Ribbon Fiber Macrobend in an Aerial Pedestal Closure
After a 24-fiber ribbon cable is spliced in an aerial pedestal, the closure is closed and the span is tested at 1310 and 1550 nm. All 24 fusion splices had splicer-estimated losses between 0.01 and 0.05 dB. At 1310 nm the worst event is 0.12 dB. At 1550 nm one event at tray 2 reads 0.85 dB, and the same event reads 0.10 dB from the opposite direction; the bidirectional average is 0.48 dB, above the 0.30 dB acceptance limit for this project. The event is not reflective. The closure was assembled at -5°C, and the ribbon in tray 2 makes a tight loop around the tray radius; a tie wrap secures the ribbon bundle. No fibers are broken. The customer needs the link turned up today.
1. What is the most likely cause of the 1550 nm event, and why is it not visible at 1310 nm? Include the significance of the reverse-direction value.
Read the worked answer
The most likely cause is a macrobend in tray 2 caused by the tight ribbon loop and the tie wrap. Bend loss is strongly wavelength-dependent: 1550 nm is much more sensitive to a tight radius than 1310 nm, which is why the event is large at 1550 nm and only 0.12 dB at 1310 nm. The event is non-reflective, which is consistent with a bend rather than an air gap or connector.
The reverse-direction reading of 0.10 dB does not prove the bend is negligible. The true event loss is the bidirectional average, (0.85 + 0.10) / 2 = 0.48 dB, which exceeds the 0.30 dB limit. Backscatter differences can make one direction read much lower, just as they can create apparent negative loss. The wavelength dependence plus the physical routing evidence points to a bend that must be corrected.
2. Describe the correction and the retest sequence before turn-up.
Read the worked answer
Reopen the closure and work tray 2 first. Remove the tie wrap from the ribbon bundle, release any tension, and reroute the ribbon so it follows the tray's specified bend radius without pinching, twisting, or crossing under the lid. Do not pull on individual fibers. Check that the splice protection sleeves are seated flat and that no ribbon fiber is trapped under a tray edge.
Close the closure and retest at both 1310 and 1550 nm in both directions. The acceptance decision is based on the bidirectional average at 1550 nm, not on the improved single-direction value. If the average is now below 0.30 dB and no other event fails, the link can be turned up. Record the before-and-after values and the routing correction so the next crew does not recreate the same bend.
3. If rerouting does not clear the event, what other causes should you investigate, and what would you not do?
Read the worked answer
If rerouting does not clear it, inspect for a fiber pinch under the tray lid, a ribbon split or crossover, a microbend at a clamp, a kinked splice sleeve, or a tray radius that is smaller than the ribbon's rated minimum. Also verify that the event is truly in tray 2 and not a launch-cable artifact by testing from the far end. A visual inspection with a fiber scope or a gentle wiggle test while monitoring the OTDR can localize a pinch.
I would not increase launch power, ignore the event because one direction reads low, or immediately cut and re-splice the fiber. Those actions either mask the loss or add splices without fixing the routing problem. The correct sequence is mechanical inspection and rerouting first, re-splice only after the bend or pinch is ruled out, then bidirectional retest.