Optical Time Domain Reflectometry System Engineering Essay

Published: Last Edited:

This essay has been submitted by a student. This is not an example of the work written by our professional essay writers.

The global aim of this project is to update an Optical Time Domain Reflectometry (OTDR) system in the laboratory and use a new fast sampling data acquisition system to improve the performance of the OTDR system. It is envisaged to implement fibre optic sensing using OTDR setup in a flexible modular system. This involves setting up lengths of glass or polymer fibre with a pulsed laser diode source and a suitable detector.

Prior to setting of the system, initial work will involve selecting and characterising OTDR components in the Photon Science Institute (PSI) laboratory, and thereafter, testing the assembled system with a new data acquisition system to trigger the light pulse and capture the light signal detected by the detector. Finally, the project will involve comparing the performance of the previously used photon counter and Photo multiplier tube (PMT), with a detector such as an avalanche photodiode (APD) and the new data acquisition system (DAQ). OTDR signal will be analysed and the parameters of data acquisition in laboratory set up evaluated.

1.2 Objectives

The breakdown of objectives of this project includes the following:

Define the performance of the existing OTDR system from Saunders PhD thesis and decide what improvement can be made [3].

Set up a new fast sampling digital signal processing DAQ system, interfaced with a fast modern computer.

Evaluate the system performance with a fibre as a distributed sensing system that senses over its whole length.

Investigate from DAQ set up: the effect of increased sampling rate and number of data on; measurement distance resolution, spatial resolution of events, number of averages, and problem of "dead event" zone, the dynamic range, and time response of the detector and the attenuation resolution of the sensor.

1.3 Introduction and scope of the dissertation

The implementation and analysis of fibre optic sensing using OTDR will be used to investigate new aspects of sensing and data acquisition with fast modern computer in the laboratory. This is a set up of OTDR in a flexible modular system (i.e. where the measurand and measurement properties required define the wavelength used, choice of specialist sensing fibre, light source, detector, etc.), with a view to being able to change and optimise components easily and carry out investigation based on fast sampling signal processing with a CompuScope 2125 card of 250 MHz or 1GHz digital oscilloscope.

This implementation and analysis enable determination of the effect of light pulses launched into an optical fibre of various lengths. The light undergoes Rayleigh scattering along the fibre length, and interaction with sensing points along the fibre, and a proportion of the light pulsed into the fibre is backscattered causing a reflected light signal to be detected at the input with total time duration for the light to travel through twice the total fibre length. It utilizes a multi-mode (MM) fibre, made of glass or polymer clad silica (PCS) or plastic optical fibre (POF) cables of various short lengths (ranging from 1m for initial test up to hundreds or 1Km). The detected backscattered light will provide information about the status of the fibre to determine whether there are any losses or break along the fibre, or the status of sensing points distributed along the fibre length.

This technique is widely used in the telecommunication industries to detect the location of fibre breaks along the link, measure attenuation and reflectance levels using OTDR equation that expresses Rayleigh backscattered signal and Fresnel reflection from the fibre end [1]. It has been practically proved that an OTDR can pinpoint the exact location of events which include reflections and losses due to bending, pressure, etc along the fibre. The timing position of the optical loss or discontinuity of the back scattered light signal could be related to the length of the fibre and used for estimating the overall attenuation and mated-connector losses.

The project involved comparing the performance of the previously used photon counter and PMT, with a detector such as a pin photo diode (HFD8000) and DAQ system. The DAQ system used (based on ADC, digitized with an appropriate clock rate) will be based around a waveform recovery instrument which records the relevant pulsed events down the fibre after the laser trigger occurred and signal detected. The recorded data will then be was then examined to extract any pertinent information. Initially, a flexible platform will be used to optimise the DAQ to the signals timing and intensity profiles. Some of these settings (user pre-settable) will include;

Sampling rate (of up to 10 GHz),

Signal sensitivity (from 20 mV with 8 bit scaling without pre-amplification),

Memory depth (Number of data points recordable up to 1 million),

Variable number of averages per acquisition (1 to 1000).

