Troubleshoot Link STM-64 using BPA and DCU in OSN 7500II

3. Distance : 96 km (with one site as OTB passthr), as like picture by OTDR below:

 

 

4. Patrol is OSN 7500II    Jatiluhur is OSN 7500II
5. Topology

This is caused of non liniear factor.
BA output is so high (+14db), the distance is so far (about 96 Kilometres), so when the fiber quality is not good, it have a big nonlinear affection on service

A. Pre-troubleshoot:
1. Please prepare:
More XFP that support > 80 km – colored XFP(1550.12nm,193.4THHZ),
More Attenuator (5db, 7 db, 10 db, 15 db) ,
More Patch cord (LC/FC, LC/LC) ,
Core cleaner (alcohol, dust free, etc) ,
2 pcs OPM (in each of site is mandatory to bring) ,
Laser / OTDR >> to get actual fiber loss and distance (mandatory to bring, enough in near site or far site only) ,
Local connection to monitor using U2000
2. Please let familiar about BPA link connection guide
3. BA and PA must be used as one pair. That mean cannot use BA or PA seperately
4. Install fiber connection each of site only for TX direction first without attenuator
5. using OTDR to check OSP/OTB firstly (note: no patch cord in opposite site connect to its OTB during OTDR)
6. Fiber check by OTDR
7. Integrate one direction from TX in each of site

B. Integrate one direction (TX from Jatiluhur to RX Patrol )
Each of site stay focus to one direction only :
First measuring the power TX from Jatiluhur to RX Patrol :
install fiber connection in TX Jatiluhur without attenuator first in Jatiluhur with step by step measure by OPM until Patrol
Write down the power in each of point, each of patch cord, the purpose for avoiding if there is bad fiber also or not
(in advance, we found out that 2 pcs fiber in Patrol was bad/cannot transmit signal, after changed it can transmit)
Put attenuator and measure power to match the range in RX Patrol

D. Step by step troubleshooting 
1. The first one we considered, the board and XFP is ok or not
So we Try to loopback itself in Patrol (SLD64-out to SLD64-in)

2. BA output is 14~15dbm, when the fiber quality is not good, it have a big nonlinear affection on service,
Then we try put 15 db at BA-out to decrease the BA output power, but the received power is -33dbm

E. The reverse direction
measuring the power TX from Patrol to RX Jatiluhur :
Follow the same steps with opposite direction (TX from Jatiluhur to RX Patrol ) ,
install fiber connection in TX patrol without attenuator first in Patrol with step by step measure by OPM until Jatiluhur ,
Write down the power in each of point, each of patch cord, the purpose for avoiding if there is bad fiber also or not
(we found out that 2 pcs fiber in Patrol was bad/cannot transmit signal, after changed it can transmit) ,put attenuator with same range attenuator with first measuring.

F. Result
then in that link used BPA and DCU in source (sample: Jatiluhur) and sink (sample: patrol) site already clear from alarm, and also can be created in NMS U2000

Suggestions and Summary:

From case study above, have conclusion that although now the link don’t have the alarm. But we notice that the received power(-31dbm) is lower than specification(-28dbm). That is mean in the later , the fiber degradation, the link don’t have any margin to work normal and must go to site to decrease the attenuator on.

It is suggested that If the received power of PA is higher than suggested value(-26), please add attenuator on BA out port firstly to decrease the nonlinear affection.