Fiber optic measurements

Performing measurements in small fiber-optic installations is not a common practice. Many installers implementing this type of networks assume that if they work properly after connecting active devices, then even general attenuation measurements are unnecessary. However, this is not a correct assumption. There is a number of situations in which the connected active devices start to work properly, however, problems will occur after some time of operation of the network or optical path.

An example of such a situation is the case where the devices operating at 1310 nm wavelength are replaced by devices using the wavelength of 1550 nm. It happens that the previously undetected increased attenuation (which is e.g. the consequence of a micro or macro bend of the cable/fiber, a very bad splice, a dirty connector or the use of a poor quality adapter) will cause a significant loss of signal on the receiver side at the higher wavelength.

Another situation is with connections permanently exposed to dirt – the operation of such a network involves the risk of increasing the attenuation of all detachable connections (connectors in adapters) with the passage of operation time and increasing number of connecting/disconnecting/switching operations, especially when a dedicated cleaning tool is not used. In such installations, it is necessary to check the initial attenuation of entire optical paths and ensure that the values do not significantly increase.

According to the above examples, it is worth doing at least basic diagnostics of optical connections - even when it is not required by the investor.

It is worth mentioning that an optical meter can be useful not only for verifying attenuation of optical connections and links. It can often be used to verify correct operation of active devices - i.e. if the signals they generate are in the declared power range.

Cleaning connectors during the implementation of fiber optic installations is obvious for some installers, or an unnecessary procedure for others who think that it only increases installation costs (tools, cleaning cartridges, time necessary to clean a number of connectors). It can be assumed that both sides are right here, but it should be noted that cleaning will never worsen parameters and in some cases can save time needed for troubleshooting.

In systems where the key parameter, in addition to attenuation, is reflectance, i.e. the measure how the connector reflects the incident signal, cleaning is extremely important. Even a small contamination of the connector can significantly worsen its reflectance and cause a significant reflection of the signal towards the transmitter creating problems with its proper functioning.

In smaller local systems, such as fiber optic LAN networks or CCTV monitoring systems based on fiber optic cabling with SFP modules and media converters operating at 1310 nm, the value of connector reflectance is not that critical. What's more, even a greater attenuation on a slightly dirty connector will not usually cause signal disappearance on the other side. The reason is that the equipment used in this type of installations usually offers much higher power budget than generally needed. Many installers performing this type of installations do not use any cleaning tools. It should be presumed that most of them are also not obliged to perform any measurements in their installations. The basic condition that they have to meet is that "the installation works".

When measuring parameters of optical networks, cleaning the connectors is no longer a matter of choice and should always be performed. For measurements, the installer usually uses the same measuring patch cords, which, despite that they are factory-made connectors, get dirty over time. Neglecting this action may result in severe distortions in the measurement results. Moreover, in the case of reflectometric measurements, each time it is absolutely necessary to clean the plug of the launch cable that is connected to the socket of the reflectometer. Repeated insertion and removal of a dirty plug is the most common cause of reflectometer defects, most of which end in a costly replacement of the measuring connector. Damage to the connector is a direct consequence of its soiling. The accumulated dirt can firstly scratch the ferrule surface of the reflectometer connector (in the best case it results in "only" incorrect results), secondly it strongly reflects the impulses generated by the device. If the installer chooses a wide pulse and a large measuring range, then serious damage to the reflectometer is only a matter of time.

Before purchasing any measuring equipment, the installer should answer the basic question: what is the purpose of the planned measurements. The selection of devices and measurement methods will depend on this. Secondly, it is worth to consider specific functions that may be useful at work.

Contrary to appearances, determining the "measurement purpose" is not always easy. Often, the installer being not much experienced in the implementation of fiber optic installations, is to provide a "measurement report" to an even less aware investor. The investor does not exactly know what to demand, the installer does not know what should be found in the measurement report.

There are three main "measurement goals":

Required device: optical power meter.

In this situation, it is assumed that at the input of the installation or on one side of the optical path there is an active device generating a signal at a certain power level (i.e. the target device that will be the source of the signal in the installation). Using the optical power meter, the installer is able to determine the signal level on the other side of the optical path, i.e. in the place of installation of the receiving device. A good analogy for this type of measurement can be measurements made in copper antenna installations, where installers verify by means of a signal level meter whether the signal on subscriber outlets is within the specified range.

An example of optical installations in which the signal level at the receiver is a key issue are so-called passive optical networks (PONs). Another case can be fiber-optic TV antenna systems based on optical transmitters and passive components such as cables, splitters, attenuators and optical receivers.

