phenomenon is caused by
diffraction, which is the change of the beam direction of propagation, or pattern due to variations
in the medium and/or interference between different beam components. The beam divergence
causes the power intensity of the beam to be decreased away from the axial center of the beam as
it travels through the atmosphere, thus a portion of the optical power would be lost if the receiver
apparatus wasnt able to collect the beam properly. The receiver apparatus is a collecting
biconvex lens, whose foci is centered on the photo detector.
The effect of the geometric attenuation for a given range can be reduced by decreasing
the beam divergence. Hence collimation is needed. Collimation is defined as the process of
decreasing the divergence angle of the transmitted beam by means of a collimating lens or a
beam expander. The divergence of a lambertian optical transmitter such as an LED can be
decreased by employing a collimating lens, namely, a biconvex lens. The divergence angle due
to the lens is expressed by [3]:
= 1
2

. . (2.5)
: is the divergence angle due to the collimating lens
1
2
: is the divergence angle of the LED
E: is the LED diameter of emission
L: is the diameter of the collimating biconvex lens
It should be noted that the LED should be placed in the foci of the lens to achieve
maximum collimation. The figure in the next page illustrates the concept of collimation by
means of a biconvex lens.
11


Figure 2. 3:Divergence reduction using collimating lens
The narrow beam attribute is important for security considerations, where intercepting the
beam is practically a very difficult task, which maintains the credibility of the system.
In this project, a LED transmitter is used. Therefore, for an optical beam generated by a
LED, assuming a lambertian emission pattern, and low divergence angle, and a perfect
alignment, an analytic treatment of geometric attenuation results in the following approximate
formula [4].

+ tan

2
. (2.6)

: is the geometric attenuation


: are the received and transmitted powers respectively


: are the receiver and transmitter apparatus radii respectively


: is the link spatial distance
: is the divergence angle of the transmitted beam after collimation.
The figure below illustrates the concept of geometric attenuation.

Figure 2. 4: Geometric attenuation
It should be noted that the relation above is valid for a beam propagating in a clear line of
sight. To ensure the validation of this assumption; there should be a clear line of sight with a
minimum radius of the Fresnel zone plus the beam diameter at the transmitter[17]:

4
+

. (2.7)
Where is the wavelength of the transmitted beam.
12

Close to geometric attenuation are the misalignment losses. There are two types of
alignment to be considered: lateral alignment, and angular alignment. Lateral alignment occurs
when the axial line of the illumination cone generated by the transmitter is parallel to that of the
Field Of View (FOV) of the receiver, while angular alignment occurs when there is no angular
displacement of both the transmitter and the receiver from the systems Line Of Sight (LOS). If
the transmitter and the receiver are allowed to rotate freely while fixed at the center of rotation;
angular alignment ensures the occurrence of lateral alignment. Ergo, attention should be drawn
to angular alignment alone. The figure below illustrates the concept of angular alignment.

Figure 2. 5: Angular alignment
A small angular displacement of the transmitter would result in a relatively large spatial
displacement of the illumination cone at large distances. This displacement will result in a
complete drop of the link if the photo-detector ceases to view the illumination cone. Power loss
will occur if the photo-detector views a portion of the illumination cone.
For an exact treatment of the problem in hand; Eq.(2.6) is not sufficient, as it doesnt take
into account angular displacement of the transmitter and the receiver from the line of sight. To
include the effect of angular displacement another relation is needed. According to [18] and
others references, the exact relation that relates the transmitted and received powers to spatial
and angular coordinates is given by:

=
+ 1
2
2



2
cos . (2.8)
0


13

Where:

: is the geometric attenuation


: are the received and transmitted powers respectively


: is the receiver apparatus radius


: is the link spatial distance
: is the divergence angle of the transmitted beam after collimation.
: is the angular displacement angle of the receiver from the axis of the transmitted beam.
g : is the receiving lens concentration gain, and is a function of the radii of the photo
detector and the receiving lens , and . It attains maximum value when = 0.

: is the angle of view of the receiver.



Its obvious from the last relation that if the angular displacement angle exceeds the
angle of view of the receiver (

) there will be a complete drop of the link, as the receiver


ceases to view the illumination cone of the transmitter. Another result - a rather trivial one- of the
latter relation is that the smaller the displacement angle, the smaller the attenuation gets.
2: Atmospheric attenuation
As the beam propagates through the atmosphere, it suffers from several atmospheric
effects that tend to attenuate the transmitted power. Scattering, absorption and scintillation are
the most influential among these effects.
The effect of scattering and absorption is governed by Beer-Lambert law for gases [5]:
] ) ( exp[
) 0 , (
) , (
) , ( R
P
R P
R

t = = ... Eq. (2.12)


) , ( r t : is the atmospheric attenuation factor due to scattering and absorption.
) 0 , ( P : is emitted power
) , ( R P : is power at a distance R from the transmitter
) ( : is atmospheric attenuation (extinction) coefficient
The Krsue relation expresses the extinction coefficient as follows:
q
nm
nm
V

= )
550
(
912 . 3
) (

dB/Km.Eq. (2.13)
V: is the visibility in Km
nm
: is the beams wavelength in nanometer.
Q: empirical coefficient that is given in the following table:


14

Q Visibility (V)
1.6 V>50Km
1.3 6Km<V<50Km
0.16V+0.34 1Km<V<6Km
V-0.5 0.5Km<V<1Km
0 V<0.5Km
Table 2. 1: The q coefficient versus visibility
Visibility is defined as the length of path in the atmosphere required to reduce the power
in a collimated beam from an incandescent lamp at a color temperature of 2700K to 0.05 of its
original value. Visibility is usually measured at airports, and weather forecast stations.
Scintillation attenuation can also be of significant contribution in a turbulent weather.
Scintillation can be defined as the spatial and temporal dependence of power intensity due to
thermal fluctuations in the index of refraction along the transmission path. These changes cause
the atmosphere to act as filled with small lenses deflecting portions of the beam. The behavior of
scintillation is of a statistical nature. The attenuation due to scintillation in dB for a plane wave
can be expressed as follows [5]:
6
11
2
6
7
2
17 . 23 2 R C
n x
|
.
|

\
|
=

t
o ... Eq.(2.14)

2
: is turbulence intensity, and assumes a value in the order of 10
14
for weak disturbance.
R: is the distance at which scintillation is measured.
It is obvious that the atmospheric attenuation due scattering and scintillation can be
reduced if longer wavelengths are used. Therefore IR beams are preferred to be used in FSO
systems. In this project, we are using a near IR beam emitter, with a wavelength of 875jnm.
The attenuation due to rain is given by:

=
1.076
0.67

. . . . 2.15
The attenuation due to snow is given by:

/. 2.16
S: is the snow fall rate (mm/hour)
a= 3.7855466+ 10.23x10
-5


, b=0.72 for wet snow
15

a=5.49588776+5.42x10
-5


, b=1.38 for dry snow
Other losses not accounted for include losses in the apparatus lenses. For instance ZnSe
lenses have losses not higher than 3%.
2.2.2 Frequency Response
For optical communication systems in general, the only phenomena that causes the signal
to possess a frequency dependent behavior is dispersion. Dispersion is defined as the phase or
group speed dependence on frequency. Dispersion causes pulse spreading, and Inter Symbol
Interference (ISI). In normal conditions, the atmosphere isnt considered to be a dispersive
medium, thus acting as an ideal filter, and doesnt impose ISI on the transmitted optical signal.
2.2.3 Induced Noise
In FSO systems, two types of induced noise should be accounted for: background noise
and receiver end electrical noise. The background noise is due to the ambient irradiance, which is
dominated by the sunlight at daytime. The ambient illumination of the sunlight generates an
additional current in the photo detector, and creates an Additive White Gaussian Noise (AWGN).
The radiation power generated by the sunlight is given by [1]:
P
sun
= F

A
Rx
. Eq. 2.18
F

: is the sunlight irradiance spectral density at earth surface level at a wavelength


A
Rx
: is the receivers effective area
is the receivers optical bandwidth
Apparently, this type of noise can be reduced by decreasing the optical bandwidth of the
receiver by employing an optical bandpass filter. The photodiode used in this project has one
with optical bandwidth of 300nm.
The noise induced by the receiver front end consists of two parts: shot noise, and thermal
noise. Shot noise is due to the random movement of the carriers in the photodiode, and thermal
noise is due to the thermal energy of electrons in feedback resistors of the photodiode amplifier.
The equivalent Root Mean Square (RMS) of shot and thermal noise current is given by:


2
= 2

+
4

+

2

2
. . (2.19)
e: is the elementary (electron) charge = 1.6x10
-19
C
16

: is the amplitude of the photocurrent generated due to the received signal.

