1. May 19, 2011

2. Bill Reynolds
Technical Support Engineer

4. Agenda
Fiber Optic Theory
Connector Cleaning
OTDR Testing
OLTS Testing
CD Testing
PMD Testing
CWDM/DWDM

5. Fiber Optic Theory

6. Spectrum Optical fiber domain 850 nm  353 000 GHz
λ (nm)=c (m/s) / f (Hz)
Units
Micrometers (mm) m
Nanometers (nm) m

7. Index of Refraction n = c vac / c mat
The speed of light = c = km/s under vacuum
Any material that can transmit light has it’s own index of refraction represented by n
Example
n = c vac / c mat
In a given index of refraction, the speed of light gets slower
Speed of light in water ≈ km/s
Speed of light in fiber optic ≈ km/s
Index (n)
1.000
1.400
1.333
Material
Air
Fiber Optic
Water

8. Fiber Optic Coating Cladding Core Acrylate, Teflon, polyimide
Glass index n2
Core
Glass index n1

9. Fiber Optic Coating diameter = 250 µm Cladding diameter = 125 µm
Core diameter = 9, 50, 62.5 µm

10. Reflection θi = θR Reflection
When a light beam I hits a material with a different index of refraction, a portion of the beam is reflected R
The angle of this reflected beam is the same as the incident beam n1 θi θR
Core I R n2
Cladding
θi = θR

11. Refraction n1 sin(θi) = n2 sin(θr) Refraction
For the same light beam hitting a different material, another portion is refracted
This occurs when a the light goes through a material with a different index of refraction
The angle of this beam changes because the speed of propagation changes n1 θi θR
Core I R n2 θr T
Cladding
n1 sin(θi) = n2 sin(θr)

12. Total Internal Reflection
Critical angle (Total internal reflection)
There is a certain angle where 100% of the light is reflected and no light is refracted, we call this angle, the critical angle
Fiber optics use this concept to propagate light
Cladding
Core I R1 θC θR R2

13. Fiber Types Singlemode Multimode ITU-T G.652D
There are 2 fiber optics types in telecommunications
Singlemode
Multimode
Core
Core
Core
Cladding
Cladding
Cladding
62.5/125 (µm)
50/125 (µm)
9/125 (µm)
ITU-T G.652D
For telecommunications applications

14. Agenda
Fiber Optic Theory
Connector Cleaning
OTDR Testing
OLTS Testing
CD Testing
PMD Testing
CWDM/DWDM

15. Inspecting & Cleaning

16. Inspection Inspection techniques:
A microscope or fiber probe can be used to inspect connectors
A microscope acts as a magnifying glass. If you inspect a connector on a live fiber, permanent damage can be done to your eyes!
Using a fiber probe is the safest was to inspect a connector:

17. Cleaning Cleaning Techniques:
The best way to clean connectors can be done by following these easy steps:
Clean the outside of the ferrule with a wet pad
Clean the ferrule using a dry pad
Inspect the connector using a fiber probe
If the connector is still dirty, repeat the 2 previous steps with a wet pad
Dirty ferrule
Clean ferrule

18. 80% of network problems are related to dirty connectors!
Why Clean?
80% of network problems are related to dirty connectors!
Permanently burnt –
combined high power and dirt
Clean
Permanent damage can occur on dirty connectors on high power systems
RF video may reach nm
MDUs now includes MPO connectors
MPO are tricky to clean
View of an multi-fiber angle-polished connector 18

19. Cleaning
Bad cleaning results
Broken surface

20. ICIC: Inspect – Clean – Inspect - Connect
Damaged Dirty Clean
ICIC: Inspect – Clean – Inspect - Connect

21. Agenda
Fiber Optic Theory
Connector Cleaning
OTDR Testing
OLTS Testing
CD Testing
PMD Testing
CWDM/DWDM

22. OTDR Testing

23. Attenuation in fiber is wavelength-dependent
Water peak
C: 1310nm = 0.34 dB/km
D: 1383nm = 0.50 dB/km
E: 1550nm = 0.19 dB/km
Optical fiber is normally tested at the same wavelength that the fiber system will be operated.
Water peak
Corning SMF-28 SM Fiber

24. OTDR Principle The OTDR can be compared to a submarine radar.
Instead of sending RF or audio (submarine) signal to detect distant objects, it sends short pulses of light to detect events in fiber.
The OTDR locates and identifies events along the fiber.

