Artificial Sunlight (official topic)

@marcdraco Thank you for the welcome! The current for that LED is basically ~3A, so nothing horrendous. But yeah, i'd have a normal outlet power cable go up into the attic and have the power source / driver live there. But i'm stymied on is where to plug this little guy into:

It's a bog standard heatsink fan connector. I guess I could run the wires for that down and plug it into a little raspberry pi that happens to also live on the shelf, but that seems a bit silly. Ideally I'd like one mains plug that powers both the LED and the fan.

Gotcha. I guess the question here is what's the voltage to the LED and what's the drive voltage for the fan.

If they are both (say) 5V, you can take a 5V tap and run the fan like that. I'll really have to bone up on this project but my spare time has been gobbled up like an Ouroboros due to my work on the mic. 😉

@marcdraco Please don't worry, you can consider buy our high quality and high CRI Ra 95+ daylight white 5600K 28x28mm COB LED to build this project, this guy, ojensen used our 100W COB LED and DC0-60V 8A Current Voltage Adjustable Knob Switching Mode Power Supply to drive 100W COB LED, it's very wonderful.

here is our 100W COB LED, https://www.aliexpress.us/item/3256802501151936.html

and here is the power supply, https://www.aliexpress.us/item/2251832788842825.html

@marcdraco this guy, ojensen used our 100W COB LED and DC0-60V 8A Current Voltage Adjustable Knob Switching Mode Power Supply to drive 100W COB LED, it's very wonderful.  Our 100W COB LED rated working voltage is DC34V-36V, working current is 2.5-2.6A.

I think the answer here is that I need to put a power strip up in the attic, into which I can plug the LED power source and a power cable for the fan -- something like this: https://www.amazon.com/CRJ-Power-Supply-4-Pin-Computer/dp/B071FNN9W7/

I'm curious how others have done it.

@ojensen That's how I would do it. Larger fans tend to be 12V as a rule anyway, but I'd figured the Vf for the LEDs would be quite high given the power. Most of what I work with is in the 10s of milliamperes. It's all the same but the numbers are bigger. 😉 If I run an LED at 10mA I feel like it's going to glow like the sun. Compared to the ones we got in the 1970s, they do!

So you'd have a mains tap to a distribution block (say a four way adaptor) which supplies mains voltage to the LED driver PSU and a separate 12V wall wart (I love that expression) for the fan.

You could tap the 36V line and drop a 12V regulator in there but that's a lot of extra work for no real gain.

My real concern for any of these things that you keep the LED and cooler well away from combustible items. In the unlikely event that the fan fails, the LED will rapidly overheat and although the PSU is protected, the LED isn't so it could cause a small fire... and you can do the math from there.

I'd probably put a thermal cut-out in there for extra protection but that requires design and testing.

@diy-led-u-home seems to have the kit and expertise for this, so thank you to them for chiming in.

I got the easy part done.

If i was gonna do this again: i foolishly believed that larger strips would be better because there would be fewer "stripes". In fact, larger strips conform far worse to the parabolic reflector, meaning you spend forever trying to deal with ripples and still can't get all of them out. Oh well. It still came out pretty decent, I think.

I tried the thing where I moved towards and away from the dish looking at the receiver thing, and it didn't stay exactly the same size, which makes me wonder if it isn't quite in the right place. Or maybe the actual receiver is embedded inside that plastic thing or something. Either way, I'll need to play for a bit to figure out where exactly the LED needs to go -- and of course how to put it there. Removing the receiver from the end of the stick is easy enough, so i'm not too worried 🙂

Next is the hard part: wiring up the LED / heatsink / powersource / etc. Then comes the harder part: figuring out how to make the scattering window. Then comes the scary part: cutting a big ol' hole in my living room ceiling 😳 

@ojensen I feel all that pain brother! But - and this is important - we learn by doing and when we're doing we make mistakes. I know, I make a lot. But persistence gets you there in the end.

Putting a hole in the ceiling is pretty easy if you're careless walking on the beams. 😉

But joking aside, my old mate told me "measure twice, cut once". And I would also suggest that you measure and mark up the cavity on the rockboard/drywall and then leave the cut for at least 24 hours. Sleep on the hole - let it sink in a bit and then measure it again.

It's surprising how easy it is to spot a mistake that you made a day or two later than when you've just done it. I've learned this the hard way on several PCB designs by firing off to JLCPCB and then checking it again in a day or two only to find some horrible cock-up!

