FOA CFOS/FC cheat sheet

Inspect at 200x-400x magnification: 200x for multimode ferrules, 400x for single-mode ferrules and for judging scratches and pits.

IEC 61300-3-35 is the reference standard that defines zone boundaries (core, cladding, adhesive, contact) and pass/fail defect sizes for each zone.

Zone A (core) must be defect-free for single-mode; any scratch, pit, or particle in the core zone is a fail.

Zone B (cladding) allows small defects; Zone C (adhesive) and Zone D (contact) permit progressively larger defects and more of them.

Always inspect the endface before mating; inspect both connectors and the bulkhead adapter, since contamination transfers from a dirty plug to a clean jack.

Dry cleaning with a lint-free wipe or one-click cleaner is the default; wet-dry (99% isopropyl alcohol then dry) is used for stubborn residue and must be followed by a dry pass.

Never use compressed air from a can (propellant residue), cotton swabs on ferrules, or the same wipe twice; a dirty wipe scratches the polish.

Inspect after cleaning to confirm removal; a connector that fails after three cleaning attempts is rejected and replaced.

Inspect patch cords, pigtails, and adapters on every handling cycle; a capped connector is not automatically a clean connector.

A visible-light source (VFL, typically 650 nm red) is used for continuity checks, fiber identification, and locating sharp bends and breaks.

Red light escaping at a tight bend or crack indicates a macrobend or fracture; the glow is the location marker.

VFL range is limited by fiber attenuation at 650 nm (high loss); useful typically to a few kilometers in single-mode and less in multimode.

Never look into a live fiber or a connector end with the eye, even with a VFL; use a fiber identifier or an inspection scope instead.

Tone generators plus fiber identifiers detect traffic direction and presence without disconnecting the fiber.

Trace from the known end: inject at the patch panel, confirm at the far end, then label both ends to prevent future misidentification.

A VFL that shows no light at the far end indicates a break, a disconnected adapter, or attenuation too high for the wavelength; confirm with an OTDR.

Use VFL only on dark fiber; injecting visible light into a live link disrupts service.

Power meters read in dBm: P(dBm) = 10 log10(P(mW)/1 mW); 0 dBm = 1 mW, -10 dBm = 0.1 mW, +3 dBm = 2 mW.

Set the meter's calibration wavelength to the source wavelength; a 1310 nm source read on a 1550 nm setting gives a small but real error.

Warm up the source and meter (typically 5-15 minutes) before recording, and zero the meter if it has a zero function.

Use the correct connector adapter and a clean, undamaged ferrule; a dirty connector can add several dB of error.

Measure reference power first (source directly into meter), then measure at the far end; the difference is the link loss.

Typical expected receive power on a healthy link: -8 to -28 dBm for most transceivers; check the module's specified receive range.

A reading near the meter's saturation limit (e.g., 0 dBm) or near its noise floor (e.g., -60 dBm) is suspect; change range or attenuate.

Record the source wavelength, meter serial, and calibration date with every measurement for traceability.

Insertion loss = reference power (dBm) minus measured power (dBm) at the far end; report as a positive dB number.

One-jumper reference: source to meter through one launch jumper, then insert the link; two-jumper adds a second reference jumper and is used for patch-panel-to-patch-panel links.

The reference method must match the link under test: one-jumper for a link with a single connector at each end, two-jumper for a link with patch panels at both ends.

Keep the reference jumper and adapters clean and unchanged between reference and measurement; any change invalidates the reference.

Test at the operating wavelength(s): 850/1300 nm for multimode, 1310/1550 nm for single-mode, 1490/1550 nm for PON.

Loss budget check: measured loss must be below the calculated budget (fiber loss + connector loss + splice loss + margin).

Use an OTDR or a power meter/light source pair; the meter method gives end-to-end loss, the OTDR gives per-event loss.

Report per-link results with wavelength, method, and date; a link that passes at 1310 nm may fail at 1550 nm if macrobends are present.

OTDR sends a pulse and measures backscattered and reflected light versus time; distance = (c/n) x t / 2.

Dead zone: attenuation dead zone (events merge) and event dead zone (events indistinguishable); a shorter pulse reduces dead zones but lowers dynamic range.

Dynamic range (dB) sets the maximum measurable fiber length; choose pulse width so the far-end reflection is above the noise floor.

Reflection (Fresnel) at a connector appears as a spike; a splice appears as a step with little or no reflection.

Set the IOR (index of refraction) to the fiber's actual value; a 1% IOR error gives about a 1% distance error.

Use launch and receive (tail) fibers to move the first and last connector out of the dead zone and to measure end-to-end loss.

Bidirectional testing and averaging removes the directional bias of backscatter coefficient differences between fibers.

A gainer (negative loss splice) indicates a backscatter coefficient mismatch, not a real gain; average both directions to correct.

OTDR cannot measure through a live PON splitter without a filter or a dedicated test port; use a 1625/1650 nm wavelength or a PON-optimized OTDR.

Optical power meter: measures absolute power in dBm or microwatts; paired with a light source for loss testing.

Light source (LED/laser): stable output at 850/1300/1310/1550 nm; must match the meter's calibration wavelength.

OTDR: locates events, splices, connectors, and breaks; gives distance, loss, and reflectance per event.

VFL: 650 nm visible laser for continuity and fault location on dark fiber.

Fiber identifier: clamps on a fiber and detects traffic/tone without bending or disconnecting.

Optical spectrum analyzer (OSA): measures wavelength content, used for CWDM/DWDM channel checks.