: The dark current of the photodiode (to be discussed in chapter 4)


B: The systems electrical bandwidth.
K: Boltzmann constant = 1.38x10
-23
J/K
T: the systems equivalent temperature in Kelvin, usually taken as the room temperature
R: the feedback equivalent resistance used in the amplifier.

: The amplifier noise current.


After considering all types of noise, SNR of unipolar bi-phase IM is given by:
=


2
+ 2P
sun
. . (2.20)
Where (r) is the responsivity of the photodiode in A/W (to be discussed in chapter 4).
It is obvious from the last relation that the SNR can be increased by increasing the
photocurrent, and limiting the receivers electrical bandwidth. The photocurrent can be increased
either by increasing the received power, or using a photodiode with higher responsivity.
Responsivity of the photodiode is proportional to its area, therefore it is appropriate to use wide
area photodiodes in FSO systems. Its customary to limit the electrical bandwidth of the receiver
to a range of 70% to 100% of the received signal bandwidth.
2.2.4 Link Budget Equation
The link budget equation is used to predict how much margin, or extra power is available
in a link under a set of conditions. The more margin the link possesses, the less the probability of
drop. Therefore, the link budget equation helps with predicting the availability of the system, in
terms of downtime as a Key Performance Indicator (KPI). The link budget equation helps design
the transmitter in terms of transmitted power, and the receiver in terms of sensitivity. Sensitivity
is defined as the minimum detectable power at the receiver end needed for a reliable
communication, that is its the minimum power required to provide an SNR value acceptable for
a reliable communication . The link budget equation is given by:

+

+

. (2.17)
) (dB M
link
: Link margin
17

) (dBm P
Tx
, ) (dBm S
Rx
: total transmitted power and receiver sensitivity respectively
) ( ), ( dB A dB A
c atmospheri geometric
: Geometric and atmospheric attenuations respectively
) (dB A
other
: Other losses, such as losses due to imperfect lenses.
Numerical analysis of the link budget equation is presented in Appendix A





18

Chapter 3 The Optical Transmitter
The transmitter aims to produce a suitable signal for transmission; the transmitter is the
stage between the interfacing circuit from one side, and the channel from the other side, the
transmitter will take its input signal from the interfacing circuit, and it will put its output to the
channel, so the design of the transmitter will greatly depend on the output of the interfacing
circuit, and the characteristic of the channel.
The transmitter mainly consists of three parts, the first part is a conditioning and driving
circuit, the main purpose of this part is to take the data signal from the interfacing circuit and
modify it such that to provide the light source with a suitable driving signal to emit the required
optical signal, also this part should compensate for any undesirable effect or noise that may
happen for some factors like the Light Emitting Diode (LED) temperature variation or any
injected noise to the system. The second part is a light source that emits an optical signal suitable
for transmission, this optical signal is an emitted light for a certain period of time with a certain
power to represent high LOGIC level, and emits nothing to represent low LOGIC level, the
optical light source could be of two types: (1) a Light Emitting Diode (LED) which was chosen
to be the light source in this project, (2) a laser diode [5]. The third part is the lenses that are used
to direct the output of the light source properly to obtain a good distance between the transmitter
and the receiver.



In this chapter, we are going to discuss the designed transmitter and its operation,
including the light characteristics and safety issues, the transmitter main blocks, physical
implementation which will include PCB implementation, finally, practical results.
3.1 Light Characteristics
Light is considered as a part of the electromagnetic spectrum, the electromagnetic spectrum
contains a vast range of frequencies with comparable nature, and each band of frequencies has its
own name due to historical reasons, this is especially true with optics, optics include the band of
Conditioning and
Driving Circuit
Light Source
(LED or LD)
Lenses
Figure 3. 1 The Transmitter Block Diagram
19

frequencies of invisible light, we are mainly interested in the IR band in the electromagnetic
spectrum, hence we use IR LED as a light source in the project.
Figure (3.2) shows the electromagnetic spectrum, as we can see, infrared wavelength lies
between (80nm-1mm) with a corresponding frequency range of (300 GHz-300 THz).

Figure 3. 2 The electromagnetic spectrum with respect to the wavelength with the usual names for the various frequency bands
Light characteristics is covered by Maxwells equations, however the analysis process
using electromagnetic theory is quite tedious and requires a tremendous mathematical
background, the modeling of light behavior using geometrical optics is much easier and it is
more suitable for our application, thus in our discussion we always assume the geometrical
model of light, for farther analysis of light behavior using geometrical optics see reference [6].
As it was mentioned in Chapter 2; IR is most preferable to be used in FSO systems as
opposed to visible light. The use of IR beams allows for a significant reduction of the
atmospheric losses in terms of absorption and scintillation. Therefore, for this project it was
decided to use an IR LED as a transmitting source.
3.2 Eye Safety
One of the most important aspects in FSO system design is the safety of the user, and this
is also the main limitation factor in the design process, for example, if the output emitted power
is increased, we can compensate for the attenuation factors mentioned earlier in Chapter 2, this
would yield in increasing operating distance, reducing the error by increasing SNR,
unfortunately the emitted power is restricted for safety operation [5]. Infrared, visible and
ultraviolet radiation can cause damage to the human eye if the power at their respective
wavelengths is increased over a specified value. The limits to the output power for safety
operation is specified by the International Electrotechnical Commission (IEC), the IEC specifies
the Allowable Exposure Limits (AELs), these limits make sure that the output power by any
20

optical source is under regulation and is safe for use and doesnt require warning labels, any
product that employs LEDs or LDs is subject to the IEC 6082s-X standard in the European
countries [5].
3.3 Driving and Conditioning Circuit
The main role of the driving and conditioning circuit is to provide the light source with
the suitable signal for transmission; in this case, the light source is a current-driven device whose
brightness is proportional to its forward current. Forward current can be controlled in two ways;
the first is to apply a direct voltage source connected with a series resistor along with the LED,
this produces a current which is directly proportional to the input voltage , this gives us a simple
designed circuit, on the other hand, the produced current greatly depend on the forward voltage
of the LED, and this voltage depends on other factors such as the operating temperature, so this
give us a large variation in the current for a small variation in the LED forward voltage, and also
this will produce a wasted power in the added resistor. The second method is to drive the LED
with a constant current source; this will eliminate the variation of the current due to the variation
of the LED voltage, producing a more stable output optical signal from the LED. [7]
For our particular design, the transmitter gets the input signal from the interfacing circuit,
this signal is a square wave with (-0.35 to 0.35) V peak to peak voltage level values, the key
function of the transmitter is to shape this signal without adding any noise, and with a stable
operation so that the light source (in our case the LED) will be driven with the appropriate
current to transmit the data.
3.4 Light-Emitting-Diodes
A Light Emitting Diode (LED) is a special type of diodes, in the following discussion we
will talk about LED electrical characteristics and optical characteristics.
3.4.1 LED Electrical Characteristics
In this subsection, we will discuss the LED electrical properties; this includes, the LED I-
V characteristic, LED modeling with discrete components and there explanation, the emission
energy and the temperature dependency of the LED voltage.
LED I-V Characteristics
The current-voltage characteristic of a PN junction was first introduced by Shockley, Shockley
equation is given by:
21

