Fiber and component metrology for high-speed communications:What the manual doesn’t tell youPaul Williams, Paul Hale, and Tracy ClementNational Institute of Standards and TechnologyBoulder, Colorado1. Polarization-mode dispersion (Williams)2. Transmitter/receiver frequency response (Hale and Clement)1Part 1. Polarization-mode dispersionPMD measurement advice for folks with turnkey measurement systemsAssumptions:A basic understanding of PMDA PMD measurement systemAn understanding of the measurement techniquesAvoid measurement traps that give false results.2Steps to a good PMD measurement1. Perform measurement “calibration”2. Understand limitations imposed by measurement conditions3. Choose measurement parameters correctly4. Be aware of measurement uncertaintiesAssumption: measurement system works correctly3Review: PMD DefinitionsOutput PSP_General Case:(slow)Output PSP+Input PSP_ (fast)Input PSP+(slow)(fast)• Birefringence affects propagation velocity and output polarization state• PMD is characterized by two Principal States of Polarization (PSP)• PSPs are wavelength-independent (to first order)• Propagation along PSPs is the fastest/slowest possible• PMD is the phenomenon, DGD (∆τ) is the magnitudeDifferential group delay (DGD): ∆τ =τ −τfast slow4Review: Time domain PMD measurementlow-coherence interferometery (INT)broadband sourceDet.∆τ Time (ps)• “Width” of delay histogram gives mean DGD• Measures only mean DGD ...
Fiber and component metrology for high-speed communications: What the manual doesnt tell you
1.
2.
Paul Williams, Paul Hale, and Tracy Clement National Institute of Standards and Technology Boulder, Colorado
Polarization-mode dispersion (Williams)
Transmitter/receiver frequency response (Hale and Clement)
1
Part 1. Polarization-mode dispersion
PMD measurement advice for folks with turnkey measurement systems
Assumptions: A basic understanding of PMD A PMD measurement system An understanding of the measurement techniques
Avoid measurement traps that give false results.
2
Steps to a good PMD measurement
1. Perform measurement calibration 2.Understandlimitationsimposedbymeasurement conditions 3.Choosemeasurementparameterscorrectly 4. Be aware of measurement uncertainties
Assumption:eremtnssyetmowmeasuyltcerrocskr
3
General Case:
Input PSP+ (fast)
pu _ In t PSP (slow)
Review: PMD Definitions
Output PSP _ (slow)
Output PSP+ (fast)
Birefringence affects propagation velocity and output polarization state PMD is characterized by two Prin cipal States of Polarization (PSP) PSPs are wavelength-independent (to first order) Propagation along PSPs is the fastest/slowest possible PMD is the phenomenon, DGD (∆τ) is the magnitude
Differential group delay (DGD):
∆
=
− fast
slow
4
broadband source
Review: Time domain PMD measurement low-coherence interferometery (INT)
∆τ
Width of delay histogram gives mean DGD Measures only mean DGD
Det.
Time (ps)
5
Poincaré sphere
S2(ω2)
Review: Frequency domain PMD measurements
∆S
S1(ω1)
(3-d representation of polarization state) Equator:linear polarization states Poles:left and right circular Elsewhere:elliptical
Measure transmitted polarization state of light - at two optical frequencies (ω1andω2) r d Sial Group ∆τ=dω)DGD(ylaDDefeifntre Polarization-based DGD definition (ω= radian frequency)
Artifact Selection Criteria: DGD approximates that of your DUT DGD can be predicted by other means Environmentally stable
Single birefringentcrystal
Polarization-maintaining fiber
PMD emulator
Calibration: Non-mode-coupled (measure a device of known DGD)
Predictable DGD∆τ=∆ngL ,L=thickness,∆ng= group birefringence c Beware of waveplate tilt, multiple reflections, and dispersion Maximum DGD limited (0.5 ps or so)
DGD predictable but less certain Beware of temperature coeff. (2-10x > quartz) Large DGD values possible
DGD predictable from geometry Beware of reflections and polarization extinction ratio Variable DGD possible
s p
9
Artifact Selection Criteria: Mean DGD approximates that of your DUT DGD(λ) looks typical Environmentally stable