Talk:Beam splitter

Amplitude vs. Intensity

Do beam splitters split amplitude or intensity? —Preceding unsigned comment added by 128.112.50.118 (talk) 19:51, 16 May 2008 (UTC)

conservation of energy would imply that intensity should be split. Ionsme2 (talk) 21:46, 5 April 2023 (UTC)

• Is it true that, for a half-silvered mirror, one half is polarized and one half isn't? If so, it should be added to the article, but I'm not sure...

Image

Can someone create some kind of image to demonstrate it?

Like?

Is it something like this?

• http://www.idquantique.com/products/quantis.htm

Beam spiltters with single mode fiber?????

>Beam spiltters with single mode fiber for PON networks is using the single mode behavior to split the beam. The splitter is done by physically >splicing two fibers "together" as an X.

How is one fiber split 32 times? Can someone explain this process? Is it done by micro-bending, prisms and colors or physically splicing 32 fibers to one?

Also how does it work when the 32 signals return and concentrate into one fiber then to the Central Office?

thank you,

cleandog

---

Here is the answer to your question "how is one fiber split 32 times": It is done by making optical waveguide circuits on a planar substrate. The professional name is "planar lightwave circuit" (PLC). This substrate is typically a silicon substrate. The optical circuit has many cascading 1:2 beam splitter elements so that one beam is split into two and two into four, etc, until 32. The 1:2 beam splitters are the equivalent of the two fibers spliced as an "X". The 32 output beams of this optical chip are butt-coupled to a v-groove linear fiber array using epoxy.

For this 32 channel splitter, the input signal is split equally so that each output fiber has 1/32 of the original signal strength. The reverse is also true: when the 32 signals return to the output end of this 32 channel splitter, only 1/32 of the signal end up at the input end of this splitter and it goes back to the central office.

Quickwin —Preceding unsigned comment added by Quickwin (talk • contribs) 01:56, 6 September 2008 (UTC)

Disagreement

Unfortunately I can´t completely agree with what is stated for the Technicolor beam splitter: The two halves of the prism block were not coated for dichroic function but with a gold layer which should behave colour neutral, yet still enhancing the longer wavelengths. The actual colour separation was arrived at through filters, a green one and one of purple hue, purple being the half-half mixture of red and blue. The third filter was a red tinted gelatine overcoat on the farther of the two films run in Sandwich manner. Orthochromatic, non sensitized, and panchromatic stock was used for the green, blue, and red portion, respectively. Filmtechniker alias 80.219.85.241 (talk) 18:32, 29 November 2007 (UTC)

In fact, a suitably deposited transparent coating of gold will preferentially reflect the yellow (red-to-yellow-green) portion of the spectrum, transmitting as much of the rest as it does not absorb. Coatings of some aniline dyes behave similarly. Frederic E. Ives patented this for use in color photography in the 1910s. Technicolor was using such a coating early on, very possibly going back to their two-color system. Anything that helped make more efficient use of the light coming through the lens was a plus, but filters were indeed still needed and used to get adequate color separation. It seems that a multilayer coating of the relatively modern type was substituted at some time before the end of three-strip photography in 1955. AVarchaeologist (talk) 07:11, 24 August 2013 (UTC)

Significance in quantum theory etc.

Many quantum theoretical experiments involve beam splitters, which makes me wonder if it is worth adding something about them to this article, which is currently rather slight. PJTraill (talk) 19:10, 28 July 2015 (UTC)

I have now done this (more generally, not just for quantum theory) PJTraill (talk) 21:09, 28 July 2015 (UTC)

Bad explanation of quantum description

You do not need to have recourse to annihilation operators, second quantification and commutation relations (which, IRONIE-ON of course, to the layman and reader of this article are "well-known" IRONIE-OFF) to deduct the conditions on the matrix. We only have to require that the matrix is unitarian (which is necessary if we want to have preservation of probability) - and the conditions in the text are exactly equivalent to having a unitary matrix. — Preceding unsigned comment added by 93.207.216.218 (talk) 14:54, 10 April 2017 (UTC)

French version

Hi,

A Beam splitter in French is

Prisme_(optique)

and not https://fr.wikipedia.org/wiki/Miroir_semi-r%C3%A9fl%C3%A9chissant which is https://en.wikipedia.org/wiki/One-way_mirror I don't know how to change that ! — Preceding unsigned comment added by Zarg (talk • contribs) 15:49, 17 August 2017 (UTC)

I don't speak French. (Although, etymologically speaking, French and English are nearly as identical as Norse and English are. Only the small words are really different.) When I look up Miroir semi-réfléchissant (semi-reflective mirror) I find the French Wikipedia article about beamsplitters. That seems correct, because a beamsplitter is effectively a semi-reflective mirror. When I look up Prisme_(optique) (prism (optics)) I naturally find an article about prisms. If these terms are incorrect, then you would have to take that up with the French Wikipedia. All different language Wikipedias operate independently of each other, so there is nothing we can do to correct those articles from here.

