Key takeaways
- The fusion splicer display calculates a theoretical geometric profile without measuring any actual transmitted optical signal.
- A difference in mode field diameter between two fibres distorts the estimate without impairing mechanical quality.
- A clean cleave with an angle under one degree is essential for fusion and prevents micro-bubbles in the core.
- Only photometric measurement or bidirectional reflectometry constitutes admissible contractual proof of the joint.
Alignment, cleaving, and optical readout: why the splicer screen is merely a geometric calculation
- PAS (Profile Alignment System) attenuation estimation
- Internal algorithm of the fusion splicer calculating theoretical loss in decibels using orthogonal video image processing. It analyses the angular offset, axis misalignment, and local deformation of the fused silica, without injecting or measuring any actual light flux in the fibre.
On a construction site, seeing 0.02 dB displayed on the screen of a core-alignment fusion splicer provides a false sense of security. The splicer is not an optical metrology instrument. Its side cameras magnify the fusion zone from two angles (X and Y axes) and compare the position of the cores before and after the electric arc. If the machine detects two perfectly coaxial axes and clean visual continuity of the glass, it applies a standardised mathematical formula and displays a flattering figure. In reality, not a single photon passes through the strand during the operation.
This purely geometric approach ignores the intrinsic parameters of the glass. When two single-mode fibres have slightly dissimilar mode field diameters (MFD) — for instance when splicing a bend-insensitive G.657.A2 riser fibre onto a G.652.D feeder cable —, the interface creates a propagation discontinuity. Even with perfect arc alignment, this transition dissipates power. The machine will see a geometrically flawless fusion and display 0.01 dB, whereas the actual attenuation experienced by the signal will exceed 0.15 dB.
- Core eccentricity relative to the mechanical cladding, invisible if the arc obscures refraction contrasts.
- The presence of microscopic dust on mirrors or V-grooves, shifting the optical analysis of the cameras.
- Electrode wear that disperses the arc energy, generating internal mechanical stresses without visible external deformation.
- Refractive index differences between distinct fibre manufacturers, undetectable by simple image analysis.
- Stepped strippingStrip the outer jacket, then the 250 µm primary coating using the stripper without scoring or nicking the silica cladding.
- Solvent degreasingWipe the bare fibre with a lint-free wipe soaked in pure isopropyl alcohol (> 99%) until you hear the characteristic squeak.
- Orthogonal cleavingSever the fibre using a precision cleaver set to achieve a cut angle strictly under 1 degree.
- Immediate loadingPlace the strand into the V-groove of the splicer without bumping the electrodes or leaving the bare silica exposed to ambient air for more than ten seconds.
Heat-shrink oven, mechanical stresses, and link certification methods
The estimate displayed by the fusion splicer occurs before insertion into the heat-shrink oven. Yet this second thermal phase is the primary cause of splice degradation in splice closures or patch panels. Inserting an off-centre heat-shrink splice protector (smoove), the presence of a trapped particle, or insufficient heating time compresses the silica during cooling. The central metallic strength member must absorb stresses: if the sleeve bends, the fibre suffers an invisible local micro-bend that severely degrades the optical budget.
| Evaluation method | Physical principle | Contractual value | Sensitivity to MFD disparities |
|---|---|---|---|
| Splicer estimation | Geometric video image analysis (PAS) | None (purely an on-site decision aid) | Blind to mode field discrepancies |
| Direct power measurement (Source + PM) | End-to-end received power measurement (Tier 1) | Valid for overall link budget | Integrates total loss without locating the splice |
| Unidirectional reflectometry (OTDR) | Rayleigh backscatter measurement | Partial (risk of false gains or excessive losses) | Sensitive: may show a negative step or a false gain |
| Averaged bidirectional reflectometry | Arithmetic mean of opposite OTDR traces | Absolute proof in compliance with Swiss standards | Fully compensates for core asymmetries |
To certify an installation according to recognised industry standards, the professional applies the measurement requirements defined by Swiss universal cabling standards. Art. 364 CO also specifies that the contractor is liable for the proper and faithful execution of the work entrusted to them. A self-test report exported from the splicer carries no legal weight in a dispute over excessive attenuation; only Tier 1 photometric measurement deliverables or Tier 2 reflectometric trace files are legally binding.