When you say "the receiver thing" are you referring to the LNB (low-noise block) which is the thing that sticks out in front of the dish?

I'm not following this one as close as I should (I love optics but I don't have enough life to learn about Rayleigh Scattering and nano-fluids). I assume the dish is being used as a collimating lens? Someone more up on this can correct my assumption perhaps.

I suspect the ripples won't cause a massive issue but again, I'm not an expert on this project and there's some really smart people in the thread.

I would also suggest that you measure and mark up the cavity on the rockboard/drywall and then leave the cut for at least 24 hours. Sleep on the hole - let it sink in a bit and then measure it again. 

This is fantastic advice, thank you! Will definitely do this.

When you say "the receiver thing" are you referring to the LNB (low-noise block) which is the thing that sticks out in front of the dish?

Yep, exactly. It can be detached relatively easily from the stick, I just need to find some way to attach the LED there 🙂 I have a couple ideas, but I think that will basically end up being some tinkering.

I assume the dish is being used as a collimating lens?

Yes! And I'm with you on the Scattering thing -- my plan there is to weaponize the sunk cost fallacy by leaving that for last and then saying "what am I gonna do, throw away all the work I've already done?" to convince myself to figure it out 🙃 

Now for the process of incessantly refreshing of tracking pages waiting for my electronics / LED to arrive.

@ojensen I feel that pain! I literally just got through burning even more money ordering some new boards from JLCPCB.

Amusingly it's an opto project too - but this is just 30 LEDs at an absolute max of 0.4 amps. 🙂

This one doesn't need (but would benefit from) a filter so perhaps I'll have to lean on someone to make me something in translucent plastic. Naked LEDs look a bit naff. 😜

I had to look up the sunk cost fallacy (I'd always heard it as sKunk cost and not even expanded into the logical fallacy). Does it stink? Maybe it does. I would imagine a huge hole in your living room ceiling would be more than enough to convince you to carry on. 🙂

I've had some interesting times with plaster board. These days I watch instructional videos before I start filling holes in the ceiling... In my case it was some bad plumbing, which I poked and discovered just how much water can get behind that stuff and how fast it comes out when the bubble is popped.

Still, man who never made a mistake never made anything.

Where development is concerned sometimes projects run over (and over and over) but I consider the sunk cost as part of doing business (or the cost of development) and consider how much closer we are to the goal.

The Varee project has cost me more than buying a fairly decent microphone - but the results mean that I can make very high quality, 48V phantom powered mics for around a tenth the cost of a professional offering. Maybe more. And because it's Open Source hardware, everyone can benefit.

In your example, you will already know you're electronics work before you break out the utility knife and you'll already have projected that on the area you're fitting the film. Everything is correct but now it's time to cut a huge hole in the ceiling.

It's like looking over a cliff edge isn't it? If I didn't rent this place, I'd be sat on my hands waiting for the LED to come too! 🙂

@ojensen someone on this forum previously discovered that waterproof inkjet film is great for creating the blue sky effect.

https://fr.aliexpress.com/item/32807226254.html?spm=a2g0o.order_list.order_list_main.49.39ef5e5bgQ450m&gatewayAdapt=glo2fra

@jordiemc Thanks! I saw that link earlier in the thread, but haven't pulled the trigger on that yet, mainly because i will probably want a window-size, i'm envisioning maybe 3' by 4'. But to get something 3' wide I need to spend ~$100 for 30 meters of the stuff, when I need only a little more than 4% of the length i'd be buying. I could get a narrower roll and do multiple stripes on the window, but then I'd have seams on the window.

So then I was thinking maybe I do that thing with the epoxy and the colloidal silica. But I guess the likelihood that I can get that done for less than $100 is... maybe unrealistic. I dunno, I'll cross that bridge when I get to it 🙂

@diyperks do you have an update on the V2 version? I want to build my own and I'm trying to figure out who's plans to follow to allow for a smaller footprint. I might just go with a grid of fresnel lenses with the inkjet film as that seems to be the easiest. I just don't like the idea of needing 10+ cob leds to get a sufficient spread of light. anyone tried a similar aproach but with less leds? any help would be great. thanks!

Are we doing a V2? I haven't heard - not that anyone tells me anything. I'm just the idiot who volunteered! 😉

@marcdraco I just saw Matt mentioning he was working on a V2 in some of the posts on this thread...

