Académique Documents
Professionnel Documents
Culture Documents
For The Certification of Infrastructure for
10GbE, 40GbE und 100GbE
MPO/MTP
Testing for 10 – 100Gb/s Readiness
Mike Kazemian
Synergix International
Mike_kazemian@synergix.ae
Agenda
• Standards: Application, Cabling, Components, …
• Field Testing
– Test Cords and Methods
– EF … Encircled Flux
– Test Regimes
• MTP/MPO Testing In The Field
– ….for 40/100Gb/s Applications
– … for 10Gb/s Applications
2
Choosing the right limit
….. It’s often not simple
Component Requirements
Standards Based Limits
Type λ (nm) Loss‐Limit Units / Comment
Applications IEEE 802.3: 10GBASE‐SR/SW 850 2.6 dB requires OF‐300
IEEE 802.3: 10GBASE‐LR/LW 1310 6.2 dB requires OF‐2000
….the transmission ISO/IEC 8802‐3: 1000BASE‐SX 850 3.56 dB requires OF‐500
path has to comply 1300 2.35 dB requires OF‐500
with the application ISO/IEC 8802‐3: 1000BASE‐LX
1310 4.56 dB requires OF‐2000
or channel
requirements ISO/IEC 8802‐3: 100BASE‐FX 1300 11 dB requires OF‐2000
ISO/IEC 14165‐111: Fibre Channel (FC‐PH) 850 4.0 dB requires OF‐500
List is not complete! @1062 Mbit/sec, 1310 6.0 dB requires OF‐2000
Channels OF‐300 / OF‐500 / OF‐2000 850 2.55 / 3.25 / 8.5
… consists of 1 or Multi M. (max.300m/500m/2000m) 1300 3.25 / 2.25 / 4.5
more links + equip. dB
OF‐300 / OF‐500 / OF‐2000 1310 1.8 / 2.0 / 3.5
/patch cords Single M. (max. 300m/500m/2000m) 1500 1.8 / 2.0 / 3.5
Links The loss limit is dependent on λ, length and connector/splice count. (min. 3 Components)
Component 850 3.5
OM1/ ‐2/ ‐3/
Fiber 1300 1.5 dB / km
OS1 / OS2 both 1.0 / 0.4
Splice all all 0.3 dB / Splice
Connector all both 0.75/0.5/0.35 100% / 95% / 50%
Requirement for TRC (Test Reference Cord) 850/1300 0.3
dB / Connection
Field Testing & Random Connector @ Link Ends 1310/1500 0.5
TRC & TRC 850/1300 0.1
ISO/IEC only dB / Connection
(periodic Check) 1310/1500 0.2
Standards Overview
Region Channel / Link / Component ‐ Requirements Field Testing Requirements
International • ISO/IEC 11801 2002 AMD2 (2009) ISO/IEC 14763‐3 (June 2006)
Information technology – Generic cabling for Information technology – Implementation
customer premises and operation of customer premises
• ISI/IEC 24702 cabling – Part 3: Testing of optical fiber
…‐ Generic Cabling for Industrial premises cabling.
• ISO/IEC 24764
…‐ Generic Cabling for Data center premises
• ISO/IEC 15018
…‐ Generic cabling for homes
Europe EN 50173‐x ”Family” Amd AB:2010 EN 50346:2002 A2:2009
Information technology ‐ Generic cabling systems Information technology ‐ Cabling
• 50173‐1…Part 1: General requirements installation – Testing of installed cabling.
• 50173‐2...Part 2: Office premises (Refers to 14763‐3 for all technical details)
• 50173‐3…Part 3: Industrial premises
• 50173‐4…Part 4: Homes
• 50173‐5…Part 5: Data Centers
USA ANSI/TIA‐568‐C.0‐2009 ANSI/TIA‐455‐78‐B‐2002, Optical Fibers Part
Generic Telecommunications Cabling for 1‐40: Measurement Methods and Test
Customer Premises Procedures – Attenuation
C.1 Commercial Building …. ANSI/TIA/EIA‐455‐8‐2000, Measurement of
C.2 Balanced Twisted‐Pair …Components Standards Splice or Connector Loss and Reflectance
C.3 Optical Fiber Cabling … Components Standard Using an OTDR
40/100 GbE Fiber Applications
40GBASE‐SR4 40GBASE‐LR4 40GBASE‐FR
Mulitmode Single Mode
Source 4 x 10Gb VCSEL 4 x 10Gb Laser 1 x 40Gb
100GBASE‐SR10 100GBASE‐LR4
Laser 100GBASE‐ER4
Optics Parallel
Mulitmode Simplex Single Mode
Connnector
Source MPO
10 x 10Gb VCSEL LC (or similar)
4 x 10Gb Laser 1 x 40Gb Laser
Wavelenght
Optics 850 nm Parallel 1271‐1331 nm Simplex (4λ
1550 @ 25Gb)
FiberConnnector OM3 OM4MPO OS2 OS2similar)
LC (or
Max Distance
Wavelenght 100 m 150 m
850 nm 10,000 m1300 nm 2,000m 1550 nm
Loss Budget 1,9 dB 1,53 dB 6.7 dB (DWDM) 4.0 dB (DWDM)
Fiber OM3 OM4 OS2 OS2
Max Distance 100 m 150 m 10,000 m 40,000m
Loss Budget 1,9 dB 1,53 dB 8.3 dB 18 dB
40/100GBase‐SR/LR/ER:
Which ISO Channel supports it ?
