This tool estimates splice loss from a total coupling efficiency η and converts it to decibels.
- Total loss:
L(dB) = -10 log10(η_total) + L_fixed - Gaussian overlap model (typical for single-mode):
- Offset:
η_off = exp(-2(d/w)^2) - Angle:
η_ang = exp(-(π n w θ / λ)^2) - Gap:
η_gap = 1 / (1 + (z/z_R)^2), withz_R = π n w^2 / λ - MFD mismatch:
η_mfd = (2 w1 w2 / (w1^2 + w2^2))^2
- Offset:
- Geometric model (useful for multimode): simple step-index style approximations for core/NA mismatch, plus conservative offset, angle, and gap terms.
These equations are widely used approximations. Real splices can also include cladding alignment errors, end-face quality, contamination, and refractive-index profile effects.
- Select a fiber type and a calculation model.
- Enter known splice conditions: offset, angle, and gap.
- For Gaussian overlap, fill MFD1 and MFD2 and keep wavelength realistic.
- For geometric calculations, ensure core diameters and NA values are set.
- Enable only the effects you want to include, then click Calculate.
- Use Download CSV or Download PDF after a calculation.
1) What splice loss represents
Splice loss is the insertion loss created when two fiber ends are joined. In link budgets it is added to connector losses and distributed attenuation. When splice counts grow, small reductions protect system margin.
2) Typical magnitudes in practice
Fusion splicing of matched single‑mode fibers often targets about 0.02–0.05 dB per splice in clean field work. Mechanical splices frequently land around 0.10–0.30 dB. Values depend on fiber type, tools, and preparation.
3) Offset sensitivity for single‑mode fibers
Single‑mode coupling is highly sensitive to lateral offset because the guided mode is only a few micrometers wide. A typical mode field diameter near 10.4 µm at 1550 nm implies a mode radius near 5.2 µm. At this scale, sub‑micrometer offsets can be measurable, so alignment and cleanliness matter.
4) Angular and gap effects
Angular misalignment behaves like a tilt between mode wavefronts, so overlap falls as angle increases. The calculator uses wavelength, effective index, and angle to estimate this reduction. End separation (gap) adds diffraction‑like spreading that can matter when the joint is not fully fused.
5) Mode‑field mismatch between fibers
When two fibers have different mode field diameters, the overlap is reduced even with perfect alignment. This is a common issue when splicing dissimilar single‑mode fibers, such as standard transmission fiber to bend‑insensitive or specialty fibers. The mismatch term helps estimate that unavoidable coupling penalty.
6) Multimode considerations
Multimode splicing is more tolerant to small offsets because the core is larger (for example 50 µm). Core diameter and numerical aperture mismatches can still reduce coupled power. The geometric option provides quick, conservative comparisons.
7) Why “selected effects” matters
Real splice loss is a combination of multiple mechanisms. This tool lets you enable only the terms you can justify from measurement or process knowledge. For example, if you are analyzing a fusion splice with high-quality end faces, gap loss may be negligible, while offset and mismatch terms dominate. If you have OTDR events or insertion loss test data, use them to decide which mechanisms are credible for your case.
8) Interpreting outputs and improving results
The calculator reports total coupling efficiency η and its dB equivalent. Use the contribution table to find the dominant sensitivity and target improvements such as better cleaves, alignment, cleanliness, and fiber compatibility. You can also set acceptance thresholds for field teams, such as a maximum allowed splice loss per repair. Rerun the same case to estimate recovered margin.
1) Should I use the Gaussian or geometric model?
Use Gaussian overlap for single‑mode fibers because it models the guided mode shape. Use the geometric option for quick multimode estimates or when core/NA mismatch is the main concern.
2) What does coupling efficiency η mean?
η is the fraction of optical power that couples through the splice into the receiving fiber. The calculator converts it to loss using -10 log10(η), then adds any fixed extra loss you specify.
3) Why does tiny offset create noticeable loss?
Single‑mode fields are only a few micrometers wide, so a small lateral shift reduces overlap strongly. That is why core alignment, proper cleaving, and clean end faces are essential.
4) When should I add fixed extra loss?
Add fixed loss when you have a measured baseline penalty not captured by geometry, such as imperfect fusion, contamination, or microbending at the splice protector.
5) Can I model dissimilar fibers?
Yes. Enter different MFD values in Gaussian mode, or different core and NA values in geometric mode. Enable the mismatch checkboxes so the calculator includes those penalties.
6) Why are some terms optional?
Because you may not have reliable data for every mechanism. Enabling only justified terms avoids overestimating loss and helps you isolate the dominant contributor.
7) Are results exact measurements?
No. They are engineering approximations that help compare scenarios and sensitivities. Validate with OTDR, insertion loss testing, and vendor splice specifications for your specific fiber types.
| Case | Model | Offset (µm) | Angle (deg) | Gap (µm) | MFD (µm) | Estimated loss (dB) |
|---|---|---|---|---|---|---|
| A | Gaussian | 0.5 | 0.5 | 5 | 10.4 / 10.4 | ~0.06 to 0.25 |
| B | Gaussian | 1.0 | 1.0 | 10 | 10.4 / 10.4 | ~0.20 to 0.80 |
| C | Geometric | 2 | 1 | 5 | 50 / 50, NA 0.20 | Varies with selection |