The recorded waveform(s) examined (off-line as a start), for further signal analysis and this could be used for the development of appropriate digital signal processing algorithms in the future. The latter could be included in the real-time acquisition system for on-line analysis of a given fibre. Optimisation of the above factors for an ample dynamic range, a good signal-to-noise ratio, and in essence, an appropriate test and measurement environment for sensing lengths of glass fibres will be discussed. The sensing system will have application for various sensing and monitoring applications such as structural health monitoring in industries (civil, aerospace, automobile, steel and power, etc).

Optical fibre sensors are classified into two main categories which include intrinsic or extrinsic. The intrinsic type of sensor uses its optical features as a sensing element which means that sensing takes place within the fibre itself but not in a region outside the fibre (extrinsic). The latter merely acts as a light transmission medium which means that the sensing takes place in a region outside the fibre. This project is based on intrinsic fibre optic sensors formed by sensitising the fibre to the measurand. Some examples of sensors are as follows:

Micro-bend sensors

Distributed sensors

Evanescent field sensors, etc.

Some advantages of fibre optic sensors include flexibility, electrical isolation, small size and weight, high bandwidth, no equalization required, ground loop elimination, electromagnetic and chemical immunity, no licensing required, low cost sensors, etc.

More over, suitable software based on VB6 program will be developed to capture and display signal from the scope and display it in the computer for easy analysis of the data acquisition and storage of recorded waveforms in the computer memory. This software will make it possible to vary the configuration of data acquisition settings remotely and perform measurement and analysis of traces of the recorded waveform with ease.

Measurement of the effect of pulse width, modulating frequency, lasing current to the detected signal and their effect on the captured signal by the OTDR based technique will be investigated and presented comprehensively using graphs or tables within the measurements and results from the experiment. Calculation of power budget, transmission loss and estimate of maximum fibre link within the coverage of the sensor will be performed to serves as a guide for project methodology and analysis of measurement.

In this work, definition and calculation of transmission losses and analysis of OTDR signals based on the parameters of the fibre sensing system (wavelength of the light sources and detector, fibre transmission characteristics, number of connectors, splices, loss margin, etc) will be presented. It will also investigate how to determine the maximum fibre link for a given optical power budget and attenuation specifications. It is important to distinguish between the OTDR for telecommunication (which has a defined light source wavelength and fibre properties due to optical telecoms standards) and the OTDR based sensing system implemented in this project. The technique here is different from telecommunication OTDR because the set up is flexible, i.e. every component (detector, new light sources, sensitised fibre lengths, etc) and parameters (e.g. variable wavelength) could be varied according to the application and sensing of pulsed signals and measurement is taken only from the other end of the fibre. However, an inclusion of a Y-Splitter in the system setup makes it an exact replica of an OTDR system and this will be clearly illustrated and explained with diagrams in chapter four.

1.4 Scope of the dissertation

Chapter two considers the background and literature review. It briefly describes what led to the development of OTDR as a fibre optic sensor and how it developed into commercial products. It captures most current OTDR available in the market according to product information and journals and technical trend of sensor in Europe.

Other works to be reviewed include some of the previous works relating to OTDR as a fibre optic sensor distributed optical fibre sensor (DOFS) based chemical sensing by PhD student from University of Manchester (UMIST) in 2006 [3] and a research group from University of Limerick (a multipoint 2 sensing element, optical fibre interrogated using OTDR and application of artificial neural network pattern recognition techniques) [6].

Chapter three describes the lab equipment available for the project. Testing the subsystems, (calibration where necessary) and discussion of components parameters which will be applied in the experiment. It covers the comparison between some OTDR equipment, emitter/receiver characteristics, and laser diode and laser diode safety awareness as prerequisites for starting the project.

Chapter four covers the methodology of the experiment, which includes the flowchart and block diagram with description of stages of measurement and system setup in the laboratory. This covers the set up of various components and their interface with each other, which involved matching the relationship between their bandwidth, spectral response, the use of different classes of light sources (class 1 to 3R) with various characteristics of the output light (infrared light and visible light) and detectors of various wavelengths, short length of glass fibre cables and so on, in a flexible setting until a satisfactory methodology of implementing a sensor based on OTDR techniques was developed.

Chapter five describes the measurements and analysis of results. This includes presentation of data sets with graphical representation of light source output and detection signals which followed the ac modulation voltage without distortion, and some waveforms recorded from the sensing system set up and analysis of different calculations based on the results and general discussion regarding the DAQ experiment.

