37.5cm fresnel lens spotlight build

Hi Djozz,

yes, the fresnel has to be covered completely. I used some kind of UV glue but don´t know exactly any more. That was about 10 years ago when beamers with high resolutions were expensive. Today there is no need to build one by yourself.

Greetings

Kenjii

What(!), we have a yearly build competition and it’s here??? I thought that “lighthouse” build I sold my precious coated aspherics for was another website completely having a “design competition”, or something like that. :open_mouth: Is it all over? Ahh, nvm, I’ll do my thread digging tonight/later for that stuff.

When Mr Kloepper Knife Works visited my home last month I brought out a decent grade of 320mm diameter fresnel lens with around 50-grooves/Inch groove density. I was thinking it was slightly lower groove count like 1 per mm, but I believe it did end up being 2/mm when I counted rings in the light with caliper and loupe.

We were putting a dedomed XP-G2 build with bezel taken off behind that lens and that was driven around 4 amps current I believe (close enough?). Focal length is around 200mm on that fresnel lens (always smooth side faces LED just like aspherical lens). I do say “FREZ-NAL” lens just like it looks and we are not supposed to say. (FRAH-NELL = “proper”, go figure.) The fresnel we were shining the G2 through 200 meters out, looked like a tiny G2, I would say. It was not rainbow colored. When the lens was quickly flipped around so grooves were facing LED, apparent brightness looked like it went down to 50%. If the lens is bending everywhere, I can imagine you’d get prisms all over the place happening. But chances are djozz, you have some very sharp grooves that aren’t rounded well in there, which would split light well into different colors like that. Good, correct grooves should display a good, semi-correct color image with some resolution reduction is all.

Fresnel lenses vary more than any other collimating lens type in their ability. Some of them are odd and have a definite FL you can see is there, but they are really intended to be more of an ‘end point’ for light to be viewed on rather than projected through to a far distance (even though light technically still passes through to be seen by the eye—the image is viewed on the fresnel itself). They are very hit-and-miss lenses; I usually don’t buy them unless I plan to experiment and hope to expect a good lens grade if I’m paying anything for one. In this case I did get a decent grade of lens, but this one’s very thin, about 1.5mm—which is a factor known to reduce throughput (not to mention rigidity). Candela can’t be predicted, as you see djozz can get a reading, but on what color is the cd measurement, green? :wink:

A good fresnel output should look like a very dirty aspheric or lens with a very light diffuser grade behind it, and prism-effect with the extreme aberration should not be there, not that bad at least if the lens will be of good use.

TV lenses are circular groove and horizontal together, so you actually get a short line (more like a bow-tie shape) when an LED is focused down and the Kcd is probably not as high as you think—not even close.

For 5-million cd, there are a few different ways I can think of, but non are as cheap as a fresnel.

Nice old arcade screen, I mean retro center speaker, I mean light box, djozz! :slight_smile:

The competition is still on, just join and you have 'till sept 30th . https://budgetlightforum.com/t/-/33821

It is nice to know that there is better quality fresnel lenses out there, and as Kenji pointed out, ways to make it flatter, so there is a lot of improvement still possible. This light for me was checking out the concept using cheap materials and I even put in more effort in the casing than originally planned (I'm glad I did, it is a nice thing now). I was already surprised that it works as well as it does (the hotspot is bright yellow btw, could be a worse colour :-) ). If there are easy fixes I will surely implement them, but to get it past 10 million candela it will take more, like using a better lens.

Not all aspherics are well made either.

This TrustFire Z5 has a lens that has a shorter focal length around the outside than near the center, resulting in this ringy spot at intermediate zoom. Others with similar problems include some cheap lenses from DX.

For a medium or decent quality fresnel lens I usually just assume about 35% loss and maybe with good fresnel (lower groove/mm count) only about 25% throughput loss against a typical un-coated lens of the same size. I only use that as an approximation for known fresnels, like I said if you haven’t held the model series it is totally unpredictable. That is basically what you pay for between thin/thick lens; all the ring valleys which do not amount to real usable area. Good luck on getting to ten million. :wink:

The unavoidable loss is due to diffraction. The diffraction blurring is important with telescopes but not with LED lights, because of the size of the LED. The diffraction loss is from an area around each corner of width comparable to the wavelength. So for 500 nm. light and 100 zones per mm., the loss is a few percent. That is comparable to the loss in the material, so a perfectly shaped Fresnel lens may have more or less loss than a comparable thick lens.
500 * 10-9 / (1/100 * 10-3 ) = 0.05 = 5%
I don’t know why LED Lenser only uses two zones. Maybe because it is prettier that way.