Visual inspection scope: 200x/400x magnification for endface pass/fail per IEC 61300-3-35.

Optical return loss (ORL) meter: measures reflected power; important for analog and high-speed links.

Calibrate or verify instruments against a known reference before a test campaign; record serial numbers and calibration dates.

Link loss budget = fiber attenuation (dB/km x km) + connector loss (number x dB) + splice loss (number x dB) + margin.

Typical attenuation: 850 nm multimode ~3.0 dB/km, 1300 nm multimode ~1.0 dB/km, 1310 nm single-mode ~0.35 dB/km, 1550 nm ~0.20 dB/km.

Typical connector loss: 0.3-0.5 dB per mated pair (0.75 dB worst case for budget planning).

Typical fusion splice loss: 0.1 dB each (0.3 dB for mechanical splices); use worst-case values in the budget.

Include a design margin (commonly 3 dB) for aging, repairs, and temperature effects.

Budget must be less than the transceiver's power budget (Tx power minus Rx sensitivity).

Example: 2 km single-mode at 1310 nm, 2 connectors, 1 splice: 0.7 + 0.6 + 0.1 + 3 = 4.4 dB.

Verify measured insertion loss is below the calculated budget before accepting the link.

Splitters in PON add their split ratio loss: 1:2 = 3.5 dB, 1:4 = 7.2 dB, 1:8 = 10.5 dB, 1:16 = 13.8 dB, 1:32 = 17.0 dB.

Characterization measures the fiber's optical properties beyond simple loss: attenuation, bandwidth, chromatic dispersion, PMD, and cutoff wavelength.

Chromatic dispersion (ps/nm/km) broadens pulses at high bit rates; zero-dispersion wavelength is ~1310 nm in standard single-mode.

PMD (polarization mode dispersion, ps/sqrt(km)) limits 10G+ links; test with a PMD analyzer or an OTDR with PMD option.

Multimode bandwidth is specified as MHz·km (e.g., OM3 2000 MHz·km at 850 nm); modal dispersion limits distance at high rates.

Cutoff wavelength ensures single-mode operation above the specified wavelength; test with a cutoff wavelength test set.

Attenuation vs. wavelength is measured with an OSA or a spectral attenuation test set, not just a single-wavelength meter.

Characterization results feed into link design; a link that passes loss testing may still fail due to dispersion or bandwidth limits.

Record IOR, numerical aperture, and core/cladding dimensions for multimode links; these affect launch conditions and measurement.

PON wavelengths: downstream 1490 nm, upstream 1310 nm, video overlay 1550 nm; test ports often use 1625/1650 nm to avoid live traffic.

A live PON carries traffic; never connect an OTDR directly to a live feeder without a wavelength-selective filter or a dedicated test port.

PON splitter loss is inherent: 1:32 splitter adds ~17 dB, so the OTDR must have enough dynamic range and a PON-optimized pulse.

Test from the OLT side through the splitter with a PON OTDR; the splitter appears as a high-loss event, and individual drops are hard to separate.

Test each drop from the ONT side or at the splitter output to isolate a single subscriber's fiber.

Power meter checks at the ONT: downstream power must be within the ONT's receive range (commonly -8 to -27 dBm for Class B+).

Use a 1310/1490 nm filtered power meter to separate upstream and downstream; unfiltered meters read the sum.

Loss budget for PON includes splitter loss, feeder loss, distribution loss, and connector/splice losses; compare to the class budget (e.g., Class B+ 28 dB).

Label and document splitter ports and drop fibers; mislabeling is a common cause of PON troubleshooting errors.

Pre-test: confirm fiber is dark, verify instrument calibration, clean all connectors and adapters, and record IOR and wavelengths.

Reference: establish the reference with the exact jumpers and adapters used for the test; do not change them mid-test.

Test: measure at all specified wavelengths, record per-link values, and compare against the calculated budget.

Post-test: clean and cap connectors, restore the link, and file results with serial numbers and dates.

High loss at one connector: clean and re-inspect; if it persists, replace the connector or re-terminate.

High loss at a splice: re-splice; check for a gainer (backscatter mismatch) versus real loss with bidirectional OTDR.

No light at the far end: check source power, connector seating, adapter damage, and fiber continuity with a VFL.

Intermittent loss: suspect a bent fiber, a loose connector, or a damaged adapter; flex-test and inspect.

A link that passes at 1310 nm but fails at 1550 nm suggests a macrobend; inspect for tight bends and re-route.

Uncertainty sources: source stability, meter linearity and calibration, connector repeatability, reference jumper quality, and temperature.

Typical power meter uncertainty is ±0.2 to ±0.5 dB; connector repeatability adds ±0.1 to ±0.5 dB per mating.

A dirty or damaged connector can add several dB of error, far exceeding instrument uncertainty; cleaning is the largest single control.

Reference method mismatch (one-jumper vs. two-jumper) introduces a systematic offset of about one connector loss.

OTDR distance uncertainty comes from IOR error and time-base accuracy; a 1% IOR error gives ~1% distance error.

OTDR loss uncertainty is affected by backscatter coefficient differences; bidirectional averaging reduces it.

Report results with an uncertainty statement (e.g., ±0.5 dB) and compare against the budget with margin.

Repeat measurements and average to reduce random uncertainty; a single reading near a pass/fail threshold is not conclusive.

Keep instruments calibrated and record calibration dates; an out-of-calibration meter invalidates the result.