Siemon recently published a paper in which it reveals results of testing it conducted on the performance of plug-and-play optical fiber cable assemblies used in 40- and 100-Gbit networking applications. The data in the paper, titled “A Closer Look at Plug and Play MPO/MTP Assemblies,” “clearly show that lower-cost generic assemblies from unknown suppliers can compromise performance in today’s high-speed fiber-optic links,” the company said when announcing the paper’s availability.
Siemon Labs evaluated random samples of multifiber MPO/MTP assemblies acquired via standard distribution from four different assembly houses that Siemon characterized as “low-cost,” and also evaluated assemblies manufactured by Siemon. “Each MPO/MTP fiber assembly was tested to TIA and IEC standards for endface geometry, cleanliness, optical performance and mechanical reliability,” the company said. “Each assembly was also tested to Siemon’s specifications, which are more stringent to ensure superior performance and application assurance. The testing showed a significant discrepancy between Siemon’s high-quality products and the generic products, with the majority of the generic assemblies failing to meet minimum standards requirements across the range of performance-critical parameters. Siemon was the only manufacturer to pass all the parameters for all tests.”
Mike Corke, fiber application engineer for Siemon, explained, “With the Internet of Things and Big Data driving the need for more bandwidth and increased transmission speeds to 40 and 100 Gig in data center fiber links that use preterminated plug-and-play MPO/MTP fiber assemblies, overall performance of these assemblies has become a critical factor. Unfortunately, not all MPO/MTP assemblies are created equal, and data center managers need to be aware that the savings from using substandard components is not worth putting critical performance at risk.”
In a post to Siemon’s Network Infrastructure Blog, the company’s global marketing communications manager, Betsy Conroy, explained that of the low-cost assembly producers, “three of the four assembly houses had assemblies that didn’t even offer an insertion loss performance that would allow for the use of a crossconnect in a 40- or 100-Gigabit channel. And all samples from one of the assembly houses completely failed cable flex mechanical reliability testing, with the cable jacket completely pulling out of the rear of the crimp sleeve.”