Effectively, a beamsplitter, an output coupler, and a one-way mirror are the same things. They operate on identical principles. The main differences are the tolerances used for construction. A laser output-coupler is manufactured to very high, extremely exacting tolerances, using complicated coatings of precise thicknesses. A typical beamsplitter is constructed with fairly high, optical quality tolerances. A one-way mirror, on the other hand, has very low tolerances (typically just wet-deposited silver) in which the flatness and coating thickness may vary considerably across the substrate. Zaereth (talk) 18:39, 17 August 2017 (UTC)

Classical lossless beam splitter Notation

What are Ea and Eb? Are these the polarization states, or are these the spatial states? Are Ea and Eb vectors within the Ein vector? Same questions apply to the output.

It seems that if we reduce to scalars, then we would need a 4-component vector. 2 contributed by polarization, and 2 contributed by spatial modes. Ionsme2 (talk) 20:27, 24 October 2022 (UTC)

These are beam A and beam B, to differentiate between the two different beams incident on the splitter. Zaereth (talk) 21:42, 24 October 2022 (UTC)

Confirming then, they are vectors? (If so, maybe we should put the vector arrows over them (or use bold text) Ionsme2 (talk) 22:31, 30 March 2023 (UTC)

I don't know. I guess it depends on what that means in English. If "vector" means the same thing to you as it does to me (the direction of an aircraft's flightpath), then no. They are not planes, they are two different laser beams. Beam A and Beam B. If you shine Beam A on one side of the splitter at a 45 degree angle, it reflects at a 45 degree angle as well, but it splits into Beam A (the reflected beam) and Beam C, which transmits straight through. Now, if you shine Beam B on the other side of the splitter at the same time, also at a 45 degree angle, the reflected Beam B can be made to combine with the transmitted Beam C. Likewise, the portion of Beam B that transmits will become Beam D, which combines with the reflected Beam A. So what this is saying is that you can use a beam splitter to combine two different laser beams, but the combined output beams will match the power of the input beams, thus no net gain of power is possible. The best this can be used for is combining beams of different colors to produce a third color. Zaereth (talk) 23:13, 30 March 2023 (UTC)

I meant vector as in electric field vector, or jones vector.

Ea can describe 2 polarization states so I would imagine it's a 2 component vector. Ionsme2 (talk) 21:48, 5 April 2023 (UTC)

Partially Silvered Mirror redirects here, but shouldn't

Partially silvered mirrors are used for a lot of things besides beam splitters. To me, that should be a separate article. Will (Talk - contribs) 04:36, 27 October 2023 (UTC)

Well, we already have an article on one-way mirrors and one on output couplers. The redirect could just as easily go to one of them. The thing is, they're all basically the same things; only the uses differ... and perhaps the manufacturing tolerances. I couldn't see any use in creating a separate article for just that term. One could argue that they're all just forms of beamsplitters. Maybe we should just redirect it to the transmissivity section of the mirror article, and let the reader choose which application they want to explore further from there? I don't see that it really matters much, though, because no other articles are using that redirect for anything, and it get an average of two views a year. Zaereth (talk) 19:26, 27 October 2023 (UTC)

It occurs to me that another alternative is to add a subsection about it in the mirror article under the types by reflective material. Need to be very careful there, though, because it once had such a section that became a mess of incorrect explanations of how a one-way mirror works. (Lot's of urban myths about that one.) You're welcome to add such a section, but I'd make it brief, just enough to add wikilinks to these other articles, with just enough explanation that readers who don't want to go on a scavenger hunt won't have to click those links to understand what it means. (Too often technical articles end up that way, but an article that relies on wikilinks to define its subject is lazily written and almost impossible for an outsider to follow. A person should be able to read the entire article and infer the meanings without ever needing to click on a link.) But that way we'd have a place to _target this redirect to. Zaereth (talk) 01:51, 28 October 2023 (UTC)

potential citations here

https://www.thorlabs.com/newgrouppage9.cfm?objectgroup_id=9028 ElectronicsForDogs (talk) 16:33, 16 October 2024 (UTC)

We have to be really careful with catalogs, because misuse can appear promotional. That said, catalog pages like this are often wonderful sources of information. They are primary sources, of course, so great care needs to be taken to make sure they are interpreted correctly, but as long they're giving the universal concepts and not promotional, then they can be used... with care.

The big problem is this only supports the most general concepts presented in this article. In other words, there are only a select few sentences in this article supported by this source. Most of this article is completely unsourced. We have a ton of math equations with no sources. Entire sections without sources, and this catalog page doesn't cover any of those. So, while this is better than nothing, it would need to be added carefully and it barely scratches the surface of the cite issues. Zaereth (talk) 17:33, 16 October 2024 (UTC)