CALIFORNIA STATE UNIVERSITY NORTHRIDGE

Transcription

1 CALIFORNIA STATE UNIVERSITY NORTHRIDGE STUDY OF ALL FIBER OPTIC CURRENT TRANSDUCER IN OPTICAL TRANSMISSION SYSTEM AND EVALUATION OF PERFORMANCE ON OPTISYSTEM A graduate project submitted in partial fulfillment of the requirements For the degree of Master of Science in Electrical Engineering By Prasanna M Thawale August 2016

2 The graduate project of Prasanna M Thawale is approved by: Dr. Jack Ou Date Dr. Xiaojun (Ashley) Geng Date Dr. Nagwa Bekir, Chair Date California State University Northridge ii

3 ACKNOWLEDGEMENT It gives me immense pleasure and satisfaction to present project report on Study of All Fiber Optic Current Transducer in Optical Transmission System and Evaluation of Performance on Optisystem towards partial fulfillment of Master of Science degree in Electrical Engineering. With my deepest gratitude I would like to thank Dr. Nagwa Bekir for her guidance towards completion of my project. I really appreciate the effort she has put into this project as much as I did. It couldn t have been possible without her constant supervision, knowledge and support. I would also like to express my sincere gratitude to Dr. Jack Ou and Dr. Xiaojun (Ashley) Geng for their valuable suggestions and guidance. Above all I would like to thanks my parents for their love and support that always motivated me towards success in life. iii

4 TABLE OF CONTENTS SIGNATURE PAGE..ii ACKNOWLEDGEMENT..iii LIST OF FIGURES vi ABSTRACT x Chapter 1: Introduction 1 Chapter 2: Introduction To Current Sensors, Basic Sensing Principles..3 & Configurations. 2.1: Fiber optic Interferometers 3 2.2: Sagnac Effect 4 2.3: Sagnac Interferometer : Faradays Effect and Verdet Constant : Current Sensors a): Electro-optic current Sensor b): Optical Fiber Current Sensor c): Bulk-Optic Glass Current Sensor d): Fiber Bragg Grating Current Sensor..13 Chapter 3: Working principle of All Fiber Optic Current Transducer : Interferometric Detection scheme : Practical Sagnac Interferometric fiber optic current sensor (SIFOCUS) a): Polarimeteric fiber current sensor s Theoretical Background b): Practical SIFOCUS Structure And Theoretical Analysis : Basic principle of AFOCT : AFOCT Practical Model. 25 iv

5 Chapter 4: Simulation and Analysis : Optisystem : Simulation Layout and Component Description (I) Analysis of the Main Components of the Layout Optical transmitter/ Light source model Transmission and reception of simulation model (II) Results and analysis of the overall system Case I. Simulation using sine wave modulating signal..34 Case II. Simulation using Square wave modulation : Simulation using Coherent Source of light.51 Case I: sine wave modulation...54 Case II. Square wave modulation Phase difference Calculations.. 59 Chapter 5: Concluding Remarks References.. 67 v

6 LIST OF FIGURES Figure 2.1: Phase shift between two counter propagating waves [8]... 5 Figure 2.2: Sensor based on Sagnac fiber Interferometer... 6 Figure 2.3: (a) Setup for Optical Gyroscope (b) Phase difference generated due to effect of angular velocity [9] Figure 2.4: linearly polarized light having Faradays Effect [7]...8 Figure 2.5: Faraday/ Sagnac closed loop Interferometer current sensor [5] Figure 2.6: In-line Sagnac Interferometer current sensor [5] Figure 2.7: a) Fiber Bragg grating sensor on GMM b) Fiber Bragg Grating sensor with MCP [5] Figure 3.1: Close loop Sagnac Interferometer detection scheme [7] Figure 3.2: In-line Sagnac Interferometer Current detection scheme [7] Figure 3.3: Polarimetric fiber optic current sensor [23] Figure 3.4: Practical Sagnac Interferometer using 3x3 coupler [23] Figure 3.5: Flint Fiber basic Structure [7] Figure 3.6: Practical AFOCT structure [1] Figure 3.7: Practical Simulation Structure [1] Figure 4.1: Optisystem simulation layout for AFOCT [1] Figure 4.2: Block Diagram of a Simple Transmitter Figure 4.3: Source Model/Optical Transmitter Figure 4.4: Spectrogram of light source [1] Figure 4.5: Analysis of the overall system Figure 4.6 a): stokes of polarization for source simulation Figure 4.6 b): stokes of polarization for source simulation after the linear polarizer vi

7 Figure 4.7: AFOCT sine signal Modulation wave form [1] Figure 4.8: Optical time domain visualizer Figure 4.9 a): Right Hand Circular polarization Figure 4.9 b): Left Hand Circular polarization Figure 4.10 a): Unmodulated wave in time and frequency domain using optical time domain visualizer_3 and optical spectrum analyzer_3 respectively Figure 4.10 b): Modulated wave in time and frequency domain using optical time domain visualizer_4 and optical spectrum analyzer_4 respectively Figure 4.11: Spectrum of light intensity Figure 4.12: Photodiode output time domain Figure 4.13: Photodiode output in frequency domain Figure 4.14: Filtered output waveform of AFOCT simulation system using sine wave modulating signal input Figure 4.15: simulation structure with Square wave modulation Figure 4.16: Improvised simulation structure with Square wave modulation Figure 4.17: Distorted output of the square wave simulation system Figure 4.18 a): Stokes of polarization before Linear Polarizer Figure 4.18 b): Stokes of Polarization after Linear polarizer Figure 4.19: AFOCT Square wave Modulation signal waveform Figure 4.20: Square wave modulated signal Figure 4.21 a): Right Hand Circular polarization Figure 4.21 b): Left Hand Circular polarization Figure 4.22 a): Modulated wave form in time and frequency domain respectively vii

8 Figure 4.22 b): Unmodulated waveform in time and frequency domain for square wave modulation scheme respectively Figure 4.23: Spectrum of light intensity Figure 4.24 a): Output of photodetector in Time domain Figure 4.24 b): Photo diode output in frequency domain..50 Figure 4.25: Rectified AFOCT simulation system output waveform in square wave signal modulation Figure 4.26: Spectrogram of Laser light source Figure 4.27: simulation structure using Laser light source for sine modulation Figure 4.28: Mach-Zehnder Modulator output Figure 4.29 a): Unmodulated carrier signal Figure 4.29 b): Modulated sine signal Figure 4.30: Output waveform of AFOCT simulation system in sine signal modulation and filtered for Coherent source Figure 4.31: Square wave modulate system with laser as a source carrier signal Figure 4.32 a): Mach-Zehnder modulator output in watts Figure 4.32 b): Mach-Zehnder output in dbm Figure 4.33 a): Unmodulated carrier signal Figure 4.33 b): Modulated square wave signal Figure 4.34: Improvised output waveform of the AFOCT simulation system in square wave modulating signal using a laser source Figure 4.35: Phase difference for 150MHz Figure 4.36: Phase difference for 300MHz Figure 4.37: Phase difference for 450MHz viii

9 Figure 4.38: Phase difference for 600MHz Figure 4.39: Phase difference for 750MHz Figure 4.40: Phase difference for 900MHz Figure 4.41: Frequency vs. phase difference ix

10 Abstract STUDY OF ALL FIBER OPTIC CURRENT TRANSDUCER IN OPTICAL TRANSMISSION SYSTEM AND EVALUATION OF PERFORMANCE ON OPTISYSTEM By Prasanna M Thawale Master of Science in Electrical Engineering Optical fiber finds its application in almost all fields and even dominant in few because of the numerous advantages offered by them. One important area of fiber optics applications used in industry is fiber optic sensors. As there exists distortion in the results obtained from a conventional current and voltage sensors due to the electromagnetic interference, there is need for sensor immune to these effects for having the proper results. Thus fiber optic sensors play a vital role as they are practically immune to external magnetic fields and currents. This project discussed in detail All Fiber Optic Current Transducer (AFOCT) modulation technique. Interferometric detection scheme have been explained which forms the base of AFOCT. A practical Sagnac Interferometer is explained to have a better understanding of the project. Depending on AFOCT working principle a simulation structure is developed using the software of optisystem 10. All the main components of the simulation structure are described thoroughly with analysis of their results. Using the direct detection structure, different signal performance was simulated. Signals were analyzed & processed using optisystem and the feasibility of the design has been validated. x

