Testing and measurement: LSPM, OTDR, return loss, inspection

A fibre link is accepted, certified and later diagnosed with a small set of instruments: a light source and power meter for the total loss, an OTDR for the loss of each event along the route, a return-loss meter for reflections, and a microscope for the connector endfaces that cause most trouble. The transceiver's own DDM is a fifth instrument, always present. This page explains what each measures, how to use it correctly (reference methods, launch cords, wavelengths), how to read an OTDR trace, what the certification standards require, and where DDM fits.

The instruments

Instrument	Measures	Accuracy	Use
Light source + power meter (LSPM / OLTS)	insertion loss of the whole link, end to end, at 850/1300 (MMF) or 1310/1550/1625 (SMF)	±0.1–0.2 dB	acceptance (Tier 1), troubleshooting total loss
OTDR (optical time-domain reflectometer)	loss and reflectance of each event (connector, splice, bend, break) versus distance; fibre attenuation per km	±0.05 dB per event, ±1 m distance (after calibration of refractive index)	characterisation (Tier 2), fault location
Return-loss meter / OCWR	total back-reflection of the link	±0.5 dB	UPC vs APC verification, laser stability
Fibre inspection microscope	endface scratches, pits, contamination	per IEC 61300-3-35 zones	before every mating
Visual fault locator (VFL)	red laser (650 nm) visible through jacket at breaks and tight bends	—	tracing, gross faults, polarity
Fibre identifier	live traffic direction and presence by bending	—	before disconnecting the wrong fibre
Optical spectrum analyser (OSA)	wavelengths, channel powers, OSNR	—	WDM (OSNR)
Transceiver DDM	Tx and Rx power at the link ends	±3 dB spec, ±1 dB typical	in-service monitoring (Reading DDM with tools)

Insertion loss with LSPM (Tier 1)

Reference: connect source and meter with the test reference cords and zero the meter (0 dB). The one-jumper reference (TIA-526-14/ISO 14763-3 method A) includes both link connectors in the result; the two- or three-jumper methods exclude one or both.
Measure: insert the link under test between the reference cords; read the loss at each wavelength, both directions if required (IEC recommends bidirectional).
Compare with the design budget and the standard's pass value: link loss ≤ (fibre length × coefficient + connector pairs × allowance + splices × allowance).
Wavelengths: MMF at 850 and 1300 nm; SMF at 1310 and 1550 nm (1625 nm to reveal bends). Multi-mode sources must have the standard encircled flux launch.

Pitfall	Effect
Dirty reference cord	0.5–2 dB error on every link
Reference cord of a different grade or MFD	offset of 0.1–0.5 dB
Wrong reference method for the standard	link connectors counted twice or not at all
Laser source into multi-mode without encircled flux compliance	reads too low or too high by up to 1 dB
Measuring only one wavelength	misses bends (1550/1625) or water-peak issues

OTDR (Tier 2)

The OTDR sends pulses and times the backscatter; loss appears as a drop in the trace, reflections as spikes.

Trace feature	Meaning
Steady slope	fibre attenuation (dB/km) — should match the cable spec
Step down without spike	fusion splice or bend (non-reflective event)
Spike then step down	connector or mechanical splice (reflective event)
Spike to the top, then noise	end of fibre or break; open connector
Step up (gainer)	splice between fibres of different backscatter (MFD); average both directions
Spike without step, far away	ghost — multiple reflection; not a real event
Rising slope near end	high-reflectance end saturating the receiver

Setting	Guidance
Pulse width	short (5–30 ns) for close events and resolution; long (1–10 µs) for range; dead zones grow with pulse width
Range	1.5–2 × fibre length
Averaging	30–180 s for splice-level accuracy
Refractive index	from the cable datasheet (1.467–1.470 for G.652); a 0.001 error = 0.07 % distance
Launch and receive cords	300–1 000 m (SMF) / 100–300 m (MMF) so the first and last connectors sit outside the dead zone
Wavelengths	1310 + 1550 (SMF), 1625 for bend hunting; 850 + 1300 (MMF)
Bidirectional test	required for accurate splice loss (averages the MFD artefact)

Event dead zone 1–3 m, attenuation dead zone 5–20 m: two events closer than that merge.

Return loss and reflectance

Quantity	Definition	Good values
Reflectance of one event	reflected/incident power at that event, dB (negative)	UPC connector −50 dB or better; APC −65; splice < −70
Optical return loss (ORL) of a link	total reflected power, dB (positive)	> 27 dB (10G), > 35 dB (analog video, DWDM); > 45 dB APC links

Reflections destabilise DFB lasers and add interferometric noise; standards specify minimum ORL for the PMD (e.g. 12 dB for 10GBASE-LR, 21 dB for 1000BASE-LX — the module's own tolerance; plant design aims far higher).

Endface inspection (IEC 61300-3-35)

Zone	Diameter (SMF)	Allowed
A — core	0–25 µm	no scratches, no defects
B — cladding	25–115 µm	no defects > 2 µm (a few ≤ 2), no scratches > 3 µm
C — adhesive	115–130 µm	not assessed
D — contact	130–250 µm	no defects > 10 µm

Inspect both sides of every mating (the adapter side with a probe tip); clean with a one-click cleaner or lint-free wipe and re-inspect. Automated pass/fail scopes apply the zones (Connectors & fibre).

Certification and documentation

Standard	Tier 1 (loss)	Tier 2 (OTDR)	Notes
ANSI/TIA-568.3-D + TIA-526-7/14	required	optional/recommended	loss budget = attenuation coefficient × length + 0.75 × pairs + 0.3 × splices
ISO/IEC 14763-3	required	recommended	grade-based connector allowances
IEC 61280-4-1 / -4-2	MMF / SMF test methods	—	encircled flux for MMF
Customer / carrier acceptance	loss per fibre at 2 wavelengths, bidirectional OTDR traces, ORL, endface images	—	delivered as a test report per fibre

Keep the reports: the difference between installed loss and today's loss is the first question in any later fault (Plant faults).

Using DDM as a meter

The transceiver reports Tx power at one end and Rx power at the other; the difference is the link loss — within DDM accuracy (±3 dB by spec, ±1 dB on good modules, better after comparing to a real meter once).

Use	How
Sanity check of a new link	expected Rx = Tx − design loss; ±2 dB deviation → measure with LSPM
Trend monitoring	log Rx power; a slow 1 dB decline = connector contamination or bend; sudden 3 dB = event (Monitoring)
Which end is bad	compare both directions: asymmetric loss points to a directional fault (bad connector on one Tx side)
Calibration	measure one link with a meter and note the DDM offset (Calibration)

In CodingBox

Bench measurement of the module itself — Tx power into a short reference cord, Rx power from a calibrated source — separates a weak module from a lossy plant before the OTDR comes out. CodingBox reads the module's DDM and thresholds and lets you record the values against the serial number (DDM in the app, Accuracy & limits).

What the OTDR is judging — the splicing process, its acceptance limits and the defects behind a bad event: Splicing and termination; these measurements as part of the maintenance cycle, baselines and restoration: Maintenance and restoration.