Testing multimode fiber cabling in high density environments requires a specialized OTDR capable of testing closely spaced connectors. Frequently, these connectors have high insertion loss and high reflectance. As a result, testing with an OTDR becomes difficult for all but the OTDRs with the highest spatial resolution.
At the heart of this type of OTDR are two components, a pulsed laser and avalanche photodiode (APD). The design of the electronics and more importantly, the type of APD selected, determines the deadzone performance.
All OTDR suppliers provide a deadzone specification. However, there are several things to consider when reviewing deadzones. First, the conditions under which the deadzone is specified must be considered. Second, how the deadzone changes as reflectance increases is important – something suppliers don’t specify. And third, what can one expect with deadzone performance in a real world fiber network.
As shown in Figure 1, the attenuation deadzone (ADZ) is defined as the distance, usually for a single “good” connector reflective event, between the rising edge of the pulse to the 0.5 dB deviation from a straight line fit to the backscatter level. The backscatter level is the sloping line on the trace that provides the fiber attenuation value. This deadzone specification is usually given under best case conditions such as the shortest pulse width and best case connector reflectance.
The purpose of the ADZ specification is to provide an indication of the distance after a connector at which an accurate loss measurement can be made. From this definition, there might be an expectation that a patch cord, the length of the deadzone, can be concatenated to a previous connector to make a loss measurement. In reality, this may not be true.
The purpose of the EDZ specification is to provide an indication of the distance after a connector at which an accurate length measurement can be made. From this definition there might be an expectation that a patch cord, the length of the deadzone, can be concatenated to a previous connector to make a length measurement. This is usually only true if both connectors both meet the criteria for the conditions under which the EDZ is specified (i.e. -45 dB reflectance). When the reflectance changes for either connector, the definition is no longer valid and the deadzone increases.
For OTDRs, one should expect deadzone specifications to be limited to near end measurements under stated conditions. Deadzones should not be expected to remain constant with measurement length. Deadzones are a function of emitted pulses of finite width that become wider as measurement length increases (wide pulses are used for longer length measurements). Deadzones increase with reflectance in all but a few cases discussed later. Deadzone specifications are provided so a user might compare OTDR performance. However, the deadzone spec is defined for a single event, not as a network test.
More sophisticated OTDRs not only display a trace and event table, they also provide a graphical “map” of the fiber cabling under test. Mapping was first introduced in Premises-based OTDRs but has now become popular by many suppliers. The mapping information is derived from the same analysis used to generate an event table but is shown as a more easy to use schematic. Analyzer software is pushed to its limits when closely spaced connectors are to be measured, especially if each have different reflectances (i.e. clean connector followed by a scratched connector).An example of an event deadzone expectation might be as follows. The event deadzone is specified as 1 meter. The fiber network has a 1 meter patch cord in the middle of two longer lengths. The user expects the OTDR to locate and identify the 1 meter patch cord and possibly make loss and reflectance measurements. The OTDR will be able to measure the length of the 1 meter patch cord only if the conditions of the specification have been satisfied; both reflectances must be within the restricted limits as defined within the specification footnote. Keep in mind that event deadzones only locate the reflectance peaks, so loss measurements are not possible.
In another example, the attenuation deadzone claim is 2 meters. The fiber network has a 2 meter patch cord in the middle of two long lengths. The user expects to be able to measure the loss of the patch cord. If there is sufficient backscatter after each reflection as shown, the OTDR will be able to make the measurement.In Figure 3, the first connector of the 1.94 meter length is identified with location, loss, and reflectance. Since two connectors are spaced close together, there may be limited backscatter after the first pulse. The second pulse may merge into the backscatter of the first pulse. As a result, the loss is measured from the backscatter of the second pulse to the end of backscatter at the front of the first pulse. Therefore, what is actually being measured is the loss of two pulses.
Figure 6 shows an example of two connectors place close together that might very well be the length of the attenuation deadzone specification. A skilled OTDR user might be able to make a manual measurement of the attenuation deadzones of both pulses. The analysis software, on the other hand, might measure the loss of the first connector (pulse) by taking the difference in backscatter from the start of the first pulse to the end of the second pulse.
The following two graphs are composed from data taken from two OTDRs using either a Si APD or an InGaAs APD. The InGaAs data, although taken at 1550 nm, would have the same type of deadzone response at any wavelength including 850 nm. Similar pulse widths were used for 850 nm and 1310 nm.The data below shows the relationship between deadzones and connector reflectance at 1550 nm using an InGaAs photodiode commonly used in OTDRs. The first graph in Figure 7 shows the 850 nm event deadzone (EDZ) and attenuation deadzone (ADZ) as the reflectance increases from a value for a typical UPC connector (-45 dB) to a connector with high reflectance (i.e. dirty connector).
The data shows that the EDZ is not affected by reflectance. This is because the measurement is made below a non-saturated peak. If the peak became saturated (i.e. “flat top”), the EDZ would increase but this is related to the design of the OTDR. For the ADZ, there is a gradual increase from 2 meters to 2.75 meters but at -26 dB reflectance, there is deflection and the ADZ increases to 4.5 meters when reflectance is -25 dB. Despite the increase in ADZ over this range, it is much better than what the ADZ might be using an InGaAs APD as shown in Figure 8.
While there might be a premium for high performance OTDRs, for the user comparing specifications from OTDR suppliers, it is not evident if a Si APD is used unless an evaluation under high reflectance is made.