11 CHAPTER 1 INTRODUCTION The main objective of this project is to introduce different types of fiber optic sensors and demonstrate in detail the All Fiber Optic Current Transducer In Optical Transmission System [1] in terms of design and simulation. Current transducers play a key role in power industry these days [2-4]. Because of the tremendous development in the utilization of electric equipment s in developed nations and the overall increment of electrical conveyance/utilization, the interest for sensors for electrical current measurement, particularly in high voltage levels, has expanded essentially in course of the most recent decade. Although these sensors are still excessively costly for low voltage applications, for example, in local locations, they are amazingly alluring for high voltage applications, for example, in electrical high voltage (EHV) substations when contrasted with traditional sensors. Because of their inherent property such as [7] immunity against electromagnetic interferences (EMI), given the high immunity of fiber-optic transducer to electromagnetic interference, they are ideally adapted to high voltage environment. Electrical isolation (sensors are made of dielectric materials), Measurement of AC and DC is made possible, No effect of saturation, Minimal power Utilization, Small size, lightweight, and relatively low cost [7]. These are the main characteristics of such sensors. Moreover these sensors normally coupled with optical fibers having huge communication bandwidths, and because of their low absorption loss it permits remote detection, high multiplexing capacity and transmission of data over long distances is possible. In optical transducers many problems such as magnetic saturation or danger of catastrophic failure are inexistent compare to their conventional counterparts. For fast protection and power quality monitoring purposes the wide bandwidth of optical fiber is very important. It is easy to integrate or install Optical transducers to existing substations like bushings or circuit breaker etc. which results in significant space saving and reduced installation cost. Also, optical current sensors measure the magnetic field generated by the electric current rather than the current itself, thus avoiding the electric hazards that the high voltage 1

12 measurements imply. This magnetic effect is measured by the fiber optic current sensors using two linear effects namely magneto-optic effect (or faradays effect) and magnetic force (Lorenzt force). The optical current transducers have attracted more attention because of the recent development of fiber communication and invention of laser. Both digital and analog signals can be produced by OCT (optical current transducer) [5]. OCT can be organized in four groups taking into account the sensing mechanism and material used: All fiber optic Transducer, bulk optic sensor, hybrid sensor and magnetic force sensor. All fiber optic current transducer have been recently developed (AFOCT). In comparison to the bulk-optics transducer it offers more flexibility in its design and applications as well as improved performance. AFOCT will be the main topic of this report. The 2nd chapter of this report will have an overview of different types of fiber optic current sensors such as Electro-Optic Current Sensor, Optical Fiber Current Sensor/AFOCT, Bulk Glass Optic Current sensor and Fiber Bragg Grating current sensor [4] to know different configurations available and their practical application. Detailed explanation of Faradays effect will be covered and verdet constant of magnetic materials will be explained. As the main configuration setup is based on Sagnac interferometer Principle, it will be discussed along with the basics of Fiber optic Interferometers. Chapter 3 will discuss the principles behind the current sensory mechanism i.e. the Interferometric detection scheme. The basic principle of AFOCT with the conditions of polarization which enable the sensor with immunity against magnetic interference will also be discussed in chapter 3. After understanding the basic principle of the Optical current transducer from chapter 3 the practical implementation of the optical current transducer using the Sagnac interferometer principle will be realized in chapter 4 using Optisystem. Chapter 5 covers summery of this report with improvisation of future current transducers. 2

13 CHAPTER 2 INTRODUCTION TO CURRENT SENSORS AND INTRODUCTION TO BASIC SENSING PRINCIPLES & CONFIGURATIONS. Different fiber optic current sensors like Electro-Optic Current Sensor, Optical Fiber Current Sensor/All Fiber Optic Current Sensor (AFOCT), Bulk Glass Optic Current sensor and Fiber Bragg Grating current sensor will be covered in detail in this chapter. Before getting into detail of the sensors, the verdet constant and the Faradays effect will be explained to have a better understanding of sensors working. The basic interferometric scheme will be discussed along with Sagnac interferometer (SI) in detail. Before talking about the SI the Sagnac effect will be explained to have a better understanding of SI working. 2.1 Fiber Optic Interferometers Fiber Optic Interferometers are extremely useful for sensing various physical parameters like temperature, pressure, strain and Refractive indices. Two or more light beams having the same frequency are super imposed on each other to measure the phase difference between them, which forms the basis of interferometry. As interference of two beams travelling through different optical paths of single or two different fibers is used by the Fiber optic Interferometers (FOI), every FOI system needs a beam splitting and combining configuration. These incident light beams are split into two or more different parts to create an interference pattern by combining them together. For the interference pattern the constructive points are the one which corresponds to the path difference between two optical paths having the number of wavelength as integer and have destructive points which corresponds to half wavelength having odd number. Minimum two optical paths are required for interferometry experiment and out of the two optical path one is kept in the reference and the other one is so arranged as to easily affect by external perturbation. A single fiber can accumulate these paths with different fiber modes. One optical path is defined by each mode and for Interferometers like Sagnac, optical path is defined by clockwise or counterclockwise mode. FOI can perform extremely well with their sensors in high sensitivity, accuracy and large dynamics as an interferometer provides with the spectral and temporal information by which mesurand can be determined by changes in 3

14 frequency, phase, intensity, Bandwidth etc. The various interferometer configurations realized with the optical fiber are the Sagnac, Michelson, Fabry-perot and Mach-Zehnder Interferometers. In this chapter the Sagnac interferometer is discussed as the principle of Sagnac interferometer will be implementing in the chapter 4 to demonstrate the working of All fiber optic current transducer direct detection technique via simulation using optisystem Sagnac Effect The light wave accumulates the phase which is linearly dependent on the path s rotation rate, as it propagates along the circular path which is rotating slowly. As a result of which a phase difference is exhibited by two counter propagating waves along the same circular path. Consider a perfect circular path for better understanding of Sagnac effect. A beam splitter splits the light entering the path into two beams: one with clockwise and other with counterclockwise propagation. Now both the light beams will return to their origin in phase, while travelling through the same path, given the system is at rest. Now consider the whole system has a clockwise rotation, with angular velocity Ω with respect to inertial system. For this case the two beams will have an optical path difference between them. This shown in the figure 2.1 below. As the waves have their origin rotating with the system an observer moving along with the system (clockwise) will notice that the clockwise rotating beam of light had to travel a distance which is longer compare to the counterclockwise rotating beam of light, in order to return back to the system. Now an Interferometer can be used to determine the phase difference. 4

15 Figure 2.1 Phase shift between two counter propagating waves [8]. 2.3 Sagnac Interferometer (S.I) Owing to their advantage of simple structure, fabrication and robustness to environment the Sagnac Interferometers are been used extensively in various sensing applications [9]. Figure 2.2 illustrates basic Sagnac optical fiber configuration. A single mode semiconductor optical source, coherent & stabilized is used or an optical fiber laser doped with Erbium is utilized. A Sagnac Interferometer is basically a device which detects the time difference for two counter rotating beams of light in optical loop. The output beam of the laser is assumed to be uniformly phased and properly collimated. This beam of laser is made to enter optical fiber coupler (OFC) of 3dB lossless in nature. The OFC splits the injected light into two beams having same intensity, and these two beams counter propagate along the single mode optical fiber coil. The Sagnac coil output is directed towards a single detector. Given this configuration of SI it is primarily utilized as a rotation sensor. The clockwise (CW) and counter clockwise (CCW) modes of non-rotating Sagnac Interferometer are in phase, whereas for rotating Sagnac interferometer configuration having rotating velocity, the optical path of one of the modes is shorten and the other one is lengthen. The inference spectrum is caused by the Sagnac effect and it depends on 5

