How to Use an OTDR to Measure Fusion Splice Loss: A Step-by-Step Guide

par Fusion Splicer 8 août 2026

Why Measuring Fusion Splice Loss with an OTDR Matters

A fusion splicer's built-in loss estimate is calculated using a core-alignment camera — it's a prediction, not a measurement. The only way to verify actual splice loss in a deployed fiber link is with an Optical Time Domain Reflectometer (OTDR). Whether you're building FTTH access networks, backbone routes, or data center interconnects, OTDR-verified splice loss is the industry standard for network acceptance testing.

This guide walks you through the complete process: equipment setup, launch cable use, trace acquisition, bidirectional measurement, and result interpretation — all aligned with ITU-T G.671 and IEC 61300-3-35 standards.

Equipment You Need

OTDR — matched to your fiber type (SM 1310/1550 nm for G.652D; MM 850/1300 nm for OM3/OM4). See our OTDR collection for professional-grade options including the EXFO MaxTester 730C PON OTDR and the EXFO MaxTester 715B.
Launch cable (dead-zone box) — minimum 100 m, ideally 300–500 m. This eliminates the front-end dead zone so the first splice is visible. We recommend the 500 m SC/UPC Launch Cable Box for single-mode testing.
Receive cable — same length as launch cable, placed at the far end to reveal the last splice before the connector reflection.
Fusion splicer — for reference, see the COMWAY C10S Fusion Splicer or the JETFIBER H5 Core-to-Core Fusion Splicer.
Fiber identifier, index gel, and IEC-compliant connector cleaner.

Step 1: Configure OTDR Parameters

Correct parameter setup is critical. Incorrect settings produce misleading traces.

Wavelength: Use 1310 nm for splice loss (lower Rayleigh scattering sensitivity reveals splice events more clearly). Run 1550 nm as a second pass to detect bending losses.
Pulse width: Shorter pulses (3–10 ns) give better resolution near splices; longer pulses (100–1000 ns) extend dynamic range for long spans. Start with 30 ns for spans under 20 km.
Range: Set to 1.5–2× the actual fiber length.
IOR (Index of Refraction): Match to your fiber spec — typically 1.4677 for G.652D at 1310 nm. Incorrect IOR shifts all distance readings.
Averaging time: 30–180 seconds for field acceptance; longer averaging improves SNR on high-loss or long spans.

Step 2: Connect the Launch Cable

Connect the launch cable between the OTDR port and the fiber under test. The launch cable serves two purposes:

It moves the front-end dead zone (typically 5–25 m depending on pulse width) away from your first splice point.
It conditions the OTDR's output mode field before it enters the fiber under test, reducing connector-induced measurement error.

Clean all connectors with an IEC 61300-3-35 compliant cleaner before mating. A dirty connector is the #1 cause of false high-loss readings.

Step 3: Acquire the OTDR Trace

Run the OTDR and observe the trace on screen. A healthy single-mode fiber trace shows:

A straight, downward-sloping Rayleigh backscatter line (the "slope" represents fiber attenuation, typically 0.33–0.36 dB/km at 1310 nm for G.652D).
Small downward steps at splice points — these are your splice loss events.
A large reflection spike at the far-end connector or fiber end (Fresnel reflection).

If you see a gain event (upward step) at a splice, this is a measurement artifact caused by different backscatter coefficients between the two fiber segments — not a real gain. This is why bidirectional testing is essential.

Step 4: Measure Splice Loss Using the LSA Method

Most professional OTDRs (including EXFO MaxTester series) offer two splice loss measurement methods:

2-Point Method: Measures the dB step at the splice event. Fast but susceptible to local noise.
LSA (Least Squares Approximation) Method: Fits regression lines to the backscatter on both sides of the splice and calculates the loss from the intercept difference. This is the recommended method per IEC 61300-3-35 and gives the most repeatable results.

Place your OTDR cursors (A and B markers) on the linear backscatter sections immediately before and after the splice — avoid placing them within the event dead zone (typically 1–4 m for short pulses).

Step 5: Perform Bidirectional Testing and Average the Results

This is the most important step that many technicians skip. Because OTDR measures backscatter — not transmitted power — a splice between two fibers with different backscatter coefficients will show different loss values from each direction.

Bidirectional average formula:

True Splice Loss = (Loss measured A→B + Loss measured B→A) ÷ 2

Measure from both ends of the fiber span, then average each splice's loss value. This cancels out the backscatter artifact and gives you the true insertion loss of the splice.

Modern OTDRs like the EXFO MaxTester 730C support SmartLink Mapper and bidirectional analysis tools that automate this calculation.

Step 6: Interpret Results Against Acceptance Thresholds

Use these industry-standard thresholds as your acceptance criteria:

Application	Max Splice Loss (per splice)	Standard
FTTH / PON access	≤ 0.10 dB	ITU-T G.984 / G.9807
Backbone / long-haul SM	≤ 0.10 dB	ITU-T G.652
Data center OM3/OM4 MM	≤ 0.10 dB	TIA-568-C.3
General field acceptance	≤ 0.20 dB	IEC 61300-3-35

If a splice exceeds the threshold, re-cleave and re-splice. Common causes of high splice loss include: contaminated fiber end-faces, poor cleave angle (>0.5°), fiber core misalignment, or arc power/duration mismatch on the fusion splicer.

Common OTDR Trace Artifacts to Know

Gainer / Gain step: Apparent gain at a splice — artifact from backscatter mismatch. Always bidirectionally average.
Ghost reflection: A repeated reflection event at double the distance of a real reflection. Not a real event.
Non-reflective event: A loss step with no reflection spike — typical of a good fusion splice (vs. a mechanical splice or connector which shows a spike).
Dead zone masking: A splice too close to a connector may be hidden in the dead zone — use a launch cable to push it into the visible range.

Pro Tips for Accurate Splice Loss Measurement

Always use a launch AND receive cable to make all splices visible, including the last one before the far-end connector.
Clean connectors before every measurement — use an IEC-compliant cleaner and inspect with a fiber scope.
Use auto-analysis mode on your OTDR for initial screening, then manually verify each splice event with LSA cursors.
Save traces in .sor format (Bellcore SR-4731) for standardized reporting and future comparison.
For PON networks, test at 1625 nm (in-service wavelength) in addition to 1310/1550 nm to detect bending losses that affect live traffic.

Recommended Equipment for Fusion Splice Loss Testing

For professional-grade OTDR-based splice loss verification, we recommend:

EXFO MaxTester 730C PON OTDR — 39 dB dynamic range, 0.6 m event dead zone, ideal for FTTH/PON splice testing
EXFO MaxTester 715B — SM 1310/1550 nm, compact field OTDR for backbone and access networks
500 m SC/UPC Launch Cable Box — eliminates front-end dead zone for accurate near-end splice measurement
COMWAY C10S Fusion Splicer — 6-motor core alignment, 5s splice time, ideal for FTTH and backbone deployments
JETFIBER H5 Core-to-Core Fusion Splicer — cost-effective core alignment for FTTH deployments

Conclusion

Using an OTDR to measure fusion splice loss is not optional for professional fiber network deployment — it is the only method that provides verified, documentable proof of splice quality in a live fiber span. By following the steps above — correct parameter setup, launch cable use, LSA measurement, and bidirectional averaging — you can consistently achieve splice losses well below 0.10 dB and deliver networks that meet or exceed ITU-T and IEC acceptance standards.

Have questions about selecting the right OTDR for your splice testing application? Browse our full range of fiber optic test equipment or contact our technical team for a recommendation.