Who 'd like test our dichoric

[The_Doctor]

Very convincing test. Those seem to work on the wavelengths for best blue and red like ideal binary switches.

One question (I'm not sure I know enough to ask any more).. To get that clean a change, do they use many coatings, and if so, do they show any non-flatness that needs focus or other change to compensate for that?

[danielbriggs]

One question (I'm not sure I know enough to ask any more).. To get that clean a change, do they use many coatings, and if so, do they show any non-flatness that needs focus or other change to compensate for that?

Yep

Does anyone remember the Semrock 640/660nm dichros? They had a massive effect on divergence; so much that you had to compensate for the apparent focal length by decollimating the red modules. It was all down to the residual stress from the several hundred reflective coatings on the substrate.

[planters]

Daniel,

Wow! Like the first time I fired up a 445 nm diode, this is VERY impressive. Just thinking, but you probably have a polarizing filter of some kind. What if you were to place this filter between the source and the dichroic and rotated it slowly to the "best" and "worst" orientation. These have to represent the S and P maxima even if you don't know which they are. Right now, assuming a non-polarized source, the results are an average and usually 1/2 way between these maxima in any case and are probably pretty good.

A test of the wavefront (far field spot) for transmission and reflection, and if decent then I am sold and c'est la vie to Semrock.

[The_Doctor]

Yep

Ok. But that does raise a question... What about coating both sides, half as much? I don't know much about coating procedure, but it seems to me that balancing forces might help. Or is the image ghosting of two layers (each side) an even worse problem?

I already think the question is a silly one with yet another 'yes' answer. But I might as well ask it because there might be a more interesting answer I don't know about.

[mixedgas]

Ok. But that does raise a question... What about coating both sides, half as much? I don't know much about coating procedure, but it seems to me that balancing forces might help. Or is the image ghosting of two layers (each side) an even worse problem?

I already think the question is a silly one with yet another 'yes' answer. But I might as well ask it because there might be a more interesting answer I don't know about.

You mean like S1 Dichroic S2 AR as it should be? Otherwise you get greater then 4% Fresnel reflection from S2.

Steve

[The_Doctor]

I didn't but I should have. AR one side is a far better idea. I'm editing a lot of audio tonight, not thinking as well as I should. But sure, if one side had a lot of AR coat layers to balance the stresses of the filter coating, is that doable as a way to balance stresses that cause wavefront distortion, divergence, etc? I'm just wondering because these dichros seem to be so good that if some small problem were found it would be worth going all the way. One fairly definitive set of dichros might mean bulk, and cost savings. And less agonising about what to use..

[planters]

Steve,

Your right, but the coatings for these two sides are likely to be pretty different and hence the stresses. I suppose that if the coating design was sophisticated enough then the stress each side introduces might be balanced.

Keep in mind that whatever the distortion, this will be a characterization that is routinely provided only for much more expensive filters.

[The_Doctor]

Ok, so what about thicker glass? I know you can't do that for first order waveplates, but can you do it in a dichro? If the greatest loss is in the surfaces, a bit of extra thickness might help.

[mixedgas]

Broadband "V" AR coat is a few layers tops. Far fewer compared to the S1 side. Thick Glass does somewhat, but not totally, reduce the surface stress. It adds new issues, such as needed to be wedged.

I work with Thick Dichroic optics in the day job, the beam translation is a issue. Ours are even more interesting, in that half of them are intracavity and exposed to power levels that scare even experienced engineers. This would drive costs up enormously, even for show use, as you go from cheap float glass substrates, at best with additional single pass grind and polish, to a expensive substrate.

A better substrate moves end user costs, even with low labor costs, to ~150$ a optic. Plus now you need a "swept" substrate, where its probed for internal defects during manufacture. Float glass is cheap to enough to reject after coating, yet still quite satisfactory from a wavefront point of view. Scratch/Dig rated substrates are not that cheap.

Most of this arises, because the planetary in the coater has to be very hot during coating laydown for good adhesion.

Steve

[The_Doctor]

Broadband "V" AR coat is a few layers tops. Far fewer compared to the S1 side. Thick Glass does somewhat, but not totally, reduce the surface stress. It adds new issues, such as needed to be wedged.

I work with Thick Dichroic optics in the day job, the beam translation is a issue. Ours are even more interesting, in that half of them are intracavity and exposed to power levels that scare even experienced engineers. This would drive costs up enormously, even for show use, as you go from cheap float glass substrates, at best with additional single pass grind and polish, to a expensive substrate.

A better substrate moves end user costs, even with low labor costs, to ~150$ a optic. Plus now you need a "swept" substrate, where its probed for internal defects during manufacture. Float glass is cheap to enough to reject after coating, yet still quite satisfactory from a wavefront point of view. Scratch/Dig rated substrates are not that cheap.

Most of this arises, because the planetary in the coater has to be very hot during coating laydown for good adhesion.

Steve

Translation, as in lateral shift from the the more greatly separated planes of refraction? I'd forgot about that... 150 bucks sounds like a lot, but if this solves enough problems, the difference might be paid for nicely this way. I'm not sure about the entire economics of a good projector, but I'm thinking along the lines of making a good meal, being only as good as the most important ingredients. Tis a hell of a lot of loot though, if three of them are needed..