Chapter six covers the discussion and concluding remarks which includes the objective evaluation of the project's achievements and failures with suggestions for future projects relating to OTDR data acquisition for fibre sensing systems.

The list of references (which consist of all books, articles, instrument operation manuals, relevant websites with date and time of visitation, etc) are included and the appendix which covers definitions of some important words, phrases and terms relating to OTDR signal analysis, which the reader may come across in this work.



Aims and introduction

Description of systems set-up in the laboratory

Flow chart of experiment stages

Block diagrams of basic steps of design implementation

Power budget and loss estimation

Fibre optic sensor and fibre splitting


Aims and introduction

This chapter describes the project implementation methodology, power budget and loss prediction to ensure that light signal reaches the detector via the optical fibre with acceptable SNR. The chapter introduces the set up with a diagram (schematic and photograph), describes the steps to assemble the optical and electronic components (to confirm that each component is interfaced with the previous and next component properly), calculates a power budget calculation based on the set up, and explains a diagram of the final OTDR system with y-splitter of 3.0 dB inserted to couple the light source, detector and extended length of fibre cable at the same end. Both block diagrams and flow chart are used to explain the methods for the project implementation where appropriate.

Description of systems set-up in the laboratory

Diagram and description of sensing system set-up in the laboratory

The diagram in figure 4.1 shows a laboratory arrangement of a fibre sensing system based on OTDR techniques. At the start, measurement of optical light power is carried out at the other end of the fibre with intention to eventually include a 50:50 splitter or y-coupler to enable detection of reflected light at the same end of fibre.

A suitable digital pulse generator is used to pulse an appropriate modulating frequency (of square voltage waveforms) into the LDC-3724B Laser diode (connected to LDM-4980 LD mount) in order to amplitude modulate its optical output signal, which is coupled through a glass fibre link (1 to 1km length of glass fibre). The light transmitted through the fibre is detected with a detector (HFD8000) via a meter length of glass optical fibre. The detector is connected to a 1GHz digital oscilloscope (via a current amplifier depending on fibre length) which is interfaced to a fast modern computer via 82357B USB/GPIB interface USB 2.0 cable. Appropriate OTDR software based on VB6 is written for fibre optic sensing and data acquisition analysis. Optical signals (waveforms) as a result of the modulating frequency and light power pulsed into the fibre are detected, captured and is displayed in the oscilloscope locally or in the computer remotely. The detected optical signal which contains the characteristics or status of the fibre cable are captured with the help of the VB 6 software in form of generated graph known as traces and provide the intensity of detected light power on the y axis versus sample rates of data acquisition (based on the number of averages) on the x axis. Figure 4.2 illustrates the photographic picture of the sensing system set up (flexible and interchangeable system):

Figure 4.2 Laboratory set up of fibre optic sensing (photographed)

The diagram in figure 4.2 includes 1km length of glass fibre in multi-mode. Methodology adopted to realise the set up and carry out fibre optic measurements is covered in sections 4.3 (flow chart) and 4.4 (block diagrams).

Flow chart of experiment stages

The first procedure was to button ON the PC, pulse generator and DAQ. Connected the pulse generator to the DAQ via a BNC connector and interface the DAQ to the PC using a plug and play cable (82357B USB/GPIB interface USB 2.0 which has the capability of 1.15MB/sec data rate), as described in chapter three (section 3.2.5). This was to ensure proper interface of the DAQ to the PC and check that voltage signals from the pulse generator could be detected and stored. This is better explained in section 4.4 and illustrated with a block diagram (figure 4.5). However, some basic stages of the methodology are represented in a flow chart in figure 4.3:

Figure 4.3 Flow chart of some basic stages in methodology

Block diagrams of basic steps of design implementation

Basic steps of project implementation are also illustrated with block diagrams in figures 4.3, 4.4, 4.5, 4.6, 4.7, 4.8 and 4.9 as follows:

Step 1: Connect laser diode to the detector (HFD8000).

Figure 4.3 Interfacing light source to detector

The first step towards implementing the setup involved connecting the laser diode optical output directly to the photo detector that has matching spectral response characteristics. The emitting wavelength of the light source was checked to ensure that it is matched to the wavelength range of the detector.

