For the test of PMD (Polarization Mode Dispersion) we use equipment of EXFO, which consist of the light source FLS 5834 at the far end and at near end the mainframe FTB 400 together with the PMD-Analyzer FTB 5500B. For analyzing PMD the equipment used is the Michelson-Interferometer-Method (GINTY). The dynamic range of this combination is 47 dB and so it’s well prepared for tests on long backbone links. A new feature is also the eliminated auto-correlations peak, tests are now possible from 0ps on, and up to 115ps.
The PMD of a link is not constant, there for, you have the possibility to repeat the tests several times and then the average is calculated. The lengths of breaks between the tests are readily available. The second order PMD will be calculated by the tester automatically.
It becomes more important to test the PMD of a link, especially if this link is established by several Carriers (maybe also different fiber types). Sometimes they are still using old fibers with a very high PMD-value and this can cause problems for a 10 or 40 GBit DWDM-system. In this case you can try to find a better fiber pair, where the PMD-value is lower. Our PMD-analyzer does not support pinpoint
faulty sections. We have to split the link and test again. The geometry of a fiber, round or oval, micro bending, temperature and movement of the cable has strong influence on the PMD-value.
It is not enough to test only one fiber of a tube to get a full evaluation of all fibers in a cable. One example, one link about 100km long, type of fiber G.655, one tube of 12 fibers have been tested and one had a bad value (PMD = 9ps at 100 km length =0,9 ps/√km). We have repeated this test several times and no changes. Then we tested the tube with OTDR at 1550/1625nm with launch lead in front and at far end, but we could not find any anomalies in the loss.