Cube beam splitter and through-objective illumination

[jnh]

I recently found a reasonably cheap Edmunds cube beam splitter on eBay (like this one: http://www.edmundoptics.com/optics/beam ... ers/54823/)

Together with my T-tube setup and some small parts I had lying around, I thought, this would make make for a neat way to implement through-objective illumination, when inserted into the infinity space between tube lens and objective.

Alas, this turned out to be a big disappointment. The beam splitter does its job sending light at a 90deg angle towards the objective, but much to
my dismay, there's also a fraction of light that is reflected towards the camera. So much so, that any image is completely obliterated.

I suspect, that this stray light is due to a reflection of the transmitted beam from the cube surface back into the cube, which is then diverted towards the objective
(see the little scheme bellow, with the offending beam in green).



I tried removing the cap that was closing the port opposite the light source, but that didn't help at all. Also, a laser pointer directed through the cube shows the same effect (so it's not caused by stray beams bouncing around inside the cube or metal cage).

That's something, I didn't take into account beforehand. So, I'd probably be better served with a plate beam splitter, as this should not be prone to those offending reflections (which, I'm aware, have been discussed here before). Btw: this is something Edmund does not mention, when when extolling the virtues of cube beam splitter vs. plate beam splitters: http://www.edmundoptics.com/technical-r ... egoryid=15.

[Chris S.]

Jnh,

I'm sorry for your difficulty, but much appreciate your post. I came close to buying that beamsplitter myself, but you beat me to it. My intended use was very like yours, except that I'd probably use collimated halogen light. Of course, collimation is probably not the issue here, since you've demonstrated that the problem exists with a laser. (BTW, I do realize that the vendor has another specimen at a higher price--I was hesitant to pay that price, and your post has significantly increased that hesitation.)

Your problem leaves me scratching my head for a couple of reasons. First, if this problem were common for this beamsplitter, it's hard to imagine that it would perform well for some of the tasks it's designed for. Second, my intuition balks at seeing sufficient reflection off the glass/air boundary opposite your light to cause the problem you're having. (I suspect Rik may at some point supply the math to refute or confirm my intuition.)

A few thoughts:

Are you certain that you're orienting the cube properly? (As you likely know, the cube consists of two prisms--one coated with reflective material, the other uncoated. And Edmund Optics recommends shining the light onto the coated surface. They say that this is keep from damaging the cement, but perhaps there is also an optical reason.)

If you take the cube out of the metal holder, shoot the laser through, and observe from above, do you see anything that helps explain things?

You might want to call Edmund Optics support engineers. I've found them to be both knowledgeable and helpful.

Good luck! And please keep us apprised--if you solve your problem, I'll likely buy one of these for myself.

--Chris

[g4lab]

You are on the right track but you have not finished the job.

What you need to do is get a couple of linear polarizers and a quarter wave plate. Beam splitters are often used in incident light microscopy but to suppress the effects you are seeing you need to polarize the light and use a crossed polarizer to suppress the unwanted light.

Also on stereo microscope coaxial illuminators which almost always have beam splitters in them there is a piece (which never makes it as far as the guy that is trying to peddle it on fleabay) that goes in front of the objective , a quarter wave plate , which will even suppress nasty reflections on opaque specimens. Some compound objectives (Zeiss AntiFlex) have these too.

Circular polarizers have quarter wave plates in them and it may be possible by selecting the correct orientation to make use of the quarter wave plate without effects of the linear pol filter sandwiched with it.

I gemmological friend of mine picked up a coax illuminator that did not have the quarter wave plate. He said "It's a piece of doodoo" He changed his mind after I told him to add a quarter wave plate.

The pol filters will undoubtedly help matters.
The Zeiss "Handbook of Incident Light Microscopy" which the Zeiss sales reps used to give away is wonderful. I had a copy of it it and it got away from me. But thanks to the internet everything is available.

http://www.microscopy-uk.org.uk/zeiss/Z ... tive-a.pdf

[jnh]

Thanks for the feedback.

Are you certain that you're orienting the cube properly?

I read Edmunds' recommendation and tried both orientations, and - as shown in the image below, with the laser pointer shining through the cube - it doesn't seem to make much of a difference:



So, I actually have two other cubes sitting around and they seem to be afflicted by the same problem (I first bought a Thorlabs cube, that was advertised as being 50:50, but it turned out to be 90:10 cube. So the seller send me a full refund and let me keep the cube, and in addition, he send me me another 50:50 cube. But that one was designed for a far red laser and it's transmission plot in the visible range is quite a roller coaster. The cube in the Thorlabs cage was glued in, though, hence I couldn't just swap in another cube).


(Pardon the less than stellar images, to see R' of the 90:10 I had to turn off the light and I shot hand-held)

If I interpret g4lab's post right, this doesn't seem to be unexpected behavior and can be rectified with polarizers and a quater wave plate. But I don't understand quite yet, how these componets are supposed to arranged. Does a polarizer go between the light source and the cube, another one between the cube and the tube lens and the quarter wave plate in front of the objective?

Also, is my above assumption correct, that this is caused by reflections from the cube face back into the cube? Instead of using polarizers and a cube, could I simply use a beam splitter plate, to solve the problem (as there are no straight faces to cause reflection problems)?

[Pau]

Instead of using polarizers and a cube, could I simply use a beam splitter plate, to solve the problem (as there are no straight faces to cause reflection problems)?

Yes, epiilluminators usually use plate beamsplitters. A cube prism BS is much more robust but I only use it for the microscope light source

I tried the same with a small cube from SurplusSed and it worked only whith cross polarization. Now with a plate BS I can get illumination without pols but still find useful the use of cross pol to deal with subject reflections and glare induced by other optical surfaces like objective lenses.

[Frank_77]

Hi all

I’m reviving this thread because I have a similar problem and I don’t understand if a solution was found.

My setup is similar to the one described by jnh but instead of a single laser dot I’m projecting a pattern.

My application is in the context of lase range scanning and my idea was to project a laser pattern through the beam splitter towards an object and capture the pattern with the camera.



The resulting image has several reflections which I don’t understand how to remove.




As for jnh also for me, when I shine the laser beam through my beam splitter, there’s an extra reflection towards the camera. This doesn’t seems to happen when I shine the pattern.





These are the components I am using:

Laser Source: https://www.thorlabs.com/thorproduct.cf ... ber=HLS635
Optical fiber: https://www.thorlabs.com/thorproduct.c ... -630A-FC-1
Collimator: https://www.thorlabs.com/thorproduct.cf ... r=F220FC-B
Diffractive optical element (DOE R-354): https://holoeye.com/diffractive-optics/ ... -plastics/
Beam splitter: https://www.thorlabs.com/thorproduct.cf ... mber=BS010
Camera: Raspberry Pi camera V2

Apologies for resposting if the problem was solved but my knowledge in optics is pretty limited so before I buy some components I wanted to make sure I understand what is the problem.

[AlxndrBrg]

Have you tried changing the material that is in contact with the surface that produces the reflection? I assume its currently air, or am I wrong? In the image with the red and green arrows, what happens if you instead stick some black tape to it, maybe that could reduce the reflections?

Alternatively you could use something like glycerol that matches the refractive index of the glass and "lead" the offending light somewhere else, or simply absorb it.