Intrinsic core disparities and limits of PAS optical evaluation
The display of a fusion splicer routinely shows a flattering figure of 0.01 dB or 0.02 dB after the electric arc. For the telematics technician in the field, this number in no way constitutes an actual photometric measurement, but merely a calculated estimate. Modern core-alignment fusion splicers use the PAS (Profile Alignment System). Two digital cameras positioned at 90 degrees observe light refraction through the silica to identify the geometric axis of the 9-micrometre single-mode cores. The device evaluates insertion loss using a geometric algorithm based on residual offset, the initial cleave angle, and visual deformation of the fusion zone. No light flux is injected into the fibre during the process.
The discrepancy between the machine estimate and final verification stems from physical parameters that are entirely invisible to the splicer's CCD cameras. The Mode Field Diameter (MFD) frequently varies between fibre manufacturers, or even between two reels of the same reference within manufacturing tolerances. If a fibre with a slightly larger effective core is spliced onto a narrower fibre, unavoidable diffraction occurs at the interface. Furthermore, internal stresses from germanium doping in the silica, residual core ellipticity, or a local change in the refractive index induced by arc temperature escape image analysis. Only a photometric measurement via direct transmission or backscatter can establish the actual optical budget.
| Inspection tool | Operating principle | Analysed parameters | On-site limitations |
|---|---|---|---|
| Fusion splicer (PAS system) | Side imaging via 90° cameras | Axial core alignment, cleave angle, visual cleanliness of the joint | Detects neither mode field diameter (MFD) discrepancies nor refractive index variations |
| Light source and power meter | Injection of a calibrated continuous wave | End-to-end overall link attenuation | Indivisible: drowns the splice loss within total connector and cable attenuation |
| Optical Time-Domain Reflectometer (OTDR) | Time-domain analysis of the backscattered signal | Point attenuation of the splice, reflectance, and precise distance | Requires a launch fibre lead and averaged bidirectional measurement to eliminate apparent gain |
Operational rigour during cleaving and mechanical splice stabilisation
The validity of the splice relies first and foremost on cleave geometry. A poorly maintained precision cleaver, fitted with a blunt blade or clogged with acrylate debris, produces cuts with angles exceeding 1 degree or presenting roll-off lips. Even if the fusion splicer attempts to compensate for the angular offset by extending the pre-fuse arc to round the silica, this thermal manipulation alters the index profile at the joint. The cleanliness of the V-grooves also remains decisive: a microscopic dust deposit tilts the fibre by fractions of a micron when the magnetic clamps engage, distorting the motorized alignment path without the software necessarily triggering a blocking error code.
- Preparation and cleaving in a controlled environmentStrip the 250-micrometre primary coating over 30 to 40 millimetres using a calibrated stripping tool. Clean the bare silica with a lint-free wipe saturated with 99% isopropyl alcohol. Perform the cleave at a constant length of 10 millimetres, verifying that the diamond blade cuts cleanly without chipping.
- Positioning and fusion arc inspectionPlace the cleaned fibres into the bottom of the V-grooves without bumping the electrodes. Start the automatic core-to-core alignment cycle. Monitor the display screen: immediately reject any cut with an angle exceeding 0.8 degrees or showing a suspicious shadow on the butting faces before the discharge arc.
- Sleeve protection and tray coilingSlide the heat-shrink sleeve centred over the splice point without rotating the strand. Position the assembly completely flat in the heating sleeve. Immediately upon removal, allow the smoove to cool on the dedicated metal cooling tray before coiling the excess slack in the cassette, strictly maintaining a bend radius greater than 30 millimetres.