=

1 . (3.1)
Is: is the saturation current
e: elementary charge
k: Boltzmann constant
T: Temperature in Kelvin
Equation (3.1) shows that the current in the LED is exponentially dependant on the
forward voltage of the LED, this means that a small variation in the forward voltage will produce
a large variation in the forward current (I), also we can note from equation (3.1) that the current
increases rapidly after a certain value of threshold voltage

, it can be shown that the threshold


voltage

, can be approximated using the band gap energy

, the band gap energy is defined as


the energy required for an electron in the last orbit to get to the conduction orbit and be out of the
nucleus attraction range, the threshold voltage is given by equation (3.2) [8]:

. . (3.2)
The band gap energy depends on the material used to construct the PN junction, we can
find the threshold voltage using equation (3.2). Figure (3.3) shows the I-V characteristic curves
for different materials along with the band gap energy.










LED Internal Resistance And Capacitance
LEDs have internal resistance due to the n-type and p-type regions, even though the I-V
characteristics is exponential; this resistance is only a model for the relationship for our
Figure 3. 3 Characteristics curves for different types of material along with a set of gap energy
22

application, we can assume a linear relationship between the current and voltage after the knee
voltage (threshold voltage), this resistance includes the effect of any metallic contact attached to
the LED and any imperfection in the junction, usually this internal resistance is given in the
datasheet of the LED manufacturer, otherwise we can find its value using experimental
approach .
The LED has also an internal capacitance, when the LED is reversed biased, the depletion
region is empty of any charges and acts as an insulator, while the p-type and the n-type of both
sides of junction act as plate with different charges, this effect can be assumed as a parallel plate
capacitor, usually LED manufacturer tend to specify the value of this capacitance.
Another type of capacitance is the junction capacitance, when the LED is forward biased,
the carrier tend to diffuse across the junction, these carriers take time before recombination and
this induced charges to be stored at different energy bands near the junction, this simulates a
capacitance effect which is called junction capacitance [9].
Emission Energy
Energy is produced when electrons and holes recombine in the PN junction; they produce
a photon with energy equals to the band gap energy given by [8]:

= . . (3.3)
h: planck constant
v: photon frequency
Temperature Dependence On LED Voltage
The I-V characteristics of LED are given by Shockley equation previously introduced:
=

1. (3.4)
We can approximate the voltage dependency on temperature by:

. . (3.5)
The band gap energy depends on the temperature, as the temperature increases, the band
gap energy decreases, the following formula describe the relationship between

=
|=0

2
+
. . . . (3.6)
23

, are constants
T: Temperature
The following figure shows curves for for different materials, the voltage depends on
temperature comes directly from the band gap energy temperature dependence.









3.4.2 LED Optical Characteristics
In this subsection, we will discuss the LED optical characteristics; this includes optical
rise and fall time definitions
Optical Rise And Fall Times
LEDs are non-linear devices, their internal resistance and the internal junction
capacitance is greatly depend on the forward voltage of the LED, in studying the rise and fall
times of an LED, these non-linearity behaviors should be accounted for, however, a simplified
analysis could be obtained if we consider the LED as a linear device, in the following analysis,
we assume a linear device behavior of an LED.
There is a good analogy between the definition of an LED optical rise time and fall time
and an RC circuit behavior [8]; we recall that the definition of a rise time is the time required for
the step response of a system to rise from 10% to 90% of its final value, and we define the
optical rise time as the time required for the optical power of an LED to rise from 10% to 90% of
its final value.
Figure 3. 4 Band gap energy dependency on temperature for various materials
24

Recall that for an RC circuit, the voltage output in the rising mode is given by:

=
0
1

1
. . (3.7)
And the output voltage in the decaying mode is given by:

=
0

2
. . (3.8)
Where the

is the final value and


1
=
2
= (time constants).
Now we can find the rise time (

) and fall time (

) using the defined time constant above:

= 2.2
1

= 2.2
2
. . . . . (3.9)
The voltage transfer function () is given by
= (1 + )
1
. . . (3.10)
Where is the frequency response of the system.
The bandwidth of a system, correspond to the frequency at which the power transmitted
through the system is reduced to half of its low-frequency value. This condition can be written
as:
||
2
=
1
2
. . (3.11)
=
3
=
1
2

0.70
(+)
. . . . (3.12)
Now enough with the RC circuit, lets turn our attention to LEDs, assume the LED rise time is

.
As a step-function input current is applied to the LED, the optical output power increases
according to

=
0
1 exp

. . . . . . (3.13)
Where is time constant, note this equation is based on the analogy between the RC circuit and
the LED operation as a linear device, particularly this equation is an analogy with equation (3.8).
Also in analogy to equation (3.10), the power transfer function is given by:

2
= 1 +
1
. . . (3.14)
The absolute value of the power transfer function is reduced to half at the 3 dB frequency of the
LED. Thus the 3 dB frequency of an LED is given by
=
3

1.2
( +)
. . (3.15)
25











Even though these equations are delivered assuming a linear device and with an analogy to the
RC circuit, still they give a good perspective for the main concept and terminology of the optical
rise and fall times, the following figure shows the analogy between the RC circuit and the linear
LED device [8].
In practice, the rise and fall time are measured experimentally, operating the LED with a
pulsating current input, and having a faster photodiode to capture the LED output power and
convert it into output current, we can approximate the rise and fall time of the LED from this
output current [8].
3.5 The Designed Transmitter
In this section, we are going to talk about the Transmitter design, simulation of the
design, practical verification and results for the transmitter, PCB design and physical
implementation.
3.5.1 Transmitter operation and simulation
As mentioned earlier in this chapter, the transmitter gets the input signal from the
interfacing circuit which is a Manchester coded signal, Figure 3.6 shows the designed
transmitter, the input signal used for simulation is a square wave (-0.35 to 0.35) V, 10Mbps, as
shown in Figure 3.6, Stage 1 in an inverting amplifier, with a gain of (10),
Figure 3. 5 THIS FIGURE SHOWS THE DEFINITION OF THE RISE TIME AND THE FALL TIME OF AN LED WITH AN ANALOGY TO THE
RC CIRCUIT. [8]
26

R1
1k
R2
10k
R3
909
R4
75
0
0
0
V1
TD = 0
TF = 1ns
PW = 50ns
PER = 100ns
V1 = -0.35
TR = 1ns
V2 = 0.35
C7
10n
R23
10k
VCC
1 2
3 4
5 6
9 8
11 10
13 12
1 2
3 4
5 6
9 8
11 10
13 12
1 2
3 4
5 6
9 8
0
VCC
5
11 10
13 12
0
VEE
-5
VCC VEE
0
VCC
VEE
C1
10n
C2
10n
0
0
+C5
10uF
STAGE 1
+ C6
10uF
STAGE 4
STAGE 3
+
3
-
2
V+
7
V
-
4
OUT
6
V STAGE 2
Dbreak
D1
R22
10
Stage 1 in an inverting amplifier, with a gain of (10), this will amplify the input signal in order to
use it to drive the LED, the output of Stage 1will be fed to Stage 2, which will work as a DC
level shifter, the Capacitor will work to raise the signal to minimize the negative values as shown
in Figure 3.7 as shown in the blue waveform, the resister next to the capacitor is a pull up
resister, which makes sure that no unidentified logic level will be fed to the inverters, finally the
output of the inverters is proportional to the input signal, which will drive the LED to produce
the suitable intensity for communication.
3.5.2 PCB manufacturing and implementation
In this subsection, we provide a summary for the PCB layout used to implement the
designed transmitter, in general we used chemical approach to manufacture the PCBs, using a
combination of HCL solution with a 33% concentration of an H2O2 solution with a
concentration of 3% and double the amount of the HCL solution, we were able to find a way to
etch the Copper layer of the PCB and leaving the desired traces, OrCAD Layout Plus was used
as the Layout program Figure 3.8 shows the Top layer of the transmitter, Figure 3.9 shows the
bottom layer.

