Before any power meter or OTDR gets connected to a fiber, the single most common source of poor test results and real-world link failures is a dirty or damaged connector end face, which is why connector inspection and cleaning sits at the very front of the testing Knowledge category rather than being treated as a minor housekeeping step. A fiber core is roughly the width of a human hair or smaller, and a single dust particle, oil residue from a fingerprint, or a small scratch sitting directly across the core can scatter or block enough light to push a connection's loss outside its budget, even though the connector appears perfectly normal without magnification.
A fiber inspection scope, either a handheld unit or a video probe that displays the end face on a screen, is the standard tool for checking connector condition before every test and before every mating. Technicians look for contamination (dust, oil, residue), physical defects (scratches, pits, chips), and improper geometry from a bad polish. Cleaning methods range from simple dry wipe cassettes that mechanically drag contamination off the ferrule tip, to wet-dry cleaning using a fiber-rated cleaning fluid followed by a dry wipe to remove any residue left behind. The IEC has published visual standards for what constitutes an acceptable end face, dividing the ferrule into zones (core, cladding, adhesive, contact) with different defect tolerances for each zone, and while memorizing every zone tolerance is less important for CFOT purposes than developing the habit of inspecting and cleaning before every connection, exam questions often test whether a technician understands that core-zone defects matter far more than defects out near the ferrule's edge.
Cable Tracing, Continuity, and Polarity
Before a cable plant can be tested for loss or characterized with an OTDR, a technician often needs to simply confirm which fiber is which, whether a given strand is continuous end to end, and whether a duplex or multi-fiber connection is wired with correct polarity so that transmit lines up with receive on both ends. Visual fault locators (VFLs), which inject a bright visible red laser into a fiber, are the workhorse tool for this kind of basic tracing and continuity check: because the injected light is visible, a technician can often see it glow through the jacket at a bend, or see it emerge clearly at the far end connector, letting them positively identify which fiber in a bundle corresponds to which position in a patch panel or splice tray.
VFLs also help pinpoint gross faults such as a full fiber break, a bad connector, or a poor mechanical splice, since the injected light will visibly leak out at the exact point of a significant discontinuity, sometimes visible right through the cable jacket depending on cable construction and the severity of the fault. Polarity is a related but distinct concern: a duplex or multi-fiber link needs its transmit fiber on one end to reach the receive port on the other end, and getting this backward, easy to do with unlabeled fiber pairs or a mis-keyed MTP connector, produces a link that will not come up even though every individual fiber tests fine for continuity and loss. Standard methods for maintaining and verifying polarity include using cables and patch panels built to one of the recognized TIA polarity methods (commonly labeled Method A, B, or C in industry documentation) and physically tracing or testing each connection rather than assuming a cable was built correctly.
Optical Power, dB, and dBm
Every fiber optic test ultimately comes back to measuring optical power, and understanding the decibel notation used to express it is foundational to every other testing topic. Optical power itself is measured in absolute terms as dBm, decibels referenced to one milliwatt, where 0 dBm equals 1 milliwatt of power, positive dBm values represent more than a milliwatt, and negative dBm values, the vast majority of what a technician actually measures on a working link, represent power levels below one milliwatt. Because dBm is a logarithmic scale, a swing of 3 dB represents roughly a doubling or halving of actual optical power, a 10 dB change represents a tenfold change in power, and small-looking dB numbers can represent large real-world power differences.
Loss and gain, as distinct from absolute power, are expressed simply in dB, a relative measurement describing the difference between two power levels rather than power referenced to a fixed baseline. A splice loss of 0.05 dB, a connector loss of 0.3 dB, and a total link loss of 4 dB are all relative measurements describing how much power was lost somewhere along the way, and they get added together (not multiplied) when calculating a total loss budget, which is one reason the logarithmic dB scale is convenient for this kind of system-level accounting. A technician who is comfortable converting between dBm readings at each end of a link and the dB loss those readings imply has the core numerical fluency the entire testing domain depends on.
Optical Power Meters and Optical Loss Test Sets
An optical power meter measures the actual optical power present at a fiber end, expressed in dBm, and is the basic tool for confirming a transmitter is putting out expected power or a receiver is getting enough power to function. Used alone, a power meter tells a technician what power is present, but not directly how much loss the cable plant itself introduced, since that requires knowing the power that went in as well as the power that came out.
An optical loss test set (OLTS) solves this by pairing a calibrated light source with a power meter, typically used as a matched pair at each end of the link under test. The source injects a known power level at one end, and the meter at the far end measures what actually arrives, letting the technician calculate the total insertion loss of everything between those two points: cable attenuation, splices, and connectors combined. This is the standard method for cable plant acceptance testing specified in most industry standards, typically performed at both 1310 and 1550 nm for singlemode fiber, or 850 and 1300 nm for multimode fiber, since loss varies by wavelength and a link that passes at one wavelength is not guaranteed to pass at another. A technician needs reference-grade test jumper cables in known-good condition to get an accurate OLTS reading, since a bad reference cable introduces its own loss that gets misattributed to the cable plant under test.
OTDR Testing and Fiber Characterization
Where an OLTS gives a single end-to-end loss number, an optical time domain reflectometer (OTDR) provides a much more detailed picture by sending a series of short light pulses down the fiber and measuring the tiny amount of light that scatters backward (Rayleigh backscatter) continuously along the fiber's length, along with any Fresnel reflection at discrete events like connectors, mechanical splices, and the far end of the fiber. By measuring the time delay and intensity of this returning light, an OTDR builds a trace showing loss as a function of distance along the fiber, letting a technician see the location and loss contribution of every splice and connector along the route, along with the overall fiber attenuation rate between events.
OTDR testing is essential for troubleshooting (locating a specific fault's distance from the test end), for acceptance testing on new outside plant construction (confirming every splice and connector performs as expected and no unexpected events exist), and for fiber characterization more broadly, since a skilled technician can read an OTDR trace to identify things like a fiber section with elevated attenuation, a bend causing excess loss, or reflectance from a poor connection. OTDR interpretation takes real practice: dead zones near the instrument can hide events close to the test end, and certain trace artifacts (ghosts from strong reflections, or apparent gain at a splice between fibers with different backscatter coefficients) can mislead an inexperienced reader if not properly understood. Dispersion test sets serve a more specialized characterization role, measuring chromatic and polarization mode dispersion on long-distance, high-bit-rate singlemode links where dispersion, not just loss, determines whether a system will perform at its designed data rate.