25. Setting Up the OTDR

26. Rayleigh Backscattering
Comes from the “natural” reflection of the fiber
The OTDR will use the Rayleigh back reflections to measure the fiber’s attenuation (dB/Km).
Back reflection level is around -75 dB (depends on pulse length)
Higher wavelength will be less attenuated by the Rayleigh Backscatter
Source
Ray of light
Dopant particules

27. Fresnel Reflections Fresnel back reflections
Will come from abrupt changes in the IOR, ex: (glass/air)
Fiber break, mechanical splice, bulkheads, connectors
Will show as a “spike” on the OTDR trace
Reflection are typically UPC –45 dB and APC –65 dB (Typical OTDR results)
Fresnel reflections will be approximately 20,000 times higher than the fiber’s backscattering level
Will create a « Dead Zone » after the reflection

28. Pulse versus resolution and dynamic range
Short Pulse : more resolution but less energy
Long Pulse width: more energy but less resolution

29. Pulse Width
Short pulses will give a better resolution but less dynamic range:
Two connectors 3 meters apart
End of link (patch panel)
Connectors are measured for distance
and marked as separate events
5ns pulse
End of fiber is not reached due to low power of short pulses
Long pulses will give a better dynamic range but less resolution
Connectors are « merged » and identified as one event
30ns pulse
End of fiber is reached and located when using a larger pulse

30. 10 us pulse

31. 20 us pulse

32. OTDR Trace Single ended measurement UPC APC OTDR Connector
Fusion splice
Connector (P.P.)
Connector (P.P.)
End of link
UPC
Reflection
Power (dB)
Loss
APC
Slope shows fiber attenuation
Distance (km)

33. Macrobend
Loss in fiber is wavelength-dependent
Acquisition at 1310 nm
Acquisition at 1550 nm
Shorter wavelengths are more attenuated by fiber’s scattering
Longer wavelengths tend to leak out of the fiber more easily due to bending

34. Agenda
Fiber Optic Theory
Connector Cleaning
OTDR Testing
OLTS Testing
CD Testing
PMD Testing
CWDM/DWDM

35. OLTS Testing
Insertion Loss

36. Insertion testing is done in pairs
Dual ended measurement

37. Measurement Units The laser output is -3.50 dBm dBm
The dBm is use to measure the output power of a light source
Instrument reading
dBm
Fiber optic
Detector
Laser source
The laser output is dBm

38. Measurement Unit 1 mW 0.00 dBm mW How to convert the dBm in mW
dBm = 10*log(mW)
The laser seen on previous page was emitting -3.50dB
-3.50 dBm so 0.45 mW
Optical Power mW vs dBm
10000 mW
dBm
1000 mW
dBm
10 mW
dBm
1 mW
0.00 dBm
500 µW
-3.00 dBm
100 µW
-10 dBm
10 µW
-20 dBm
1 µW
-30 dBm
100 nW
-40 dBm
10 nW
-50 dBm
1 nW
-60 dBm
100 pW
-70 dBm

39. Measurement Units dB (relative power)
dB is the difference between 2 power measurements
Take the dBm laser shown previously
Laser output = dBm
There is an event on the fiber and the detector reads dBm
-3.50 dBm
To calculate this difference:
(-3.50 dBm) – (-4.25 dBm) = 0.75 dB
We have lost 0.75 dB
So the insertion loss is -0.75dB
-3.50 dBm
-4.25 dBm

40. Reflectance [dB] = Preflected [dBm] - Pincident [dBm]
Will come from abrupt changes in the IOR:
Fiber break, mechanical splice, bulkheads, connectors, etc.
We use the term « reflectance » when speaking of the amount of energy returned by specific points within the network
Expressed as a negative value
Connector 1 = -40dB
Connector 2 = -50dB
Q: Which one has the best reflection value?
Reflectance [dB] = Preflected [dBm] - Pincident [dBm]
Patch Panel
Fiber section
Connector
Fiber section
Fiber section
Mec. Splice
Patch Panel
Connector reflectance: -55dB
Connector reflectance: -45dB
Mechanical splice reflectance: -45dB

41. Optical Return Loss (ORL)
Comes from the amount of energy lost within components and fiber due to back reflections
We use the term « ORL » when speaking of the amount of energy returned by a section or an entire link
Expressed as a positive value
Link ORL = 35dB
Section 2 & 3 ORL = 45dB
Patch Panel
Fiber section
Connector
Fiber section
Fiber section
Mec. Splice
Patch Panel

42. Rayleigh Backscatter + Reflectance = ORL
Optical Return Loss (ORL) represents the sum of all of the light returned to the source from the fiber link under test.