16 angular frequency of the setup. The output frequency of detector is the beating frequency that means there is a periodic variation between the two CW & CCW modes. Now the phase difference in the modes of counter propagating waves is given by [10]: = 8πNAΩ (λ)(c) (2.1) This phase difference occurs when the axis of rotation is aligned with the axis of optical fiber coil. Here λ is the optical wavelength of free space, A is Sagnac coil area i.e. the area enclosed within the coil, Angular velocity is Ω and N being the number of turns of the fiber loop. c is the velocity of light in free space. The sensitivity S of the system is given by [10]: S = 8πNA (λ)(c) (2.2) Which is the ratio of phase difference to angular velocity Ω. The sensitivity can be increased by increasing the number of turns N of the coil, also the length of the fiber and frequency of laser. Time varying and nonreciprocal phenomenon can also be measured by employing Sagnac Interferometer configuration. Hence SI becomes significant tool for current, strain, acoustic wave and temperature detection. This phenomenon of SI is used in chapter 3 for explaining current sensing mechanism. CW Laser OFC CCW Detector Fig.2.2 Sensor based on Sagnac fiber Interferometer One of the most extensively used application of Sagnac effect is the fiber optic gyroscope. The setup has a counter propagating light guided by fiber optic and the beam splitter is replaced by fiber coupler. The effective area is multiplied if the fiber is coiled in number of loops and as the phase shift and the time difference are dependent on the enclosed area the effect gets multiplied too. The phase difference is given by 6

17 = 2 LD λc Ω (2.3) Ω = angular velocity λ = wavelength C= speed of travelling light L = total length of fiber optic coil D = Diameter of the coil. (a) (b) Fig.2.3 (a) Setup for Optical Gyroscope (b) Phase difference generated due to effect of angular velocity [9]. As it s been extremely sensitive to rotation, fringe shifts detection in its interference gives a precise calculation on the rotation. Fiber optic gyroscope is also resistant to external perturbations like vibration, shock and acceleration. Because of its solid construction and scaling this device is useful in technologies which utilizes Sagnac effect. We are going to use this principle in determining the phase change in the main setup in the chapter 3. 7

18 2.4 Faradays Effect and Verdet Constant For right handed and left handed circularly polarized light there is a difference in refractive indices of glass, induced in them due to external magnetic field. This effect was discovered by Michael Faraday in In 1854, Emile Verdet presented, that rotation angle of light which is polarized linearly is proportional to magnetic field strength and cosine of the angle between the light wave direction of propagation and the field. Mathematically it is expressed as [7]: 1 L θ f = VB. dl (2.4) Where, V is Verdet constant of material, which is both temperature-dependent and dispersive [7], B is the magnetic flux density vector and dl is the differential vector along the direction of propagation. Thus optical sensors can be build using this effect, called Faradays Effect or Magneto optic Effect. Figure 2.4 shows external magnetic field causing rotation in polarization in magneto optic material, such as glass. Figure 2.4 linearly polarized light having Faradays Effect [7] The angle of rotation of polarization per unit propagation length per unit magnetic field is indicated by the verdet constant. Sometimes, magnetic field strength is expressed in units called oersteds symbolized Oe. The oersted is a larger unit than the ampere per meter and 1 Oe = 79.6 A/m [4]. Now V is dependent on refractive index dispersion dn where n is the refractive index λ is the wavelength, and it is given by [8]: 8 dλ

19 V = 1 2 e λ dn m c dλ (2.5) Here e m is the ratio of electronic charge to mass of the electron & C is velocity of light. Optical fiber have pretty small Verdet constant, deg/oe cm to deg/oe cm as compared to the bulk optic devices. To compensate for low verdet constant the fiber length i.e. the optical path can be increased. This increased length fiber can be wounded around the current conductor thus increasing the number of turns and sensitivity of the system. Better stability mechanically and compact size are the merits of bulk optic current Transducers (CTs) but CTs of optical fiber type have simplicity to form closedloops, flexibility to adjust dynamic range & sensitivity, small insertion loss & a comparatively high signal to noise ratio. There is a difference between circular birefringence inherent in some materials and Faraday Effect. In optical cavity/circular Birefringence, its sign relays on magnetic fields direction with respect to the propagation direction of the light. Practically it is non-reciprocal. A cumulative result in rotation of polarization will be obtained, if the same light is made to travel through the same medium, but having opposite direction of propagation. Up to certain extent, Faraday Effect is prevalent in all materials and depending on material s magnetic property their characteristic differ. The variation in Faraday Effect due to temperature is more prominent in paramagnetic as compared to diamagnetic and ferromagnetic materials. 2.5 current sensors With development in electric power system capacity there a vital role played by protection relaying systems. A sudden system failure like surge may occur in the system separating the failure part from it. Current sensors are required for these relaying systems and referred as current transformers or transducers (CTs.). Currently used CTs used are mostly electromagnetic devices, and these devices undergo residual field effect & magnetic saturation effect. Moreover, in power industry with the increase in voltage in distribution system, the insulation of electromagnetic CTs has become expensive as well as difficult. 9

20 Hence, the optical current sensors (OCS) or optical transducers are good replacements for conventional CTs that has no effect of electromagnetic interference on them [2-3]. With the progress in fiber communication and laser invention, the optical current sensors have drawn more attention and has been the prime concentration of the current sensing field these days. OCS is primarily developed in US by introduction of fiber to a communication signal. The proposal for replacing all conventional current sensors by all fiber optic current sensor was made. Sensing element using bulk glass was employed to overcome linear birefringence induced by fiber bending and intrinsic birefringence within optical fiber. Bragg grating based OCSs are recently interrogated and are becoming the prime field of OCTs having great potential. Classification of OCSs depending on different sensing element used is as following: Electro-Optic Current Sensor, Optical Fiber Current Sensor/AFOCT, Bulk Glass Optic Current sensor and Fiber Bragg Grating current sensor. The section that follows below gives the synopsis of these four types of OCSs, which provides the readers with a comprehensive introduction about OCSs. All fiber optic current sensor will be discussed in detail in chapter 3 for basic working principle and the same principle will be employed in chapter 4 for simulation purpose. This is the main topic of this Master project. 2.5 a) Electro-optic current Sensor: Given the performance of Electro-optic Current sensors, they are developed for high electric power industry [13]. Based on the schemes for optical modulation various types of Electro-optic Current sensor have been developed. These modulation schemes comprise of chromatic modulation, optical phase modulation frequency & intensity modulation. Now a sensor implementing intensity based modulation technique comprise of a standard detecting head, set of fiber optic cable and photo- electronic transformer. The photoelectronic transformer converts the modulated signal output to a modulated optical signal. This optical signal is then send to high voltage regions on ground through optical fiber link for advance processing. A frequency modulation scheme is proposed in [14]. Conversion of output signal coming from the sensing head to a frequency modulated signal is the most important procedure of this scheme. This way there is no dependency on temperature. Electro-optic Current sensor are facing sensing head power source problems and various 10

21 methods to overcome this problem has been presented. A different current sensor is used to supply power. This sensor draws power i.e. the electric power from applied current conveyed by the wire [15]. Rechargeable batteries is another approach to this problem, and these batteries can be recharged using solar panel [16]. This structure is easy to implement has low cost and a higher insulation level, due to the combination of optical fiber and traditional current sensor. These sensors can be the base formation for all optical current sensors to be accepted by the power industry. 2.5 b) Optical Fiber Current Sensor: Faradays Effect is used by OCSs implementing fiber optic cable for sensing. There is a change in polarized light wave velocity because of the magnetic field produced by the current in the conductor. It is possible to obtain precise current measurement by measuring the velocity change. This current measurement is been carried out using Sagnac interferometer loop current sensor since long time. Figure 2.5 illustrates its setup. Figure 2.5 Faraday/ Sagnac closed loop Interferometer current sensor [5] Change in the magnetic field because of the Faradays Effect varies the velocity of two circular polarization, hence induces a non- reciprocal phase shift between the two waves given by [17]: s = 2φ F (2.6) Where φ F =VNI, V being the Verdet constant, N is the number of turns of the fiber coil and I is the electric current. It can be seen that phase shift measurement results in availability of current. Optical phase shift induced by current can be measured using gyroscope 11

