NEMI Works to Improve Fiber Splicing

Paula M. Powell

A little-appreciated economic tenet of any market downturn is that vital resources become available for critical projects that previously sat on the back burner. By redeploying these resources to precompetitive research areas, it may be possible to boost manufacturing competitiveness when the market rebounds. Such is the thinking that is fueling a fiber optic splice improvement project being conducted by member companies of the National Electronics Manufacturing Initiative (NEMI) in Herndon, Va.

Participants in the project range from manufacturers of fusion splicers and optical fibers to electronic manufacturing service providers, all of whom are looking to capture a larger share of the optical component manufacturing market when it rebounds. Their goal is to develop industrywide splice quality criteria and to test methods that will allow the systematic investigation of variability and comparison of equipment and procedures. Ideally, these would lead to improved yield and lower costs.

One important result may be the development of a functional draft standard that could be submitted to the standards-making bodies -- something sorely needed, according to Eric Mies of Vytran Inc. in Morganville, N.J., which makes products for fiber splicing. He noted that, although a variety of current standards apply to fiber splicing, such as Telcordia, they are designed primarily for field applications. Manufacturing support is limited.

Peter Arrowsmith, chairman of the project, agreed, adding, "There currently are no widely agreed-upon acceptance criteria for fiber splices, and specifications often differ significantly between OEMs, electronic manufacturing services providers, suppliers and customers." The advisory scientist for Celestica Inc., an electronic manufacturing services provider headquartered in Toronto, sees the splicer insertion loss estimate in particular as being a critical area to target for improvement.

"Different splicer vendors use different techniques for estimating insertion loss, and the accuracy of these estimates is often insufficient to know whether the splice is acceptable, particularly when really low losses -- less than 0.05 dB -- are required."

According to Arrowsmith, specific project activities will include:

Developing and validating test methods for insertion loss, strength and extinction ratio.
Making test splices using different equipment.
Defining splice acceptance criteria more precisely based on real splice data through the use of an established gauge repeatability-and-reproducibility approach.
Identifying major causes of splice variability.
Developing fiber-handling requirements for preparation, splicing and assembly.
Developing methods for testing to predict splice reliability.
Investigating the performance of different splicing techniques and insertion loss estimators.
Developing a cost model for splicing as well as cost-of-ownership estimations to compare manual and automated splicing techniques.

Besides Celestica and Vytran, project participants include 3SAE Technologies of Woburn, Mass. (fiber preparation equipment); Aurora Instruments Inc. of Philadelphia (fusion splicers, fiber alignment systems); Jabil Circuit of St. Petersburg, Fla., Sanmina-SCI of San Jose, Calif., and Solectron Corp. of Milpitas, Calif. (contract electronic manufacturing services); and Sumitomo Electric Lightwave Corp. of Research Triangle Park, N.C. (fusion splicers and cabling products).

With all this on their plate, the companies expect to see preliminary results sometime next year. They also are amenable to the possibility of adding team members.

For more details, contact the National Electronics Manufacturing Initiative project coordinator David Godlewski at mailto:dgodlewski@nemi.org