43. Consequences of ORL
• Less light is transmitted
• Causes interference with light source signals
• Creates higher bit error rate (BER) in digital systems
• Reduces signal-to-noise ratio (SNR) in analog systems
• Causes fluctuations in the light source’s central wavelength
• Causes fluctuations in its output power
• Damages the light source permanently

44. UPC Connectors SC/UPC FC/UPC Ultra Physical Contact LC/UPC

45. APC Connector SC/APC FC/APC Angle Physical Contact LC/APC

46. Mitigating ORL Backscatter Reflectance
Is an inherent property of fiber and therefore there is nothing you can do about it.
However backscatter will have minimal affect on ORL
Reflectance
Dirty connectors!!!
Damaged endfaces
Improper mating
APC-UPC etc.
Clean connectors, clean connectors, clean connectors!!

47. Agenda
Fiber Optic Theory
Connector Cleaning
OTDR Testing
OLTS Testing
CD Testing
PMD Testing
CWDM/DWDM

48. Dispersion

49. Data rate on long fibers! Limits length on high data rate fibers!
Dispersion is the Fundamental limiting factor in transmission links and determines the:

50. Dispersions Multimode Fiber Optical Paths Optical 1
Frequencies 2
Difference in Arrival Times
Chromatic Dispersion
Polarization Modes
Polarization Mode Dispersion
Input Pulse
Output Pulse

51. A light pulse will spread and lose peak power as it travels down the Fiber.

52. The Pulse Broadening Causes intersymbol interference and bit errors
Each light pulse representing a data bit is smeared out and runs into the time slot allotted to the next bit, giving bit errors

53. Dispersion was never an issue because:
Low Transmission Rates (DS-0 to OC-48)
Switching equipment (OEO)
Regens were used, No EDFAs
Single transmission wavelength was near zero dispersion point.
(SMF nm = Lambda 0)

54. PMD Testing

55. PMD is Stochastic (random)
System Tolerance
Average PMD

56. PMD for % probability that the tolerable broadening will correspond to a mean power penalty of 1 dB
SONET-SDH
Bit rate
(Gbit/s)
Average DGD*
(ps)
2.5 (OC-48)
10 (OC-192)
40 (OC-768) 40 10
2.5

57. Birefringence Speed varies with Index of Refraction (density)
Fiber cross-section can have variable Index, which is called Birefringence

58. Birefringence Fiber manufacturing Fiber Geometry Internal Stress
Core Concentricity
Lateral Pressure
Environmental constraint
Heat
Bend
Wind (aerial fibers)

59. Birefringence Medium PMD High PMD The causes
Asymmetries in fiber core geometry and/or stress distribution create fiber local birefringence.
A "real" fiber is a randomly distributed addition of these local birefringent portions.
Medium PMD
High PMD

60. Visualizing PMD
Let’s visualize a light pulse traveling into a fiber

61. Polarization Mode Dispersion PMD
z,t

62. Polarization Mode Dispersion PMDh
Coupling length β
Bi-refringence

63. t
z,t

64. T0 T t Dt
fast axis
z, t
slow axis Dt

65. Dt t
fast axis
z, t
slow axis Dt

66. PMD Impact When the detector receives this:
It sees: Clear 1 and clear 0
When the detector receives this:
It cannot differentiate 1 and 0, this is when we get BER!!!
? ?