Figure 3. 6 The Designed Optical Transmitter
27

Time
0s 50ns 100ns 150ns 200ns 250ns 300ns
V(R4:2)
-400mV
0V
400mV
-350.000mV
350.000mV
V(U31:OUT)
-4.0V
0V
4.0V
SEL>>
-2.7802V
3.4665V
V(U26A:A)
0V
2.5V
5.0V
-523.643mV
5.4046V
V(U27F:Y)
0V
2.5V
5.0V
466.576uV
4.7496V
I(D1)
-200mA
0A
200mA
400mA
345.819mA
-156.338nA



Figure 3. 7 simulation results of the optical transmitter for different stages, the first waveform from above is the current fed to the LED, the second waveform from
above is the output voltage of the inverters, the third waveform from above is the input to the inverters, the fourth waveform from above is the output of the amplifier,
finally, the last waveform is the input waveform
28


Figure 3. 8 the Top layer of the Transmitter

Figure 3. 9 The Bottom layer of the Transmitter
3.5.3 Practical Results
In this subsection, we are going to state the result of the transmitter, these results guarantee a
healthy operating transmitter that is able to transmit a 10Mbps without adding any distortion
optically, the following set of pictures shows practical results of the transmitter, Figure 3.10
shows the input signal of the transmitter 10Mbps, this signal will be passed to the LED after
certain amplification and modification, then transmitted optically to the other end the optical
receivers.

29

In this chapter, we managed to design, understand and implement an optical transmitter which is
able to transmit a 10Mbps signal optically without changing and affecting the transmitted data.






Figure 3. 10 The figure shows a 10Mbps transmitted signal and the received signal, they are almost identical
30

Chapter 4 The Optical Receiver
4.1: Introduction
A receiver should be able to detect any incoming optical signal and transform it into
digital meaningful electrical signal. The initial detection of the signal is done by a Photo Detector
(PD), the PD output signal shall be modified using a conditioning circuit to give us the required
digital signal shape for the interfacing circuit to understand. Figure 4.1 shows the block diagram
of typical Rx; we will go through each block in details in the coming sections.

Figure 4. 1: Receiver (Rx) block diagram
PDs are capable of supplying current (instead of voltage) that is proportional to the
received optical power, so our first task is to transform this current into voltage by means of
Transimpedance Amplifier (TIA), this voltage is reshaped and boosted by means of Post
Amplifier (PA) and Limiting Amplifier (LA), now the signal being ready is transferred to the
interface circuit.
Our system shall be able to discrete signals from noise and maintain constant logic levels
at its output whatever was the detected signal level.
4.2 Photodiodes
4.2.1 Introduction
Photo-detectors are divided into two types, Positive-Intrinsic-Negative (P-I-N)
photodiodes, and Avalanche Photodiode (APD). Photo detector produces current that is relative
to the power intensity it receives. PDs have a detection range of signal wavelength with a peak
wavelength at which it achieves its maximum responsivity, i.e. current to power ratio.
Responsivity is proportional to the PD surface area. To be able to generate electrical current;
electrons should gain energy above the band gap energy value thus having enough energy to
become in conduction mode.

31

4.2.2 Positive-Intrinsic-Negative (PIN) Photodiode
Earlier we stated that PDs are divided into two types, Positive-Intrinsic-Negative (PIN)
and Avalanche Photodiode (APD). PIN photodiodes are similar to regular diodes except that they
contain an additional intrinsic layer as the name suggests. The added layer results in larger
depletion region thus increasing the spectral bandwidth. The PIN response is somewhat selective,
that is: photons penetration depends on the wavelength, and only photons that release carriers
within or near depletion region do contribute in the output current. To fully benefit from the
intrinsic layer, reverse biasing is normally required, this insures that the depletion region covers
the whole intrinsic layer; this extends the electric field of the depletion region along the intrinsic
layer and more accelerated carriers are being generated. The added layer increases the transit
distance for carriers but the low doping of the layer assures most of the carriers complete
transition. Figure 4.2 illustrates the layout of a PIN photodiode, and the effect of the intrinsic
layer.

Figure 4. 2: Effect of intrinsic layer on the depletion region in the PIN [10]
On general PINs have wider spectral bandwidth than APDs, while APDs have superior
performance in terms of sensitivity and response time, they are a lot more costly and need very
high reverse biasing voltages [10].

Due to price it was chosen to use a PIN as a PD for this
design.

4.2.3 PIN characteristics
The PIN PD can be modeled using a circuit similar to the one seen in Figure 4.3, where
the current source represents the current generated by the incident photons, the diode represents
32

the PN junction, and the junction capacitance and resistance are modeled using the capacitance
C
j
and resistance R
sh
respectively. R
sh
is ideally infinity, in practice its the slope of the current-
voltage curve of the photodiode at origin, the higher the shunt resistance the better is the PD
performance. The capacitor results from the two boundaries of the depletion region which
happens to be parallel-plate like, thus the capacitance is inversely proportional to the depletion
region width and is proportional to the surface area; it is used to determine the response speed.
The series resistance R
s
results from the undepleted semiconductor material, it affects the rise
time and the linearity of the photodiode.
PIN photodiodes are generally limited by two factors: the parasitic junction capacitance
C
j
and the transit time. As we mentioned earlier C
j
is proportional to the surface area cross
section, yet reducing the surface area will badly reduce the responsivity and hence the resultant
output current, it will reduce the spectral efficiency too. C
j
is also inversely proportional to the
depletion region length (L
i
), but likewise increasing the L
i
will increase the spectral efficiency
and increase the response time. So a compromise should be taken between current (I),
capacitance (C
j
), depletion length (L
i
), efficiency (), response time ().

Figure 4. 3 PIN model Figure 4. 4 PIN Photodiode
When operating we inversely bias the PIN so that the output current is the leakage current
in the inverse region, we can notice from Figure 4.5 that leakage current will happen even
without light but that would be the normal break down voltage curve, also we can notice that the
higher the biasing voltage, the more the current generated.
33


Figure 4. 5: Current-voltage characteristics
Photo detectors as any other electronic device have an output voltage fluctuation with
temperature; even though its not so serious since it is only measured in mV/K, it should also be
taken care of in the design and the specifications of the system operating temperature range.
Also Photo detectors create shot noise and some other additive noise, this last one is
usually measured in noise equivalent power and it is mostly in pW. The Noises generated by PIN
are usually combined in the Noise Equivalent Power (NEP) and is given in the below equation.
=
()
(/)
. . . (4.1)[11]

4.2.4 Responsivity
Not all the dropping photons are collected; we define Quantum Efficiency as the ratio
between falling photons and the ones transformed to current. Typically its between 0.8-0.9.
Table 4.1 shows the relation between the PIN fabrication material, its corresponding band-gap
and the responsivity at the peak response wavelength.

