As a follow up to a blog from last July that introduced the latest encircled flux (EF) method required for testing multimode fiber in 40 and 100 gigabit applications, I wanted to take minute to provide an update on TIA TSB-4979, Practical Considerations for Implementation of Encircled Flux Launch Conditions in the Field.
Published in August 2013, TSB-4979 addresses the fact that a lot of legacy test equipment remains in service in the field and describes two different methods to achieve EF compliance.
Let’s first review EF testing.
To achieve more accurate testing results, the EF method uses launch controllers and conditions that launch the right amount of light into the core of the fiber that more closely matches the laser launch condition of the actual transceiver.
As defined in TIA-526-14-B, the launch power is specified as a function of the core radius. For example, at a distance of 22 ?m from the center of the core of a multimode fiber, the cumulative light power must be within the range of 96.45% to 97.67%.
A value less than 96.45% means more power than specified in the higher order modes away from the center of the core, causing an overfilled launch condition that can lead to pessimistic readings and false failures. A value greater than 97.67% means more power in the lower order modes near the center of the core, causing an underfilled launch condition that can show artificially low loss values and lead to unforeseen network problems.
Using an EF-compliant light source reduces measurement uncertainty from greater than 40% to less than 10 %. We have actually witnessed as much as 0.5 dB variability when testing a 2-connector channel with non-EF-compliant testers. EF-compliant sources reduce the variability to 0.1 dB, which is much more ideal when certifying links with a tight loss budget of 1.9 dB or less.
The problem with specifying EF compliance for testing fiber links in the field is the amount of remaining legacy equipment and the cost to upgrade. That’s where TSB-4979 comes in.
The first method described in TSB-4979 uses a universal controller that works with legacy sources. This device is comprised of a “black box” with a fixed input cord and output cord. The “black box” works with any light source and redistributes the modes to provide an EF-compliant launch at the output.
One drawback is the cost and size of the universal controller, as well as the limited number of connector matings. Re-terminations of the output cord may be possible if a connector becomes damaged, but the universal controller would then need to be recertified for EF-compliance by a capable supplier.
The second methods uses a simpler matched controller that works with specific equipment. The source is matched with a launch cord, which may or may not have a “black box” controlling device. Interoperability is achieved by matching the source model number to the launch cord model number.
An advantage of Method 2 is that a launch cord may be replaced without requiring EF verification.
TSB-4979 also provides a final word of advice for extending the life of launch cords. Launch cords are subjected to stress and will not remain EF compliant indefinitely. It is important to inspect, clean and re-inspect before every connector insertion. Doing so will extend the life of equipment and launch cords and reduce the incidence of bad test results.
When it comes to pre-terminated fiber solutions, choosing manufacturers like Belden that use EF-compliant factory testing will help ensure that you get the insertion loss you think you’re getting. Click here to find out more about Belden Pre-terminated FiberExpress Solutions that are all subjected to EF testing in a factory environment.
Looking back at his 42-year career in the cabling industry, Paul Kish was one of the founding fathers of the industry. Retiring from Belden in 2015, Paul was recognized as an expert in cable transmission. He served as a role model, an innovator and a thought leader. Paul was a key contributor to the development of cabling standards with TIA, ISO and IEEE, and also served on the BICSI Technical Information & Methods Committee.