Customers for fiber optic cable installations usually require documentation of test results before accepting and paying for the work. Testing needs to be done carefully to ensure the measurements are accurate and reliable. A definitive repertoire of tests, known as the "essential six," can benefit the inexperienced system engineer. Six basic test procedures measure distance, fiber loss, event loss, link loss, event-return loss and link-return loss. These procedures are implemented at all four levels of fiber operation, including pre-installation, installation and acceptance, maintenance and restoration.
An OTDR can be used to accomplish the six test procedures. In addition, clean connections on the fiber media under test are imperative. All six test procedures will prove inaccurate or impossible to accomplish if the fiber-optic connectors are dirty.
The optical distance between one point and another depends on definition. For example, the distance could be the fiber-cable length between a transmitter and a receiver, or it could be the fiber-cable length between two splices. An OTDR is the test instrument used worldwide to measure optical test events either automatically or manually. Events are detected as disturbances in the OTDR's relatively linear trace display.
To measure optical distance between two points, the OTDR launches a laser-generated light pulse down the fiber at the transmission end of the cable. The instrument then detects the backscatter returned from the fiber and any reflections from shiny surfaces. It measures the time taken by the light pulse to make the round trip on the fiber and calculates that time into distance.
One minor deviation in this test is the difference between real and apparent distance. The optical, or apparent distance, is the distance reading registered on the OTDR, and is always longer than the real distance. One reason for the distance difference results from the undulation of fiber as it resides within loose-tube cable, which adds to its length. Another reason involves buried cable as it winds within a trench, thereby producing a longer optical length.
The backscatter trace is a representation of the fiber itself. The slope of the backscatter trace discloses that less and less light is being reflected back as the length of the fiber increases. This slope represents fiber loss, a manufacturer's specification. Typical fiber-loss measurements are given as the amount of light (in decibels) lost per kilometer. For example, a long-haul telephone fiber might lose 0.15 dB/km, whereas a multimode local area network fiber could lose 3 dB/km. Fiber loss is always measured along a featureless section of backscatter with no events to skew the calculation.
A test event is a disturbance that occurs above or below the backscatter baseline. Splices, connectors, bends and cracks are typical events that produce trace disturbances on the OTDR display. Normally (but not always), an event results in a loss of light. There are two types of events--reflective and non-reflective. The spikes along the baseline indicate a reflection. Because more photons appear and thus exceed the normal backscatter level, a mechanical splice or the end of the fiber is revealed. Other causes of reflections are connectors and fiber cracks.
Events that occur along the fiber become important when a fiber-loss budget is calculated. Only a finite amount of light is launched by the transmitter. Consequently, if the receiver does not receive enough light, a major cable problem has occurred.
Link loss is the total amount of light lost between two points. A link can be the distance between events or between two end points. Total link loss is typically specified when it directly affects the loss budget. If the link loss is a high value, then specific events are consuming light.
Return loss is essentially the light lost because of reflections back toward the transmission or source end. The shiny surfaces of connectors and mechanical splices reflect light. Some of this reflected light returns to the source. Any transmitted light that does not reach the end of the fiber is lost. An OTDR trace displays return loss as the height of a reflection.
Return loss is defined as the ratio in dB of the incident power to the reflected power. Return loss is always expressed as a positive number:
In contrast, reflectance is defined as the ratio of reflected power to the incident power or the inverse of the return-loss formula. When expressed in decibels, reflectance is a negative number. In addition, reflectance can be expressed in terms of density or as a percentage.
In reality, these terms mean noise. The reflected light travels back to the source, reflects off the input and makes another round trip. To a digital system, the reflected light looks like a bit error. To an analog system, such as cable TV, reflected light creates sparkle. The higher the reflection value, the more dramatic the noise level becomes.
Link-return loss is similar to link loss. It is the total amount of reflected light in the link. Therefore, link-return loss is often used as an acceptance test. If the total amount of return loss is below a certain level, the link is assumed not to contain a single event reflecting above specification.
These six essential tests should be used to test fiber during pre-installation, installation and acceptance, and for maintenance andrestoration. A pre-installation test should be performed when fiber-optic cable arrives from the vendor. This receiving type of test is important because it quickly and easily determines product acceptance or rejection before system usage.