Material Band-gap, eV
Wavelength range
(nm)
Wavelength of peak
response (nm)
Responsivity
(max) (A/W)
Si 1.17 300-1100 800 0.5
Ge 0.775 500-1800 1550 0.7
InGaAs 0.75-1.24 1000-1700 1700 1.1
Table 4. 1: Photodiodes materials and their characteristics [12]
34

When detecting light at its peak spectrum response wavelength, the silicon PIN
photodiode will leak about 0.5A of current for each microwatt of light striking it for 800nm Si-
PIN. This relationship is independent of the size of the detector. The PIN photodiode size should
be chosen based on the required frequency response and the desired acceptance angle with the
lens being used. Large PIN photodiodes will have slower response times than smaller devices.
For example, 1 cm X 1 cm diodes should not be used for modulation frequencies beyond 200
KHz, while 2.5 mm X 2.5 mm diodes will work beyond 50MHz. If a long range is desired, the
largest photodiode possible that will handle the modulation frequency should be used [13].

4.2.5 PIN and the datasheet
The datasheet supplies us with a set of curves that explain the operation of the PIN. The
main curves supplied by most datasheets are shown in the figures below, the data sheet of
SFH203 was used as an example. The first characteristic curve is the Spectral sensitivity curve
shown in Figure 4.6, this curve resembles how much is the detector sensitive to different
wavelengths. The narrower this curve the better, also it should have its peak value on the
frequency that we are using in the transmitter.
The second curve is the Dark current curve, Figure 4.7, which shows how much current
the detector supplies even if it is not detecting any signal (complete darkness), this current should
Figure 4. 6: Spectrum sensitivity
Figure 4. 6 Spectrum sensitivity Figure 4. 7 Dark current relative to reverse voltage
35

be dealt with to be below any level affecting value (it would be barely recognized since it is
valued in pA). Dark current is considered to be a noise source thus affecting the SNR value.
The third curve is the capacitance curve, Figure 4.8. Earlier it was stated that the response
time is directly proportional to capacitance, from the curve we can see that the higher the inverse
voltage the less the junction capacitance hence faster response. So we prefer to operate at the
lowest possible voltage but with care so that we dont exceed the reverse voltage break down
limit.
The fourth curve is the directional characteristic curve, Figure 4.9, shows the
normalization of detection ratio with the beam incidence angle, it shows the trivial fact that in
incidence was 100% direct (at angle zero) we will get 100% of the power detected. The
directional characteristics vary with the package covering used. For highly directed LOS systems
we need the directivity to be as narrow as possible, while for non-LOS applications good
detection along the 180
o
is required.


Figure 4. 9 Responsivity ratio relative to light falling angle (Directional curve)
4.3 Receiver Stages
4.3.1 Transimpedance Amplifier
As mentioned before, PIN PDs produce a small current that needs to be transformed into
corresponding voltage levels for further processing. Transimpedance amplifier (TIA) is a circuit
Figure 4. 8 Reverse voltage vs. capacitance
36

used to accomplish this task. The simplest form of TIA is merely a shunt resistor. A better way is
to build it using a current to voltage converter employing a low noise Operational Amplifier
(Opamp). Figure 4.10 below illustrates this configuration.

Figure 4. 10 Transimpedance Amplifier
The feedback capacitor is inserted to add a pole in the transfer function of the amplifier in
order to phase compensate the amplifier, and prevent gain peaking and oscillations introduced by
the series combination of the photodiode junction capacitance and the inductive resembling load
of the amplifier. This configuration is still prone to oscillations if the capacitor in the feedback
circuit isnt chosen carefully .To ensure a stable and oscillation free operation the following
relation should be satisfied:
1
2

. . . (4.2)
GBP: OpAmp Gain Bandwidth Product.
R
F
is the feedback resistor.
C
F
is the feedback capacitor.
C
D
is the shunt capacitance contributions from the photodiode and OpAmp .
After phase compensating the amplifier, the transimpedance gain of the TIA is:

1 +2

. . . . (4.3)
Gain is an important factor of the TIA, high gain ensures the amplification of the small
photocurrent (I
photo
) to the voltage levels needed. Therefore, gain determines the sensitivity of the
receiver. In addition, the resistance used to set the transimpedance gain should be large enough
37

to reduce the thermal noise discussed in chapter 2. By setting the gain impedance we can set the
expected output voltage levels and noise reduction ratio according to the following equation.

=


. . (4.4)
It is obvious from equation (4.3) that the TIA has a Low Pass Filter (LPF) characteristic
with a cutoff frequency that is inversely proportional to the gain impedance R
f
, therefore
compromise between bandwidth and sensitivity takes place. A TIA should have minimized
bandwidth to reduce the integrated noise, but limiting the bandwidth very much will introduce
ISI. Its a usual practice to set TIA 3-dB bandwidth to 70% of the detected signal bandwidth; this
will drop the noise by a factor of 0.7 = 0.84 (1.55dB). For our particular design, with a signal
of a bandwidth of 20MHz, the bandwidth of the TIA should be between 14MHz and 20MHz. But
this will decrease the sensitivity of the receiver significantly, and increase the thermal noise.
Therefore, it was decided to maintain the TIA at a low cutoff frequency around 0.5MHz or less,
and cascade it with a post amplifier with a High Pass Filter (HPF) characteristic which will
compensate for the limited bandwidth of the TIA, and produce a Band Pass Filter (BPF)
characteristic, and amplify the signal further more.
The TIA must also be tolerant to overloading. High received power at the photo detector
such as a shot pulse due to the channel nature will cause high photocurrent, and thus high voltage
at the TIA output, a more severe result may occur in the feedback circuit where poor output
behavior happens due to driving the circuit into saturation. Overloading and current variations
are usually compensated by integrating an Automatic Gain Control (AGC) but its design is
complicated and beyond the scope of this project [15].
4.3.2 Post Amplifier
Post Amplifiers (PA) are generally used after the TIAs for the purpose of increasing the
gain, another purpose for the PA is to transform the signal into pulses, since the TIA has a LPF
characteristic, it integrates the signal thus requiring a stage to compensate for that by
differentiation in order to recover the signal. Differentiating amplifier is achieved using active
High Pass Filters (HPF). A HPF has a desired effect of adding positive phase which will
compensate for instabilities in the TIA integrator [16]. A Differential video amplifier was used,
38

this supplies an additional feature of amplifying the signals selectively depending on their
respective frequencies, that is as we know signals get attenuated more at high frequencies and a
compensation for that is needed.
The video amplifier gain equation is:
=
1.4 10
4
+31
. . . (4.5)
4.3.3 Limiting Amplifier
The amplitude of the signal received from the post amp is variable within the previous
stage dynamic range, yet the interface circuit needs definite signal swing that will not go beyond
the maximum logic level and form it as required. A Limiting Amplifier (LA) amplifies the signal
but clips it beyond certain limits.
LAs should have very high slew rate so that the rise and fall times are very small to
achieve high speed level transition.
The LA should exhibit enough high gain to maximize the SNR. LA can be designed of
multiple stages, but it is preferred to be done with the least number possible to minimize levels
noise contribution. Limiters are usually incorporated with offset cancellation to avoid saturating
the following stage [14].
The configuration used acts a current source where the current switching is controlled
using a transistor, and then voltage difference is taken around a resistor which current passes
through. The design reduces miller effect and reduces noise due to the high Common Mode
Rejection Ratio (CMRR).

4.3.4 Output Stage
To transfer the detected signal to the interface circuit with maximum power transfer, we
will need an output impedance of the receiver to be equal to the characteristic impedance of the
transmission line. To achieve the required result we set the LA output impedance to 75.
4.3.5 Received Signal Strength Indicator (RSSI)
The Received Signal Strength Indicator (RSSI) is a vital part for highly directed systems
that operate at large distances; it is used to aid in the system alignment at installation. The RSSI
provides DC voltage relative to the received signal at the front end. The optical heads would be
directed for the maximum RSSI voltage output. In the new design we made use of the
39

differential amplifier that supplies two outputs, one was used as a input for the LA, and the other
was used as an input to the RSSI circuit.
The RSSI circuit is simply a voltage doubler followed by RLC filter to assure pure DC
output signal. A voltage doubler is a circuit that inputs AC signal and outputs a DC signal with
the DC level being roughly twice the input peak signal value.
4.4 System design
4.4.1 Design Revision 0 problems

Figure 4. 11 The Receiver (Revision 0)
The receiver design in Figure 4.11 was the design used to produce the first PCB, yet, the
design even though fully functional and working in simulation, refused to work past the TIA and
the only thing visible after the Post Amplifier (PA) stage was noise. Also the dark noise was
noticed to be extremely high. Even the PCB circuitry itself turned out to have some problems
such as missing wires or faulty routing.

