|
|
Lesson 4Radars and Light Sabers - How Fibers are Tested |
 |
For those of you who
may aspire to become a Communications Technician or Engineer, this is a fun and interesting
lesson. Pay particular attention to the Lab
Experiment(s). Let There Be Light! Probably the most enjoyable task associated with the installation of fiber optic cables is performing end-to-end quality tests on the completed optical link. It is almost like being in a space war movie. The tools and equipment for this work are pretty awesome. Some of the tools for performing this work are: 1. Visible Light Source (a glorified flashlight with bright red LED…. resembles a light saber) 2. Multi-Wavelength Optical Light Source (Emulates the actual signals used in optical communications) 3. Multi-Wavelength Optical Power Meter (measures the optical signals from the unit above) 4. Optical Time Domain Reflectometer (OTDR- a radar for analyzing optical fibers from the inside) 5. Optical Feature Finder (poor man’s version of the OTDR) To connect the test equipment listed above to the actual fibers requires a short optical "jumper," also called a Fiber Optic Patch Cord. These were covered in Lesson 3 as well, as they are also used to connect fibers to optical transmission systems. However, test jumpers are generally in the neighborhhod of 1 meter to 5 meters in length. It is also critical that the right type of test jumper be utilized, depending on whether the fiber cable is single-mode or multi-mode. Since fiber jumpers get handled on a regular basis, they have layers of protection to keep the optical fiber safe from damage. However, as we discussed in Lesson 3, it is important to avoid short-radius bends while testing, as these will impact our measurements.
Optical test jumpers can become damaged by frequent use if not properly maintained, so it is not uncommon for them to fail to pass light signals (fiber breaks). Thankfully a tool was developed to aide in quick checks for optical jumpers to provide a viual method for testing. The Visible Light Source (VLS) is what I call a light saber, as it is a glorified flashlight with a colored LED (Light Emitting Diode) as a bulb, and a slender shaft that allows mating to a wide range of connector types. In the photo below, we are using a VLS to perform a quick check on an optical test jumper before employing it for testing purposes. While the visual method does not reveal if the jumper has high loss, it does confirm light is getting all the way through. During this test it is critical that the fiber is not aimed at the face or eyes, as the test signal can damage the eyes and lead to blindness. The open end of the jumper is always pointed toward another surface such as a tabletop or the side of a cabinet, etc.
The photo below provides a better view of the VLS and mating of the connectorized jumper to the slender shaft/barrel on the end of the instrument. The VLS is about the size of a small flashlight and a critical tool when the need arises for quick checks of optical jumpers, especially if the ends are 30 to 40 meters away from each other.
When testing optical fibers, it is always good to have many test jumpers with a variety of connector types. Over the years a high number of optical connector types have been installed and there are also a wide variety of connector types on optical test equipment. It is rare to arrive at a test site and have all the same connectors, so it may require a dozen or more types of optical jumpers to connect test equipment to the fiber optic patch panel connectors.
What Did You Expect? The purpose for performing measurements on fiber optic cable links is to validate that the original design criteria was met in the installation phase. When testing communications links, there are two basic metrics of concern: 1. Expected Measured Loss and associated measures 2. Actual Measured Loss and associated measures In the design phase for communications links, engineers will calculate certain parameters to ensure they are within usable limits for the desired transmission equipment to utilize the fiber link. The actual measurements are compared with the expected (calculated) measurements and if they are within allowable deviations the link is qualified for service. Therefore, it is important to have all the calculated measurements on hand when performing the actual tests. For example, if the Expected Measured Loss (EML) was 22dB, but our measurements resulted in an Actual Measured Loss (AML) of 29dB, that variance is so large further investigation is required. If the EML was 22dB and we measured anywhere from 21dB to 23dB, that would be an allowable variance. The key is to know what to expect before making the measurements. This same principle applies to any communications media (copper, fiber, coax, wireless). Everybody Line Up Obviously the first test is to determine not only if light passes from one end to the other, but if the fibers are correctly identified at each end. For example, transmitting a signal from Fiber 1 at the near end should result in a received signal on Fiber 1 at the far end. If we transmit on Fiber 1 at the near end but receive a signal on Fiber 2, that is an error in most cases. There are occasions where a large fiber cable feeds many smaller fiber cables along the route and a detailed splice plan will tell the technician which fiber in the main cable splices to a smaller fiber cable, but for now lets just test a cable that is straight from one end to the other.
The above represents an optical fiber cable that is correctly terminated at both ends. Fiber 1 - 6 are same at both ends. Therefore, connecting a signal to fiber 3 on the left will result in a signal on the patch panel connector labeled 3 on the opposite end. The fibers are tested by simultaneously moving the transmitter and receiver to the same fiber position at each end. Obviously the technicians must coordinate (talk to each other) to make sure they are testing the same fiber end-to-end, but will presume the fiber labels are correct unless proven otherwise.
