Insertion loss testing
This domain tests Tier 1 insertion loss measurement with an optical loss test set (OLTS) or source and power meter. The exam expects correct reference method, wavelength, launch conditions, direction, and averaging so the reported loss matches the standard rather than the tester’s convenience.
Reference methods
Insertion loss is the change in received power after the link is inserted, expressed in dB. The standard reference method may be one-jumper, two-jumper, or three-jumper; the choice changes which connector pairs are included in the result. Use the method specified for the job and keep the reference cords clean and stable. For multimode testing, wrap the launch cord several turns around a mandrel (commonly five turns on a 25 mm mandrel) to strip cladding modes and produce a launch condition closer to the cabling standard. Test single-mode at 1310 and 1550 nm, and multimode at 850 and 1300 nm, unless the specification says otherwise.
Direction and averaging
Many cabling standards require testing in both directions and averaging the two results because connector loss is direction-dependent. Record both values, not just the average. Compare the result with the link loss budget, not with a generic “good fiber” number.
Worked example
A single-mode link is tested at 1310 nm with a one-jumper reference. The reference reads 0.00 dB. After the link, the meter reads −2.7 dB in one direction and −3.1 dB in the other. The average is (2.7 + 3.1) / 2 = 2.9 dB. If the budget is 3.5 dB, the link passes with 0.6 dB margin. Reporting only the better direction would overstate the margin.
Traps
- Referencing through a dirty connector, which can make the link loss look lower.
- Using the wrong reference method and including or excluding the wrong connector pairs.
- Omitting the multimode mandrel wrap and launching too many modes.
- Testing only one direction when the standard requires bidirectional averaging.
- Comparing insertion loss with a power budget in dBm instead of a loss budget in dB.