Initially, a tiny laser diode module (of the type used in a laser pointer pen) from Maplin electronics with an output light power 0.7 to 0.99mW, wavelength of 670nm, 5Vdc power supply and narrow beam aperture of 2.8mm (diameter) was used to interface with BPX series photodiode silicon detectors. BPX 65 fibre optic detector is an ultra speed photodiode with bandwidth of 140MHz and spectral response in the wavelength range of 400nm to 1100nm and responsivity of 0.5A/W (at wavelength of) 850nm.

The output of the BPX65 photo detector was connected to a DVM, while it is reverse bias using a battery as shown in figure 4.4. The initial test was to switch the light source on and off and look for a response from the detector using a DVM. Since the light was visible you could see whether it was emitting or not and where it was being projected.

Figure 4.4 BPX65 photodiode circuit

Pin Designations

1 - Dout-

2 - Cathode

3 - Dout

4 - Ground

5 - Anode

6 - Vout (5V)

Step 2: Connect laser diode and detector to digital oscilloscope, figure 4.5

Figure 4.5 Connection of LD, detector and oscilloscope

When it was confirmed that the BPX65 photo diode could detect light from the Maplin LD module (output light power 0.7 to 0.99mW) as indicated by the DVM, the output of the detector was transferred to the scope via a BNC connector.

Subsequent light sources and detectors including ANDO AQ-4112 laser diode stabilised light source (emitting wavelength of 1550nm) and the detector ANDO AQ-2105 (wavelength range of 750nm to 1700nm) was interfaced, with the output of the detector connected to the scope using a BNC connector. Signal level was noted as the detected photocurrent by the optical detector is converted to voltage via the impedance lead of the scope.

The ANDO AQ-4112 laser diode and ANDO AQ-2105 photo detector (optical power meter) was latter on interchanged with Lightwave laser diode and optical photo detector (HFD8000) via the pigtail and ST type connector, detail of their transmission characteristics and matching is detailed in chapter three sections 3.2.

The light power detected by the photodiode is converted to a photocurrent by the detector. The oscilloscope measure voltage, via Ohm's law that current flowing through a conductor varies directly proportional to the potential difference (V) across the terminals and inversely proportional to the resistance (R) between them; i.e. I = V/R using the impedance of the scope leads to obtain voltage from current through equation (4.1) below:

V = IR ………………………………………………….. ….. (4.1)

The optical photocurrent detected by the photo detector could also be converted to a voltage by the current amplifier (via trans-impedance amplifier) with a gain of +10 as shown in figure 4.2.

Step 3: Connect the LC534Al digital oscilloscope to a PC for DAQ, figure 4.6

Figure 4.6 Interfacing to a fast modern computer

Next stage involved connecting the oscilloscope to the fast modern computer (with Windows XP operating system, very fast RAM and an Intel processor). It automatically has an inbuilt standard Universal Serial Bus like every modern PC.

The PC and the scope interface were possible via 82357B USB/GPIB interface USB 2.0 cable USB 2.0. This is an Agilent fast and easy connection instrument USB 2.0 (IEEE 488 compactable) cable with data transfer rate of 1.15 MB per second. It made the scope accessible from the computer and controlled through the window.

In order to perform DAQ, a software program development was needed, hence VB6 based program was written and installed. Check on the compatibility of the VB6 program with hardware system set up went successfully. The application of the software in this work is better described with the measurements and results covered in the next chapter.

Step 4: Connect laser diode through a meter fibre length to the detector and confirm that the all components are matched and interfaced as would be indicated in the scope if the photo detector is detecting any signal through the fibre optic cable.

Figure 4.7 Connect a meter length of glass fibre

Then Lightwave laser diode was connected to optical photo detector (HFD8000) via the pigtail and ST type connector, detail of their transmission characteristics and matching is detailed in chapter three sections 3.2. The usable spectral transmission range of the fibre cable and the spectral transmission response of the photo detector had to be matched (as discussed in chapter three sections 3.2).