Low Loss MPO …The Drivers
• “Zoned Data Centers”
Courtesy:
TE Connectivity
Fiber Optic Test Methods
Standard Methods
TIA‐568‐C
Tier‐1 Tier‐2
ISO 11801 AMD.1 / ISO/IEC 14763‐3
BASIC Test Regime EXTENDED Test Regime
LSPM: Light Source & Power Meter OTDR: Optical Time Domain Reflectometer
• The two methods are complimentary !
– OTDR based method does not replace the LSPM based solution
• Both methods have advantages and limitations
LSPM versus OLTS
LSPM … Light Source & Power Meter OLTS … Optical Loss Test System
• Measures the insertion loss of a fiber • Measures a duplex fiber in both directions
path 1. Measures the length
• Length Dependent Limits and Margins 2. Calculates the length dependent loss limit
need to be calculated manually 3. Measures the insertion loss
4. Calculates the margin
• Optional 2nd step: Measure both fibers in
the duplex link in both directions
Sources of Attenuation: Coupling Losses
Losses at splices and connectors, including test equipment
connectors:
• Area mismatch
• Spacing loss
• Axis misalignment
• Angular misalignment
NOTE: Coupling loss is the dominating contributor to the the
overall link loss in premise and Data Center cabling
Reference Methods
1 Jumper #1
Method B
2 Jumper #1 #2
Method A
3 Jumper #1 #3 #2
Method C
#1
Link
Under Test #2
• Test Reference Cords are not mandatory by TIA
• Most manufacturers demand the most accurate 1 jumper method
• The one jumper method always require a “non critical” connection on the meter
• Other methods need to be used if the meter does have an ICA or if
the link different end‐faces on either end
– There is a benefit to use a “Modified Method B” overcome the lack of an ICA.
Accuracy of the three reference methods
1 Jumper #1
Methode B
2 Jumper #1 #2
Methode A
3 Jumper #1 #3 #2
Methode C
Cords /Ref.
1 Jumper 2 Jumper 3 Jumper
Method
Reg. Patch
n
Cords
Hi risk of negative loss results !
Test
Reference
Cords
Encircled Flux ‐Reducing multimode uncertainty
• Encircled Flux (EF) is a way to define launch conditions…the
final piece in the puzzle to reducing measurement uncertainty
in the field.
• This is important because uncertainty is the enemy of ever
tightening loss budgets!
Optical Source launch conditions impact the consistency and
repeatability of Loss Measurements!!
Source 1 Source 2
Over filled Restricted or Under filled
Why the NEW EF STANDARD is NEEDED
• Different light sources may have different launch conditions
• Previous standards for launch conditions, in the past CPR
(Coupling Power Ratio) and since 2006 MPD (Modal Power
Distribution) (like MFM2) allowed for too much variation
How is the launch condition verified ?
CPR (Coupling Power Ratio)
Pre 2006
MPD (Mode Power
The power is compare with and without a 9μm cord
with 30mm mode filer inserted Distribution) Since 2006
EF (Encircled Flux) New
Principle:
• A filter makes the light “visible” for the camera
• The image is saved as a TIFF file
• SW is used to evaluate the distribution of light
within the fiber
MPD vs. EF Relationship
Two EF compliant
light sources
+ #1 @ Under Filled Limit
+ #2 @ Over Filled limit
… may cause a
variation of 10%
Two MPD compliant
light sources
+ #1 @ Under Filled Limit
Area of Interest + #2 @ Over Filled Limit
Subject to
Fiber Miss‐Alignment
may cause a
variation of 40%
How to make
The Field Test EF compliant
• Use an external mode conditioner instead of
mandrels
Encircle Flux
Common Questions From Customers
Q&A: What is the variability of DTX‐MFM2 &
Mandrels ?
