Post on 18-Mar-2018
transcript
Fiber Certification: The Impact of New Standard forThe Impact of New Standard for Multimode Fiber Measurements
Dr. Andre Girard and Dr. Gang HeEXFO Electro-Optical EngineeringEXFO Electro Optical Engineering
OUTLINE
Introduction- Introduction
- Experiments performed on various Linksp p
- Experimental results
- Conclusion
Measurements on MMF LinksMeasurements on MMF Links
Multimode Fiber Link
Multimode Connectors
andPatchcords
(CW) Light Source (LS)or
O
Powermeter (PM)
OTDR
How Many Acronyms Do We Need to Remember?
• CPR: Coupled Power Ratio• EF: Encircled Flux• FF: Far Field• HOMP: High Order Mode Power• MMF: Multi-Mode Fiber• MPD: Mode Power Distribution• NF: Near Field• OFL: Over-Filled Launch• RP: Relative Power (%)RP: Relative Power (%)• SMF: Single-Mode Fiber
9 Types of MLCs (!?#$%)1. OFL + 1 km of MMF
2. 70 / 70 launch
3. OFL + Mandrel
4 CPR4. CPR
5. HOMP
6 NF / FF with diameter6. NF / FF with diameter requirements for RP at – 5 % or– 75 %, 15 % and 5 %
7. MPD Template
8. EF Template = target with upper and lower limits
Review of Current MLCs:Review of Current MLCs: (1) 1-km MMF after OFL
(2) 70/70 launch
f ¨ ¨
(2) 70/70 launch(3) Mandrel after OFL
• Define ¨How to do it¨; but…– Better define what you get than how to do it!
• Practically under-fill launch
Methods (1) and (2) not practical in the field• Methods (1) and (2) = not practical in the field
• OFL definition = “fuzzy” MLC = “fuzzy”
Review of Current MLCs:(4) Coupled Power Ratio (CPR)
F MMF t SMF• From MMF to SMF
• Defined at source output
• Simple/easy-to-do NF power distribution measurement
W k l if ll ll d NF PD– Works only if well controlled NF PD
• But with little information on high-order modes (HOM)g ( )– Loss = dependent on HOM power distribution
Review of Current MLCs:(5) High-Order Mode Power (HOMP)
• Power difference (in dB) with and without HOM filterPower difference (in dB) with and without HOM filter acting as a mandrel
• Works with OFL = but causes bad measurement• Works with OFL = but causes bad measurement reproducibility, because:
Size of fiber core– Size of fiber core– Macro-bending dependence – Too many transient losses
• Not widely used in the industryNot widely used in the industry
Review of Current MLCs:(6) N d F Fi ld (NF / FF)(6) Near and Far Fields (NF / FF)
NF Power Distribution 850nm
FF Power Distribution850nm 50μm
• Good control• Need both NF and FF test equipments• Need both NF and FF test equipments• Used internally as reference metrics
Review of Current MLCs: (7) MPD T l t d (8) EF T l t(7) MPD Template and (8) EF TemplateMPD Template EF Template
e po
wer
5 um, .0688 - .1061EF RADIAL BOUNDS
20um 8948 924715 um, .5838 - .697710 um, .2690 - .3820
F
Rel
ativ
e
OFLTARGETTEMPLATE
22 um, .9645 - .976720 um, .8948 - .9247
mal
ized
EF
Normalized mode group number Nor
mBoth obtained from NF power distribution:
MPD bt i d f d i tiRadius
• MPD obtained from power derivation• EF obtained from power integration
Advantages of EF Template• Not sensitive to noise fluctuations• Reproducible and repeatable measurements• With EF template, slight
under-filled MLCs: goodgcorrelation between loss and connector lateral
ffoff-set• In a multi-connector link
using EF target MLCs: the loss of a connector does not depend on itsdoes not depend on its location
OTDR versus LSPM
• Many studies done with LSPM
• Not much done with OTDR
• For instance: compared to LSPM does the OTDR:• For instance: compared to LSPM, does the OTDR: – Measure the same loss?– Have the same limits for obtaining 10% loss
accuracy? – Give the same ORL?