Step 5: Connect laser diode through a fibre to detector and connect to DAQ systems, figure 4.8

Figure 4.8 Flexible and interchangeable sensing system

Step five as illustrated with block diagram in figure 4.8 comprises the complete sensing system set up. This is also represented in figure 4.2 (photograph of the lab set up). Appropriate modulating frequency by the pulse generator is connected the to laser diode driver (via a BNC connector from the pulse generator main out of 50Ω/600Ω load to the laser diode external of the laser diode module, and another BNC connector from the pulse generator AUX output of TTL/CMOS connected to Ch 2 of the scope). If the length of fibre optic cable increases, the signal level detected by the fibre optic photo detector may become too low and interfacing the detector output via current amplifier (with gain *10) would improve the signal level.

The complete set up as presented was to help to compare the optical pulse signal delivered to the laser diode driver with the optical signal detected by the detector.

Furthermore, the tests to be performed with this flexible modular system set up are aimed at the following:

To achieve a laser output and detection signal that follows the ac modulation voltage signal without distortion or clipping.

The characteristic performance of the laser diode (and other components) in the set up and relate it to data sheet

Evaluation of the effect of changing DAQ parameters like; number of averages, sampling rate, and sampling points on the signal to noise ratio at a given fibre length.

Also the rise and fall time of the detected signal and how this is affected by the bandwidth of the pulse generator, oscilloscope, DAQ/sampling rate, and the rise and fall time of the laser diode and the photo detector.

Evaluating all these parameters would help to predict the optimum spatial resolution (explained in appendix) of an OTDR setup.

Step 6: Keep fibre length constant and perform data acquisition and sampling.

The next stage is to keep the fibre length constant and explore data acquisition with the 1GHz digital oscilloscope capable of 2 GS/s 8 Mpt in single mode and 500MS/s 2 Mpt in QUAD mode. Basic parameters of the software to investigate include the effect of number of averages and sampling rate to the output waveform characteristic.

Carryout measurements and analyze result with graphs to describe the equipment transfer characteristics relationship (laser diode bias drive current, optical signal detected, amplitude of optical signal modulation and modulating frequency). Threshold lasing current value is defined as the value of the laser diode parameter (e.g. current) at which the laser diode switches from strictly spontaneous emission (acting like LED) to lasing emission. At this point if represented in graph as illustrated below in figure 4.9, the direct proportionality slope between an output power (light) and current starts to rise. The rising indicates that the laser diode has been biased above the threshold current hence; it is in lasing emission region (not like LED).

Figure 4.9 Definition of threshold value graph [19].

Figure 4.9 graph is typical CQF939/400 Output power versus the laser bias current drive, and it illustrates that the output power of the laser diode is directly related to the laser diode driving current. It is aim in the measurement to relate the laser diode drive current to the signal level detected by the photo detector, and evaluate the effect of the laser diode drive current stability to the level of signal detected by the fibre optic detector in this set up. The effect of the amplitude modulation and modulating frequency required to set or bias the laser diode drive current above the threshold lasing current and modulate the laser current around this bias current at the pulse generator frequency.

Power budget and loss estimation

This is simply defined as the minimum transmitted power (under worst scenario) and the maximum (under worst scenario) receiver input power required. This is better explained with a simple diagram as in figure 4.10

Figure 4.10 Power budget estimation

Power budget calculations are carried out for an optical transmission link in order to confirm the power requirement to get to the optical receiver, under all conditions of power losses along the link. Some factor that are considered in power budget include: output power from the transmitter, Pt (dBm), the receiver sensitivity, Pr (dBm), number of connector, Nc, loss per connector, Lc (dB), average distance between connector, number of splices, Ns and loss due to splices, Ls (dB), attenuation due to fibre, Lf (dB), power margin, M (dB), etc. Mathematical relation for power budget and power loss estimation is in equation (4.2) below

Pt - Pr = Loss Budget = zLf + NsLs + NcLc + M ……………………………….. (4.2)

The calculation of losses based on the flexible modular system setup (figure 4.2) will cover the loss along a 1 km fibre length (applied in the experiment) with diameter of 62.5/125 µm core/cladding, three connectors, no fibre splitter used yet, fibre attenuation/km of 1.5dB and HFD8000 detector of wavelength of 1300nm (under multi-mode fibre). Total transmission losses are estimated as below:

Sum of all losses = zLf + NsLs + NcLc + M


Safety margin (M), = 3.0 dB

No of connectors (Nc) = 3

Losses due to ST type multi mode fibre connector (Lc) = 0.75 dB

No of splices (Ns) = 0

Losses due to splices (Ls) = 0 (however, it is always 0.1 dB for both single and multi mode fibres)

Fibre attenuation (Lf) = 1.5 dB/Km (values are per TIA/EIA and industry specification as used in all transmission network for link calculation) at 1300nm.

