FTTH Testing
What this domain tests
This domain covers proving a link is good: connector inspection, Tier 1 loss/length testing, Tier 2 OTDR testing, testing through live PON splitters, and the test documentation an acceptance package needs.
Inspect before you measure
Use a bench microscope and the IEC 61300-3-35 zones: A (core, 0–25 µm), B (cladding, 25–120 µm), C (adhesive, 120–130 µm), D (contact, 130–250 µm). Any defect in zone A fails. Clean with a one-click cleaner or lint-free wipe and inspect again - a dirty reference connector invalidates every number that follows.
Tier 1: loss and length
An OLTS measures insertion loss at 1310, 1490 and 1550 nm against TIA-526-7 practice. Set the reference at the start of each job or whenever jumpers change, and record the reference method (one-, two- or three-jumper) because mixing methods creates false passes and failures. Compare results to the design budget, not to a generic number.
Tier 2: OTDR
Enter the correct IOR, choose a pulse width that balances dead zone against dynamic range, and average. A launch cable of 100 m or more lets you see the first connector; a receive cable shows the last. Event and attenuation dead zones limit what a short pulse can resolve. Splice loss should be measured from both directions and averaged per IEC 61300-3-6, and ORL must be reported.
Testing live PON
Use 1625 nm with a WDM filter, or a splitter-tolerant PON OTDR, so the test signal does not disturb subscribers. A power meter at the ONT verifies received power against the ONT specification.
Worked example
A splice reads −0.42 dB from the CO and +0.12 dB from the subscriber end. The second value is a gainer caused by differing backscatter coefficients. Bidirectional result = (−0.42 + 0.12) ÷ 2 = −0.15 dB, i.e. 0.15 dB loss - a good splice. Reporting the single-direction −0.42 dB would wrongly flag a re-splice.
Traps
Referencing through a dirty connector; forgetting to re-reference after swapping launch cables; reporting a gainer as a loss; testing 1310 nm through a splitter and calling the splitter an anomaly; leaving IOR at the default; and passing a link on loss while ignoring ORL.
FTTH Troubleshooting and Certification
What this domain tests
This domain tests fault isolation logic and the paperwork that closes a job. You should be able to read PON alarms, interpret ONT receive power, localize a fault with an OTDR, and produce the acceptance record that certifies the link.
Read the alarm first
LOS means no light reaching the ONT. LOF and LOAi point to frame or acknowledge problems rather than optics. Dying Gasp means the ONT lost local power, not fiber. RDI indicates a downstream defect signalled upstream. Check whether one subscriber or every subscriber on the splitter is down: a whole PON points to the OLT port, feeder or splitter; a single subscriber points to the drop, connector or ONT.
Optical symptoms
A Class B+ ONT typically expects roughly −8 to −27 dBm. Below sensitivity gives no service; above overload also fails, which is why a short link with too little loss is worth checking. Loss that grows sharply at 1550 or 1625 nm while 1310 nm looks normal is the signature of a macrobend. A localized step of several dB at one distance is a connector or splice event; a gradual rise is cable or bend loss.
Work in order
Power at the ONT, then connector inspection and cleaning, then power meter reading, then OTDR from the OLT, then splitter port assignment and ONT serial number provisioning. Replacing an ONT before measuring optical power is the classic wasted truck roll.
Worked example
An ONT reports −31 dBm where the design predicts −22 dBm, a 9 dB gap, and other subscribers on the same splitter are fine. An OTDR at 1625 nm shows every event within 0.3 dB of the as-built except a 7.2 dB event at 1.15 km with high reflectance, at the splice closure. Because the loss is localized and the rest of the link matches design, the fault is in the closure - a contaminated mechanical splice or a damaged connector - not in the drop or the ONT.
Certification and acceptance
A link is certified when measured loss and ORL meet the design budget, results are recorded per wavelength and direction, connectors are inspected, and labels follow TIA-606 administration. The package includes test reports, as-builts and the design budget so a future technician can compare. Technician credentials in the FOA program build from the CFOT technician certificate to specialist CFOS designations.
Traps
Swapping hardware first; assuming LOS always means a cut cable; accepting a dirty connector as a bad splice; testing from one direction only; and skipping splitter port and duplicate-serial checks.