4.4.2 Design Revision 1
V-
V+
0
C13
100n
C14
100n
0
0
0
0
C4
13p
0
0
C3
2.2p
0
R5
100k
0
C6
2.2p
2V-
+
-
V
+
V
-
U1
OPA657
1
2
3
4
5
R4
100k
0
U2
MAX4305/MXM
+
3
-
2
V+
7
V
-
4
OUT
6
R15
30
R17
1k
2V-
V
V
V
I
COAX
75
0 0
R13
80k
C2
270p
R6
10k
0
R12
210
U4
MAX4202/MXM
in
3
V
+
4
V-
2
OUT
5
0
D1
SFH203
I1
TD = 0
TF = 1n
PW = 49n
PER = 100n
I1 = 1nA
I2 = 4.06u
TR = 1n
V6
-12
2V-
0
V+
V7
5
V-
V8
-5
0 0
C8
100n
V+
C9
100n
C10
100n
C11
100n
C12
100n
V-
C5
1n
V
+
V
-
+
-
U3
OPA699
3
2
7
4
6
8
5
R1
6k
R2
1k
0
0
R9
6k
R10
1k
R8
500
R7
5k
40

Since the original design up to the TIA was correct it was kept as is, after that the PA was
substituted with a differential video amplifier, and to keep good amplification levels a discrete
Limiting Amplifier (LA) design with automatic gain control (AGC) was used from Twibright
Labs.
The addition of the differential amplifier allowed for the addition of a Received Signal
Strength Indicator (RSSI) that will be connected to the negative output of the differential
amplifier. Design revision 1 can be seen in Figure 4.12.

Figure 4. 12 The Receiver (Revision 1)
G 2 A
G 2 B
V - O 2
I 1 G 1 A
G 1 B
V + O 1
I 2
U1
Dif errential amp
L2
1.5uH
1 2
R23
100k
0
D5
1N5406
L1
1H
1
2
R24
75
C22
3.3n
0
Q1
Q2
R13
1k
R14
1k
R16
270
R18
75
R19
6.8k
R20
6.8k
C11
100n
C12
100n
C13
230n
0
C14
1n
R25
12
R26
22
0
0
C31
100n
0
0
0
5
-5
100u
C33
12
-12
0
0
0
R21
1k
D2
SFH203
C30
100n
C32
2.2p
R17
100k
0
C27
2.2p
+
-
V
+
V
-
U5
TIA
3
2
6
7
4
R15
100k
0
0
0
C15
110n
C5
2.2n
C6
2.2n
0
C7
270p
C8
110n
C9
110n
C10
100n
C16
230n
C17
230n
R6
680
R7
680
R8
18
R9
18
R11
6.8k
C19
2.2n
R12
6.8k
0
R22
1k
D3
BAT46
D4
BAT46
C20
100n
C21
1n
41


Figure 4. 13 Receiver (Revision 1) Simulation
4.4.3 The PCB
The PCB layout design is somewhat cumbersome, where the trace-to-trace distance
should be maximized, the length of each trace should be minimized, and the trace width is as
high as possible. The design shown in Figure 4.14 and Figure 4.15 is the final PCB layout that
was implemented. The bottom layer can be noticed to be ground plane; this is desirable to reduce
noise as much as possible.
Time
129.6us 129.7us 129.8us 129.9us 130.0us 130.1us 130.2us 130.3us 130.4us 130.5us 130.6us 130.7us
V(C32:2)
-4.0V
-3.8V
-3.6V
V(U1:OUTA)
3.96325V
3.96330V
3.96335V
3.96340V
SEL>>
V(R24:2)
-200uV
0V
200uV
42


Figure 4. 14 The top layer of the receiver

Figure 4. 15 The bottom layer of the receiver
4.5 Practical results
The resulting PCB can be seen in Figure 4.16, starting from the left is the PIN, the TIA
amplifier in SOIC package, then the Video Amplifier in D14 package, and finally the LA stage,
the red cables are the power cables.
43


Figure 4. 16 The Receiver
The Rx output can be seen in both Figure 4.17 at 1MHz and in Figure 4.18 where a signal
with various pulse lengths is sent and received correctly. In Figure 4.17 Ch1 is the Rx channel
while Ch2 is the Tx channel, the strange frequency and duty cycle is caused y the pulses
oscillations. Figure 4.18 shown clearly that any sequence of data bits can be sent and received
correctly, this figure can consider as data-sending simulation, the correct reception of the signal
assures that the system works correctly and that it can detect whatever signal however its shape
is, provided it is within the 10Mbps data range.
44


Figure 4. 17 Rx and Tx signals at 1MHz

Figure 4. 18 Rx and Tx signals at various signal shape
45

Chapter 5 The Interfacing Circuit
The interfacing circuit serves the purpose of mediating between an Ethernet interface-
such as a computers Network Interface Card (NIC) - and the optical transceivers. The
interfacing circuit can be thought of as an encoder and a decoder, where it encodes some special
Ethernet signals fed by an Ethernet interface for the sake of an appropriate optical transmission,
and decodes a special received signal at the optical receiver for the sake of appropriate data
retrieval. The figure below illustrates this configuration

Figure5. 1 The interfacing circuit
A proper understanding of the interfacing circuit requires a brief description of the
Ethernet signaling.
5.1 Ethernet Signaling
The 10Mbps Ethernet links over copper twisted pairs are termed 10Base-T Ethernet. In
10Base-T networks, data is transmitted in discrete frames, each of a length that varies between
64 up to 1518 bytes. The data in 10Base-T links is encoded with Manchester line encoding, with
a minimum -350mV as low level, and 350mV as high. 10Base-T Ethernet physical interfaces
have 8 pins, two for transmission, two for reception, and 4 reserved for other applications.
Synchronization between the transmitter and the receiver of an Ethernet link is accomplished by
means of a segment in the Ethernet frame called the preamble. The preamble segment consists of
7 Bytes, each having the following sequence 10101010. When an Ethernet interface is in idle
mode, i.e doesnt transmit data, it generates a special type of pulses called Normal Link Pulses
(NLP). They serve the purpose of maintaining link integrity by informing the next Ethernet node
in the network that an Ethernet interface is still connected, but is in idle mode. NLPs are unipolar
positive only pulses with duration of 100 nanosecond generated periodically at intervals of 16
46

milliseconds. The figures below illustrate the form of the Ethernet frame, and the signaling
scheme of the 10Base-T Ethernet .