We will now answer these questionsWe will now answer these questions
Loss Issues with MMFLoss Issues with MMF
3 main loss contributors + an effect in a MMF li k b d b OTDRlink probed by an OTDR:
1 Connector loss due to lateral offset1. Connector loss due to lateral offset
2. Long MMF
3. Macro-bending
4. The OTDR pulses traverse the Link twice with additional mode scrambling effectswith additional mode scrambling effects from backscattering
Experiments Performed With MMF LinksE i l S Thi i i !Experimental Set-up - This is it!
Multimode Fiber Li k
Variable Mode LinkMode
Conditioner
Launching diti
Speckle Scrambler Launching
Cable conditions
OTDR Light Source =
(50m)
850nm / 1300nm Powermeter
850nm / 1300nm Fabry-Perot Laser
Experiments Performed With MMF LinksLi k T T d
Variable M d
Link Types Tested50/125µm fiber
Mode Conditioner
Speckle Scrambler Launching 2km 1km
OTDR Light Source =850nm / 1300nm
ScramblerCable (50m) Splices Long Link
850 / 300Fabry-Perot Laser PM
20m 50m 50mShort Link
18mm-10 turns Mandrel
M b di Li k25m 25m
Macro-bending Link
Experiments using EF Launch Conditions p gLoss Measurements at
850 nm 1300 nm1
0.80.9
1
UFL
EF LBEF TargetEF HB
0.50.60.7
aliz
ed E
F UFL EF HBMLC0MLC1
C2
0.20.30.4
Nor
ma
OFLMLC2MLC3MLC4
00.10.2
5 10 15 20 25
MLC5MLC6
5 10 15 20 25
Radial (um)
Long LinkExperimental results Loss vs. MLCs
850 nm 1300 nm
Short Link
OTDR(Total)LSPM(Total)LSPM(Total)
Macro-bending Link
17
Macro-bending Link
Experimental results ORL MLC55
L Li k ORL vs. MLCs
45
50
)
Long Link
Short Link
35
40
OR
L(dB
25
30
35
250 1 2 3 4 5 6 7ORL(Total)_850nm
ORL(C1)_850nmORL(C2)_850nmORL(C3) 850nm
ORL(C1)_1300nm
ORL(C3)_850nmORL(Total)_850nmORL(C1)_1300nmO (C )ORL(C2)_1300nm
ORL(Total)_1300n
ORL(C2)_1300ORL(C3)_1300nmORL(Total)_1300nm
OTDR Experiment results vs. MLCsConnectors and Macro-bending Losses
New Proposed MC ArtifactMandrel
2 5
S1 18mm10 turns 50/125 fiber
25m 25mC1 C2
Mandrel50m
2
2.5
ectio
n
C1+S1C2C2+M d
1
1.5
ding
/con
neos
s (d
B)
C2+MandTotalC1+S1+C2
0.5
1
Mac
robe
nd Lo C1+S1+C2Mand
00 1 2 3 4 5 6
Mode Launching Condition(MLC)
Conclusions
Experiments were performed =
Compare losses of 50/125 MMF at 850nm and 1300nm by OTDR and LSPM for 3 types of links
1. A Link with a few km length (Long link)
2 A li k ith h t l th (100 ) d2. A link with a short length (100 m) and many connectors (6) and splices (3)
3. A short link with macro-bending
4. All with various launching conditionsg
Conclusions
With Encircled Flux Template, it was shown that:
• Loss varies less in Long Links than in Short Links
• In Long Links, loss does not vary much at 850nm
In general loss varies less at 850 nm than at• In general, loss varies less at 850 nm than at 1300nm
• The loss measured by OTDR or LSPM exhibits good agreement
ConclusionsConclusions
With Encircled Flux Template, it was shown that:
• OTDR and LSPM loss = same behavior for any 3 Link types3 Link types
• 70/70 MLCs do not work with OTDR and LSPM i ll i Sh t Li k ” ithLSPM, especially in Short Links” with multiple connectors and macro-bending
• ORL is insensitive to MLCs
ConclusionsAnd Finally…
• At last, the Encircled Flux Template provides At last, the Encircled Flux Template provides MPD maintaining launch
• Reconciles MMF link• Reconciles MMF link loss measurements using either LSPM and OTDR
• A new physical MLCs artifact is proposed andartifact is proposed and recommended for OTDR– VerificationVerification– Qualification, and – Calibration
Thank you very much for your attention; andThank you very much for your attention; and…