High-performance optical return loss measurement - new return loss module for HP 9153A lightwave multimeter - Technical

Hewlett-Packard Journal, Feb, 1993 by Siegmar Schmidt

Although high-performance optical return loss measurements pose some tough technical challenges for fiber optics engineers, careful selection of appropriate test equipment and correct setup make precise measurements readily achievable. A new return loss module for the HP 8153A lightwave multimeter simplifies these measurements.

High-speed digital systems and analog cable television systems using fiber-optic media need to employ laser sources with narrow linewidths, such as distributed feedback (DFB) lasers. The natrower the linewidth of a laser, the more sensitive to backreflection it is. If a component reflects too much light back to the laser transmitter, the modulation characteristics and the spectrum of the laser change. This degrades performance in both digital and analog systems. Fabry-Perot lasers can also be affected, depending on their quality and the application. Therefore, reflection measurements on such components become more and more important in R&D and manufacturing.

Reflections in optical systems can come from a variety of sources. Fresnel reflections occur at connectors, splices, fiber ends, bulk optic interfaces, and detector surfaces. For example, reflection from a noncontact fiber connection (a Fabry-Perot resonator) can theoretically vary from 0% to 15% depending on the distance between the fiber endfaces, because of constructive and destructive interference.

In fiber optics it is common to use a logarithmic quantity called the return loss for measuring such optical reflections. It is defined as ten times the negative logarithm of the reflectivity R, which is the ratio of the backreflected optical power to the incoming optical power at a component's input.

RL in dB -- -101ogR -- -1010g[(P.sub.back]/[P.sub.in]). (1)

This article describes a method for measuring optical return loss that is both highly accurate and easy to perform.

Return Loss Measurement Method The following equipment is needed to measure the return loss of an optical component:

A laser source

A fiber-optic coupler

An optical power meter.

The light emitted by the laser source is guided to the test port using the fiber-optic coupler. The light power that is reflected back is guided to the detector of the power meter. Fig. 1 shows the experimental setup. The test port for connector measurements is a master connector to which the connectors under test are connected. For measurements on pigtailed components, the test port is the bare fiber end, onto which the components are spliced. The slanted connector and the splice in Fig. 1 represent discontinuities between the coupler and the test port that contribute to the parasitic reflections. The parameters Ki and K2 are the coupling coefficients of the fiber coupler in the forward and reverse directions, respectively. PL is the output power of the laser source, and PD is the power reaching the detector.

Three power levels have to be measured to determine the return loss of a device under test. The first value is the detector power Pref for a known reference reflection Rref attached at the test port. The second value is the detector power caused by parasitic reflections of the setup itself (Pp). The third power level is the detector power with the device under test attached (Pmeas). Fig. 2 shows the first three steps of the return loss measurement.

Step 1. Measuring the detector power with reference reflection attached (Pref). A known reflectance of Rref at the coupler output port is used to effect an absolute calibration. The optical power at the detector is:

[P.sub.ref] = ([P.sub.L K.sub.1 K.sub.2])[R.sub.ref] [P.sub.p] (2)

Pref is sum of the detector power caused by the reference reflection and the power offset Pp at the detector caused by the parasitic reflections.

Step 2. Measuring the detoctor power of the unwanted parasitic reflections (Pp). The unwanted parasitic reflections Pp can be measured by terminating the fiber close to the front of the port. This can be done by wrapping the fiber five times around the shaft of a screwdriver or similar object with a diameter of approximately 5 mm. This creates some 80 dB of insertion loss in the fiber and therefore a two-way loss of some 160 dB. Since the reflectivity at the test port is zero under this condition, all the power reaching the detector now comes from the parasitic reflections. Pp includes the coupler directivity (direct coupling from port 1 to port 2), reflections from events between the coupler and the test port, and the light scattered back by the fiber itself. The last item should not be forgotten, because 9 meters of singlemode fiber corresponds to a return loss value of approximately 60 dB, and even 1 meter corresponds to some 69 dB of return loss.

Step 3. Measuring the detector power with the device under test attached (Pmeas)- In this step, the device under test is attached to the test port and the fiber is terminated closely behind it. The detector power is now determined by the sum of the power reflected from the device under test and the parasitics.