The most important goal in such networks is to ensure that optical power at the input of the receiver falls within the range specified by the producer. Interestingly, the power often proves too high rather than too low. It is a perfect example of a situation in which there is a little sense to report the attenuation of the optical path alone, since despite excellent passive elements and appropriate measurements, the system may not work. The only exception is a situation when the cabling is laid first and the devices are mounted later. In this case, both parameters (optical path attenuation and receiver signal strength) should be tested separately.

Measurement of optical signal power at the inputs of optical receivers in a SMATV system

Required equipment: optical power meter, stable light source, measurement patch cords, intermediate adapter.

This is by far the most common case - the installer wants to verify that the optical path has been made correctly in terms of permissible total attenuation. The use of an optical power meter and a stable light source is a measurement performed with "transmission method". The measurement report should include the total attenuation of the fiber optic connection for selected wavelengths.
Using the transmission method, the installer should select the appropriate wavelength(s) - depending on the transmission window in which the link is to operate, or according to accepted standards. Without specified requirements, the most comprehensive testing should cover all the wavelengths specific for the type of the fiber used, i.e. 850 nm and 1300 nm for multimode fibers, and 1310 nm, 1550 nm, and 1625 nm in the case of single-mode fibers.

Measurement of optical path attenuation with the transmission method

This method involves connecting a signal source of constant and known power on one side and an optical power meter on the other. The measurement result is simply the difference of the indications of both devices. The measurements most often use the logarithmic scale.
Example: source power level of -5 dBm, meter indication of -10 dBm, so the attenuation of the optical path is 5 dB.

Looking at the diagram above, it can be seen that the measurement result also takes into account the attenuation of the patch cords used. The longer and more extensive the optical path (i.e. containing a large number of splices and detachable connections), the smaller the influence of the patch cord attenuation on the final measurement result. However, when measuring a short fiber optic link terminated on both sides with connectors, it may turn out that the impact of measurement patch cord attenuation will be significant. In the extreme case – when patch cords are dirty or damaged, their attenuation may be responsible for the major part of the total attenuation.

The solution to this problem is the implementation of the so-called calibration of the measuring system (reference measurement). It is worth noting here that some devices may not offer such a possibility – this usually applies to meters for which the measurement of optical power is one of additional functions (multi-purpose meters, power meters built into fiber optic splicers, etc.).

For this purpose the installer can use two patch cords (with connectors compliant with the source of light and power meter, and suitable intermediate optical adapter. The attenuation value obtained in this way on the meter is reset by pressing the "REF" key (L5815 meter) or with the use of other means responsible for this function (e.g. by pressing "dBm" key on the L5816 meter). Then, the normal measurement of the tested optical path will not contain the component caused by the attenuation of the measurement patch cords.

The last issue is the interpretation of the results obtained, which should be compared with theoretical calculations of attenuation for a given optical path. The calculations must take into account the attenuation of the optical fiber for a given wavelength, fusion or mechanical splices and separable connectors. The following values are assumed for individual elements:

Estimating attenuation for the optical path shown the diagram above, we get:

So, the estimated total attenuation is [dB]: 1.0+ 0.4 + 6.7 + 0.52 = 8.62.

The result of the actual measurement should not differ significantly from the theoretical calculations. Of course, some deviations in this respect are acceptable. The measurement performed by the cheapest meters is burdened with an error of +/-0.35 dB or +/-3–5% of the measured value (this data can be found in the meter datasheet). In this particular example, a measurement with the use of the L5816 meter will have a +/-3% error. Therefore, for values close to 9 dB, this error will be 0.45 dB. It means that any final result in the range of 8–9 dB is fully acceptable, as close to the theoretical value. In practice, the margin for accepting results as satisfactory may be even a bit wider.

Determination of attenuation of links and connectors, reflectance of connectors, line length, distances between events in the fiber line, the impact of micro- and macro-bends on transmission parameters

Required equipment: OTDR reflectometer, launch cable.

In a situation where the installer is required to provide more detailed set of parameters than the total attenuation of the optical path, the application of a reflectometer is a must. This advanced device is designed for measuring reflectance (the ratio of reflected optical power to the incident optical power) of connectors and determining the attenuation of all events (disturbances) in the transmission path.

Grandway FHO3000 L5828 reflectometer. Devices of this kind are irreplaceable when looking for the causes of optical path failure, as well as precise location of the failure.

Even though reflectometers are advanced devices that offer many diagnostic advantages, the total attenuation of the optical path should be measured with the transmission method, i.e. using an optical power meter and a stable light source.

Firstly, the method guarantees greater accuracy (as opposed to reflectometry, it measures actual attenuation). Reflectometers average out and analyze a series of measurements. Rather than being measured directly, attenuation values are calculated by the software. More importantly, OTDR measurements do not include the attenuation of the last event in the tested fiber line, i.e. typically the meter connection to the line, which may have a significant impact on total attenuation. This dead zone can be eliminated with the use of a launch cable, which means, however, that the measurement will include the influence of this additional cable-to-line connection.