22 configuration. This configuration processes electronic signal and utilizes open loop minimum configuration. There are various advantages of implementing this scheme like compact size, quick response, light weight, accurate and wide dynamic range. The biggest problem with the sensing mechanism is that the sensor undergoes from Sagnac Effect. As Sagnac Effect induces non-reciprocal phase shift, it is difficult to distinguish from the one induced by current causing internal error. Hence, a Sagnac Interferometer current sensor with In-line configuration is been proposed to eliminate Sagnac Effect s influence. These sensors have same configuration as that of the close loop S.I as shown in figure 2.6. Due to the reflector or mirror at the end of sensing fiber, the polarized light is reflected back and send via same fiber. Thus the phase difference shift induced by Faraday Effect doubles, while travelling back in the fiber. Figure 2.6 In-line Sagnac Interferometer current sensor [5] Faraday Effect is the only factor responsible for phase shift between two beams, because of the polarization swapping. The non-reciprocal phase shift is two times as large as in close loop S.I current sensor. s = 4φ F, whereφ F = VNI (2.7) Close loop and In-line S.I configurations will be explained in more detail in chapter 3. 12

23 2.5 c) Bulk-Optic Glass Current Sensor Is seen that the sensitivity of AFOCT can be increased and their implementation is also easy but they have a comparative low Verdet constant & higher linear birefringence. Thus Bulk Optic sensors are used in most of the Industrial applications. There is a rotation in angle for the polarization plane, with respect to intensity of magnetic field, when the direction of polarized light is same as that of magnetic field in a magneto-optic material. Magnetic intensity can be obtained from this rotation angle. Hence, current is obtained [18]. There are few obvious advantages of Bulk Optic sensors as compared to AFOCT. As they are made from a single glass the Verdet constant of these sensors is 2 times greater than AFOCT. They are compact, light, stiffer or robust mechanically. Also given their low co-efficient of elasticity the linear intrinsic birefringence is very small, which provides sensor with high sensitivity. One more excellent advantage is that the sensor need not to be made up of single piece of material around the conductor. Hence without interrupting the flow of current in conductor, the OCS system can be installed easily. Internal reflection is one of problem that has an impact on the sensor s sensitivity. In the glass block the internal reflections are imperative as the incoming beam of light is essentially needed to propagate around the conductor. This leads to a phase difference among the orthogonal components of linearly polarized light which then changes the state of polarization of the light [19]. Different sensing schemes are proposed to overcome this problem. These schemes are triangular shape sensing element, ring-shaped & square shaped sensing element. There is also a sensing element which is openable and is based on principle of maintaining state of polarization of light beam by critical angle reflection [20]. Moreover, as the sensing element is fragile & demands high accuracy, it is difficult to manufacture them. Hence phase difference induced by the internal reflection & manufacturing defaults are the main issue for research in these sensors. 2.5 d) Fiber Bragg Grating Current Sensor There are distinguished advantages of Fiber brag Grating (FBG) compared to others given the advancement in Manufacturing of Optical fiber. There are many sensors based on FBG. Central refractive wavelength of FBG is given by [5] 13

24 λ B = 2n eff (2.8) Where n eff the core index of refraction, Λ is spacing among the fringes of the FBG. Change in the fringe spacing & core refractive index both aid to vary central refractive wavelength. If the temperature is kept constant, the ratio of strain to shift in wavelength is given by: λ B λ B = (1 P e )ε ax (2.9) Where λ B denotes the Bragg wavelength of the FBG, P e is the efficient photo-elastic coefficient, ε ax represents the axial strain of the FBG. As the FBG are temperature and strain sensitive, current measurement can be obtained from the sensing head by the induced temperature and strain changes. Several sensing head are now available. As shown in figure 2.7a on the surface of a giant magnetostrictive material (GMM). A different sensing head is illustrated in the figure 2.7b. For this sensor the two magnetic collection pieces (MCP) are arranged close to the current carrying wire in such a way that FBG are able to get the force between the two MCP. A wavelength drift of FBG can be obtained using various structure of sensing head. The drift in wavelength is relative to the current measured. Figure 2.7 a) Fiber Bragg grating sensor on GMM b) Fiber Bragg Grating sensor with MCP [5] Since the sensing information is encoded in the Bragg wavelength [21], the demodulation is independent of light level which is an outstanding advantage of the current sensor based on FBG over other schemes. Thus light intensity has no effect on these sensors. These sensors are still at experimental stage. 14

25 CHAPTER 3 WORKING PRINCIPLES OF ALL FIBER OPTIC CURRENT TRANSDUCER This chapter will discuss in detail the basic working principle of AFOCT and its theoretical model. Before that an interferometer detection scheme will be discussed as the AFOCT principle is based on Interferometric detection scheme. A practical model of Sagnac Interferometric fiber optic current sensor (SIFOCUS) will be proposed and will be explained in detail. The close loop and In-line S.I detection technique which were discussed briefly in chapter 2 will be covered in depth in this chapter. 3.1 Interferometric Detection scheme The rotation in plane of polarization of linearly polarized light, in terms of circular polarization for right & left handed circular polarization can be analyzed. The analysis is with respect to the phase difference of two circularly polarized modes. Interferometric detection scheme can be implemented for analysis purpose by generation of frequency modulated carrier, and the time delay introduced in the interferometer arms that modulates the optical phase [7]. An unbalanced Mach-Zehnder or Michelson Interferometer or a Sagnac Interferometer configuration can be used to generate the phase carrier [6]. A close loop S.I is shown in the figure 3.1 used commonly in gyroscope & the system is sensitive for non-reciprocal effect [22]. Figure 3.1 Close loop Sagnac Interferometer detection scheme [7] In this detection scheme the light obtained from the light source is polarized linearly using a fiber polarizer and is inserted into the Sagnac loop. The S.I loop has quarter wave plates i.e. crossed λ 4 plates, which are placed at 45 for the upper plate and 45 for lower plate with respect to the polarization plane of the input light. In this arrangement each of the counter propagating waves are converted to orthogonal circular states. The Sagnac loop is travelled by these light waves with different velocities different to each other, because 15

26 of the circular birefringence induced by the magnetic field. After travelling through the S.I loop the two wave s crossover & are converted back to linearly polarized modes. They interfere at the output of linear polarizer. On Mach-Zehnder Interferometer, which is used for generating phase carrier, a phase modulator is introduced to retrieve the phase information. In the Sagnac loop the accumulated phase proportional to magnetic field which is produced by the conductor current is recovered using heterodyne processing scheme. The Sagnac close loop current detection scheme is implemented in chapter 4 for simulation purpose. An improvised version of close loop S.I detection scheme is the In-line S.I as shown in the figure 3.2 [7] Figure 3.2 In-line Sagnac Interferometer Current detection scheme [7] There is a slight change in the In-line S.I configuration as compared to close loop S.I. with reflector at the end of the fiber and polarization maintaining fiber introduced between the polarizer and a quarter wave retarder. Quarter Wave retarder s retard (alter) the state of polarization or phase of light traveling through them. They are birefringent in nature having a slow and fast axis. When the polarized light is made to pass through quarter wave retarder, the light that travels through the fast axis moves swiftly through retarder compared to the one that travels through the slow axis. In the case of a quarter wave retarder, the wave plate retards the velocity of one of the polarization components x or y one quarter of a wave out of phase from the other polarization component. As a result the light coming out of the quarter wave retarder is circularly polarized. This function of quarter wave retarder is at times referred as rotating or twisting of polarized light. It should be noted that retardation of particular polarization component determines whether one will have a right handed or left handed circular polarization. In the In-line S.I setup the light is made to pass through to PMF at 45. A λ quarter wave 4 retarder plate oriented at 45 with respect to PMF s birefringence axis, converts the 16

