Re: Beamsplitter Recommendation
- From: Phil Hobbs <pcdh@xxxxxxxxxxxxxxxxxxxxxxxxxxxx>
- Date: Fri, 08 Feb 2008 16:19:49 -0500
Jarvis wrote:
I've been experimenting with different types of beamsplitters in anJohn,
application where uniformity of splitting function vs. incidence angle is
key. I haven't exactly found what I'm looking for so I'm looking for
insight and/or a recommendation.
My optical system has a 45 degree beamsplitter mounted in a f/# = 1.4
converging beam. The purpose of the beamsplitter is to direct a portions of
the converging beam to two different optical sensors.
Basically, the uniformity of split varies across the beamsplitter. It is
particularly bad in the direction that includes the 45 degree angle of the
beamsplitter partially reflecting film. For this case, the external angle
of incidence with the
beamsplitter partially reflecting film is about 45+20 degrees for the worst
of the extreme rays while for the extreme ray on the opposite size it is
45-20
degrees. Split ratio changes by ~+/- 10% from one side to the other. While
the exact composition of the partially reflecting film is
not known, it all seems pretty well explainable by Fresnel reflection vs.
angle effects.
Also FWIW, I'm working in the telecom portion of the NIR (~1.5 - 1.7 um)
Here are some thoughts:
1) I'm presently inclined to use a cube beamsplitter with non-polarizing
film because of ease of
mounting and lack of lateral displacement effects. These all seem to be
designed for collimated light however. Plate beamsplitters have the same
problem but also have lateral displacement of the transmitted beam which
complicates setup.
2) I'm experimenting with using an optic to collimate the converging beam
prior to impinging on the beamsplitter. Would like to eliminate if possible
so as to reduce design complexity.
3) Several vendors sell a polka-dot beamsplitters (small, closely spaced
reflective metal dots
on a silica plate). This would seem to be a technology that would address a
good
portion of the angle sensitivity issue. I suspect that there would be some
residual splitting non-uniformity due to angle sensitive Fresnel reflections
off of the silica but this might be manageable with AR coatings. Anyway,
this seems to be calculatable. On the negative side of things,
design complexity increases as a result of a fancier mount and also must
take into
account the lateral beam displacement for the transmitted beam.
Soooo... The questions that I can think of are:
1) Has anyone had the same issue and successfully resolved it ?
2) Seems like a polka-dot beamsplitter in a cube configuration would be
ideal for me. Does anyone make such a thing? An hour or so on Google
hasn't turned up any suppliers of stock optics.
Thanks in advance for any insight you can give me.
John
This is hard. With an aperture that wide, you're going to get a bunch of astigmatism from anything thicker than a pellicle, even at that wavelength. There's also the phase vs angle of the coating to consider.
Depending on what you're trying to do, a polka-dot beam splitter might be OK, but watch out for moire effects between the TX and RX beams--if it's a regular array, then tilting the sample a little will give you bright and dark moire fringes in the received light.
You might also want to consider doing the beam splitting before the objective lens, where it's much much easier.
Cheers,
Phil Hobbs
.
- Prev by Date: Re: candle spectra
- Next by Date: Re: Beamsplitter Recommendation
- Previous by thread: candle spectra
- Next by thread: Re: Beamsplitter Recommendation
- Index(es):
Relevant Pages
|
Loading