Secondly, using the transmission method, the system can be calibrated to eliminate the attenuation at the start and the end points of the optical fiber. Every reflectometer is equipped with a power meter module, so there is no need to buy two separate devices.

The results of reflectometric measurements are always presented in two forms: in the form of a reflectogram (figure above) and the so-called "event table". A reflectogram is a graph that shows the power of the optical signal along the entire length of the tested fiber line. It allows the technician to determine what events occur in a given optical path and measure their transmission parameters (such as attenuation and reflectance). The same events are also listed in a table. The image above illustrates the typical events represented by a reflectogram:

Example of a real measurement performed with the Grandway FHO3000 L5828 reflectometer.
The pdf file is available here.

As visible in the diagrams, this type of measurement can provide maximum information about the tested transmission line. It is also the best method of identifying the causes and locations of optical line failures.

In this particular example, the fiber-optic line consists of the following segments and components that generate the detected events:

• 150 meters (150 m and 2x 1.5 m patch cords, total: 153 m) of launch fiber (the 161 m distance to the first event results from the existence of a dead zone – the distance is correctly calculated starting from the 8th meter).
• event on 161 meter, marked by the reflectometer as a splice with 0.26 dB attenuation. In fact, it is a connection of the launch cable with the measured line. This high quality SC/APC connection with very low reflectance and attenuation typical for a splice has been identified just as a splice. In such situations, some installers "improve" the reflectogram by specially using a dirty connector, so that it shows the reflective character and becomes visible on the graph.
• event on 333 meter (172 meters of fiber from the previous event) – a detachable connection with attenuation of 0.14 dB and reflectance of 48.36 dB (the event also includes one splice)
• event on 363 meter (30 meters of fiber from the previous event) – detachable connector with attenuation of 0.32 dB and reflectance of 47.85 dB (the event also includes one splice)
• event on 404 meter (41 meters of fiber from the previous event) – the end of the analyzed optical line.

The amount of information provided by this method is its biggest advantage. To correctly interpret it, the user must have the appropriate knowledge and experience in the field of measurement preparation (setting the appropriate options in the device depending on the type and length of the measured line), as well as in analyzing the results obtained. Unfortunately, automatic modes are in practice not very useful.

It is also worth adding that reflectometric measurements should be performed in two directions. Only such "cross" measurements guarantee obtaining correct values of event attenuation.

Installers starting the adventure with reflectometric measurements must be aware that this method also has its limitations. Not every event will be on the reflectogram. Consecutive events - e.g. a connector and a splice, or two splices made at a short distance from each other can be detected as a single event. This is due to the so-called dead zone – for some time after each event the reflectometer is not able to recognize subsequent events. The zone extends with the rising power of the light pulse injected into the fiber and depends on the quality of the reflectometer.

As an example, the attenuation dead zone (independent of pulse width) of the FHO3000 L5828 reflectometer is 6 meters. The event dead zone depends on the pulse width and in the case of testing short links with pulses from 3 ns to 20 ns, it changes from 1 meter to about 6 meters.
Therefore, in theory the reflectometer will not be able to detect events located closer than 7–12 m apart. However, these are theoretical values - in practice the distances may be slightly larger due to deviating from the manufacturer's assumptions on event reflectance (the worse the reflectance, the larger the dead zone after the event).

The occurrence of the dead zone is the main reason for which launch cables are used. They are connected between the reflectometer and the measured line, so that the correct measurement of the line is possible from its very beginning.

The 150-meter-long L58415 launch cable with SC/PC and SC/APC connectors is dedicated for measuring short- and middle-distance fiber-optic links (up to 20 km) do 100–200 ns.

No events on a reflectogram may result from a very good quality of the individual components of the optical path. Considering the fact that the reflectometer detects events based on the amount of light reflected from the components (reflectance), perfectly made APC connections can be unnoticed by the device. The minimum reflectance value for this type of connector is 55 dB. Good quality components feature the value at 60 dB, or even over 70 dB level. This means that the connector reflects one millionth (60 dB) of incident light, and the reflectometer must register and interpret this situation! Even the highest class equipment can have problems with detecting such small values. A similar situation applies to fusion splices. The fusion-splicing technology is so advanced today that the attenuation of a well-made splice is often close to 0 dB. Such small attenuation changes may not be recorded even by the most sensitive measuring devices.

Inexperienced installers can have problems with interpretation of such cases. They may think that the lack of certain events is due to the use of inadequate measuring equipment or insufficient skills in preparing the measurement. Although such reasons are possible, paradoxically, such a "strange measurement" can be a sign of a well-made installation. The basis of judgement here should be the awareness of actions being taken, knowledge of parameters of the equipment used, and the ability to interpret the results obtained.