27 incoming light from PMF, which has two modes orthogonally polarized into circular polarized modes having different direction of rotation. A converse process takes place at the fiber end when reflection takes place due to a mirror or any other reflector material. This reflection swaps the polarization states and each component goes through the optical path generated by the other. Towards the end, same optical path is travelled by these two polarization compensating all the reciprocal effects. A phase difference is still maintained, which is proportional to measured electric current. Because of the reflection the sensitivity of this detection scheme is twice compared to previous one, having the same number of turns around the conductor. Small current generate small phase difference and because of the introduction of PMF in front of 45 splice the system ability to detect these phase differences is enhanced. Even though a resistance to reciprocal perturbations to a certain extent is provided by the above scheme, the two waves propagating in different directions face birefringence alteration of the PMF due to perturbations varying with time, which are not compensated [22]. In addition a steady light source is required as the system is dealing with interferometers and the system bandwidth is limited by the process of modulation. Also there are small divergences introduced in the system output because of the λ 4 sensitivity of system [6]. quarter wave retarder, which affects the 3.2 Practical Sagnac Interferometric fiber optic current sensor (SIFOCUS) As discussed earlier in chapter 2, one of the major inevitable problem that hinders the performance of fiber optic current sensor (FOCUS) is the presence of linear Birefringence. Linear Birefringence affects the FOCUS in two ways. First the total usable length of the fiber is limited to half of the beat length of linear Birefringence, thus reducing sensors sensitivity. Secondly, as the linear birefringence is sensitive to environmental perturbations like temperature changes, the stability of the sensor response is affected. There are several approaches proposed to overcome the effect of linear birefringence, such as using a highly circular birefringent fiber obtained by twisting the fiber, or using a spun fiber [24] having low linear birefringence. However coiling the fiber leads to introduction of inevitable bend induced linear birefringence and these cannot be overcome by twist induced circular birefringence. If the coil fiber is annealed, the effect of bending induced birefringence can be removed but this makes the fiber fragile. Annealing is the process in which a metal or glass is heated and then allowed to cool down slowly removing the stresses in the glass or metal and toughens it. Moreover both the twist induced and bend induced birefringence are temperature sensitive and hence the Faraday rotation effect is masked. Despite of all, there still remains the problem of linear birefringence. The purpose is to have a sensor such that the sensitivity is not affected by any of the induced birefringence. Fiber optic current sensors are basically of two types, namely polarimetric FOCUS (PFOCUS) and Sagnac Interferometric FOCUS (SIFOCUS). This chapter will be concentrating on SIFOCUS. There are numerous advantages of SIFOCUS over PFOCUS. Which comprise of an uncomplicated structure of All-fiber without using the polarizer, 17

28 independence to input polarization and flexibility to use low-coherent source of light. Most importantly the effect of twist induced circular birefringence and bend induced linear birefringence can be reduced to minimum. To understand why SIFOCUS is more suitable as a fiber optic current sensor, an example of polarimetric fiber current sensor is presented in the section 3.2 (a) and the following section 3.2(b) will demonstrate practical SIFOCUS structure. 3.2 a) Polarimeteric fiber current sensor s Theoretical Background Based on Faraday Effect, a theoretical background of fiber optic current sensor having a single mode is introduced. Conventional polarimeteric fiber current sensor response (T) in existence of linear and Faraday effect induced circular birefringence is discussed for the comparison purpose between PFOCUS & SIFOCUS and to know why SIFOCUS is better. The setup of the system is shown in the figure 3.3. Figure 3.3 Polarimetric fiber optic current sensor [23]. S: optical source, I e : current carrying conductor, PBS: polarization beam splitter. P x, P y : Power measure in x and y polarization component. For given linear birefringence and Faraday Effect induced circular birefringence, the response of the sensor T from [25] & [26] is given by, T = 2 F sin{ [δ 2 +(2 F ) 2 ]L} [δ 2 +(2 F ) 2 ] (3.1) T is circular birefringence per unit length which results due to the intentional fiber twisting, δ is the linear birefringence per unit length (L) and 2 F is the circular birefringence per unit length induced by Faraday Effect. In conventional single mode fiber usually δ 2 F, then Eq. 3.1 becomes, T = 2 F L sin(δl) δl 18 (3.2) Equation above implies two important factors, the system sensitivity varies periodically along with the length (L) of the fiber. The sensitivity is affected and reduced by the factor sin(δl) δl also with the increasing length there is a decrease in the amplitude. And secondly as the linear birefringence is sensitive to environment, having a stable sensor will be an issue. Now if a twisted fiber is used which has a twist induced circular birefringence 2 T

29 δ, the sensitivity issue due to linear birefringence can be solved. In that case Eq. 3.1 with T + F becomes [23], T = (2 T + 2 F ) sin{ [δ2 +(2 T +2 F ) 2 ]L } (3.3) [δ 2 +(2 T +2 F ) 2 ] ( F = circular birefringence induced by Farady effect, T twist induced circular birefringence of twisted fiber) Thus it becomes possible to use longer fiber length to increase the sensitivity of the sensor. But it can be seen from Eq. 3.3 that any changes in T and δ can be confused with the circular birefringence induced by Faraday Effect. It is nearly impossible to make the measurement as F is much smaller than T and δ. Even a small variation in values of T and δ changes the sensor response significantly. Thus it can be seen that there are certain issues in implementing the PFOCUS system. Next section will discuss thoroughly the practical SIFOCUS structure which will help to reduce and overcome the twist induced circular birefringence and the bend induced linear birefringence. 3.2 b) Practical SIFOCUS Structure And Theoretical Analysis (I) Proposed Structure: Figure 3.4 shows the SIFOCUS proposed structure [23]. The sensing coil is formed by twisting the fiber uniformly so as to make the twist-induced circular birefringence much greater and it swamps the fiber s intrinsic birefringence. The rotating fiber loops form two state of polarization (SOP) controllers. This is a new feature of SIFOCUS and is used as indicated below. As shown in the figure 3.4 the transfer matrix M 1 & M 2 which lies between port 2, 3 and the sensing coil are made equivalent to λ 4 plate with orientation of fast axis angle of + π and 4 -π to x-axis respectively in the x-y plane. 4 The main reason to have this configuration is to have an additive result of nonreciprocal phase shift obtained between counter-propagating optical wave s induced by Faradays effect and the intrinsic nonreciprocal phase shift of the 3 x 3 coupler, which serves as a constant bias. It is possible to have a λ plate replace the two SOP controllers, if an ideal 4 isotropic condition is met by the connecting fiber which connects the fiber coupler to sensing coil. But practically it is impossible not to have any birefringence introduce in the fiber leads of the sensing coil and coupler. A matrix that can be defined for polarizing optical components which allows the nature of the polarization to be easily propagated, is the Jones Matrix [23]. 19

30 Figure 3.4 Practical Sagnac Interferometer using 3x3 coupler [23] I e is the current carried by the electric conductor in the center of the fiber coil. Input optical wave vector is E I. E II & E III are the output optical wave vector emerging from port II and III re θ) is the θ with respect to X-axis. (II) Theoretical Analysis: Above section describes the Jones transfer matrix for the fiber having loop in local (X, Y, Z) co-ordinate system with Z-axis along the fiber length. In laboratory system with fixed co-ordinate (x, y, z) the Jones vector of the optical wave filed is given as [23], E i = K 3i RJ c K I2 + K 2i J A RK I3 ]E I (3.4) (where i symbolizes port II, III from the fig. 3.4) For the above equation, between the ports of 3 3 coupler i & j, K ij forms a Jones transfer matrix and due to the fiber folding effect [27], R= [ 1 0 ] conversion matrix is used. 0 1 E i & E i are optical wave vectors. For the fiber forming the loops, J A & J C describe its Jones matrices in the local co-ordinate system for clockwise and counterclockwise optical paths respectively. J A & J C for a lossless fiber can be given as follows [23], J C = [ J XX J YX J YX J XX ] and J A = [ J XX J YX J YX J XX ] (3.5) With J XX 2 + J YX 2 =1, J XX 2 + J YX 2 = 1 & complex conjugate is indicated by *. For the analysis purpose the losses in the fiber are not considered, which makes the analysis simple. T T If the reciprocal case prevails then simply using relationship J A = J C gives J A. Where J C is the transpose matrix of J C. Now considering the case with Faradays effect, if J C is represented as a function of faradays rotation F i.e. J C = J C ( F ), then J A can be written as [23], J A = J C T ( F ) (3.6) 20