The above represents an optical fiber cable that is incorrectly terminated at the right end fiber distribution panel. Fibers 1, 4 & 6 are correct end-to-end, but in error, fiber 2 is connected to position 5, fiber 3 is connected to position 2, and fiber 5 is connected to position 3 in the fiber distribution panel. Therefore, when transmitting a signal from the left fiber 3, no signal is available on position 3 on the right. The technician would presume that fiber 3 is broken (open) between both ends when in reality it is merely out of place in the fiber distribution panel. In this case, the technician has several options for identifying the problem: 1. Test all remaining fibers to see if any other positions appear faulty. If more than one fiber exhibits this same trouble (no signal), it is possible the fibers are merely in the wrong positions at one end or the other. 2. While a signal is still applied at the far end fiber 3, move the power meter to each fiber on the near end to see if a signal appears on another position. In this case, a signal would appear on position 2, so the technician would know the fiber is good end-to-end, but is not inserted into the correct position in the fiber distribution panel. 3. Utilize an Optical Time Domain Reflectometer (OTDR), a radar instrument to measure the length of each fiber. If the measured lengths are nearly identical, any break is very close to the far end, or the fibers are just out of position in the fiber distribution panel. We will view actual OTDR tracings later in this lesson. Send Me Some Light! Once we have verified that all fibers are labeled correctly at both ends (1, 2, 3, 4, etc.), we will conduct a series of tests where we transmit a series of signals from the near end and receive them at the far end to be recorded in a spreadsheet. This test is called the “Insertion Loss Test”, meaning the actual fiber link inserts a loss between the transmitter and receiver. For example, assuming any field splices exhibit minimal loss, singlemode fiber presents a loss of 0.25 dB per kilometer and fiber connectors can also add up to 0.5 dB at each end. Rather than connecting the same test sets to the fiber’s multiple times, the Insertion Loss Test may serve to satisfy the labeling test described above (Fiber 1 is Fiber 1 at both ends, etc.). To perform an Insertion Loss Test, two test sets are required: 1. Multi-Wavelength Optical Light Source (a transmitter) 2. Multi-Wavelength Optical Power Meter (a receiver)
The purpose for using Multi-Wavelength meters is to validate several “windows” of bandwidth in the fiber link. For example, your car radio allows you to select one of several potential radio stations for listening. In other words, you are tuning the radio to a specific frequency assigned to the radio station of choice. Similarly, the fiber optic link offers a few differing frequency bands and we must transmit and measure signals in all the normal bands. Fiber optic cables normally pass optical signal bands just fine, but if a fiber has a microbend or macrobend of enough severity it can attenuate some frequency bands more than others, so we must compare the measured results for at least the three most common bands. Mr. Hertz, What is my Wavelength? Let’s get our terminology straightened out. In the late 1800's a researcher named Heinrich Rudolf Hertz was able to confirm early theories concerning electromagnetism, so in his honor the term Hertz was designated to represent wave signals. A wave is energy which has a repeated motion that increases over time in one direction, decreases over time in the opposite direction and again returning over time to the level of the original starting point as shown below. The point of reference is how many revolutions of enery occur within one second? In other words, how frequently does the energy reverse itself? Therefore, the signal shown below has a Frequency of 1 Hertz, meaning it completes one revolution in positive, then negative direction in a single second.
The number of revolutions per second is the Frequency of the wave, expressed in Hertz. Therefore, if the wave repeats 10 times per second as shown below, the signal has a Frequency of 10 Hertz.
This same convention is applied to all periodic signals (signals that repeat) and the Frequency is associated with either sound waves, radio waves, light waves and also invisible waves like X-Rays. For example, frequencies between 30 Hertz and 17,000 Hertz are generally considered to be the range of human hearing.
When discussing signals in the audio or radio range, we refer to them as Frequencies. However, once we get into frequencies that are so high they appear as visible light and above, we start referring to them by their wavelengths. To aid our feeble minds, its just easier. Here’s how it works: Wavelength is the distance a signal travels when transmitted. For example: A Frequency of 1 Hertz has a Wavelength of 300,000,000 meters (actually 299792458 meters) A Frequency of 1,000 Hertz has a Wavelength of 300,000 meters (actually 299277.458 meters) A Frequency of 100 megahertz has a wavelength of 3 meters (actually 2.99792458 meters) So, you see that as frequency increases, wavelength decreases, and when we get to frequencies that are so high they become visible light and above, frequencies are incredibly high numbers so we express by their wavelength. In fiberspeak we use the term “nanometers" (nm) for wavelength, meaning 1 Nanometer = 1 billionth of a meter (10 to the minus 9th meters).
