Hi Physics StackExchange community,
Today I had to re-do the entire optical path for the delivery of a 830 nm laser from our laser table to the experiment table. For the fiber coupling, I usually use a back-coupling device to quickly adjust the collimator. Since I didn't want to touch the fiber collimator on the experiment table, as everything is precisely aligned, I put the back-coupling device on a test fiber, with nominally the same core diameter (5 um) as the fiber going to the experiment. With this I got ~80% efficiency.
Once this is done, I pull out the test fiber, re-insert the experiment fiber to the fiber coupling setup, and tweak the mirrors and the collimator. After much beam-walking and adjustments, I still couldn't get the coupling efficiency above 60%. Now the test fiber and the experiment fiber are from different manufacturers (the former from Coastal Connections, the latter from OzOptics), it could be that tolerances are different and hence the actual core diameters as well. But I still do not understand why the adjustment of the collimator won't solve that problem. Any ideas on what could cause the difference in efficiency?
Fiber coupling setup: there are two mirrors for beam walking. The collimator (aspheric lens) is on a XY translation stage on cage rods. The fiber adapter is on a Z translation stage on cage rods.
Either it's a strongly charged dust that I couldn't remove with methanol, or some surface damage. But since it's not directly on the fiber core, would this explain the drop in coupling efficiency?