31 The optical power transmittance T i can be found by using Eq. 3.4 & 3.5. this power from port I to i (where i represents port II, III from fig 3.4) is defined as the ratio of output to input power & is given by [23], Where T i = E i 2 E I 2 = R oi k i R s cos( s CAi ) (3.7) R oi = (k 3i k I2 ) 2 + (k 2i k I3 ) 2 (3.8) k i = 2k 3i k I2 k 2i k I3 (3.9) CAi = 3i + I2 ( 2i + I3 ) (3.10) (where i symbolizes port II, III from the fig. 3.4) and Where R s = (A 2 + B 2 ) 1 2 (3.11) s = tan 1 ( B A ) (3.12) A = Re(J XX J XX J YX J YX ) & (3.13) B = Im (J XX J XX J YX J YX ) The real and imaginary parts are denoted by Re & Im respectively. For this system despite of the polarization properties of fiber loop the optical output is independent of the input polarization, as long as the coupler coupling ratio & fiber loss are independent of polarization. Thus even there is a change in SOP of input light, the controller needs not to be readjusted. This is because only the birefringence in the loop determine the controller setting as they are input SOP independent. For simplification purpose the 3x3 coupler is assumed to be ideally symmetric then k ij = 1, 3 CAII= π & 3 CAIII = π [28, 29]. From the 3 port II and port III the optical transmittance of the output T II,III is given by [23], T II,III = 2 [1 R 9 s cos( s ± π 3 )] (3.14) The + and sign in the above equation correlates to the transmittance subscript II and III, respectively. Using the following differential output, Faraday rotation can be obtained [23], 21

32 T = T II T III = 4 9 R s sin ( π 3 ) sin( s) (3.15) In 3.15 R s is the scale factor & value of R s is found using Eq & 3.13 by the elements of matrix J A &J C. The matrix element of J A & J C are also used to determine s which is the phase bias. Now the linear and circular birefringence, which are associated to fiber loop are used to determine the matrix element. The merit of the proposal SIFOCUS structure is having minimum effect of variation in bend and twist birefringence. A case with twist-induced circular birefringence fiber is consider to see its effect on SIFOCUS and bend induced circular birefringence is completely neglected. In order to determine the response of the sensor, the transfer matrices J A &J C of the fiber loop must be found first. The transfer matrix is represented by M (,θ) of the wave plate with retardation and fast angle orientation θ with respect to x-axis, J C can be obtained as [23], Where J C = M 2 J coil M 1 (3.16) And M 1 = M ( λ, π ) = 1 [1 j j 1 ] (3.17) M 2 = M ( λ, π ) = 1 [ 1 j j 1 ] (3.18) M 1 and M 2 are the Jones matrices for the quarter wave λ plate with the orientation axis at 4 π 4, π 4 respectively, in accordance to X-axis and for the fiber coil the Jones matrix J coil which is affected by Faradays effect is given by [23], J coil = [ cos( T + F )L sin( T + F )L sin( T + F )L cos( T + F )L ] (3.19) Where 2 T is the twist- induced & 2 F is faraday induced circular birefringence per unit length respectively. τ is twist rate & related to T by T = gτ, where g = 0.07 for single mode conventional fiber. The scale factor and phase bias can be obtained by using Eq and and are given as [23], The output of sensor response which is differential is given by, R s = 1 & s = 2 F L (3.20) T = 4 9 sin(π 3 ) sin 2 FL (3.21) Eq shows twist-induced circular birefringence has no effect on the response of the sensor, owing to their reciprocal property. Here the optical path cancels out the reciprocal perturbations whereas the faradays effect which is nonreciprocal is doubled. 22

33 Also λ quarter wave retarder plays a very important role as seen in above equations. To 4 know their importance, consider a case without using them. The scale factor and phase bias can be obtained by using J C = J coil and Eq as [23], Thus the differential output becomes R s = cos(2 F L) & s = 0 (3.22) T = 4 9 sin( π 3 ) cos(2 FL) (3.23) For this case the very small Faraday rotation sensitivity tends to zero. In practice, CAi which is the intrinsic nonreciprocal phase shift, cannot be efficiently added to the phase shift induced by the Faradays effect to serve as a bias. As discussed earlier the bend induced linear birefringence will be introduced inevitably in a real system where the sensing coil is formed by wounding the fiber. Thus there exist linear and circular birefringence simultaneously in the system. Now if the bend induced and twist induced birefringence are assumed to be distributed uniformly throughout the system/fiber. Then the Jones matrix can be written taking faradays effect into consideration, Eq and by using the result in reference [30] as, J XX = cos ( βl ) j sin 2 (1+ 2 ( βl ) (3.24a) ) 2 2 J YX = 1 (1+ 2 ) 1 2 j sin ( βl 2 ) cos(2θ b) 1 (1+ 2 ) 1 2 Where β = δ b [1 + 2 ] 1 2 in which = 2 ( T+ F ) δ b sin ( βl 2 ) sin(2θ b) (3.24b) and δ b is the bend induced linear birefringence oriented along the fast axis θ b. Eq. 3.6 can be used to obtained matrix element J XX & J YX of the counter propagating optical wave. Now by using Eq & 3.24 with some mathematical manipulations the scale factor and phase bias are obtained to be [23], R s = { 1 ( δ b ) 2 [1 + ( δ b ) 2 1 ] sin 2 (2 2 T 2 F L [1 + ( δ b ) ] T 2 )} T (3.25) s = tan 1 {[1 + ( δ b ) ] 2 T tan (2 F L [1 + ( δ b ) ] )} (3.26) 2 T Putting the above two equations in Eq.3.15, sensor response can be found. It can be seen from Eq & 3.26 that the scale factor, the phase bias and ultimately the sensors 23

34 response is related to twist induced circular birefringence and bend induced linear δ birefringence by the ratio b only, instead of [δ 2 2 b + ( T + F ) 2 ] 1 2 factor in T conventional PFOCUS. In this particular case only Faraday induced circular birefringence produces phase bias s to be measure and this phase bias s is independent of twist induced circular birefringence. Practically it can be seen from Eq & 3.26 under the condition T δ b there is a minimal effect on scale factor and phase bias due to variation of bend birefringence and twist induced circular birefringence. Eq & 3.26 can be simplified further for small signal case under the condition 2 F 1 as [23], R s 1 (2 F L) 2 ( δb 2 ) 2 T 2[1+ ( δ 2 b ) ] 2 T s 2 F L [1 + ( δ b ) 2 1 ] 2 T 2 (3.27a) (3.27b) By further using the scale factor and phase bias values from E.q 3.27a & 3.27b, a more simplified sensor response can be obtained. Thus for the proposed structure an enhanced sensor response can be obtained. 3.3 Basic principle of AFOCT In AFOCT the fiber can act as the current transducer itself. For the light travelling through the fiber, the polarization angle is rotated by utilizing Faraday Effect, corresponding to the magnetic field. This fiber is wound around conductor carrying current which provide it with immunity against extraneous magnetic force and current. This makes the AFOCT configuration very simple. And by changing the number of turns of the current carrying conductor, the sensitivity of the system can be improved [7]. A basic flint fiber current sensor is illustrated below [7]: Figure 3.5 Flint Fiber basic Structure [7] 24

35 Two circular polarizations with the same sense of rotation are counter propagating in the sensing coil to the Sagnac configuration of the sensor. Non-reciprocal phase shift is introduced due to the magnetic field generated by the current. Before entering the Sagnac coil the linearly polarized light is converted to circularly polarized light using quarter wave retarder. When the two beams of circular light return from the Sagnac loop, they are again converted back to the linear waves. The linear waves going back and forth are transmitted by PMF. These linear waves coincide having polarization direction parallel to core axis. 3.4 AFOCT Practical Model The figure 3.4 [1] shows theory model of AFOCT based on Sagnac principle. There is an incoming light from stable light source which is made to pass through the polarizer (linear polarizer). The light signal from the polarizer is split into two waves of equal intensity with the help of 1:1 coupler. These two light beams/waves are then injected into an optical fiber λ but before that they are converted to circularly polarize light beams by means of plate. 4 The current in the conductor generates magnetic field thus producing Faraday Effect. Faraday Effect provides a non-reciprocal phase shift to the waves. This shift in the phase has some positive and some negative changes. This circularly polarized light is converted back to linear polarized light again by using λ plate and the intensity of output optical power 4 can be measured using Photodetector. Figure 3.6 Practical AFOCT structure [1] For the simulation purpose the same Sagnac principle is implemented, but here an unbalanced modulator is used to generate a modulated carrier signal. Comparison between the modulated and unmodulated light wave is carried out & analyzed. The practical simulation model is shown below in figure

36 Figure 3.7 practical Simulation Structure [1] The above practical model forms the base for the actual simulation model which will be formed using optisystem 10.0 in chapter 4. The results will be obtained and verified by simulation in chapter 4. 26

