I had this question for years. I frequently (like almost always) find optic designs that expand collimated SMF output into larger diameter beams (something like two achromats after a small asphere). It makes sense if a system requires beam diameter reconfiguration occasionally, but this also happens in many clearly finalized optical designs. My question is, why don't people collimate SMF output directly into a bigger diameter beams using larger diameter collimators?
I did some quick qualitative test before using a beam profiler and found no obvious difference. There is also little difference with optical simulation. I understand those larger aspheric lenses are more expensive in general, but definitely not by a lot when they are within 1 inch.
There are probably many reasons depending on the requirements, but here is one common reason. The output of the fiber is pretty fast (0.12 - 0.14 for SMF. 0.22 for multi mode step index fibers. Even larger for others). To collimate the fiber well requires a thick aspheric lens (thick in relation to the diameter). As you scale up the diameter you also scale up the thickness. A large diameter asphere is expensive, and a large thick lens is heavy. On the other hand, if you use a modest a sphere to collimating the beam, then you can make a very low aberration paraxial telescope out of two cheap, thin spherical lenses.
That being said, you can buy single aspheric lens collimators over a pretty large range of focal length/diameters. Also, when we want very large diameter we will often design systems with a fiber at the focus of a parabolic mirror. That’s effectively just like having single lens with a large aperture
I think Mike is right that cost is definitely a major factor. Also, while the beam may appear to be the same when collimated directly to a larger diameter, I'd be really surprised if the aberrations are not worse due to the larger NA, even if it is aspherical, which may have implications for the subsequent application that the system is being used for e.g. imaging.
Hi Xiang. Actually The aberrations do not change. Picture a 2D ray trace of a collimator that produces a, lets say, 3 mm beam. On the the scale of 3mm the 10um fiber mode is a point source. if you wanted to make a 6 mm beam or for that matter 6 m beam, you wouldn’t change the drawing at all. You can just scale the drawing. From the point of view of geometric optics (from which all macroscopic aberrations derive) nothing changes. The mode size and the wavelength do not scale, so the diffraction and the divergence get better, but the aberrations across the aperture remain the same.
Well, sort of the same. They scale, too. But scaled aberrations on a scaled aperture give the same far field effects.
Hi Mike, thanks for clarifying. Yes, the aberrations will scale with the aperture size, which is why I do not really understand your point about nothing changes? The ray aberration will become larger and makes it more difficult to achieve diffraction limited focusing afterwards, which is my point about the application. I guess the real question is how does this compare with the beam expander set-up, for if a single, albeit large apherical lens can achieve the same optical performance as the alternatvie combined system, at a similar cost, then I do not understand why not either haha
Double the transverse dimension and double the phase function = the same local slope in the wave front. As far as geometric optics is concerned the local rays point the same direction = the same place in the far field. Only the relation to wavelength = diffraction changes.
Thank you for the discussion.
Regarding the discussion about aberration, I am pretty sure (in practice and modeling) that a large aspheric lens can handle abbreviations as good as a small aspheric lens plus a beam expander. In fact, a single large aspheric lens is probably better in practice considering extra alignment errors and manufacturing errors introduced by the extra elements.
A larger aspheric lens is more expensive than a smaller one, but with a small aspheric lens you need quite a lot of extra optics, optomechanics, space and photon budget to expand the beam bigger. The actual cost easily adds up to way more than a big aspheric lens unless you need to expand the beam to a really really large spot.
When one get small aperture free space beam, he have to use shorter distance beam travelling path (due to higher divergence) but can use shorter distance optics to focus it. And contrariwise, if one have large aperture beam, he can use longer beam paths but have to use longer focal distance optics to focus it. In some cases shorter beam paths and shorter focal distance optics mean more compact design, in other cases one may need more space between focusing optics and focal point to contain some input window or protection glass. So small aperture collimated output is more universal - it is designed to be used in compact devices, but if one have some special demands he can allways apply beam expander.
- Badges
- Science topic
- Similar topics
- Physical Sciences
- Optics