Length of fibre (Z) = 1 Km

Transmission loss = [1 Km * 1.5 dB/Km] + [3 * 0.75 dB] + 3dB

= 6.75 dB

Therefore, with a laser diode power output of 1mW (0dBm) at constant power setting of LDC-3724B (3mW) and a typical pin receiver sensitivity of 1.0µW (-30dBm):

The power budget = 0dBm - (-30dBm)

= 30 dBm

This implies that 30dBm of attenuation is possible over the link before failure occurs. So, in order to calculate the maximum glass fibre attenuation, the total link loss calculated above (based on the system set up for this experiment) is subtracted from 30 dBm power budget.

Therefore, total attenuation allowed = 30 - 6.75

= 23.25 dB

The power margin of 3 dBm included in the loss estimation will take care of splices (if included), extra fibre length, etc. Laser diode used for the experiment can operate in either constant current mode or constant power mode.

4.6 Fibre optic sensor and fibre splitting

Figure (4.11) shows a schematic diagram of OTDR with a Y-splitter (3dB). It is the splitter that makes fibre optic coupling possible, i.e. to couple the power excitation output into the fibre cable link under investigation and simultaneously connect the reflected (backscattered) power into the optical detector. The main different between the sensing system set in figure (4.2) above is that the optical detector in an OTDR as illustrated in figure (4.11) is at the same side of the fibre cable under test (made possible by the use of splitter).

Relating the fibre optic sensing system set up as covered in the methodology above to telecommunication OTDR implementation in the lab is better illustrated with a block diagram.

OTDR includes a Y-splitter as illustrated in figure 4.11, and the wavelength of telecommunication is specific whereas, the sensing system covered so far in this project is flexibly set (variable components and parameters of transmission characteristics including wavelength matched to optical properties of measurand that the fibre is sensitised to, as deemed important for sensing application in the laboratory).

Power budget due to y- (50:50) splitter

OTDR provides a view of fibre optic link from one end of the fibre cable. As the length of the fibre optic link increases, there is limitation to the measurable light intensities an OTDR can provide. This depends on its dynamic range. The dynamic range (illustrated in appendix, figure 1) of an OTDR describes the maximum and minimum measureable light intensities, which helps to find out the loss in fibre optic link which an OTDR can analyze, i.e. the ratio of light energy launched into a fibre to the minimum detectable reflected power. OTDR can analyze a fibre optic link either by Rayleigh backscattering (calculating the loss level in the fibre optic link as a function of distance) or Fresnel reflection (this is the type of reflection as a result of physical events along fibre optic link that is under investigation).

The important point in relating OTDR (telecommunication standard with specific parameters like wavelength, etc) to the system set up in figure 4.2 (OTDR; flexible and interchangeable modular system) with regards to attenuation and power budget includes but not limited to the followings:

The y-splitter (3dBm) reduces the light power into the fibre optic cable (e.g. if 8dBm is pulsed into the fibre optic link, the splitter reduces the power down to 5dBm) and OTDR equation relating to losses due to Rayleigh scattering mechanism (y-splitter) is contained in appendix, equation 9.

The fraction of reflected light power into the link (this is 0.04 for most common glasses), as some light power are refracted (non magnetic surfaces).

The pulse width, which determines when the light source is on (figure 4.7 requires short and sharp pulse for 1m - 1km fibre length while an OTDR with y splitter would require a longer pulse which depends on the length of the link so that the reflected light power would be detectable by the optical detector), etc.


Simple step by step approaches followed to implement OTDR (flexible and interchangeable system) for fibre optic sensing have been covered. Power budget estimation along the fibre optic link with the system discussed and calculated based on the optical components used in the project. Power budget for an OTDR (with 50:50 splitter) as illustrated with a block diagram (section 4.6), including light power reflected from the fibre link via a Fresnel reflection or Raleigh backscattered OTDR analysis have been discussed and related to flexible interchangeable system set up. Measurement and results based on the set up as presented in this methodology will be covered in the next chapter.