37 CHAPTER 4 SIMULATION AND ANALYSIS In this chapter the simulation model will be generated based on the practical model from chapter 3 fig 3.5. This simulation model will be tested and analyzed using optisystem 10. This chapter will give a brief introduction about optisystem and will go in detail regarding all the components, which will provide the reason why they have been chosen for the simulation purpose to imitate the real world system of All Fiber Optic Current Transducer. 4.1 Optisystem 10 Optisystem is an innovative simulation software from Optiwave in the field of optical communication, which enables the user to design, test, analyze and optimize. It provides environment for excellent simulation capabilities, a distinctive classification system and extensive range of devices. Wide range of optical components practically used can be found in the simulation software. Interesting feature of optisystem is it comes with addition extensions to which user devices can be attached. This increases the simulation utility and makes it a widely used simulation tool in fiber optic system. 4.2 Simulation Layout and Component Description Figure 4.1 depicts the simulation structure of this project. The structure is based on the practical model described in chapter 3 in figure 3.4 & 3.5. As mentioned earlier in chapter 2 section 2.2 the Sagnac principle is put to use. This same principle is used for the theoretical model in chapter 3. Briefly in the practical application, the light wave travelling through the fiber which is wound around the current carrying conductor experiences changes in the angle of polarization due to the Faradays effect corresponding to the magnetic field. This generates a non-reciprocal phase shift in the waves. The waves are linearly polarized before entering the fiber loop and then with the help of quarter wave plates they are made circularly orthogonal with respect to each other as explained in chapter 3 section 3.1. The waves exiting the fiber loop are again linearly polarized and the intensity of optical power is measured to know the change in electric current which is basically proportional to the phase difference between the two waves. 27

38 For the simulation purpose (Fig. 4.1) the optical signal is linearly polarized before entering the power splitter/coupler then after the coupler it goes through an orthogonal circular polarization before it passes through the fiber. After exiting the fiber it changes back to a linearly polarized wave. In the lower link an optical carrier is modulated by an external signal like a sine/square wave that represents the change in current. It is then mixed with an unmodulated optical signal in the upper link through the optical coupler. The output of the coupler carries a phase change that is proportional to the original modulating current signal. Figure 4.1 Optisystem simulation layout for AFOCT [1] 28

39 4.2 (I) Analysis of the Main Components of the Layout Optical transmitter/ Light source model Optical transmitter consist of a LED or LASER as the light source. The information needed to be transferred is sent from source to transmitter via electrical signal. The function of optical transmitter is to convert the electrical signal into optical form and launch or emit this resulting signal into the optical fiber. That is the transmitter takes the binary data and converts it into a light signal. A simple block diagram is shown in figure 4.2 which demonstrates the above principle. Laser or LED Coding & Scambling Driver for optical source Fiber Figure 4.2. Block Diagram of a Simple Transmitter Similarly for the simulation purpose also, the optical transmitter has 3 main components: First an optical source having a user defined bit sequence generator, a NRZ pulse generator and a spatial LED which converts the binary digit to light signal. Spatial LED is a subsystem build using a LED component and a multimode generator. Second, an electric pulse generator, a sine/square wave generator and optical modulator which is a Mach- Zehnder modulator at the end. Mach-Zehnder modulator is used for electronically controlling the output amplitude or phase of the signal. The source model is shown in the figure 4.3 below having a center wavelength of 1550nm and a spectral width of 40 nm [1]. Figure 4.3 Source Model/Optical Transmitter An optical spectrum analyzer is used to measure the power generated by the source model versus wavelength as shown in figure 4.4 below [1]. The power launched is very important 29

40 parameter for any design as it indicates how much power loss can be tolerated by the fiber. The power is expressed in units of dbm with reference level of 1 mw. The same power can be observed throughout the system using different spectrum analyzers (SA) from SA_1 through SA_5 a seen in figure 4.5. Figure 4.4 Spectrogram of light source [1] Transmission and reception of simulation model The optical power generated from the source model is to be fed to fiber for transmission. Before that, the random unpolarized light from the Spatial LED is linearly polarized first and then divided into two linearly polarized light waves using a coupler or a power splitter. Operating with the same frequency of the signal both the coupler and 1 2 power splitter are equivalent. One of the output of the 1 2 power splitter is given directly to Right handed circularly polarized (upper wave) and the second output is given to one of the input port of a Mach-Zehnder Modulator, which is then given to a left handed circular polarizer. The other input port of Mach-Zehnder modulator is provided with a sine/square wave 30

41 modulating signal. The output from both the circular polarizers are then fed to two identical fibers. At the receiver section a coupler or 2x1 power combiner combines or mixes both the signal coming from the two fibers and passes it through a linear polarizer. The output signal of linear polarizer is inputted to a photo detector (PIN diode in this case). The photo detector involves the function of detecting the modulated signal and distracting the modulating information in form of electric current. The output of the photodetector i.e. the current is given to a trans-impedance amplifier (TIA) which is a current-to-voltage converter and also amplifies the signal. This amplified output of the TIA is then fed to a low pass Butterworth filter. As it is known a filter allows certain frequencies while blocking other, a low pass filter will allow lower range of frequencies and stop the upper range of frequencies. That is any frequency above the cutoff frequency will be blocked. The cutoff frequency used in the filter depends on input signal bandwidth. Butterworth filter is the most commonly used filter because its frequency response is known to be maximally flat i.e. no ripples. This is the result of a good design of the pass band filter which has a frequency response as flat as mathematically possible [31]. The final result is then analyzed by an electric oscilloscope visualizer at the output of the filter. 4.2 (II) Results and analysis of the overall system The power output of the light source model is observed and analyzed at various ports as shown below in figure 4.5. Every figure (graph) will refer to the figure 4.5. Stokes of the polarization before linear polarization is shown in figure 4.6 (a) and is observed using polarization analyzer. Stokes parameter of electromagnetic wave shows the state of polarization of the wave. After the polarizer almost all transforms into horizontal polarization [1]. Spectrogram of the light obtained from Spatial LED after the linear polarizer is shown in the figure 4.6 (b) using polarization analyser_1. 31

42 Figure 4.5 Analysis of the overall system 32

43 Figure 4.6 a) stokes of polarization for source simulation Figure 4.6 b) stokes of polarization for source simulation after the linear polarizer 33

44 Case I. Simulation using sine wave modulating signal. The output of modulating signal generated using a sin wave generator with a frequency of 150MHz is observed using oscilloscope visualizer is shown in figure 4.7. Figure 4.7 AFOCT sine signal Modulation wave form [1] The output of the Mach-Zehnder modulator is analyzed using optical time domain visualizer as shown in figure 4.8. It can be seen the signal is modulated in a sine wave. Figure 4.8 Optical time domain visualizer 34

45 Two circular orthogonal modes can be seen in figure 4.9 and are obtained by using circular polarizers. Right handed polarization for the upper using circular polarizer and left handed polarization using circular polarizer_1 for the lower. These orthogonal modes are obtained to achieve the phase difference in the waves [8] and are observed using polarization analyzer_3 for upper one and polarization analyzer_2 for lower one. Figure 4.9(a) Right Hand Circular polarization Figure 4.9(b) Left Hand Circular polarization 35

46 After both the waves have travelled through the fibers the resultant unmodulated and modulated waves can be observed in time and frequency domain using an optical time domain visualizers_3, optical spectrum Analyzer_3 and Optical time domain visualizer_4, optical spectrum Analyzer_4 respectively as shown in figure 4.10 and a) Unmodulated wave in time and frequency domain using optical time domain visualizer_3 and optical spectrum analyzer_3 respectively. 36

47 Figure 4.10 b) Modulated wave in time and frequency domain using optical time domain visualizer_4 and optical spectrum analyzer_4 respectively. 37

48 As described in section above after the light travelled both the fiber and obtained at the output of the coupler or 2x1 beam combiner, it is again linearly polarized and the spectrum of the light signal can be observed using spectrum analyzer_5 as shown in figure 4.11 before it is given to the photodiode where the conversion of signal from light to electric takes place. Figure 4.11 Spectrum of light intensity. 38

49 The photodiode output is observed using oscilloscope visualizer_1 as shown in figure Figure 4.12 Photodiode output time domain The signal can also be observed at the output of photodiode in frequency domain using a RF spectrum analyzer as shown in figure Figure 4.13 Photodiode output in frequency domain 39

