Artificial Sunlight (official topic)

@shavings That looks soooo  goood! Nice job!

While everyones focused on creating the mixture for the Tyndall Effect, I've been thinking about the setup for the lighting source. I'm not a fan of the satellite dish because it takes up so much space so I've mulling over the idea of a grid of (eg. 3 x 3) cheap reading/magnifying glasses made using Fresnel lenses you could get at any convenience / corner store like the one below:

https://www.amazon.co.uk/TIESOME-Fresnel-Magnifier-Plastic-Magnifying/dp/B0BFQ3X316

Then making a grid of matching LEDs which line up behind the Fresnel Lenses placed at the focal length to create a bigger panel.

Does anyone see a problem with this idea? What would you use as the material to solder the LEDS or Fresnel Lenses onto? For the Fresnel Lenses I was thinking about a big sheet of PVC or Acrylic. But I'm not sure about the LED's.

Any suggestions are welcome.

Hi, this is an amazing project. I can't believe this isn't a mainstream product yet as it would have a huge market in the UK and Ireland I think.

Just wondering if anyone knows of any other company who are making these, other than CoeLux?

DIY Perks should start manufacturing this.

Would it be possible to combine the setup from the "Smashed TV"-video with the soap mixture to get blue skylight without the artificial sun but with the white sunlight and the blue sky?

Would love to make a flat version of the skylight with a blue sky.

Is there some type of acrylic with infused nano particles that can mimic the soap effect?

@shavings

Been following this forum since i discovered the video last year and really tempted to try and diy one of these. Yours is amazing, congratulations!

I will attempt mine first with a simple projector for photo/video that i can get for cheap. Im hoping the lense that comes with it will work sufficiently well (i agree with @nolo that i dont care if it's perfectly collimated).

If not, i will give a try to the vintage headlight reflectors

Anyway, Im based in France and would gladly buy a few meters of that waterproof inkjet sheet off of you if you are still willing?

For prototyping and trial and error, 4 meters would be ideal for me. Is this doable?

New to the forum, so not sure how we can discuss this privately but let me know!

Cheers

I'm excited to share an idea I've been mulling over, but unfortunately, I don't have the means to test it out myself.

I've been inspired by the DIY Perks video I watched a while back, particularly because I struggle with seasonal mood declines, especially during winter. Although I'm not ready to dive into making it just yet, the concept keeps drawing me back.

I've noticed there's been a lot of discussion about the dispersant for achieving the Raleigh scattering effect, with micronized TiO2 being difficult to obtain and work with in epoxy. I've come up with an alternative idea, although it comes with some hazards and requires careful handling and protective gear (if it hasn't been attempted before, forgive me).

I'm thinking of using fumed silica, which is highly compatible with epoxy and commonly used to reduce its density. It has an incredibly fine particle size, lending it light-diffusing properties often used in cosmetics. I believe that by adding a tiny amount of fumed silica and dispersing it well, we might achieve a similar effect to the soap and water technique. Then, a thin layer can be poured, and any bubbles easily removed with a blow torch.

The best part is, that fumed silica is affordable and relatively easy to find, with many sellers offering it on eBay. I think the hydrophobic grade might be the most suitable, but input from someone with experience using it would be invaluable. With careful handling and proper protective equipment, followed by thorough vacuuming of surfaces afterwards, it could be a game-changer, especially for producing multiple epoxy panels simultaneously.

If anyone has the resources and safety measures in place to experiment with fumed silica, please do so and keep us updated. I first learned about its potential from a fantastic video by Tech Ingredients on YouTube, titled "Super Strong Epoxy with Diamonds and More!" around the 07:17 mark.

Wishing you all the best of luck, and I'd love to hear about any successes or setbacks you encounter.

Sorry for the long text.

@anikode is fumed silca different from colloidal silca? I've posted previously of my attempt using colloidal silca, the results were pretty good.

https://yujilighting.com/products/skyline-30120-rooflight-72w

i bought one, and received it today, anyone interested in the guts?

basically a aluminum heatsink with an attached linear "parablolic" mirror and a LED strip shining on to a blueish plastic "glassfront"

Yes!!! Interesting that it's a linear parabolic mirror. We'd done some experiments with that but not gone far, though I do have a few ideas on how to use LED strips with a linear trough. It would be interesting to see how close the plans are.

