Basic DIY optic design?

[RoGuE_StreaK]

Not sure if this is the right spot, couldn't find an area dealing with optics.

After finding a few blogs on the net dealing with DIY optics stamped from acrylic, I'm wondering if "crystal clear" polyurethanes could be moulded into suitable optics. I'd imagine the light losses would be a bit more than commercial optics, but for specific designs it may prove worthwhile if you can manage to focus more of the light where you want it to go?

Anyway... as a basic concept, was contemplating what the affect would be if a lens was moulded to "extend" a Cree dome; ie., it's inner surface fitted snugly over the dome, and it had a universal thickness (say 1cm for argument's sake). Would this resullt in the same spatial distribution as the dome, but sans some losses? Or would reflection be a major issue?

Could be a cheap way of making custom lenses for specific distribution needs? I'm thinking the "bottom plane" of the attached image could potentially be brought up into a parabolic using TIR, which feeds back into the top curve design for focusing/scattering?

XT-E shown for reference:

 

[ma_sha1]

This design will not focus any light at all, except introduce addtional loss at the junction due to refraction index difference between the two materials

 

[TEEJ]

I made a free hand aspherical lens for my Zebralight SC600W, which is normally a flood light. The above pictures are what the projected die looks like on a wall shot.

 

[RoGuE_StreaK]

ma_sha1, yes I know that design will not focus, I was asking if it would theoretically give the same spatial distribution, ie. essentially extending the thickness of the dome. Primarily asking if it's viable to design an optic that hugs the dome, then from there I can start working on what the final beam should look like and how exactly to achieve that aim.

TEEJ, what materials where you using? I'm more concerned about getting an specific distribution than the intensity, I think I could afford to lose a few lumen if it means better control over where it ends up. My application is very specific and unique, so off-the-shelf optics don't quite fit the bill.

 

[TEEJ]

ma_sha1, yes I know that design will not focus, I was asking if it would theoretically give the same spatial distribution, ie. essentially extending the thickness of the dome. Primarily asking if it's viable to design an optic that hugs the dome, then from there I can start working on what the final beam should look like and how exactly to achieve that aim.

TEEJ, what materials where you using? I'm more concerned about getting an specific distribution than the intensity, I think I could afford to lose a few lumen if it means better control over where it ends up. My application is very specific and unique, so off-the-shelf optics don't quite fit the bill.

I used the lens of a 30 x 21 mm magnifying loupe - and adjusted the spacing/alignment until I could project the die. My camera is terrible for this, in real life, the projection was razor sharp...with the donut looking similar in appearance.

I did the above as a proof of concept experiment.

:D

 

[bshanahan14rulz]

What is it that you want to do? There is a lens for it.

 

[Obijuan Kenobe]

You are aware of aspheric lenses and their properties?

I smell wheel reinvention.

 

[TEEJ]

The aspheric lenses require spacing and so forth to work out to get the focus, and one lens may nor be right for all similar diameter bezels, etc.

This uses two inexpensive magnifying glass lenses with a threaded mount, to allow simple adjustment in/out to attain focus. Obviously, the transfer properties, etc, will be much better for a real purpose built lens (Ahorton, etc)...but this allows a poor man's aspheric to project your die image...as an experiment in this case to see how it worked. In practice, the projected die was razor sharp...the camera added fuzziness that wasn't there....but the giant donut corona was very much present, projecting the OP reflector image as well.

:D

 

[Harold_B]

Perhaps this: http://www.breault.com/resources/kbasePDF/wp_spie_032_axisymmetrical_concentrators.pdf

Or this: http://144.206.159.178/ft/CONF/16419618/16419639.pdf

 

[RoGuE_StreaK]

The concern is that I'm trying to get an even light intensity at points at set distances and angles, rather than throwing. Most lens or reflector solutions give me uneven distribution at these points, I need to tailor it to focus more light at certain points and/or decrease light to other points. I know I'm being vague about my purposes and I apologise, but a lot of it's theoretical at this point.