The windows of interest in single-mode fiber links are primarily 1310 nanometers, 1490 nanometers, 1550 nanometers, and 1625 nanometers which are out of the visible range to the human eye. If we were calling 1310nm by frequency, we’d have to say 228849204.580153 megahertz. See why expressing in wavelength is easier? The Insertion Loss Test is performed at all the desired wavelengths, then the transmitter and receivers are swapped, reversing the direction of transmission on each fiber. Optical fibers have an interesting trait that in some cases the measured loss at various wavelengths will be different in opposite directions of transmission. This is caused by marginal fusion or mechanical splices and/or microbends or macrobends (crimps) in fibers at splices, or damage to the fiber cable itself (shovel damage, etc). Prior to conducting the actual Insertion Loss tests at various wavelengths, the transmitter and receiver must be connected to verify the exact output of the transmitter. For example, in the photo below, the received 1310nm signal is measuring -8.28 dBm, so that becomes our reference.
When testing with a technican at the far end a measurement of -19.44 dBm @ 1310nm is recorded, but to reveal the true loss of the optical fiber we must subtract our transmitter value. In other words, we will consider our transmitter output of -8.28 dBm to be "zero", so the formula for true loss is: Received Measurement -19.44 dBm minus Transmitter Output -8.28 dBm True loss of fiber @ 1310nm 11.16 dB I know what you are thinking....that is bad math, but the dBm identifies the reading as an absolute, whereas dB is just an ratio of two values. For example, if I were to say to you, "One foot", what comes to your mind? Is it the body part at the end of your leg, or do you envision a certain distance? Without additional information, the term One foot is not very clear in meaning. Now if I say, "Go to the wall and make a mark one foot down from the ceiling", that is an absolute location that we can all see and agree on. It is a specific point on the wall. Likewise, the term dB (deci-Bel) represents the ratio of two differnt power levels. The fiber loss shown above is 11.16 dB, but it does not establish the absolute Power Level (loudness) of the signal, it only means whatever the Power Level (loudness) of the signal transmitted into one end, that signal will lose 11.16 dB by the time it reaches the opposit end of the fiber.
To establish a standard Power Level (loudness level) for test instruments, a signal with 1milli-Watt of power dissipated into a 600-Ohm load is considered 0.00 dBm. All that means is that we have established an absolute quantity for 0.00 dBm so that any meter in the world can quantify the power of an optical signal and achieve the same numerical result. Just like every Voltmeter in the world will measure the exact AC RMS Voltage in the wall outlet in your home and display the same numerical result. When you see the m tacked onto the end of dB (dBm), it is expressing an absolute power level, whereas the simple dB merely expresses the difference between two quantities.
Therefore, when we transmit optical test signals, we send them at a specific, absolute quantity such as -8.28 dBm, and when we measure the received signal, it is also an absolute quantity such as -19.44 dBm. However, to express the loss of the fiber we don't use the m, we just say the fiber has an end-to-end loss of 11.16 dB. Again, that means whatever the level of the transmitted signal, it will experience a loss of 11.16 dB by the time the signal arrives at the far end. If we were able to transmit a test signal at a power level of 0.00 dBm, the meter at the far end would read -11.16 dBm. Over the years most optical light sources transmit signals somewhere between -8 dBm and -12 dBm. Unlike measuring voltage levels where 0.00 Volts equals no signal at all, a 0.00 dBm does have signal power. If you were to measure 0.00 Volts on an oscilloscope it would result in a straight, flat line, meaning there was nothing there. However, if you were to measure a 0.00 dBm signal on an oscilloscope, it would look like this:
Some optical light sources do allow slight adjustment of the output power, but for safety reasons, many do not allow the ouput to be adjusted all the way up to 0.00 dBm, as that is a very strong signal and presents a safety issue for technicians. Strong LED (Light Emitting Diode) and Laser transmitters can easily damage the eye, so never direct an optical test signal toward the eye, not even the Visible Light Source. Damage to the eye from a LED or Laser light source can be permanent and lead to blindness. This is why you should never look into the end of an optical fiber or optical jumper, since normal optical system signals are unseen to the human eye and damaging levels of energy may be present, harming your eyes. When discussing or documenting test signals, it is also best to identify the wavelength of the test signal, so to describe the loss of the optical fiber we would say it has an Insertion Loss of 11.16 dB @ 1310nm (nano-meters). As we continue testing the remaining fibers end-to-end, we would record the Insertion Loss for each wavelength of interest (the wavelengths that are intended to be used by the connected optical transmission equipment). These results are then sent to the Enginnering Department to make sure the proposed optical transmission system will work properly over the intended fibers. Every optical fiber transmission system has a "Loss Budget" which describes the maximum Insertion Loss the system can endure before errors will occur in the optical signals. Engineers design the fiber cable lengths with an additional margin meaning they never install the system at the absolute maximum allowable loss, but give themselves several dB of safety net for optimal system performance. Therefore it is critical that the Insertion Loss measurements are highly accurate, or the system could be installed beyond the recommended loss budget. I See You on My Radar One of the most interesting tools for analysis of optical fibers is the Optical Time Domain Reflectometer (OTDR), a radar for looking inside the length of the optical fiber. Early OTDR's were very large, cumbersome units requiring extreme care for continued proper operation. However, modern units are very compact and the latest OTDR's feature hand-held models.
