Fibre Guide
An OTDR trace tells you exactly where loss happens along a fibre — if you know how to read it. This guide walks through every feature you will see, from the launch dead zone to the fibre end.
Tier 2 OTDR testing at a data centre patch panel.
In one line: an OTDR sends short pulses of light down the fibre and plots the light reflected and scattered back against distance. A smooth downward slope is fibre attenuation; steps and spikes are events — connectors, splices, bends and the fibre end.
An Optical Time Domain Reflectometer injects a pulse of laser light into one end of the fibre. As the pulse travels, a tiny fraction of light is scattered back (Rayleigh backscatter), and larger amounts are reflected at connectors, mechanical splices and the far end. The OTDR times the returning light to calculate distance and measures its power to calculate loss. The result is the trace: power in dB on the vertical axis, distance on the horizontal.
A typical trace: ① launch and dead zone, ② reflective connector, ③ non-reflective splice, ④ bend loss, ⑤ reflective fibre end followed by noise.
The first spike is the reflection at the OTDR’s own connector. For a short distance after it, the detector is “blinded” and cannot see events — the dead zone. That is why a launch cable (typically several hundred metres to a few kilometres) is connected between the OTDR and the link: it moves the first real connector out of the dead zone so it can be measured.
Between events the trace slopes gently downward. The steepness is the fibre’s attenuation in dB/km. Single-mode fibre loses far less per kilometre at 1550 nm than at 1310 nm, so testing at both wavelengths gives a fuller picture.
A connector shows as a spike (reflection) followed by a drop (loss). The spike height relates to reflectance; the step across it is the connector’s insertion loss. Dirty or damaged connectors show higher loss and stronger reflection — another reason to inspect and clean every end-face before testing.
A good fusion splice appears as a small step down with no spike, because a fused joint has almost no reflection. The size of the step is the apparent splice loss.
A macro-bend also shows as a non-reflective step, but it is usually wavelength-dependent: the loss is noticeably larger at 1550 nm (or 1625 nm) than at 1310 nm. Comparing wavelengths is the quickest way to tell a bend from a poor splice.
The far end of the fibre gives a large reflection if it is a clean connector, followed by a drop into the noise floor. A break in the middle of a route looks similar — the trace simply ends early. The distance to that event is your distance to fault.
Sometimes a splice appears as a step up — a “gainer”. The joint has not created light; the two fibres simply have different backscatter coefficients, so the OTDR misreads the loss from one direction. Testing from the other end shows the opposite error. Averaging the two directions gives the true splice loss. This is why proper Tier 2 certification uses bi-directional OTDR testing with launch and receive cables.
A ghost is a false event caused by a strong reflection bouncing back and forth — often appearing at exactly double the distance of a real reflective event, sometimes beyond the fibre end. If an event sits at a suspicious multiple of a real one, shows no loss and changes with test settings, it is probably a ghost.
| Setting | Effect |
|---|---|
| Pulse width | Short pulses resolve closely spaced events; long pulses reach further but widen dead zones. |
| Range | Should be set beyond the fibre length so the end and noise floor are visible. |
| Averaging time | Longer averaging reduces noise and reveals small events. |
| Wavelength | 1310 and 1550 nm together reveal bends and give a fuller loss picture. |
| Index of refraction | Must match the fibre so distances are accurate. |
Modern OTDRs analyse the trace automatically and list each event with its distance, loss and reflectance, plus pass/fail against the thresholds you set. Treat the event table as a starting point, not the final word: always check the trace for missed events, ghosts and gainers — especially on short data centre links with many connectors close together.
FAQ
The OTDR cannot measure events inside its own dead zone. A launch cable moves the first connector of the link beyond the dead zone so its loss can be measured; a receive cable does the same for the far-end connector.
A gainer is a splice that appears to add power. It is caused by different backscatter properties of the two fibres and disappears when the results from both directions are averaged.
Tier 2 certification combines Tier 1 loss and length testing with OTDR traces, usually bi-directional, so loss can be located and attributed to each connector and splice.
For single-mode links, 1310 nm and 1550 nm are standard. Comparing them helps identify bends, which cause more loss at longer wavelengths.
On Site
Our engineers provide Tier 1 and Tier 2 fibre certification and OTDR fault location across London and the South East.
OTDR Testing & CertificationCall 07555 454 919
