So I made a cardboard-and-blutack prototype of the followspot reflex sight with heads-up display idea that I’ve been thinking of, with a very nice beamsplitter cube loaned from my makerspace friend. Too bad they’re quite expensive — might not be so practical to use one in the end (rather than a semi-silvered mirror), but they’re so nifty.

@s0 that's so cool! Also I'm now wondering whether it's possible to DIY a beam-splitter :blobthinking:

@aearil the answer seems to be basically no, as some back of the envelope maths told us you need the interface layer to be on the order of single microns thick, with consistency down another OOM. It’s cleanroom territory afaict.

@s0 Ouch indeed :/

My optics classes are a few years behind now, so I might be completly wrong, but wouldn't it be possible to skip the interface layer by using poured resin for the second half?

@aearil I don’t believe so. It’s not based on simple refractive index change but frustrated total internal reflection and the evanescent wave. A wave hitting the transition at greater than critical angle will have its energy split between the TIR portion and a portion from the evanescent wave on the other side of the gap, dependent on the width of the gap section compared to the wavelength iirc

@s0 thank you! I was never that great manipulating the Maxwell equations though ^^'

But even reasoning only with Fresnel's law, it makes sense that as soon as your interface layer grows, you'll get some mix of front side and back side reflection against that layer.

That said, depending on the application, it might not be that noticeable?

(also I just realized that both of the solutions I proposed were doomed because I failed to take into account the refraction angle, oups ^^)
@s0 Thinking about it some more, you can even simply skip one half of the splitter and have just air/glass interfaces
Sign in to participate in the conversation
Cathode Church

A place for trans makers, coders, tinkerers and dreamers.