50 Lastly the signal is amplified using TIA (Trans-impedance amplifier) and filtered using low pass Butterworth filter and the final result is obtained as shown by oscilloscope visualizer_2 in figure 4.14 Figure 4.14 Filtered output waveform of AFOCT simulation system after using sine wave modulating input signal. 40

51 Case II. Simulation using Square wave modulation The system can also be realized using a square wave modulation technique as shown in figure The simulation system remains the same except for the change in the modulating signal being square wave instead of sine wave. Figure 4.15 simulation structure with Square wave modulation 41

52 Also at the receiver a low pass Bessel filter is used instead of a Butterworth low pass filter as shown in figure Because when a Butterworth low pass filter is provided with a non-sinusoidal input it results in distortion. Resulting output wave form have overshoot and ringing as seen in figure This output belongs to the circuit in figure This happens because of fact that the frequency components, with respect to each other shifts in time. This is because, if there is a linear increase in phase with respect to frequency, there will be a resultant delay in the output with certain constant period of time [32]. Figure 4.16 Improvised simulation structure with Square wave modulation 42

53 Figure 4.17 Distorted output of the square wave simulation system After several trials, a Bessel filter is used instead and the problem mentioned above is rectified. The filter behaves as a delay line by introducing linear phase shift in accordance to the frequency while having low pass characteristics. Thus there are not any overshoots and ringing in the output wave form [32]. The important set of simulation results for the square wave case are observed below and they all are referred to figure The stokes of polarization before and after linear polarizer using polarization analyzer and polarization analyzer_1 respectively are shown in figure

54 Figure 4.18 a) Stokes of polarization before Linear Polarizer Figure 4.18 b) Stokes of Polarization after Linear polarizer 44

55 The modulating signal i.e. square wave in this case is generated using a NRZ Pulse generator and a user defined bit sequence generator as seen in figure Figure 4.19 AFOCT Square wave Modulation signal waveform The square wave modulated signal from Mach-Zehnder Modulator is shown in 4.20 observed using Optical Time Domain Visualizer. 45

56 Figure 4.20 Square wave modulated signal. The two circular orthogonal modes obtained from circular polarizers, right hand and left hand for the phase difference is obtained using polarization analyzer_2 and polarization analyzer_3 respectively. These orthogonal modes can be seen in figure Figure 4.21(a) Right Hand Circular polarization 46

57 Figure 4.21(b) Left Hand Circular polarization For the square wave modulation scheme too, the modulated and unmodulated waves can be observed in time and frequency domain using optical time domain visualizers_3, optical spectrum analyzer_3 and optical time domain visualizers_4, optical spectrum analyzer_4 respectively as seen in figure

58 Figure 4.22 a) Modulated wave form in time and frequency domain respectively. 48

59 Figure 4.22 b) Unmodulated waveform in time and frequency domain for square wave modulation scheme respectively. For square wave modulation scheme also the spectrum of the light wave can be observed in frequency domain before it is given to photo diode using optical spectrum analyzer_5 as shown in figure 4.23 as described in section before. Figure 4.23 Spectrum of light intensity 49

60 The output of the photodetector can be observed using oscilloscope visualizer_1 in time domain and using RF spectrum analyzer in frequency domain as seen in figure Figure 4.24 a) output of photodetector in Time domain. Figure 4.24 b) Photo diode output in frequency domain 50

61 Finally the rectified output of the filter can be seen in figure 4.25 using oscilloscope visualizer_2 referred to figure Figure 4.25 Rectified AFOCT simulation system output waveform in square wave signal modulation As it can be seen from Figure 4.25 the square wave is not retrieved completely but in parts. This may prove cumbersome to get the phase difference. Hence the system is further improvised by changing the source of light from non-coherent to coherent in section 4.3. Section 4.3: Simulation using Coherent Source of light AFOCT sensitivity is deduced by change in the phase, i.e. by calculating the phase difference. As explained in the section 4.2 above, different modulation schemes like sine/square are used to obtain the phase difference which represents the current change. But even though the sine modulation scheme works just fine, there is a difficulty in retrieving the square wave modulated signal while working with non-coherent scheme. To overcome this problem and get a proper phase difference which is proportional to the current information, I have changed the source of light which has the source information 51

62 from non-coherent to coherent. For this purpose the Spatial LED is replaced with a Measured Laser working at the same 1550 nm frequency and 10dBm power as shown in the figure The light source model and the spectrum of this light as seen in figure 4.26 are observed using spectrum analyzer which refers to figure The spectrum of light can be observed using optical spectrum analyzer_1 through optical spectrum analyzer_5. The simulation structure remains the same. Only the results that have been changed due to the change of light source will be presented in this section. Figure 4.26 Spectrogram of Laser light source 52

63 Figure 4.27 simulation structure using Laser light source for sine modulation. 53

64 Case I: sine wave modulation The output of Mach-Zehnder modulator is observed using optical time domain visualizer as seen in figure Figure 4.28 Mach-Zehnder Modulator output Figure 4.29 shows the unmodulated and modulated waves at end of fibers and can be observed using optical time domain visualizer_4 and optical time domain visualizer_3 respectively. Figure 4.29 a) Unmodulated carrier signal 54

65 Figure 4.29 b) Modulated sine signal The retrieved sine signal after TIA amplifier and the Butterworth filter is shown in the figure 4.30 below. It can be observed with the use of coherent signal the complete sine wave is retrieved as it is with a phase shift. Figure 4.30 output waveform of AFOCT simulation system in sine signal modulation and filtered for Coherent source. 55

66 Case II. Square wave modulation For the square wave modulated signal also, the system remains the same as discusses in section 4.2 (II) case II. Just with the change in the source, now being coherent measured Laser. This system is shown in figure The spectrum of laser light remains the same as in figure Figure 4.31 Square wave modulate system with laser as a source carrier signal 56

67 The modulating signal being square wave, there are changes in the output of Mach-Zehnder modulator as shown in figure 4.32 and the output of the two fibers for unmodulated and modulated wave as shown in 4.33 respectively. Figure 4.32 a) Mach-Zehnder modulator output in watts Figure 4.32 b) Mach-Zehnder output in dbm 57

68 Figure 4.33 a) unmodulated carrier signal 4.33 b) Modulated square wave signal 58

69 And the filtered output for the square wave modulated signal after the amplifier for laser source model is shown in the figure Figure 4.34 Improvised output waveform of the AFOCT simulation system in square wave modulating signal using a laser source. Section 4.4: Phase difference Calculations for sine wave case As discussed in the section 4.2 above, a practical simulation model is proposed where the change in the phase can be observed at the output of the system by comparing the output of the oscilloscope visualizer_2 and the oscilloscope visualizer. This phase change is calculated by using the formula = 360 F t (4.1) Where, = Change in the phase F = Frequency of the Sine Generator t = Time difference between the two sine waves 59

70 The sine generator frequency is user defined and an appropriate frequency is selected to get desired results for e.g. 150 MHZ as shown in Figure 4.1 above of the Opti-system simulation layout for AFOCT. Corresponding to the sine generator frequency the cutoff frequency of the filter is selected. Now to determine the sensors sensitivity and to draw a linear relationship between the input frequency and the phase change, I have considered 6 different cases. For all the cases the frequency is changes in intervals of 150MHz and the corresponding time difference is obtained using Opti-system as shown in figures below. Cutoff frequency remains same in all cases. After all required quantities are obtained the relative phase change is calculated using the above equation 4.1. Case 1: F=150 MHz ( t) = A-B = Figure 4.35 Phase difference for 150MHz 1 = =

71 Case 2 F = 300 MHz ( t) = A-B = = = 97.2 Figure 4.36 Phase difference for 300MHz Case 3 F = 450 MHz ( t) = A-B = = =

72 Figure 4.37 Phase difference for 450MHz Case 4 F = 600 MHz ( t) = A-B = Figure 4.38 Phase difference for 600MHz 62

73 4 = = Case 5 F = 750 MHz ( t) = A-B = Figure 4.39 Phase difference for 750MHz 5 = =

74 Case 6 F = 900 MHz ( t) = A-B = Figure 4.40 Phase difference for 900MHz 6 = =