67. Typical Thresholds Average PMD Bit rate (ps) (Gbit/s) 2.5 40 10 10 40

68. PMD in a cable Cable 1 (18.6km): Fiber 1: 7.82ps Fiber 2: 23.15ps

69. Dual ended or Single Ended measurement

70. Cumulative PMD PMD adds quadratically: PMDTOT =N(PMDN)2
Example: 15ps + 2ps + 1ps + 6ps = 16.31ps

71. Mitigation of PMD

72. CD Testing

73. Dispersions Multimode Fiber Optical Paths Optical 1
Frequencies 2
Difference in Arrival Times
Chromatic Dispersion
Polarization Modes
Polarization Mode Dispersion
Input Pulse
Output Pulse

74. Material Dispersion
Each wavelength as it’s own speed in the glass material therefore it refracts at a different angle, separating the colors

75. Material Dispersion DFB Laser source Zoom
Central Wavelength
dBm
DFB Laser source
Zoom
Wavelength
Spectral width
The pulse is made out of all laser frequencies
Amplitude
Amplitude
Time
Time

76. Waveguide Dispersion N1 N2
With increasing wavelength, more power is traveling in the cladding section of the fiber
Mode Field Diameter N1
Wavelength 1 N2
Wavelength 2
Wavelength 3

77. Waveguide Dispersion
Index
Index profile of step index singlemode fiber G.652
Diameter
Index
Index profile of dispersion shifted fiber G.653
Diameter

78. 1. Light will propagate both in core and cladding
2. Core and Cladding have different refractive indexes
3. Spectral width of source consists of several wavelengths
4. Different wavelengths will propagate at different velocities this is because the refractive index is wavelength dependant.

79. Pulse Spreading occurs because different wavelengths of light travel at different velocities
Sources are not truly mono-chromatic (one λ) so each wavelength component will travel at a different speed 1 2
Single Mode Fiber

80. The Laser has finite line width
Spectral width is measured at FWHM (full width at half maximum) really consists of several wavelengths. nm
Wavelength (nm)

81. Therefore:
Within the spectral width of the laser the different wavelengths will propagate at different velocities within the core and at different velocities between the light in the core and the cladding.
This is the effect of both material dispersion and waveguide dispersion, which combines together into chromatic dispersion

82. Chromatic Dispersion (CD) specified in ps/nm*km
Total Chromatic dispersion of a link is specified in ps/nm
Lambda Zero or zero dispersion is the wavelength in which there is no Chromatic Dispersion
SMF-28 has about 17ps/nm*Km at 1550nm
CD Limits:

83. Chromatic Dispersion
CD is an addition of 2 chromatic dispersion types

84. Fiber Types +17 dispersion unshifted G.652 Chromatic Dispersion
+
+
Chromatic Dispersion
(ps/nm-km) 
dispersion shifted G.653

non-zero dispersion
non-zero dispersion shifted G.655

(nm)







85. Chromatic Dispersion (ps/nm -km)
Corning LEAF
Lucent
TrueWave
Balanced +
Lucent
TrueWave +4
Reduced Slope +2
G.653
Corning LS
Chromatic Dispersion (ps/nm -km)
Lucent
TrueWave
Balanced -
- 2
Corning
MetroCor
- 4
S-band
EDFA C-band
EDFA L-band

86. Threshold

87. Threshold 2.5 Gb/s Time slot 125 us 10 Gb/s Time slot 125 us 40 Gb/s
Cause #1
2.5 Gb/s
Increasing the speed of transmission by 4 will decrease the threshold by 16. This is due to 2 reasons
Time slot 125 us
10 Gb/s
Time slot 125 us
40 Gb/s
Faster bit rate means less space between pulses
Time slot 125 us

88. Threshold The chirp effect @ 2.5 Gb/s modulation @ 10 Gb/s
Cause #2
The chirp effect P
@ 2.5 Gb/s P 
modulation P 
@ 10 Gb/s
Pulse before modulation 
Faster bit rate means broader pulses P
@ 40 Gb/s 

89. Threshold
Mask Threshold
Increases BER

90. Dual ended or single ended measurement
FLS-5800 & FTB-5800
Dual ended or single ended measurement

91. Results Table

92. Mitigation of CD: Compensation could be preformed:
Per span: Negative dispersion fiber
Per channel: DCM
Per path: Tunable Dispersion Compensator

93. Summary: CD & PMD PMD and CD can increase BER and distortion.
PMD and CD are expressed in ps
PMD depends on the cable type and age and may change with environmental conditions, so it is important to test PMD on a regular basis.
To correct PMD, faulty fiber sections must be identified and changed
CD is a stable phenomenon as it is relative to refractive index of the fiber used
To correct CD, compensators may be used