@diyperks https://photos.app.goo.gl/SLUU3byDjEjGbyfTA

@calixt very interested! Any chance of getting some photos? Particularly curious about the geometry of the parabolic trough

That's quite fascinating - thanks for the pics! I had a very similar design sketched up for mine but I must admit that the parabolic mirror on mine was much much larger in profile. How do you find its light output? Does it feel like real daylight or more like an interestingly pleasant light?

@calixt 👍 cool, very nice.

@jackh I've been thinking the same about making a slim version with fresnal lenses in a grid.  I've been experimenting with some lenses I had and i think they'll need to be doubled up to make the focal length short enough to keep the finished design from being too big.  Depends on how much space you have though.

@calixt Man that is REALLY sweet! I'm so jealous!

Hello everyone. I've been working myself on trying to make nanoparticles in order to obtain Tyndall effect, with marginal success up to this point.

However, I've recently been thinking that maybeNightHawkInLight has the solutions. In his last series of video he's been making nanoparticles for various projects. In the last one he makes synthetic opal with a solution containing nanoparticles and observe how even a thin layer of solution is able to scatter light similarly to sunlight in atmosphere.

It's not the same recipe he used to make nanoparticle for the radiative sky cooling experiments but with his method he seems to be able to obtain relatively uniform and controllable particles sizes. He literally makes bags of nanoparticles powder in no time.

So I think it could be used to create artificial sunlight scattering by capturing nanoparticles in resin, for example.

@hans23 Hey there, I just discovered this youtube video, this forum, and this community, and your post - apologize for the very late response, but curious if you ended up buying this as a group order?

Just saw a post on Wired UK about this: https://www.sundaylight.cc/

This might be a really stupid question but would it be possible to combine the smashed tv concept with this one as @hutje said - considering that the tvs contain a fresnel layer in them that should collimate the light even if there is no artificial sun created. So essentially that project with different temp LEDs and a tyndall/rayleigh scattering layer over the top. Would there be any chromatic aberration caused by the fresnel layers? I shudder in thinking of the cost of a apochromat lens array or a catadioptric system large enough for a skylight.

Another question is just to ask why use epoxied fumed silica or tio2 or other nanoparticles over the waterproof inkjet sheets discovered a couple pages back in the forum? Are there significant improvements in the effect over the sheets by using these nanoparticles instead?

Apologies if these questions are terrible, I'm quite new to this space

@devsw1 8 grand for that is daylight robbery 🤔 

Eight THOUSAND pounds ?! This is so ridiculously overpriced, it's literally a water-cooled led at the focal point of a parabolic reflector. That's insane.

Also, there's something wrong with their specs. They announce 30,000 lumen and 10,000 lux, but the reflector is very close to 1 square meter so lumen and lux values should be similar (1 lux=1 lumen/m²). The only way it can works is if the led array is 30 klx but the reflector absorbs 2/3 of the flux and therefore reflects only 10 klx over one square meter which indeed gives 10 klm. But that would be a huuuuge amount of lost energy absorbed by the reflector (something close to 200W).

EDIT : OK I think I got it, I guess that the light emitted by the leds is not colimated towards the reflector so 2/3 of the 30 klm are "lost" directly lighting the ceiling.

Would a cheap LED projector mimic this parallel ray phenomenon?

@eliam That's pretty neat, I haven't come across the use of nanoparticles to replicate atmospheric scatter. Although I did read somewhere that scatter isn't down to particulates in the upper atmosphere, rather the actual molecules in the air which is pretty far out.

Wish I had the time to dedicate to one of these, my place is very dark. Can't even move without having some lights on.

Atmosphere is Tyndall scattering, opals and other nanoparticle-based scattering is Rayleigh scattering.   They look the same, but at completely different scales.

That's really interesting @fhr. That's another gap in my knowledge I can consider narrowed (but now I need to understand those two, too).

Would you agree that, at least what we're trying to achieve here, it's the end result that matters? I mean we aren't trying to make a sun to any significant scale for fusion (checks notes, yes fusion) reactors yet. Not being flippant, I'm just looking for an appropriate analogy.

