I have been asked this question about fiber optic terminators more than any other. While different companies have written studies that justify their own recommended techniques, I decided to conduct my own investigation using graduate students of mine some years ago. This investigation considered the five major techniques for connector termination – (a) epoxy, bake and polish, (b) fast non-heat activated cure, (c) adhesive in ferrule, (d) UV-curable, and  (e) crimp and cleave.

Our investigation found that in terms of general performance, the termination methods that immerse or surround the fiber in an adhesive always offer the best solution. This method not only offers improved mechanical strength, but also increased fiber survivability in the ferrule. By far the greatest mechanical integrity was provided by the traditional "epoxy, bake and polish" method. However, it is necessary to choose the proper epoxy for the oven temperature and the type of manufacturer’s ferrule material (plastic, metal, or the several kinds of varying ceramic composites). 

The fast non-heat activated cure method worked well and required no heat, and therefore no oven. It had a mechanical strength of about 80-85% of that offered by the best "epoxy, bake and polish" connector. The "epoxy, bake and polish" and "fast non-heat activated cure" techniques can be used on many connector types across the industry depending only on ferrule material compatibility and connector construction. While the "adhesive in ferrule" method was similar in mechanical performance to the "fast non-heat activated cure" method, its use is limited to the types of connectors only offered by an "adhesive in ferrule" manufacturer. Obviously since the adhesive is supplied with the connector, and exists along the interior ferrule surface, the adhesive’s shelf life also forces the connector to be used before an expiration date.

The last two methods are not my personal favorites. Both have the disadvantage of being manufacturer dependent. While the "UV-curables" do bathe the fiber in a UV-curable epoxy, the cure time can vary with UV-lamp voltage. As the battery voltage varies, the UV lamp spectrum shifts which causes the cure time to shift as well. Not all connectors are compatible with this termination technique since the ferrule must be basically UV-transparent in order to allow the UV curing to take place. This makes the use of a connector dependent on the manufacturer’s product offerings, as is the case with the "adhesive in ferrule" technology. Even so, when properly done, the mechanical strength produced by this UV technique was only about 40-50% of that produced by the best "epoxy, bake and polish" connector.

Finally, the poorest performer of all was the "crimp and cleave" termination. This method is also manufacturer dependent and cannot be used as a generic termination technique across the connector industry. The mechanical strength produced ranged between 5 and 15% of the best "epoxy, bake and polish" connector. With these "crimp and cleave" connectors, I see an early problem from the days of plastic cladded silica (PCS) fibers called pistoning. Mechanical movement of the cable allows the core to slide in and out of the ferrule endface. Even when the kevlar material is captured, kevlar is known to stretch. I have seen these connectors demonstrated by a manufacturer’s representative on 250m loose tube fiber. When the optical fiber was turned upside down and shaken, the connectors fell off. In the past, our company actually had a contract to replace several hundred "crimp and cleave" connectors that had been improperly installed on loose tube cable in a paper mill. I do not suggest, recommend, or specify these connectors for any customer. It is unfortunate that many electrical contractors’ first exposure to fiber optic installation depends on these "crimp and cleave" connectors.

Kim Jovanovich

President, Omni Technologies, Inc.

Last modified: July 26, 2000