In a nutshell, I'm trying to funnel then throw to a narrow beam, but the beam needs ideally to be tailored for x intensity at y angle. Note that I'm planning on using Cree XP-G/XM-L/XT-E emitters, so a lot of their output is going out at quite wide angles; I want to grab as much of this as possible and funnel it generally forwards, then figure out how to get the right amount of light going where I want it to go.

From my (admittedly limited) experiments, parabolic reflectors appear to give me better "funelling" than TIR lenses, but lenses seem to give a more suitable final distibution. I've toyed with the idea of custom-CNC'ing reflectors, but I'd still need a lens over the front to strategically shape the light. As the project involves lots of CNC'ing and moulding of clear urethanes anyway, I was wondering why no-one seems to have gone down this path?

I've also got a very limited width to work with; less than 20mm. I'd read recently about a lens/reflector manufacturer that was talking up their use of polymer "filling" in their reflectors, as the refractive index meant they could make much narrower parabolics with the same result, or something along those lines.

Thanks for those links Harold B, might give me an insight into why TIR lenses are designed the way they are rather than hugging the dome, and some of the design work seems more in line with my thoughts (pseudo para with multi-profile lense)

I'm not trying to reinvent the wheel, just wondering if there's a way to DIY a specific wheel rather than having to get 10,000 of them made, when you're the only person who desires said wheel!

 

[videoman]

From what I understand, you have a very specific application and exacting requirements needed to achieve it. If you desire to fabricate a DIY lens/optic/reflector, then much time, effort, cost and frustration may be expected. I know. I have tried many times to make a specific reflector, as an example, and after all the time and sweat, I found one that closely was what I was looking for. There are many to choose from and can be individually purchased, and modified slightly here and there for some slight control. I would recommend you check into Edmund scientific ( not cheap) for stuff you didn't think existed, for lenses, prisms, large and small. Perhaps fresnels is what you may need. Check out Khatod as they have a large variety of stuff that few know about. From "I want to grab as much of this as possible and funnel it generally forwards, then figure out how to get the right amount of light going where I want it to go." I may also recommend perhaps anamorphic lens element to position the beam at various places without having to move the light source itself, from what I understand. There are many ways to achieve the objective. Please be less vague and perhaps a correct solution will be obtained. Unless it's a patent thing. Good luck.

 

[Obijuan Kenobe]

The vagueness of the whole concept is frustrating here.

I am sure the expertise exists at CPF. However, until you spell out what you are doing...what can you really expect for answers?

Funneling light (?) with a reflector or optic to get x distribution at some distance y is what flashlights, theater lights, showroom lights, car headlights, lasers, and optic fibers all generally do.

I am still not convinced you are attempting to accomplish anything new per se. You are just very cryptic about how you describe whatever the application is.

obi

 

[bshanahan14rulz]

Are you talking about using the light that the reflectors cant catch and funneling it to add to the hotspot? Because that is how many standard off-the-shelf TIR works.

You can use liquid lenses, dunno if they come in sizes big enough for you, but they will not do anamorphic focusing, they only change focal lengths.

Have you tried an aspheric with the emitter inside the focal point instead of at it?

What kind of light distribution are you looking for? You haven't really given us any details except for asking us if there is something that you can shine light through and get light on the other side.

 

[TEEJ]

BTW - you mentioned that the LED's you wanted to use throws the light at wide angles...but really, I think it emits pretty much straight ahead like most LED's. Its the incan bulbs, etc, that emit light in a ball.

:D

 

[SemiMan]

BTW - you mentioned that the LED's you wanted to use throws the light at wide angles...but really, I think it emits pretty much straight ahead like most LED's. Its the incan bulbs, etc, that emit light in a ball.

:D

Is there basic merit in your concept? Perhaps, but without delving into the properties of clear urethane which I am not familiar with, I expect that may be part of the issue.

Most new lenses have pretty complex shapes that would not lend themselves to simple molding techniques. Even prototyping with CNC may not work. To that end, you design your lens, then get a mold made. It's not that simple even though, as you need to consider shrinkage, etc. during whatever curing/cooling is going to happen (not an issue with CNC). So ..... while conceptually the idea may be sound, practically it may not.

That said, if you could shoot me anything on the optical properties, I have some thoughts on how to make it viable.

Semiman