94. Overview

95. Agenda
Fiber Optic Theory
Connector Cleaning
OTDR Testing
OLTS Testing
CD Testing
PMD Testing
CWDM/DWDM

96. CWDM/DWDM Difference testing

97. WDM Wavelength Division Multiplexing
WDM technology supports simultaneous
carriers/wavelengths/Protocols traveling along one fiber

98. WDM: Wavelength Division multiplexing
In order to increase the number of bits transmitted per fiber, multiple different wavelengths could be transmitted in the same fiber or system.
Mux/ demux and filters will combine and select the wavelength

99. WDM: Wavelengh Division Multiplexing
Class of WDM :
Wide WDM Devices: Access
channel spacing > 50nm
1310 , 1490 , 1550,
Coarse WDM : Metro
channel spacing > 200 GHz
but < 50 nm (typical less than 18 lambdas)
Common spacing 20nm
Dense WDM : Long haul
channel spacing  200 GHz

100. WDM bands

101. CWDM Application
CWDM is a WDM technology using 20nm between channels for a total count of 18 Channel in the O, E, S, C,L bands.
This technology is deployed in Access and metro networks where the cost per bit is critical.

102. CWDM Specifications
Chromatic dispersion transmission rates are a LOT lower than for DWDM
CWDM is for now limited to 2.5Gbps
Reason: Direct modulation causes CHIRP
CD is important to test, even for 2.5Gbps
CHIRP is induced by the modulation, the higher the bit rate the wider the signal mw
Ghz

103. DWDM technology
O-Band
E-Band
S-Band
C-Band
L-Band
1270
1290
1310
1330
1350
1370
1390
1410
1430
1450
1470
1490
1510
1530
1550
1570
1590
1610
These are only a small number of the ITU-T DWDM channel plan. There are hundreds more!

104. DWDM technology
DWDM is usually deployed where high bit rate and long reach are required

105. CWDM OTDR Channel Verifier Optical Spectrum Analyzer
Testing
CWDM OTDR
Channel Verifier
Optical Spectrum Analyzer

106. CWDM Channel Verifier

107. OSA Optical Spectrum Analyzer

108. System Critical Parameters
Channel Spacing
Power Level
Center Wavelength
OSNR
Peak Power
Laser transmitter output power level, which should be as high and as stable as possible to increase the transmission span length. Puisque plusieurs éléments optiques varient en fonction de la puissance il est primordial que la puissance de chaque canaux soit connu et monitoré.
Central Wavelength
The central wavelength of each channel in the signal is probably the most important single characteristic, given that its accuracy ultimately determines the ability of the source to communicate with the receiver. These values must also be monitored during maintenance programs to detect unacceptable drift. The accuracy of central- wavelength measurement increases in importance as channel bandwidths and spacing are reduced. Évidemment toutes les composantes WDM sont fonction de la longueur d’onde, alors il est important que celle-ci soit précise et constante.
Channel spacing
The channel spacing capability in GHz (ITU-T has defined a standard channel spacing of 100 GHz—about 0.8 nm—in the ITU grid; 50 GHz is considered in the grid—about 0.4 nm—and eventually lower values will be added.)
SNR
Although the BER is the best single parameter to characterize the performance of a link, it is determined principally by the optical signal-to-noise ratio (OSNR). It is an indication of the readability of the received signal; a parameter of increasing interest as the limits for long distance applications are pushed farther and farther. Plusieurs system vendors demandent de mesurer seulement les SNR pendant le commissioning de leur system, pas de BER. SNR est finalement la somme de la puissance, du bruit, de l’amplification, du center wavelength et drift. De plus, le SNR pourra donner un bonne idée du problème selon la valeur obtenues.

109. A power meter will mesure the POWER versus wavelenght
Power meter versus OSA
Sensitivity
Sensitivity
Wavelength
Wavelength
A power meter will mesure the
TOTAL POWER
+5dBm
An OSA will mesure the
POWER versus wavelenght

110. CWDM Solution for Multiple Users
Head-end
Each customer has a dedicated wavelength
Each customer has different distance from head-end
Conventional OTDR cannot test through CWDM system
Need CWDM OTDR for system installation, maintenance and trouble shooting

111. Troubleshooting CWDM

112. Testing CWDM
Specialized test equipment with wavelengths that match the CWDM filters are capable of testing end-to-end loss and optical return loss thus assuring the user that all requirements are met.

113. Thank You