I think you’re saying one thing that makes sense to you, that I’m not exactly following to the fullest extent. There’s all kinds of diffraction occurring in there, not just wavelength restriction at angular problem areas like groove transitions. Each valley is “spraying” photons all over the place into the next grooves begin, and on and on it goes. 5% loss? OKAY, when NASA is building one, maybe. :wink:

A fresnel lens suffers from higher losses because of the obvious I would say; you have hundreds or potentially thousands of plastic grooves molded and the area cannot be used with the same average resolution (not even close). Due to the size of the grooves, things like arc centering highly change per groove peak and valley; not to even highlight on arc centering like that’s its primary downfall, but it’s now like 1000s of little tiny lenses are managing the light diffraction, not one smooth ground surface. It’s much harder to make a high resolution lens that is 0.5mm diameter (times a few million of them like a bug eye) than to make one lens at 200mm diameter that’s ground by CNC from a glass type. I think that’s the part I see sticking out as obvious to me with fresnel lenses.

For a point source, it is possible, and not hard with fairly simple software, to design a Fresnel lens that is perfect, to a geometric optics approximation. The Wikipedia “Fresnel lens” article shows ray diagrams. The finite size of the LED causes some shadowing of off axis rays that don’t line up with the steps, but I think that is a small effect. The precision needed is similar to that needed for an aspheric, but the process of making the mold may be more difficult.

But light is a wave and does not go exactly in straight lines, as shown here.

Diffraction pattern from a slit of width equal to five times the wavelength of an incident plane wave. From Wikipedia “Diffraction”.

But the light that is lost to the main beam is only that which comes through within a wave length or so of the edge of the step.

Nice box. Dah da da Dah dah, Dah da da DAH dah, Dah da da DAH! dah, Dah da da dah. It loses something in print. :bigsmile:

It never ceases to amaze how little response some truly interesting threads get compared to whine and cry threads.

I'm sure you know the answer to your own question but wont answer it here due to repercussions. Either that or your to polite.

I read what is going on here but dont understand a lot so dont respond. The same in a lot of technical threads. Keep up the good work guys.

That's clearly die Walküre von Wagner!, what else? :bigsmile:

This is pretty awesome.

For the next one, you going to use the Fresnel from an old big screen TV?

Like MRsDNF said: I read everything in this thread with great interest but I do not understand “some” of it.
“some” being quite a lot, actually. :wink:

If I cannot add anything usefull, I do rather not post at all.

Grtz
Nico

Same here, i subscribe to interesting threads like this but if i can’t add anything of value i often don’t post.

But i have to say good work djozz :slight_smile: It looks great both the beam & box.

I wish i had the willpower to keep modding when the world is so beautiful outside :bigsmile: but i haven’t modded a single light since may, It is just too bright & nice outside in summer here, and so few dark hours at night i hardly get to use them at all :frowning:
I can’t wait until the darker months so i can find time to mod again :slight_smile:

In that case: use the long summer days to be prepared for the long winter nights! :wink:

Grtz
Nico

Makes perfect sense, go modding when the light's out. Looking forward to the new batch this winter :-)

About me saying 'smart' things, as a biologist and not into physics I only know some basic optics (unlike a few here on BLF who are way more knowledgable than me), but I'm fairly good in thinking those basics through and find a practical use for it in this hobby. And I can not help myself boring the BLF community with that. (As long as it comes with beamshots I think I can get away with that ;-) )

Now now Djozz, no fals modesty please :wink:

I personally think you are one of the members truly contributing to this foruym so pls keep it up!

Grtz
Nico

Djozz, I think you have a lot more fans than you think. Like others have said, sometimes I very much enjoy a post, but fail to comment. Usually because I can’t think of something constructive (or witty) to say — other than maybe the overused “awesome.”

So……. “Awesome”, it is.

Oh wait. I just thought of something almost constructive. Did you know almost everybody mispronounces “Fresnel”?

(at least here, in the backcountry.)

“Augustin-Jean Fresnel (French pronunciation: [ɔgystɛ̃ ʒɑ̃ fʁɛnɛl]; pronounced freɪˈnɛl fray-NELL in English) (10 May 1788 – 14 July 1827), was a French physicist who contributed significantly to the establishment of the theory of wave optics. Fresnel studied the behaviour of light both theoretically and experimentally.”
BTW he was born at Broglie. Not many french people know how to pronounce correctly “Broglie”

Two outside beamshots in this thread: