A technician records −9.0 dBm at the far panel. Can the link pass a 2.5 dB loss allowance? The answer depends on what reached the meter during the reference measurement. Two crews can record exactly the same far-end power and calculate different losses. The reference setup is what makes their results comparable.
Table of Contents
ToggleStart with two power readings
A meter in dBm reports optical power relative to one milliwatt. Zero dBm corresponds to 1 mW. A negative reading simply means the power is below that level. A passive link’s insertion loss compares two power levels at the same wavelength and is expressed in dB.
Using the positive-loss convention, subtract the received reading from the reference reading. Keep the minus signs. If the instrument displays relative power as received minus reference, attenuation appears with the opposite sign. Record which convention the instrument uses.
Set the reference planes before testing
Identify the two ends of the installed segment and which end connections the project requires the test to include. Select the specified reference-cord method and compatible leads. Different reference arrangements can include different connector losses. A nominal transmitter output taken from a data sheet cannot substitute for the reference reading at the meter.
Field check. Keep the launch connection at the source undisturbed after referencing. If it is disconnected, repeat the reference procedure. Select the meter calibration wavelength that matches the source and record separate references for each wavelength tested.
Work through a borderline result
Assume the correct reference is −6.0 dBm and received power is −9.0 dBm. The loss is −6.0 − (−9.0) = 3.0 dB. Against an illustrative project limit of 2.5 dB, that exceeds the allowance by 0.5 dB. If the reference was inadvertently stored at −7.0 dBm, the same far-end reading yields 2.0 dB and appears to pass.
The one-decibel difference came entirely from the reference. These numbers illustrate arithmetic, not an instrument specification or a universal limit. Apply the project’s decision rule where measurement uncertainty matters.
Investigate a suspect reference
If a passive link appears to show gain, or several fibers shift by a similar amount, check the setup before disturbing the installed path. Compare the recorded wavelength and source setting, the identity of the launch lead and the method used to set zero. Inspect the test interfaces and check the leads against the project procedure. Record the original result before resetting anything.
Re-establish the reference after correcting the setup and repeat one identified fiber with the same endpoints. If the discrepancy follows one lead when it is substituted and referenced again, investigate the lead. If it remains with the installed fiber, inspect that route. An end-to-end power reading alone cannot locate a poor splice.
Leave a reproducible test record
For each result, record the fiber ID, direction, endpoints, wavelength, reference method, reference power, received power, calculated loss and applicable allowance. Record when the reference was set and which reference cords were used. If a reference check fails at the end of a batch, identify results collected since the last successful check and repeat them under the agreed procedure.
Related field procedures
Review the fiber loss measurement reference for the basis of attenuation measurement. For suspect interfaces, follow the connector inspection and cleaning guide. If testing points to a localized splice event, the bidirectional splice measurement example explains how to interpret the trace.