Figure5. 2 10Base-T Ethernet frame

Figure5. 3 10Base-T Ethernet signaling
5.2 Interfacing Ethernet to optical transceivers
The interfacing circuit can be divided into two main parts: (a) a part that interfaces with
the transmitter (b) a part that interfaces with the receiver.
The transmitter interface accepts Ethernet Manchester encoded frames through the
transmitting pins of the Ethernet interface, by means of a Shielded Twisted Pair (STP) and
replaces the link integrity pulses with a 1 MHz square wave and passes the Ethernet frames to
the optical transmitter via a coaxial cable after eliminating the silence periods between the
frames. The receiver interface accepts the electrical signal generated by the optical receiver via a
coaxial cable, and replaces the 1 MHz square wave with locally generated NLPs, then
multiplexes the received Ethernet frame with the locally generated NLP, and passes the resultant
signal via an STP to the receiving pins of the Ethernet interface.
The NLP pulses are replaced with a 1 MHz square wave because the optical transmitter is
not designed to accept unipolar positive-only pulses, such as the NLPs. It is designed to accept
bipolar pulses, such as Manchester encoded bits. In addition to that, the NLPs are separated with
47

16 millisecond intervals, which is a quite a long period of silence, in which the average optical
power of the LED in the transmitter would decrease significantly if the NLPs were to be passed
to the transmitter, thus causing the LED to appear to flicker.
At the receiver end, transmitting NLPs would ruin the DC level of the signal due to the
relatively long periods of silence, which in turn would cause the receiver to take few
microseconds to retain the steady state level, and detect the Ethernet frame pulses after the idle
mode period has ended. This would increase the BER due to the loss of some of the Ethernet
frame bits. The 1 MHz bipolar square wave resembles Manchester encoded data, but doesnt
convey actual information; it only preserves a definite DC level through the receiver stages. In
addition to that, the 1 MHz square wave keeps the SNR at almost a constant level at the receiver
front end when no actual data is being transmitted, and prevents unwanted signals from being
amplified on the expense of the transmitted data. Most importantly, if no signal was transmitted
during the silence periods, random noise would deceive the Ethernet interface at the receiver end
into interpreting it as a preamble segment of an Ethernet frame, a rather corrupted segment,
therefore the Ethernet interface wouldnt accept the data transmitted immediately as the false
preamble segment of the Ethernet frame is corrupted, and synchronicity is lost. This technique is
used in 10Mbps Ethernet over fiber links (10Base-FL) in compliance with the IEEE 802.3
standards.
Although Media Converters (MC) are commercially available in an integrated form , it
was decided to adopt a discrete design based on logic gates , ripple counters , and shift register in
order to understand the functionality at low level . The chosen MC for this project was developed
by twibright labs for the RONJA project [17]. The figure in the next page illustrates the
interfacing circuit. The whole circuit is designed around two pulse extending cascade of the shift
registers (U63-65), and (U51, U52). A pulse extender is a device that extends pulse duration to a
certain level. The pulse extender (U63-65) accepts Ethernet frames fed by an Ethernet interface
through a line receiver (U62). C66 and R52 are there to replace the leading edge of a pulse into a
short impulse. The resulting impulse controls the shift registerU63 as it is connected to its master
reset which is active low, hence the inverter U52. When an Ethernet frame is transmitted the
pulse extender extends the duration of the pulses, and merges them into a low steady level at the
output. When NLP are transmitted the output of the pulse extender is a high steady level. The
48

output of the pulse extender (U63-65) is used to control the multiplexer U54, U56, which is the
coaxial cable driver at the same time. When Ethernet frame is transmitted, the multiplexer passes
it to the optical transmitter, when NLP are transmitter, the multiplexer passes the 1MHz square
wave generated by the binary counter U59. On the reception side the interfacing circuit accepts
the data from the optical receiver, and a cascade of shift registers (U51, U53) which form a pulse
extender. The pulse extender determines whether an Ethernet frame or a 1MHz square wave is
being received. The ripple carry counters (U60, U61) generate local NLPs with 125ns duration,
and repeat every 16.384ms. Again, the pulse extender controls a multiplexer (U54-U56) which
passes the received Ethernet frame when being detected or the locally generated NLPs when the
1 MHz square wave is being detected.


49

Chapter 6: The power supply
As illustrated previously; the optical transceivers and the interfacing circuit are externally
supplied with +12V, -12V, +5V, and -5V power rails. The interfacing circuit is supplied with
+12V, the transmitter is supplied with +5V and -5V, the receiver is supplied with -12V, +5V and
-5V. Therefore, the power supply should be able to provide these values. For this purpose a
multistage power supplied was designed. The figure below illustrates the block diagram of the
power supply.

Figure 6. 1 Block diagram of the power supply
The first stage accepts +12V from an external regulated power adapter , and a protection
circuit is inserted between the input voltage and the inverting regulator in order to filter the input
voltage, and prevent damage to the regulators in case of short circuiting or reverse current. The
protection circuit is an LC filter capable of withstanding short circuit current, and a freewheeling
diode for creating an alternative path for reverse current. The figure below illustrates the
protection circuit.
50


Figure 6. 2 Protection circuit of the power supply
In the first stage a regulator splits the output into +12V and -12V using a structured
inverting regulator. The inverting regulator PTN78060A from texas instruments was chosen for
this task. The figure below illustrates the configuration of the inverting regulator circuit.

Figure 6. 3 Inverting regulator

SET
R is chosen such that it sets the regulators output to -12V . From the manufacturer
datasheet
SET
R should be set to 2KO to obtain an output of -12.006 V. C1 and C2 are input
voltage filters, or input voltage decoupling capacitors. Their values are chosen such that high
frequency noise is bypassed to the ground. The capacitor bank C2 consists of three shunt ceramic
capacitors in order to reduce the Series Equivalent Resistance (SER) of the capacitors, hence
enhancing the filtration practice. C4 is inserted to reduce the output voltage ripple and noise . C3
is an output decoupling capacitor , its value should be relatively high in order to be able to reject
as much high frequency noise as possible , hence its chosen to be an electrolyte capacitor with a
value of 100uF .
51

This inverting regulator can withstand an output current up to 1.25A, which is
sufficiently larger than the amperage needed by the optical transceivers and the interfacing
circuit.
The second stage of the power supply consists of +12V to +5V, and -12V to -5V
regulators. For +12V to +5V regulation, LM78S05 regulator was chosen, its a high current
version of the LM7805 regulator. It is connected to the +12V power rail as illustrated in the
figure below:

Figure 6. 4 High current +12V to +5V regulator
The 0.33uF capacitor is an input decoupling capacitor for high frequency noise rejection,
and the 0.1uF at the output serves the same purpose, and reduces the ripple of the output voltage.
For the -12V to -5V regulation, LM7905 was chosen. It should be noted that the negative
supply is not to provide as high current as the positive supply; therefore a conventional negative
regulator is chosen. The figure below illustrates the -12Vto -5V regulation practice.

Figure 6. 5 -12V to -5V regulator
The 0.22uF capacitor is an input decoupling capacitor for high frequency noise rejection,
and the 0.1uF at the output serves the same purpose, and reduces the ripple of the output voltage.
52

The PCB design of the power supply took consideration the relatively high currents
passing through the traces .Hence the trace width was set to 40 mil, in order to be able to
withstand a temperature rise of about 10
0
C . The PCB design didnt incorporate a ground plane,
as there is no high frequency signal in the power rails that needs to be filtered inherently, as the
decoupling capacitors are sufficient enough to filter the power signal .
The figures below show the power supply PCB . The first figure is the top layer of the
PCB , and the second figure is the bottom layer .

