Link loss of duplex fiber connectors is easily obtained with the right software and knowledge of cabling standards.
By Eric Leichter
 |
Small-form-factor (SFF) connectors such as the MT-RJ and LC duplex are quickly becoming the common connector types placed into premises systems today. These connectors allow for the termination of two fibers in the same space normally occupied by single-fiber connectors. The higher density and cost savings associated with these new connector types make them a compelling option for data centers and LAN systems.
Installers experienced with testing standard single-fiber-connector optical links may not be familiar with the methods for testing these newer connector types. Fortunately, the current test procedures make it just as easy to test dual-fiber connectors as it is to test single-fiber types. Using the proper optical test units, the link loss of both fibers can be measured and recorded at one time, cutting the testing time per fiber in half.
The preferred test methods are compliant with the Commercial Building Telecommunications Cabling Standard, TIA/EIA-568-B.1, which details the link loss values expected for horizontal, backbone, and other types of optical links. During testing, it is important to validate the quality of the test jumpers to ensure the testing components do not affect the test results. Other important factors of optical testing, such as the proper cleaning of pinned connectors and mandrel wrapping of test jumpers during multimode testing, will ensure that test results are accurate and repeatable.
In local-area-network telecom applications, SFF connectors have gained wide acceptance because they exhibit excellent optical performance in a high-density package. These occupy about half the switch-port space of a standard SC duplex connector. The ANSI/TIA/EIA standards group developed intermateability standards for several of these connector types to ensure compatibility among manufacturers. Furthermore, many of these connectors meet or exceed the performance requirements outlined by the relevant standards groups: TIA/EIA, IEC, Fibre Channel, and IEEE.
The MT-RJ connector houses two fibers in a single ferrule and is the most common SFF connector installed in premises environments due to its early acceptance by electronics and components manufacturers. The MT-RJ connector is aligned by pins in the ferrule of the connector and requires a pinless-to-pinned connector mating. The LC and MU connectors are single-fiber connectors with a 1.25-mm-diameter ferrule, compared to the 2.5-mm ferrules used in FC-, SC-, and ST-compatible connectors. Two LC connectors can be duplexed and will fit into the space of one SC connector. Both the MT-RJ and LC duplex connectors provide Gigabit and 10-Gigabit performance once installed and tested properly. Generally, connector type (shape) does not affect system performance or significantly change the test method.
Testing connectors
Optical links built with LC duplex and MT-RJ connectors are commonly constructed of a pair of fibers for transmit and receive and can be tested with standard optical source and meter units. The testing procedures for SFF connectors are similar to those used when testing SC duplex connectors. The optimal method uses two optical testers, each containing a source and meter, which can measure the loss of both fibers in the pair. That decreases the testing time and measure polarity as well as optical power loss.
The optical meter port of the test unit should not be a fixed connector-adapter type. The meter port should be changeable to allow insertion of the proper connector type as needed. That allows for the TIA/EIA-568-B.1-required one-jumper reference to be obtained without disconnection from the source port. Using one jumper between the source and meter shows the actual output power of the source. Using the power level with one jumper as a baseline (referencing), the loss of a patch-panel-to-patch-panel system can be accurately measured.
A common test scenario uses test jumpers with unlike connectors on each end. Using patch cords with SFF connectors on one end and standard simplex (SC) connectors on the other end, a simplex connector is inserted into both the transmit and receive ports for testing two fibers simultaneously. Two optical test units, each containing a source and meter, are used to measure the loss in an MT-RJ system link (see Fig. 1). Two SC-duplex-to-MT-RJ pinless jumpers are used during the reference process; MT-RJ optical adapters are required at the meter ports to obtain the one-jumper reference at each unit. Then the MT-RJ meter adapter is replaced with an SC adapter, and the other SC leg of each test jumper is added.
A special pinned MT-RJ-to-MT-RJ jumper can be placed in between the two jumpers as a check step; another reference should not be taken here. A loss above 1.5 dB—two connector pairs at 0.75 dB maximum each—would show that the test jumpers need cleaning or replacement. If the pinned jumper does not deliver a passing value at this point, inaccurate test results may result. After a passing result has been achieved, the operator can insert the pinless MT-RJ test connectors into the first optical link to be tested.
Testing of LC duplex connectors can be achieved in a very similar fashion. Hybrid (for example, LC to SC) test jumpers can be used to connect the testers to the system. An optical meter adapter that can accept the 1.25-mm ferrule of the LC connector will be required during the one-jumper referencing process. The loss of two runs can be measured at one time, and system testing can be achieved at a rate twice as fast as for single-fiber testing.
 |
Acceptable link loss depends on its location in the system and the length of the run. The TIA/EIA-568-B.1 guideline allows a maximum loss of 2 dB for horizontal runs of 90 m, and testing is only required in one direction at one wavelength. For centralized optical links of 90 to 300 m, a maximum loss of 3.3 dB is allowed, and testing is again only required in one direction and at one wavelength. For backbone runs, testing is required at both wavelengths in one direction; the maximum acceptable loss can be determined by the sum of each connector pair plus the cabled fiber attenuation based on length and wavelength. TIA/EIA guidelines do not require bidirectional testing with power-meter measurements. The TIA/EIA-568-B.1 guideline also gives link attenuation and maximum distance values for various fiber types and applications (see Table).
If testing results in high optical loss, the first correction step would be to clean the test jumper and system connectors. Connectors should be cleaned with 90%+ isopropyl alcohol, then wiped with a dry lint-free tissue. With pinned connectors such as the MT-RJ, proper cleaning of the fibers between the protruding pins is more difficult. The use of a cleaning cassette designed for pinned connectors is recommended (see Fig. 2). Adapters are cleaned with lint-free swabs and/or compressed air. After cleaning all components, replace the connectors and retest to verify the optical loss improvement.
Test reporting
During optical testing, each test unit displays test values from one fiber of the fiber pair under test; most quality test units can record and store these data points. Today's software used for single-fiber-connector testing is not optimized for dual-fiber or duplexed connector test results. New software has been developed to store and retrieve files in an efficient manner. This PC software combines the test values of each test unit and shows the test data in one report.
In single-fiber testing, optical test units are set to count sequentially (1, 2, 3, and so on). This is problematic in duplex testing if two test units are used to record data and the odd and even fiber values are stored on different units. To solve this problem, the software in duplex test systems must account for transmit/receive fiber pairs. Such software inside the test units should be able to record data with the proper numbering sequence (see Fig. 3).
After the data has been recorded in the field, it can be downloaded directly to a PC for report generation. In this example, the measurements from the odd-numbered fibers were stored in test unit A, while the even-numbered fiber measurements were stored in test unit B. The software combined both sets of data into one report. The software creates a table showing loss values for all runs tested on a single page. Although bidirectional results are easily incorporated into this spreadsheet, they are not generally required by TIA/EIA standards.
Using test units—each with a source and meter port—can double the speed of testing while maintaining accuracy, provided the test-port adapters match the connector type. Upgrading to software that can appropriately record the test results during testing and coordinate the data after testing will simplify the testing process and make the results understandable for the customer.
ERIC LEICHTER is senior applications engineer at Corning Cable Systems, PO Box 489, Hickory, NC 28603. Tel: (828) 901-5530; email: eric.leichter@corning.com.
