Testing of Outside Plant and Premises Fiber Optic Cable Plants
This domain covers how testing differs between long outdoor spans and short building cabling: standards, sampling, direction, and which instrument is primary in each environment. Exam questions often hinge on applying the wrong environment's rules.
Outside plant
Outside plant work covers aerial, buried and underground cable, splice closures, and long spans. The relevant test standards are TIA-526-7 for single-mode and TIA-526-14 for multimode, alongside the owner's acceptance specification.
- Acceptance is normally 100% of fibers, not a sample.
- Take bidirectional OTDR traces and average each splice loss; single-direction splice readings can be wrong by tenths of a dB because of backscatter differences.
- Measure span attenuation from the trace slope, then confirm end-to-end insertion loss with an OLTS.
- Include every splice, closure, WDM and mid-span access in the loss budget.
- Long-haul and high-rate routes need chromatic dispersion and PMD characterization as well.
- Test to the temperature and environmental conditions in the specification; closures that take on water show rising loss.
Premises
Premises cabling follows the TIA-568 series and ISO 11801. Links are short, connector count is high, and moves, adds and changes happen constantly.
- Test permanent link or channel and state which; their limits differ.
- Multimode testing requires an encircled-flux or mandrel launch condition.
- Use an OLTS as the primary instrument; an OTDR's dead zones are comparable with typical link lengths, so it cannot resolve the connectors.
- Verify polarity for every duplex pair and label to TIA-606.
- Test after installation and after any move, add or change; a kinked jumper adds real loss.
Worked example
A 1.2 km outside plant span shows a splice at 0.6 km reading 0.9 dB in one direction and 0.1 dB in the other. Bidirectional average = 0.5 dB, which exceeds the 0.3 dB specification, so the splice is re-done. If the technician had tested from one end only, the 0.1 dB reading would have passed a splice that is genuinely bad. Reasoning: splice loss direction dependence is real, and OSP acceptance requires averaging.
Traps that cost marks
Applying premises limits to an outside plant span; testing OSP in one direction only; forgetting the pigtail splices at each equipment end; using an OTDR alone to accept a short premises link; skipping polarity; testing while other crews are patching, which moves jumpers mid-run.
Key points
- Outside plant acceptance is 100% of fibers with bidirectional OTDR and averaged splice losses.
- Premises acceptance uses an OLTS with encircled-flux launch and a stated link definition.
- OTDR dead zones make it unsuitable as the sole test for short premises links.
- Splice loss is direction-dependent, so single-direction readings can pass a bad splice.
- Polarity and TIA-606 labeling are part of premises acceptance, not optional extras.
- Test after any move, add or change, not only at initial installation.