Figure 6. 6 The top layer of the power supply

Figure 6. 7 The bottom layer of the power supply

53

Chapter 7 Conclusion and Problems
In this project we were able to design an FSO communication system employed to extend
a 10Mbps LANs for relatively wide geographical area. The design can theoretically support a
range of 1Km. The performance of the system is measured in terms of BER, which doesnt
exceed 10
-9
. The credibility of the system is maintained due to the narrow optical beams
employed. The availability and reliability of the system is a function of the link margin, which is
kept at maximum in normal weather conditions, however, the availability condition is violated
severely due to fog. The design is reasonably priced.
The system was implemented at the hardware level. We were able to obtain all of the
components. PCB fabrication has been a problem; the material required for etching (FeCl
3
6H
2
O)
is missing, instead we used an itching material made of H
2
O
2
+HCl, this chemical combination
was fairly cheap and easily available in the local market with outstanding performance. Also, the
use of double layer PCB layout introduced through hole plating in the interfacing circuit which
we couldn't replaced by any mean, so we had to order a manufactured PCB for our use. The
mechanical part of the system shall include casings for the optical transceivers, which was done,
but an installation mechanism wasn't done.
System testing was done for individual stages, the interfacing circuit was tested
successfully and we were able to transmit files successfully through them with the use of a cross
cable to simulate the operation of the optical heads. On the optical heads side, the operation of
data transmission was verified using one pair of the system (single Tx and single Rx), for the
other pair the Tx is correctly operational, but in the Rx one of the ICs broke down, thus a
complete system transmission couldnt be done. The faulty IC is rare and no substitute for it is
available currently in the market.
Lots of problems have been faced, starting from the unavailability of some components
which had to be ordered from overseas, then the manufacturing process where some PCB boards
were corrupted, thus some PCB boards were also ordered from overseas, then the lack of
FeCl
3
6H
2
O which was impossible to order due to the insanely high price, thus delaying the
fabrication process of the PCB two months until the H
2
O
2
substitute was discovered. Finally we
faced the noncooperation from the department in using the labs which also caused lots of delay
in our work.
54

REFERENCES
1. Hranilovoc, Steve, Wireless Optical Communication systems. Springer. 2004

2. Morelos-Zaragoza, Robert, San Jose State University
http://www.engr.sjsu.edu/rmorelos/ee296Wf07/solhomework4.pdf.Cited on December the 9
th

2009

3. Khumsat, P. Wattanapisit, N. Kulhavey, K , Optical Front-Ends for Low-Cost Laser-Based
10-Mbps Free-Space Optical Transceiver , IEEE Circuits and Systems Asia Pacific
Conference, 2006.

4. Johnson, David,Handbook Of Optical Through The Air Communications. Imagineering E-
zine, 2000.

5. Ramirez, Roberto. Idrus, Sevia. Ziran, Sun, Optical Wireless Communications IR for
wireless connectivity. CRC press. 2007

6. Olivier Bouchet, Herv Sizun, Christian Boisrobert, Frdrique de Fornel, Pierre-Nol
Favennec. Free-Space Optics Propagation and Communication. PP. 16-17, PP.38-42.

7. Michael Day. LED-driving consideration, Texas instrument Incorporated.

8. E. Fred Schubert. Light-Emitting Diodes. CAMBRIDGE UNIVERSITY PRESS.

9. Albert Paul Malvino. Electronic Principles. Second Edition.

10. Graeme, Jerald. Photodiode amplifiers. McGraw-Hill, 1996, Pg. 2

11. A Primer on Photodiode Technology.
http://home.sandiego.edu/~ekim/photodiode/pdtech.html. Cited December the 9th

12. Professor Cantrell, Cyrus. OPTICAL DETECTORS AND RECEIVERS. Erik Jonson School
of Engineering & Computer Science, AugustDecember 2003, Pg. 3

13. Johnson David. Handbook of optical through air communications. Pg. 44
14. Razavi, Behzad. Design of ICs for optical communications. McGraw Hill 2003, Pg. 57, 64,
123.
15. Photodiode Characteristics and Applications. OSI Optoelectronics.
http://www.osioptoelectronics.com/application-notes/AN-Photodiode-Parameters-
Characteristics.pdf Cited December the 9th 2009.
55


16. Current-to-Voltage Amplifier. http://www.ecircuitcenter.com/Circuits/opitov/opitov.htm.
Cited December the 9th 2009

17. Kulhavy, Karel , Twibright Labs . http://ronja.twibright.com . Cited December the 9th 2009
18.John R. Barry. Wireless Infrared Communications. The Springer International Series in
Engineering and Computer Science. Vol 280 .1994 . Pg.83
A

Appendix (A) The Link Budget
In order to determine the numerical results of the link budget equation, first we need to
determine the sensitivity of the receiver. The sensitivity is defined as the minimum received
power needed for a reliable communication. The sensitivity is determined using the SNR
equation Eq.(2.17)-. The minimum SNR required for a BER of 10
-9
is 15.6dB according to the
approximation:
1
2

2
4

Therefore:


2
+2P
sun
= 36
Assuming a daytime operation, the electrical noise is very small compared to the ambient
noise due to sunlight illumination. Hence

2
2P
sun
36
Or

2
P
sun
72
Setting B=17MHz , and r=0.55 from the datasheet of the (BP10NF) PIN photodiode

2
1.077 10
10


2

Using Eq.(2.18) to determine the solar irradiance :

2
1.077 F

A
Rx
10
10

2

The photodiode used in this project has an optical bandwidth of 260nm , and the
receiving lens is of 100mm diameter . Therefore :
B

2.966F

10
5

But :

=




5.39F

10
5

The average received power =
1
2


2.69F

10
5

The solar irradiance spectral density F

assumes a maximum value of 0.9 W/(m


2
.nm)
when the sun disc is normally directed to the receiving lens in clear weather conditions , and
midday time . F

greatly depends on the orientation of the receiver with respect to the sun as in
the figure below [1]:

Figure.(A) : solar irradiance on the receiver
Assuming an orientation that prevents normal incidence of the sunlight on the receiver,
F

can be significantly reduced. Nevertheless, the sensitivity of the receiver is still affected by
the fluctuations of the solar irradiance. Unfortunately, calculating the actual F

is not a
straightforward task , and the literature lists the tables for European and north American
countries .Therefore we have to assume that F

assumes a value measured for a different system ,


and compensate for the differences by approximation. According to the last consideration F

was
found to be = 0.01 W/(m
2
.nm).

= 2.69 25.7 , which is receiver sensitivity.


The receiver was designed to be able to detect smaller values than -25.7dBm.
C

Inserting this value in Eq.(2.11) and Eq.(2.12) , and assuming 90% transmittance of the
lenses used ,we can get the maximum distance .

= 0.9

+ tan

2
] ) ( exp[ R
The IR LED chosen for this project was chosen to transmit a maximum power of 50mW.
Assuming that the LED will transmit only 50% of its maximum radiated power

= 25 ,
and:
2.177 10
4
=

+ tan

2
] ) ( exp[ R
According to Eq.(2.5) , and by using a transmitting lens with a radius of 100mm , and
form the datasheet of the IR LED used (TSFF5210) which has an initial divergence angle of 10o
, and a virtual transmitting diameter of 3.7mm , the divergence angle due to the collimation lens
is given by :
=

= 10
3.7
100
= 0.37
0
= 6.4
Assuming a visibility of 4Km , which is considered as a medium visibility range in
Palestine , the atmospheric extinction coefficient is given by Eq.(2.13):
m dB
V
q
nm
nm
/ . 10 . 24 . 6 )
550
(
912 . 3
) (
4
= =



Solving for R using MATLAB, R=1.4Km. It should be noted that the link should be
installed with a range less than the maximum range in order to insure a reasonable margin, and
compensate for effects not accounted for in the link budget equation. We have chosen a range of
1Km.
The link margin at R=1km is given by

+

+

= 14.7722 +32.6 = 4.17


D

Which means there is 2.61 times the power needed for a reliable communication at the
receiver end at a range of 1Km. Finally, it should be noted that at nighttime the margin will
increase significantly, and also shall SNR, which will reduce the BER.
The graphs below illustrate some key parameters in the link such as BER vs range ,
geometric attenuation vs range , atmospheric attenuation vs visibility , and atmospheric
attenuation vs range .

Figure A. 1 Geometric attenuation vs range

Figure A. 2 Extinction coefficient vs visibility
E



Figure A. 3 Atmospheric attenuation vs visibility

Figure A. 4 Total attenuation vs range
F


Figure A. 5 SNR vs range

Figure A. 6 BER vs range

Vous aimerez peut-être aussi