In much the same way a police or airport radar transmits a series of radio frequency pulses and measures the "events" or reflections to identify speeders or airplanes, the OTDR also transmits a series of optical signals, measuring the reflections and displaying them onto a screen for visual interpretation. The amplitude, duration and wavelenghth of the optical pulses may also be varied for optimal results or to zero in on specific events for closer analysis. Most modern OTDR's also have automatic modes where events are identified and analyzed, displaying pages of results and eliminating guesswork. As light energy travels along the optical fiber, most of the signal moves in a forward direction, but as the signal encounters changes in the impedance or physical attributes of the fiber core, some of the energy is reflected back toward the OTDR. These reflections are then analyzed and if the event is of significant amplitude, further analysis may be required. Normal reflections called "backscatter" are usually low in amplitude and look like noise. Larger events such as the end of the fiber generate large events that are easy to spot. The screen below indicates the fiber length is 24.099Km (79,200 feet!). The scribble along the bottom of the screen after the end of the cable is just noise.
The OTDR screen displays an XY graph where the X axis (along the bottom) is distance in meters or kilometers, and the Y axis along the left-hand side is Amplitude or Power. If there was a perfect fiber that exhibited zero loss of signal, the trace on the screen would be a straight line from left to right at the Amplitude or Power Level at which the pulse was launched. Of course, there are no perfect fibers that exhibit zero loss vs distance, so the trace on the OTDR will trend downward as it continues to the right, and will display non-linearity (rate of change is not constant) when it encounters an event such as a poor splice or connector. Yes, the OTDR can see through fiber distribution panels all the way to the optical transmission system or if the fiber is cross-connected to a distant site. However, the OTDR cannot perform measurements if it is testing toward an optical transmitter that is turned on. The optical transmitter must be disconnected prior to attaching the OTDR to the fiber under test. Older OTDR's had a "dead zone", meaning it could not "see" the first several hundred meters of optical fiber, so technicians used to carry an "Optical Fiber Build-Out" or Dead Zone Kit" which in reality was a small case with several hundred meters of optical fiber. The Build-out was connected with fiber jumpers between the OTDR and the fiber under test. This allowed a technician the ablity to view the area of the fiber closest to the near end. Newer OTDR's have very short dead zones and do not usually require build-out kits.
In the OTDR trace below, there is an event just prior to 1 km (Event A), and the fiber end is just before 3 km (Event B). There is no way to know if the event is a bad splice or a macrobend in the fiber, the loss is not significant so if the end-to-end losses are acceptable, this would likely be left alone. The OTDR has cursors which can bracket the event to display the exact loss of the event which we will see in additional tracings. The cursors also display the distance to the event.
In the trace below, the event has significant loss of 6.67 dB. If this is not a fiber optic patch cable connection, it should definately be examined further. If it is a fiber patch panel, the optical cross-connect jumpers and all connections should be cleaned and retested.
The OTDR is a critical tool for inspecting optical fiber links, revealing potential fault conditions that may eventually disrupt communications. Sign Me Up! Installing, splicing and testing optical fiber networks is a very enjoyable profession with interesting and exciting high-tech tools. When constructing optical networks, you are impacting the lives of millions of customers for voice, video and data communications. This completes Lesson 4 Radars and Light Sabers - How Fibers are Tested OK, now crank up your investigative skills and conduct the Lab Experiment (s) for this lesson by clicking the button at the top right of this page, then proceed to the Lesson Test by clicking the button also at the top right of this page. Don't forget you can order a sample fiber from a real telephone cable for use in some experiments by clicking ORDER FIBER or the button on top right of the Home page. Prices from $1.00 per fiber. Great way to earn extra credits in your next Technical Presentation, Science Fair or Merit Badge! |
  |
| ©2012 ©2019 dB Levels, Inc. All Rights Reserved |
