The Fiber Connection

Even though many had been selling an advanced OM3 fiber equivalent since about 2005, the formal approval of OM4 fiber, by the TIA in August 2009, created some publicity surrounding this latest optical transmission media. We began to hear all about how this new fiber would meet the demands for higher bandwidth in data centers, offering an effective modal bandwidth (EMB) of more than double the IEEE requirement for 10 gigabit/sec over a 300-meter distance. This was followed by more talk of OM4’s ability to better support future 40 and 100Gb/s applications. But lately the buzz surrounding this new fiber type seems to have subsided. Is it because fiber with this bandwidth has technically been around for nearly six years, or is it because most users will be just fine with the less-expensive OM3 for 10, 40, and 100Gig applications?

Just like OM3, OM4 is a 50-micron, laser-optimized multimode fiber for use with VCSELs. The key difference is in the refractive index of the fiber, which is more precisely graded to better equalize the arrival time of the light modes traveling at various speeds along the core of the fiber. This translates to higher bandwidth and a 550-meter reach for 10Gb/s (with some vendors claiming a 600-meter reach) and a 150-meter reach for 40/100 Gig, compared to 300 and 100 meters respectively for OM3. This also translates to a higher price tag. But do you really need it?

Let’s consider that about 70% of premise fiber links are less than or equal to 100 meters, and virtually all premise backbones today run transmission speeds of 10 gigabits per second or less. With its ability to support 10Gb/s to 300 meters, OM3 is more than able to do the job. In the data center, the links are even shorter, with 85% or more data center links at less than 100 meters. Again, OM3 more than fits the bill, even for future data center speeds of 40 and 100Gig. In short, OM3 continues to be the most cost-effective option for fiber trunk cables and even fiber patchcords. But that’s not to say that OM4 doesn’t have its place.

Given the premium for single mode transceivers, OM4 fiber is the best option for the small percentage of users needing to run 10Gb/s over links between 300 and 550 meters (or the even smaller percent who anticipate running 40 or 100Gb/s between 100 and 150 meters). Some of these users may have already installed fiber cables with an advanced OM3 equivalent. That, combined with an extremely limited need, perhaps explains why OM4 isn’t getting a lot of hype. In any case, it’s still a step in the right direction—at least until single mode transceiver prices come down considerably.

There’s no doubt that Power over Ethernet (PoE) is a very useful technology that offers tremendous benefits to designers, installers and end users. The ability to power devices and equipment over twisted-pair copper cabling eliminates the cost associated with delivering AC power to every device—especially in locations where providing access to AC power is complicated. That’s precisely why we’re seeing wireless access points, phones, clocks, surveillance cameras and other devices powered by DC power over the same copper cabling that transmits data packets across the network. Now the IEEE 802.3at-2009 PoE Plus standard delivers even more power—up to 24.5 Watts. This has opened the door for a wider range of PoE-enabled devices with higher power requirements, including laptops, pan-tilt-zoom cameras, routers, videophones and a variety of industrial controllers.

Welcome to ‘The Fiber Connection’. Along with a new corporate website, we are launching a new blog where we intend to post information about fiber optic technologies and trends, industry happenings and anything else that might be of interest. You will see posts from people here at Fiber Connections and several guest posts as well.

Within ‘The Fiber Connection’ we will look at the markets and applications that we know best – power over Ethernet (PoE), security applications, fiber to the desk, data communications, telecommunications, Fiber to the Home and more. We encourage you to contact us if you have a relevant topic and would like to write a guest post – we certainly welcome your ideas and contributions.

So, please check back soon. We look forward to providing useful posts and adding to the fiber optic industry discussion.

by Lee Kellett – guest blogger

In the spring of 2009, Corning Cable Systems announced that they would discontinue the production of the MT-RJ Unicam connector. Corning has stated that this was done because of low sales volume.  Although they continue to offer a ‘bag of parts’, it has made some of us wonder if this might signal the beginning of the end for the somewhat maligned duplex connector.

When the MT-RJ was introduced by AMP in the late 1990s, it was touted as a low cost, easy to use workstation connector.  After all, many liked the concept of the MT-RJ because of the large degree of comfort with a plug that looked like a copper-type RJ45. By May of 2000, AMP NETCONNECT announced that shipments of its popular MT-RJ fiber optic connector had surpassed the 2 million mark. This product was really taking off – but not without some issues. When giving his opinion about which SFF (small form factor) connectors will win, Jim Hayes from Fotec said, “The MT-RJ is another elegant design, but flawed …remember you have to make TWO perfect cleaves… it will be a factor, but not without teething problems.” Although many agreed with this thought, the connector showed it would have some legs.

By the early/mid 2000s, AMP NETCONNECT had granted approximately 30 licenses for others to manufacture the MT-RJ. Millions of connectors were being installed in data networks. The problem remained that this was not a high performance connector. Since the beginning, you were hard pressed to find someone who thought the MT-RJ was an overall better choice than the LC connector. The ‘feature’ of the 2 fiber ferrule proved too often be a detriment instead of a benefit.  And, when it came to field installations – well, let’s just say – it was not always pretty. In 2005, Rodney Casteel at Commscope published an article: LC VERSUS MT-RJ  CONNECTORS: WHICH IS THE BETTER PERFORMER – and concluded: “The classic one-fiber-at a-time approach of the LC connector, coupled with its tenability, makes it the right choice for the low loss, high performance SFF connector of the present and the future.”

As we approach the middle of 2010, is the data communications market ready to leave the MT-RJ behind? Some already think the end is here. Mike Hines, former VP at Ortronics tells me, “With increased data rates came dramatically reduced connection loss budgets…this exacerbated the practical issues associated with the MTRJ…less fault tolerance…then it was game over.” As far as Corning goes, they recently surveyed many of their customers who purchase the MT-RJ to make pre-terminated assemblies. They found that about 10% of connectors sold into that market were MT-RJs – keeping it in play for now. On the other side of that, Fiber Connections has seen a significant reduction in orders for the MT-RJ. When it is ordered, the patch cord most often has an MT-RJ on one end and an LC on the other.

Finding current information on the MT-RJ, beyond spec sheets and companies selling patch cords and such, is quite difficult. The reports and papers discussed here were from the late ’90s and early ‘00s.  I found several other studies from that time period, including a case study by Methode Electronics (in 1999) about the connector relative to transceivers.  Let’s just say that relative to the geometry differences of the LC and MT-RJ, the MT-RJ did not win in that study either.

The lack of more current published articles and press of any kind, support the idea that the end may be in sight. Nobody is talking about this product anymore (except for me, I guess). I did find something that may be as telling as any industry veteran’s opinion might be – an active RFQ to fiber a Navy base in Virginia. The project calls for MM cable with LC connectors at one end and MT-RJs at the workstation. My history with Military and Government optical communications projects has taught me that they are often far from the leading edge of technology.  After all, many military projects were installing 100/140um fiber years after it was replaced with much better fiber options (50um and 62.5um). So, I’m no soothsayer, but I say yes, the industry is more than ready to move on.