A mainstream module vendor sorts through its options

Most of the developers of dispersion-tolerant modules are young companies whose business plans focus on these devices. However, mainstream module companies also have noted the benefits of dispersion tolerance and have experimented with several technologies in hopes of playing in the same game.

For example, while Opnext Inc. (www.opnext.com) isn’t very old, its lineage runs through Hitachi Corp. The company has focused on 10-Gbit/sec transmission technology using NRZ transmission formats without integrated compensation. However, Opnext recognized the desire for dispersion-tolerant modules and has investigated the various approaches, according to Ed Cornejo, director of marketing.

“We at one time had hoped that we could use electronic dispersion compensation. In fact, that was our first plan with our second-generation tunable transponder,” he says. “But we had to scrap the EDC part of it because we found some issues with the technology in terms of error floors and some instabilities that we didn’t like. We just wouldn’t take it to product at this time.”

The company then turned to duobinary technology, which it likes much better. “We thought that the optical duobinary coding scheme was a much cleaner solution from our perspective, and it seemed to work pretty well,” Cornejo explains.

As a result, Cornejo expects that Opnext will release for sampling a dispersion-tolerant 10-Gbit/sec tunable module this month that leverages duobinary technology. The company was far enough along in the module’s development that it was able to participate in the OIF interoperability demonstration at OFC/NFOEC. Cornejo reports that the subsystem demonstrated dispersion tolerance of ±3,000 psec/nm, tops in the demonstration. “We have some special techniques that we use with that technology to have such a wide tolerance,” he says.

The time has definitely arrived for such a product offering. “It’s kind of weird to say this, but we can’t hurry enough for our customers right now,” Cornejo offers. Opnext has no current plans to pair duobinary with electronic or optical dispersion compensation in the same module; Cornejo reports that he hasn’t seen a demand for such a combination.

“For a while, we’ve had a need for extending beyond 80 km,” he concludes. “Our second generation [of modules] was trying to use EDC, and it just wasn’t making the grade. That’s not to say they won’t in the future, but at least for the timeframe that we were looking at releasing product, it was failing. So the duobinary came at an appropriate time; there was a definite need fairly quickly. That’s why we jumped on it-because it worked.”