WEBINAR: Testing MTP®/MPO Assemblies: Setups, Troubleshooting, and What to Expect

MTP®/MPO cable assemblies are prevalent in fiber networks and their quantities are growing as parallel communication increases.  These assemblies introduce unique testing challenges, whether it is a standard 12f MPO to 12f MPO assembly or one of the various different fanouts and fiber count configurations.  In this presentation, learn more about equipment identification and setup, as well as proper testing during production. We also explore common pitfalls and ways to identify errors and inaccuracies during the testing process. US Conec will review terminated MPO connector standards required for test leads including polarity methodologies and how configurable connector types can be used to make the test setups more versatile.

OptoTest Logo
us conec logo optimized


Is the OTDR technique useful for multimode fiber as well?

The OTDR-style test method for return loss used by the majority of RL meters in the OptoTest product line is equally effective for single mode and multimode fiber.

How can you test a jumper with an MPO connector on one end and single fiber connectors on the other end?

Please refer to this timestamp (37:50) to skip to the section in the webinar that discusses fanout cable testing and visit the resource section for more examples.

I thought bend-insensitive fiber would be most preferred as it would withstand the significant bends that might be encountered in rack-to-rack cabling. Could you clarify on why a non-bend-insensitive fiber would be preferred in multimode?

It is important to clarify that non-BIMMF (bend-insensitive multimode fiber) is the recommended fiber type for the launch cords of the IL and RL test setup. Non-BIMMF fiber better maintains an EF compliant launch from the instrument and through connections.  If the instrument is EF compliant, connecting a patch cord with reference grade fiber and reference grade connectors will better maintain the encircled flux compliant launch.  Bend insensitive fiber in the equipment and launch cords can cause the launch to fall out of EF compliance more easily.

Is the OP415 connectorized internally? What about contamination of panel connectors?

The internal connectors of the OP415 are designed to minimize the risk of contamination and damage while also being accessible for cleaning and replacing as needed. Visit the OP415 product page or contact our team for more information.

With the potential issues that dirt can cause, what are your thoughts on the importance or necessity of maintaining a "clean room environment" for MTP/MPO production and testing?

Working in a clean room environment can only positively benefit the production and testing process, but it is not necessary or practical in most cases. Diligent cleaning of connectors and adapters is sufficient to address most cleanliness concerns.

How to account for the fiber loss in a long assembly when performing unidirectional testing?

There are a number of ways to factor attenuation into your test results depending on the equipment and software combination used. Unidirectional testing is usually implemented to test the entire fiber link across the DUT.  Typically, you do not want to subtract out the fiber loss for this case.  If it is necessary to factor out fiber loss, then you just need to know the length of the DUT and subtract out the loss of the fiber as calculated by the attenuation and length. Contact the OptoTest team to learn more.

Can your instruments only be used in the lab?

Our test equipment is used in a wide range of industries and applications extending far beyond the production environment. Click here to learn from our customers about their experiences and contact our team to discuss your needs and applications.

Are the terms reference cable and launch cable used interchangeably?

Yes, these terms are often used interchangeably.

Please explain how the type A, B and C polarities affect the referencing setup? Do they require different types of reference cables?

This scenario is an excellent example of the benefit of separating the polarity verification and IL/RL testing processes when testing by using a dedicated polarity analyzer like the OP415. The OP415 can verify whether the patch cord is type A, B, or C. Then simply use an integrating sphere or large area detector as the optical power meter for the IL/RL testing.

What is the shortest length of a jumper that you can test for RL?

OptoTest instruments can measure DUTs as short as 1.7m.

What is the best way to measure a DUT which shorter than 1.7m?

The best method is to measure both sides via the one-side method with some type of mandrel or matching block on the back side of the DUT. Note that the resulting measurement will always be a worst case. The return loss value will either be a slight combination of the two reflections (additive) or only the reflection of interest.  In the worst case scenario, you will obtain false fail results rather than false passes.

Is it possible to obtain reliable RL measurements for each of the connectors in a DUT which is MTP to 12 LCs with a length of 0.35m?

It is possible to get the combined RL for both connectors.  You could try to get the RL of each connector by mandrel wrapping the DUT AND terminating the end of the DUT with a matching block.  Doing both at the same time will eliminated the back-end reflection completely.

How can you evaluate the combined RL values from both sides of a short cable (0.35m) and what is considered a passing value for the combined RL result?

If the pass/fail for RL is the same on both sides, like 50 dB and 50 dB, then the combined pass fail is 47 dB.  The effective reflection is twice as large (3 dB).  A proper method to evaluate the pass/fail for combined return loss is to convert both values to a linear scale, add the two values, and then convert them back to dB.

When looking at IL and RL, what determines the limit for what the shortest measurable DUTs are?

OptoTest ILRL meters measure IL in the standard light source power meter method (CW). RL measured by the pulse-based method has a DUT length resolution of approximately 1.7m.

