Guidelines On What Loss To Expect When Testing Fiber Optic Cables
To be able to judge whether a fiber optic cable plant is good, one does a insertion loss test with a light source and power meter and compares that to an estimate of what is a reasonable loss for that cable plant. The estimate, called a "loss budget" is calculated using typical component losses for each part of the cable plant - the fiber, splices and/or connectors. If the measured loss exceed the calculated loss by a significant amount (remembering the inherent uncertainty in all measurements), the system should be tested segment-by-segment to determine the cause of high loss.
A loss budget estimate can also be used to compare results from OTDR testing, but the inherent uncertainties of OTDR testing make the estimate less accurate. See OTDR Measurement Uncertainty in the OTDR page .
Some standards refer to the loss budget as the "attenuation allowance" but there seems to be very limited use of that term.
The calculated loss budget is an estimate that assumes the values of component losses and does not take into account the uncertainty of the measurement. Be aware of this because if measurements are close to the loss budget estimates, some judgement is needed to not fail good fibers and pass bad ones!
Cable Plant Loss Budget
The cable plant "loss budget" is a function of the losses of the components in the cable plant - fiber, connectors and splices, plus any passive optical components like splitters in PONs.
Thus the loss budget of the cable plant is a major factor in the power budget of the fiber optic link and is what one calculates to compare against tested insertion loss (and even compares to OTDR loss measurements) to determine if the cable plant is properly installed.
FOA has a free app for iOS smartphones and tablets that will calculate loss budgets for the cable plant you are designing or testing. See the Apple app store for your device for details.
How to calculate a loss budget. Connector Loss For each connector, figure 0.3 dB loss for most adhesive/polish connectors. The loss spec for prepolished/splice connectors will be higher (0.75 max per EIA/TIA 568 ) When testing cable plants per OFSTP-14 (double ended), include connnectors on both ends of the cable. When testing per FOTP-171 (single ended), include only one connector - the one atached to the launch cable. Splice Loss
For each splice, figure 0.3 dB (0.3 max per EIA/TIA 568 ) Fiber
For multimode fiber, the loss is about 3 dB per km for 850 nm sources, 1 dB per km for 1300 nm. (3.5 and 1.5 dB/km max per EIA/TIA 568 ) This roughly translates into a loss of 0.1 dB per 100 feet (30 m) for 850 nm, 0.1 dB per 300 feet(100 m) for 1300 nm. For singlemode fiber, the loss is about 0.5 dB per km for 1310 nm sources, 0.4 dB per km for 1550 nm. (1.0 dB/km for premises/0.5 dB/km at either wavelength for outside plant max per EIA/TIA 568 )This roughly translates into a loss of 0.1 dB per 600 (200m) feet for 1310 nm, 0.1 dB per 750 feet (250m) for 1300 nm.
So for the estimated loss of a cable plant, calculate the approximate loss as: (0.5 dB X # connectors) + (0.2 dB X# splices) + (fiber attenuation X the total length of cable) For more information see calculate a loss budget. What about OTDR testing?
OTDRs are used for verifying indivudual events like splice loss on long links with inline splices or for troubleshooting. All standards require an insertion loss test for qualification of the link loss. In MM fibers, the OTDR will underestimate the loss considerably - as much as 3 dB in a 10 dB link - but the amount is unpredictable. In long distance SM links, the difference may be less, but there are other measurment uncertainties, like connector or splice loss, where the OTDR can show a gain. What happens when you test with an OTDR with its limited distance resolution? Specifically, if you have singlemode fiber terminated with fusion spliced pigtials, you cannot see the both splice and the connector losses. Or what if you have a patch panel with connections using short patchcords?
For insertion loss testing, you simply sum up all the loss contributors and get a total for the cable run. In the case of an OTDR, you are analyzing each event.
So if you have a connection point where both fibers were terminated with spliced-on pigtails, you should analyze the event as the sum of 2 fusion splices and one connection, not each individually. A patchcord termination would be two connection losses, plus splices if the termination was by splicing on pigtails.
Will the network run on that link?
Here is a table showing the loss margin for most fiber optic LANs and links. If the loss of the cable plant is less than the maximum loss allowed for the link, it should run (but you really want a little bit of margin!)