“The DTX‐MFM2 is MPD compliant which allows for a variation
of 40% in test results. Does that also apply for the DTX‐MFM2 &
Mandrel”
1. The DTX‐MFM2 uses tuned light sources and 850nm/1300nm combiners
2. The same instrument used to verify EF compliance is used to ensure a very
narrow band of variation in modal distribution
3. The variation shown is less then between two compliant but marginal EF
compliant light sources
Q&A: Can I replace the plug on an EF controller ?
“ Can I replace the plug at the end on an EF controller and how
many test can I perform ”
• When that LC connector breaks or
wears out, it cannot be re‐terminated
in the field
• Re‐terminations need to be
verified for EF compliance
• No minimum number of mating can be quote and
wear tear will greatly depend on treatment and
handling
– Abrasive dust can destroy a brand new plug
Q&A: Can a EF Controller Be Integrate Into The Light Source
?
“ Why is the function of the EF controller not integrated into the light
source? ”
Sacrificial cord
1. For the same reason the use of “Sacrificial Cords” dose not work EF is
not assured at the end of the test cord
2. What would be needed is an “EF Transparent Cord” for this to work
Q&A: Could / Should one test wit a VCSEL?
• Some believed the source should be a VCSEL –
– Some standards used to allow them…
– Some out of date project specifications require them…
• Reasoning
– Use the same source type as the active equipment
• Problem
– The optical loss limits in IEEE 802.3 are based on test equipment using
LEDs, same for ANSI/TIA and ISO/IEC
– VCSELs are under filled – results in optimistic “best case” readings
– Launch condition varies greatly from source to source
Q&A: What if I Use a DTX‐MFM2 & Mandrel Instead Of EF
Controller ?
Retest
1)
() Retest
Retest
( ) 2
• Note 1): Within the uncertainty of a EF based measurement
• Note 2): Only if the FAIL was with a slim margin
Wish LIST FOR EF COMPLIANT FIELD TESTING
? ?
1. A Light Source with an EF compliant output
2. An “EF Transparent TRC”
• ... at an affordable price
• … with TRC performance
• … without a clumsy and heavy package
40/100 Gb/s
IEEE 802.3ba
MPO - Future Proof Technology …
Migration 10 – 100Mbps
Reuse the same
MPO Trunk Cable
1G or 10G Cabling
40G Cabling
100G Cabling 28
Polarity ‐ MPO/MTP System Type A
29 29
Polarity ‐ MPO/MTP System Type B
30 30
Polarity ‐ MPO/MTP System Type A, B & C
31 31
Polarity ‐ MPO/MTP System “Other”
32 32
How were MPO fibers being tested until today?
Set Reference
Install Fan‐out
Complex Test 12 times
1. Testing of all fibers in the link in one go
2. Non critical MPO connector on the meter to
support the most accurate one jumper
method
3. Switchable MPO source to verify polarity
– System A, B, C, ….
– Field terminated connectors
4. MPO Test Reference Cords with reference
grade connectors
– A definition for reference grade MPO connectors !
5. EF Compliant Light Source
Trunk Cable Installation: Loss and Polarity Measurement –
Setting the Reference (1 jumper if possible)
Trunk Cable Installation: Loss and Polarity
Measurement
MPO/MTP Power Meter
MPO optical interface for 12
fibers
Indicate if loss limit is
exceeded.
Power or loss on 12 fibers
simultaneously
Report polarity per
TIA‐568‐C.0
Individual power or loss
on single fiber
Save power or loss
measurements to LinkWare
37
MPO/MTP Light Source
MPO optical interface for 12
fibers
Stabilized optical signal
on each fiber Scan All automatically
sources all fibers
Scan All enables polarity
per TIA‐568‐C.0 Auto encoding
User selectable individual
channels
38
10 …100 Gb/s
Trouble Shooting
Trouble Shooting
• Phase 1: The test with a Light Source &
Power Meter identifies marginal/failing
links and channels
– MPO: The problematic fiber is
identified
• Schritt2: An OTDR identifies the
problematic connection
– A “Fan‐Out” cable adapts SC to MPO
Example: Localizing the “Bottle Neck”
• The total loss of 1.98dB is within the limit
• The bottle neck is after 260m
Example : Correcting The Problem & Checking The
Result
• The problem was solved by cleaning or swapping components
• The bottle neck after 260m was eliminated
Thank you for your attention!
Questions?
43