Investigation and rework protocol in case of measurement discrepancy
When an optical link exhibits excessive attenuation during photometric or reflectometric acceptance testing while the splicer displayed a flattering value of 0.02 dB, the installer must not dispute the test results. The internal estimate provided by the automatic core-alignment splicer relies entirely on a geometric model calculated by its video cameras. This algorithm does not quantify any actual light flux. It completely overlooks mode field diameter variations between fibres from different manufacturers, micro-bends in the cassette, and compression stresses exerted by the heat-shrink splice protector. Art. 364 CO subjects the contractor's liability for the execution of the work to the same rules as that of an employee, which requires handing over a network whose actual physical performance complies with the project specifications.
- Validate the measurement using bidirectional reflectometryReflectometric analysis must be carried out from each end of the link at 1310 nm and 1550 nm. A geometric core disparity causes false amplification in one direction and an exaggerated apparent loss in the other. The bidirectional arithmetic mean constitutes the only indisputable physical value of the splice.
- Inspect the splice protection and coilingGently lift the heat-shrink sleeve from its slot in the cassette. An off-centre steel rod or cooling too rapidly prior to insertion into the holder causes permanent axial stress. This mechanical stress produces local light leakage that is particularly pronounced at the 1550 nm wavelength.
- Clean alignment grooves and opticsThoroughly clean the V-grooves and mirrors of the fusion splicer using a lint-free swab moistened with isopropyl alcohol. Microscopic dust deflects the mechanical seating of the glass, misleading the image analysis and causing the machine to display perfect alignment that is nevertheless physically misaligned.
- Completely redo the jointCut out the suspect splice, strip the primary coating afresh over 35 mm, clean the fibre with the appropriate solvent, and cleave at an angle under 1°. Restart the fusion cycle, then shrink the protective sleeve without twisting the strand when placing it in the oven.
Frequently asked questions
Why does the splicer display 0.00 dB or 0.02 dB when the OTDR measures 0.25 dB?
The splicer does not measure transmitted light; it observes the outer silhouette of the fibres using cameras and computes a theoretical geometric core alignment. If the two joined fibres have different mode field diameters, or if the cleave angle is imperfect, the fusion introduces optical diffraction that is invisible on the screen. Only a bidirectional optical time-domain reflectometer (OTDR) analyses actual backscattered optical energy and accurately quantifies the physical attenuation passing through the splice.
Why does an apparent gain (gainer) sometimes appear at a splice on an OTDR trace?
This phenomenon occurs when joining two fibres whose backscatter coefficients or core diameters differ slightly. When the signal travels from a narrower core into a wider core, the instrument captures more backscattered light and plots a positive ascending step. This is not an energy gain, but a measurement optical illusion. The true value of the splice is obtained solely by taking the arithmetic mean of the bidirectional measurements.
What role does cleave quality play in the divergence between estimated and actual loss?
A poor cleave angle exceeding 1 degree forces the splicer to extend the electric arc discharge to compensate for end-face irregularities during core alignment. Although the machine algorithm recalculates an apparently correct silhouette after fusion, this excessive melting of the glass produces internal geometric deformations and microscopic bubbles. These physical defects deflect the light path and create high insertion loss that the splicer's geometric estimate cannot detect.
How can installing the splice protector sleeve degrade a successful splice?
The heat-shrink sleeve protects the bare fibre by means of a stainless steel rod. If the protective sleeve is twisted when loaded into the heat-shrink oven, or if it is coiled in the cassette before fully cooling, the sleeve exerts a bending force on the silica. This mechanical stress creates a permanent micro-bend. The splice, though displayed at 0.01 dB during fusion, then exhibits critical insertion loss, particularly at 1550 nm.
Texts and standards cited
- CO art. 364obligations of the contractor: liability subject to the same rules as that of an employee; they execute the work in person or under their personal supervision; unless customary or agreed otherwise, they procure the necessary means, machinery, and tools at their own expense.
