07 Apr 2013
April 7, 2013

ORL Measurements in field applications

Today’s high-speed systems are comprised of many highly sensitive components, and great care must be taken to ensure that proper operating conditions are created and maintained. Failure to control optical return loss (ORL) in systems will cause high bit error rates resulting from multipath interference, degraded optical signal-to-noise ratio (OSNR) and transmitter instability. This application note will focus on a detailed description of this phenomenon and on the importance of accurately measuring ORL and identifying its main components.

What is ORL?
When light is injected into a fiber-optic component, such as a connector, a multiplexer or the fiber itself, some of the energy is transmitted, some is absorbed and some is reflected. The total light that comes back (i.e., reflected) is what we call ORL.

ORL is caused by two fundamental effects:

  • Rayleigh scattering effect; specifically, the part that goes back to the source point, known as backscattering
  • Fresnel reflections

Rayleigh scattering is intrinsic to the fiber itself. It consists of the light interacting with the density fluctuations of the fiber. It can be caused by a variation in the material density and composition that give rise to a variation of the fiber’s refractive index. This causes part of the wave to escape from the waveguide. The phenomenon is called scattering when the size of the defect is less than one tenth of the wavelength’s incident light, while backscattering refers to the part that is captured in the fiber and that propagates in the backward direction.

Figure 1: Representation of Rayleigh scattering effect

Figure 1: Representation of Rayleigh scattering effect

Because it is intrinsic to the fiber, backscattering cannot be elimated. The intensity will depend on many factors such as the incident light wavelength and the refractive index and length of the fiber, etc. Over long distances, ORL caused by Rayleigh backscattering can be as high as 32 dB. Therefore, it is very important to consider this phenomenon during network design.

Figure 2: Example of Rayleigh backscattering as a function of fiber length for different wavelengths (typical SMF-28 fiber)

Figure 2: Example of Rayleigh backscattering as a function of fiber length for different wavelengths (typical SMF-28 fiber)

As another important component of ORL, Fresnel reflections are also related to a variation in the index of refraction. This penomenon will typically occur at discrete interfaces (connectors, adapters, etc.) as a result of air gaps, misalignment, and mismatched refractive indices. Because it results
from discrete interfaces, Fresnel reflections have to be optimized during fiber and system component installation in order to ensure proper conditions.

Figure 3: Example of Fresnel reflection

Figure 3: Example of Fresnel reflection

Reflected optical power is undesirable for numerous reasons:

  • It contributes to overall power loss
  • High-performance laser transmitters like those used in DWDM systems are very sensitive to reflected light, which can significantly degrade the stability of the laser causing:
    – higher bit error rate (BER) in digital systems
    – lower signal-to-noise ratio (SNR) in analog systems

Analog transmission such as the 1550 nm CATV signal used in FTTx systems (i.e. analog video in a PON) is extremely sensitive to backreflection. All fiber-optic components contribute to ORL in one way or another, but the quality of these components has improved over time. Operators working with older fiber plants will undoubtedly be faced with more ORL challenges than those with newer plants. The table below illustrates the contribution of components to the ORL of a system.

Table 1: Example of the return loss of high-performance connectors

SM-UPCReturn LossSM-APCReturn LossMM-PCReturn Loss
SC55 - 57SC65SC20
ST55 - 57ST65ST20
FC55 - 57FC65FC20
LC55E200065

ORL vs. Reflectance
ORL is generally defined as the ratio of the incident power (Pi) on or entering a DUT to the total power reflected (Pr) by the DUT.
ORL is expressed in dB and is a positive number.

ORL_dB

When measuring discrete reflective events, the refleted light is then called reflectance. Although the definitions of reflectance and ORL are practically the same, reflectance always refers to a single event and, by definition, the value is negative.

ORL Measurement Methods for Field Applications
ORL is a threat to today’s DWDM and high-speed systems. For cable plants, ITU-T G.957 recommends a minimum ORL (including any connectors) of 24 dB and a maximum discrete reflectance of —27 dB for an STM-16/OC-48 system. The ITU-T recommendations G983.3 and G984.2, for PON systems, indicate 32 dB as a minimum ORL value. Of course, the real tolerance is particular to each system manufacturer. Also, since total ORL at one point is dependent on the distance of the different reflective events to this point (because reflected light will be attenuated during its travel back), it is very difficult to establish a specific reflectance standard. As a result, measuring ORL is paramount for network service providers to ensure quality of service (QoS).

Installation and Commissioning
During fiber-span installation and commissioning, ORL should be measured to ensure compliance with internal specifications. These specifications will vary from carrier to carrier and should be known to the technician before performing the measurement.

System Turn-Up
In order to verify compliance with system manufacturer specifications during system turn-up, ORL must be measured from the equipment transmitting to the equipment receiving. It is important not to rely solely on the fiber-span commissioning reports, as these tests do not include any crossconnects that may be present between the multiplexer/demultiplexer and the OSP demarcation.

Troubleshooting
ORL problems are generally identified by system alarms; an ORL measurement can be performed to quickly and accurately verify this information. Also, in order to locate and fix the problem, a technician must be able to locate each reflective event on the span and quickly identify the connection(s) most responsible for the high ORL.

There are many different ways of testing ORL. The method chosen will depend greatly on the scope of the test. For example, the way a field technician tests when commissioning a new span may vary greatly from the way a central office technician tests during system turn-up or troubleshooting. Here are four methods supported by the IEC 61300-3-6 standard:

There are many different ways of testing ORL. The method chosen will depend greatly on the scope of the test. For example, the way a field technician tests when commissioning a new span may vary greatly from the way a central office technician tests during system turn-up or troubleshooting. Here are four methods supported by the IEC 61300-3-6 standard:

  1. Optical Continuous-Wave Reflectometer (OCWR): OCWRs directly measure the incident power and reflected power. This method is very accurate and provides the nearest value to the theoretical definition of ORL. However, it cannot spatially resolve many different reflections on the line. Backreflection meters are based on this approach.
  2. Optical Time-Domain Reflectometer (OTDR): OTDRs measure return loss from reflection points on the optical line with nanometer spatial resolution. Today, most OTDRs also allow an operator to extract an ORL measurement from the OTDR trace.
  3. Optical Low-Coherence Reflectometer (OLCR): OLCRs measure reflection profiles of singlemode optical devices with micrometer spatial resolution.
  4. Optical Frequency-Domain Reflectometer (OFDR): OFDRs measure the return loss of a single optical devices with a centimeter spatial resolution.

The best suited methods for field applications are the OCWR and OTDR.  The general procedures are explained below:

OTDR Test Setup
OTDRs identify and specifically locate individual events along a fiber span. An OTDR test is a single-ended test performed by one technician. The unit
transmits pulsed light signals along a fiber span in which light-scattering occurs due to discontinuities such as connectors, splices, bends, and faults. The OTDR then detects and analyzes the parts of the signals that are returned by Fresnel reflections and Rayleigh backscattering.

These signals, which are detected by the OTDR’s avalanche photodetector (APD), enable technicians to draw a trace of the signal power received versus the distance along the fiber. This measurement is evaluated according to the time since the pulse was launched into the fiber. From this trace and all the discrete reflective events, an OTDR can estimate the total backreflection or ORL of the fiber span.

As the OTDR is designed with a different objective in mind (i.e., locating reflective events), backreflection can be roughly evaluated with the returning pulse. As a result, OTDRs are less accurate with respect to backreflection measurements. They are affected by noise, distance and pulse width. The typical uncertainty of OTDRs in terms of ORL measurements is ±2 dB, and with some manufacturers, this value can be as high as ±4 dB.

Source: EXFO