OTDR Testing
This domain covers optical time-domain reflectometry: how the trace is built, which settings change the answer, dead zones, ghosts and gainers, and bidirectional measurement. The exam likes questions where a plausible-looking trace is wrong because of one setting.
How the trace is made
The OTDR launches a pulse and measures Rayleigh backscatter and Fresnel reflections versus time. Distance comes from d = c·t/(2·IOR). The slope of the trace gives attenuation in dB/km; a reflective spike is a connector, mechanical splice, or break; a small non-reflective step is a fusion splice or a bend.
Settings that change the answer
- IOR (group index): a 1% IOR error produces a 1% distance error - 100 m in 10 km.
- Pulse width: short pulses resolve close events but lose dynamic range; long pulses reach farther but widen dead zones.
- Range and averaging: set range just beyond the link; average long enough to smooth noise.
- Wavelength: 1310 nm for shorter spans and higher loss; 1550 nm for long haul; 1625 nm for testing through live-PON splitters; 850/1300 nm for multimode.
Dead zones, launch and receive cables
The event dead zone is the minimum spacing at which two reflective events can be separated; the attenuation dead zone is how far after a reflection loss can be measured. A launch cable (typically 100–300 m) pushes the first connector out of the dead zone, and a receive cable reveals the last connector. Without them, the two end connectors are simply not measured.
Ghosts and gainers
Ghosts are multiple reflections that appear at multiples of a real event distance; they look like events but carry no real loss. Gainers are apparent negative splice losses caused by backscatter differences across a splice. The correct splice loss is the bidirectional average: (L₁ + L₂)/2.
Worked example
An OTDR is set to IOR 1.4600 but the fiber's group index is 1.4680. The displayed length is 12.000 km. True length = 12.000 × (1.4600 / 1.4680) = 11.935 km, so the trace overstates distance by about 65 m. Reasoning: a low entered IOR makes the instrument believe light travels faster in the fiber than it does, stretching every distance.
Traps that cost marks
Testing without launch and receive cables; reading a ghost as an event; quoting single-direction connector loss as final; using a 1 µs pulse on a 300 m premises link; expecting per-ONU traces when firing through a splitter; ignoring a splice gainer instead of averaging bidirectionally.
Key points
- Distance = c·t/(2·IOR); a 1% IOR error gives a 1% distance error.
- Short pulse improves resolution; long pulse improves range but enlarges dead zones.
- Launch and receive cables are required to measure the end connectors.
- Splice loss must be averaged from both directions to cancel gainers.
- Ghosts are reflection artifacts and carry no real loss.
- 1625 nm is the wavelength for testing through splitters on live PONs.