As an aside (not to derail this thread):

I do hope we get fusion working soon. Fission would have kept us out of the mess we're in now if it hadn't been for well-meaning, but ultimately ignorant people protesting. And every single "disaster" we've had, three major ones by my count, 3-Mile Island, Chernobyl, Fukushima, has hurt far fewer people than are killed every single year by coal fired plants.

Hard to wrap your head around, but that's true. We "think" we understand oxidisation of carbon because we've "controlled" fire for thousands of years but the reality of it is the chemical reaction releases huge amounts of unburned hydrocarbons and other crud into the atmosphere. It's a case of familiarity breeding contempt.

@fhr It's the other way around. Atmosphere is Rayleigh scattering while opal is Tyndall effect. Both are actually the consequence of anisotropic scattering of light by particle.

Only the scale of the particles change (atmospheric Rayleigh scattering concerns particles smaller than scattered wavelength, when Tyndall effect in opals concerns particles roughly the same size as scattered wavelength). This of course result in different impacts on the spectrum.

And also the fact that Tyndall effect is mostly used to describe the effect of particles in colloid, i.e particles in suspension in a liquid or gel, although some people also qualify aerosols as colloids. So really not that different.

But in theory, in order to recreate the perfect atmospheric scattering in a thin layer, we need particles smaller than the scattered light but much more concentrated than in air. Using larger particles is much easier to craft and will result in a convincing effect but it will look more like a hazy blue sky.

Posted by: @marcdraco

↑

@eliam That's pretty neat, I haven't come across the use of nanoparticles to replicate atmospheric scatter. Although I did read somewhere that scatter isn't down to particulates in the upper atmosphere, rather the actual molecules in the air which is pretty far out.

Wish I had the time to dedicate to one of these, my place is very dark. Can't even move without having some lights on.

I really hope I'll be able to make nanoparticles at home with DIY components, because you can see on the picture that nanoparticles in suspension are able to recreate the whole palette of sky colors : sunset redish when the light shines through a "thick" layer of colloid, blue scattered light from indirect lighting, and whitish blue when particles are in very high concentration, similar to smog or Mie scattering in the atmosphere, as you can see with his sample rotated and compared to a sunset viewed from space :

Pretty amazing.

Hello

For those who are interested. This is video of a light I made. These are stacked boxes. Each one has a cob LED and a fresnel lens. I used two layers of print film for the tindal effect. A metal grille gives a window effect. It is not a 100% continuous effect but it allows a thickness of 20cm.

video

Nice one! That's similar to what I'm planning for V2, though I was going to try using many small parabolic mirrors. Fresnel lenses work a treat though!

What material have you made the individual boxes from? Just be careful not to let real sunlight hit it, as it could cause a fire if the boxes are flamable. 

I used fresnel lens mounting, because of its ability to hide what is behind it. The epoxy or sheet used for the tyndal effect is not a good diffuser. With a parabolic assembly, I was not able to hide the different mechanical elements, which compromised the sky effect. The assembly with fresnel lens does not have this problem, the mechanical assembly is hidden by the optical aberrations.

But I identified several problems with this type of assembly:
-The different prototypes were 3d printed, but the alignment of the different boxes and lenses must be excellent so that the light transition is realistic (<1mm). So I built a laser-cut metal chassis to mount the cobs and lenses precisely, and thus be within tolerance. it is painted blue on the inside which reinforces the bluish effect of the diffused light. The cod and fresnel are screwed directly to the chassis.

-Approximately 5% of fresel lenses were not qualitative, which caused poor light transitions.
-The Frenel lenses must be on the same plane for the illusion to be good. This solution therefore only allows vertical lighting . To try to solve this problem I made requests to manufacture linear lenses in order to stack them with the first fresnel lenses, and deflect the light rays by 30°. But making a specific mold is too expensive.

@remitchou That's awesome, I am planning to go with this approach as well. May I ask you how many lenses you put side by side to form your grid? I am still debating if I want to keep the focal length short with smaller fresnel lenses or go with fewer lenses but a thicker device due to the longer focal length