Proportional Beam Splitter?

[Wolfman29]

Hey everyone.

For a project I am working on, I essentially need a way to split a beam into two parts.

That's not hard - just use a PBS cube, right?

But, the kicker is, it has to be polarization independent. Is there any type of glass which is equally translucent for the spectrum of lasers we use, i.e. from roughly 300nm to 1100nm?

The idea is that then I can have one part of the beam be, say, 10% of the beam's total power and the other part of the beam be 90% of the beam's total power.

Do any such filters/optics exist? Remember, this has to be polarization and wavelength (for the most part) independent.

I'm thinking just some sort of half-silvered mirror or something? I dunno.

EDIT: And preferably, not that expensive!

 

[chipdouglas]

AR coated pbs cubes are what you want. you just rotate the laser unti you get close results. but i don't think they will be 50/50 or 90/10 but probably close.

michael.

 

[Wolfman29]

The idea is that they need to be polarization independent, otherwise I would have nabbed that idea right there Now, if there were NPBSs, that would be great, but there aren't. I'm looking at ThorLabs and all of their optical devices (even the economy beamsplitters, which aren't very stable over the normal wavelength we use) are pretty expensive. Obviously, buying one would be expensive, but buying one shouldn't cost 300+$....

 

[Kmor2004]

If I am reading your post right, you want a beamsplitter that will pass 90% and reflect 10%, if I am correct then I think you may want something like->

http://thorlabs.com/NewGroupPage9.cfm?ObjectGroup_ID=913

 

[Wolfman29]

I was looking at that - it's all good, except preferably, the range would be 400-700nm, at the very minimum.

I was thinking, I could work with something that is *relatively* flat over the spectrum, as long as it has a defined function for the reflectivity (i.e. not just datapoints, but a function).

Does this look doable? http://www.thorlabs.com/NewGroupPage9.cfm?ObjectGroup_ID=914

It's cheaper, and if I can determine a function for that curve, then all would be good, especially the 70/30 one.

And then there is THIS one, which is exactly a 50/50 pass through which is good....

http://www.thorlabs.com/NewGroupPage9.cfm?ObjectGroup_ID=1110

 

[Kmor2004]

The 1st ones you linked do 450-650, so using 405's and possibly 445's may not work the way you are hoping for, also they are 30/70 or 50/50, and the second one is polka dot, which to me sounds like it would reflect a percentage as a star pattern, but then again i'm no expert on optics, just my 2 denars

 

[Wolfman29]

Yeah, that's why I was worried about that 450-650, although it would WORK for 405 (looked at an extended graph), it's just not as flat over there.

Regarding the star image - that doesn't matter

 

[Cyparagon]

Use a piece of glass at 45° to the beam. That's about 1:9 reflection:transmittance.

 

[Wolfman29]

I could actually try that, will just need to conduct some tests.