During the building up a free space Optical Coherence Tomography system, we normally want to align the beam splitter cube as perpendicular to the incoming light direction as possible, because it will make the alignment of the reflected back easier to align and then easier to couple into fiber of the detection part. However, when this beam splitter cube is very perpendicular to the light, there will be light leaking out through the fourth surface. When you block the trans-through surface and the reflect surface, there is still detectable light on the fourth surface, and this amount of light is causing noticeable noise in the interfere fringes. Is there better method to use during the align of a free space OCT system?
One way that worked for me in with more or less similar problem was to put quartet wave plate on the path of interest. Upon incidence and reflection, the beam is pathing the plate twice so has it polarization rotated by 90 degrees. So you could use a polarizer to filter out the reflection from optical surfaces.
To add to Vincent's suggestion - you could also use a wedge window as the beam-splitter. The two reflections come off at different angles, so you can avoid the overlap.
The caution here is that the setup becomes sensitive to polarization, since angular reflectance depends on polarization. If you tightly control polarization in either V or H state, you should be alright. But for a mixed-polarization signal, you may observe additional intensity variability.
As I read your question, you want to use the reflection from the incoming edge of the cube to help with alignment. If this is correct, I think the answer must be that the reflections inevitably have consequences, albeit these can be minimised by cleanliness, coatings etc. You could in principle either use a block where the fourth surface is non-parallel (though this too would have alignment consequences), or have a box constructed for your pellicle that includes a removable alignement reflector, but I suspect that this is more complexity than you desire.
This may have no relevance to your experiment, but when interference from unwanted reflections are a problem. it is sometimes possible to eliminate them by "decohering" the unwanted reflections over the measurement interval. If the measurement were sufficiently slow (unlikely), one might vibrate the offending surface. Alternatively if the optical frequency can be swept during the measuremnt interval, one can average out the interference, or differentiate the unwanted reflections via the differentially detected offset frequency versus distance.
Thanks for the answers. As for the wedge window, I thought about it, but because the light trans-through it has equal power and 0 phase delay for V and H polarization state, so I would concern about the impact to the polarization state.
About the Pellicle BS, I saw it on Thorlabs website, but I am still wondering how good it is. Thanks for your encouraging feedback about it. I will consider to replace my BS cube with it.
I would suggest to perform a spatial filtering of the reflected light in order to eliminate scattering from reflection
- Badges
- Science topic
- Similar topics
- Physical Sciences
- Optics