Can you please speak about negative IL values when testing MM assemblies (directly after referencing) and how to avoid them?

Negative values usually point to something incorrect in the IL referencing process or it can come from exchanging adapters between reference and measurement.

For example, when testing IL on a fanout, you need to reference using the MTP adapter and measure using the simplex adapter. Adapters can have different reflectivity. The difference in the reflectivity could potentially lead to negative IL values if the DUTs being measured are low loss. If your are using an OptoTest system, contact our team if you think your adapters could be causing the negative IL values.   Also, with some short BIMMF DUTs, it has been shown that they have such low IL values due to their ability to maintain and carry lossy modes in the core/cladding interface and due to the low loss, small variations in source drift or OPM inconsistencies could lead to the negative IL values.

Can you go over how to create the Encircled Flux launch condition?

This is different for each source. Sources that overfill a fiber can typically be mandrel wrapped to create a slightly restricted launch that is EF compliant.  For sources that are underfilled, it is difficult to create an EF compliant launch and typically requires a mode scrambler followed by a mode filter (mandrel wrap).  However, there is no definitive method for creating an EF compliant launch and it is best to contact your equipment vendor to supply a modal conditioner or an EF compliant system.

Are "core dips" a standard feature in fiber connectors? Is it a best practice from manufacturers or rather a lack of performance?

Core dips are a result of the polishing process and typically only affect MM fibers.  In the polishing process there are polishing films that remove material from the core faster than from the cladding which creates the core dips.  There are polishing methods and processes that reduce core dips on multimode fibers.

If the DUT is made with BIMMF fiber, how can we test (launch cable, maybe receive cable, etc.) to avoid Near Zero Loss?

It is recommended to use a “tail cord” in the setup.  This tail cord removes the lossy modes that should not propagate in the DUT, but due to the DUT being BIMMF they will propagate all the way through the DUT.  The tail cord can be constructed with either BIMMF or non-BIMMF.

Are the NMC (non-contact) Single Mode connectors a good solution to reduce connector damage?

These would be good for frequently used connections, but for reference connectors they are not advisable.  The reference connector is intended to mimic installed connections.  A non-contact connector does not accurately represent a mating in the field.

Have you ever measured MM fiber with 1310 nm or 1550 nm? Is it a laser source or LED source?

This is not a typical requirement.  Typically, with any multimode IL application we suggest using an LED source. However, there are special circumstances where a restricted launch might be necessary and a laser could be used.

How many mating cycles do these connectors usually support? And if they fail after this many mating cycles, what are typical fail modes? What are ways to extend number allowable mating cycles to 500-1000?

MTP® PRO connectors have been qualified to TIA 568.3-D which require a mated pair to withstand 500 durability cycles. Guide pin hole and end face debris are a very common culprit to deteriorating the optical performance of an MPO connector. Properly cleaning the end face with an optical connector cleaner is the best practice to improve the life of an MPO connector.

For the MTP Pro connector, is there a known point of failure for the interchangeable properties (e.g. gender swap, polarity change)?

Mechanically, the MTP® PRO connector can swap gender and change polarity hundreds of times without degradation to the interchangeable properties.

What kind of MTP ferrules do you recommend using for reference cables? Can the Elite performance be used, or you suggest using a maybe even better ferrule which can be much more expensive? For how many measurements do you think it can work?

Our belief is that testing in the factory is a spot check to ensure the population meets the random intermate requirements versus a guardbanded approach. US Conec specifications are based on random mate, so we use ferrules from standard production for reference jumpers. If a guardband methodology is used for reference testing in factories, an MT Elite® ferrule is certainly ample but we do have a Super MT Elite® ferrule available.

For MPO IL tests, do you need to have a master reference connector that has "excellent" geometric properties on each fiber?

We recommend using MT ferrules that meet IEC 61755-3-31 and IEC 61300-3-35 standards for a master reference connector.  It is the combination of fiber tip contact, fiber tip geometry, and fiber alignment that contribute to the optical performance.  The quality of MT end face geometry, visual cleanliness, and ferrule should all be considered.


IEC-61754-7 and TIA-604-5-F (FOCIS 5) 

These standards govern the intermateability of MTP/MPO connectors. Two MTP/MPO connectors should be intermateable so long as they dimensionally meet either of these standards. 

Under the IEC standards, the one-row connector variant is governed under IEC-61754-7-1 and the two-row variant is governed under IEC-61754-7-2. TIA standards cover all versions under FOCIS 5. 


This standard covers additional details of the MTP/MPO connector such as polarity. This standard specifies the opposing key adapter as a Type A adapter and the aligned key adapter as a Type B adapter. 

TIA TSB-5069 

This recently released standard (March 2020) provides general guidance on maintaining polarity of cables and other components within telecommunications systems.