Uncertainty of measurement for a fiber optic link using the 1 jumper method

As optical fiber networks handle more traffic and transport speeds increase, testing attenuation to loss budgets as low as 1,9 dB places more demand on the accuracy of test equipment. When measurements are critical and high accuracy becomes a premium, questions around measurement uncertainty are asked. Measurement uncertainty has always been the domain of metrology experts responsible for the calibration of optical test equipment such as an optical power meter (OPM). However, in the last several years, more fiber optic Standards added uncertainty guidelines to their documents. When making attenuation measurements on a fiber optic link, one should expect a measurement uncertainty from the tester supplier. This article provides the measurement uncertainty for a multimode and singlemode attenuation measurement of optical cabling using the DTX-EFM2 and SFM2 modules using the 1-cord reference method.

A fiber optic channel is an optical fiber cabling configuration that includes a longer fiber trunk and relatively short cables attaching to equipment. For testing and certifying attenuation of this network, a light source and power meter can be used. When using a light source and power meter, to measure loss, the first step is to reference the power meter to the light source as “0 dB”. After this, the light source and power meter are connected to opposite ends of the fiber cabling under test.

For both singlemode and multimode testing, the contributions to uncertainty include:

• Light source stability at 0,06 dB
• Power meter linearity at 0,04 dB
• Power meter polarization error at 0,005 dB
• Power meter resolution at 0,005 dB
• Connector insertion loss at 0,02 dB (assumes reference to a large area detector)

There is an assumption that the power meter calibration and wavelength have no effect on uncertainty because measurements are relative not absolute. Also, multimode launch conditions are not included in the calculations.

Using the following work table, the measurement uncertainty can be calculated for both singlemode and multimode. Calculations show a 2 % uncertainty which is 0,09 dB.

In summary, one should expect up to 2% or .09 dB uncertainty when performing fiber optic loss testing per industry standard procedures using the one-cord reference method. In the example of a loss budge of 1,9 dB, the measurement could fall between 1,99 dB to 1,81 dB and will have minimal impact when compared to loss budgets. However, it is important to minimize additional uncertainty introduced by ill advised reference methods and dirty or non compliant test reference cords.

The difference in launch conditions between multimode sources can represent an additional uncertainty of up to 40%. This can be reduced to 10% by using an Encircled Flux compliant test solution such as the DTX-EFM2 fiber adapters.