Gotcha. I know he's been held up on a major job recently - that has gummed up the works. The V2 USB C mic has taken way longer than I'd hoped and isn't really on topic here but Matt sets the bar so high we have to really push hard to improve things. Sure we *could* just do a "0.1" evolution but V2 in Matt's world tend to be a major upgrade.

I wish I had better news but we'll just have to wait to hear from the Horse's Mouth (so to speak).

@calixt I just had a look at the photos you posted of the inner workings of that light. I've been thinking to try create something like that. I imagined the linear parabolic mirror would be bigger. Have you any idea the wattage of the LEDs used?

Hello again from me,

I had a longer break for my thesis because I was waiting for the materials to arrive. I hope some of you read this fast to help me with the theoretical work! I am able to produce some PMMA with TiO2 nano particles tomorrow and I am currently looking for the concentration of the TiO2 in the mixture. I allready found some informations, but I am not sure if it is for the same tasks. As I am only able to use the machines once I really want to get the best results, so If somebody has some informations with a source for me to citation, let me know!

Thanks and I hope to be able to share the results soon with you guys!

I like the stacked boxes idea created by @remitchou. Are there any existing products on the market that have a similar compact design approach?

@smjedison Awesome! Does this also simulate the sun being at an infinite distance? I'm really curious to see how it looks with Rayleigh scattering, have you had time to finish this project?

https://photos.app.goo.gl/5DM82yi77AnFNGNV6

I've been working on a modular system with 3d printed lamp cones with fresnel lenses.

I ended up using two layers of the water proof inkjet film to get a more natural sunlight colour.

I noticed the one lamp cone i printed black gives a better blue sky appearance, not that you can really look into much anyway with how bright it is.

@jordiemc sweet setup ! Don't know if it's the camera settings or the power setup of the LEDs, but it looks a bit dim in the video. The black background does look much bluer, but I think the output is more important than the look of the lamp, a comparaison shot of something lit by a white lamp then a black lamp would be very interesting.

Posted by: @fhr

↑

@jordiemc sweet setup ! Don't know if it's the camera settings or the power setup of the LEDs, but it looks a bit dim in the video. The black background does look much bluer, but I think the output is more important than the look of the lamp, a comparaison shot of something lit by a white lamp then a black lamp would be very interesting.

Agreed I looked at that and went "sweeeet" (so long as the heatsinks don't melt the PLA etc.).

The brightness and even colour rendition from a camera is nowhere near what our eyes can see. I learned this (mostly the hard way) doing photography. Even with "live view" and the sensors on modern cameras, I'm not aware of any that are capable of human eye levels of contrast.

I rather suspect that our eyes actually work like HDR and might even do two or more "exposures" in real time, combining the final image. We know this is how our focusing "system" works to keep everything in focus...

Perhaps the "best" way to shoot this is to show it illuminating something else - after all that's the whole point. Making it look like a skylight as was, I think, Matt's intention is beautiful but eve "point" lights don't matter if the thing you're lighting up receives soft "sky" light.

I've really, really, really got to pull my thumb out of my own fundament and try this because it looks fun (and very useful).

what about this Modern Blue Sky 6500K LED Ceiling Light  Ultra-thin Lighting 3/4/6/8 inch?

https://www.diyledu-home.com/products/diy-led-u-home-led-ceiling-light-modern-blue-sky-6500k-ultra-thin-lighting-3-4-6-8-inch-for-corridor-balcony-bedroom-aisle-living-room-home-decor-lamp

@jordiemc So cool :O! I'd love to see the mounted result.

@marcdraco I should have taken photos of the light illuminating a subject, my garage is such a mess at the moment I didn't want to take photos of anything in it.  The phone camera adjusts for the brightness a lot.  It's way too bright to look straight at with your eyes.  With the lamp cones being modular in design, I plan on having 9 connected together.  I have also ordered reflectors to fit to the LEDs to hopefully capture a bit of the light which is being lost off to the sides.

I'll upload photos when I make some improvements.

It looks great from here. It’s one of those things best experienced in person. 🤣 

I had a thought, it should be possible to get a similar effect much cheaply by using retro-reflective tape.

* spotlight aimed at ceiling from pendant-style light

* ceiling above covered with retro-reflective tape

* additional layer of inkjet film for softness

It would give the illusion of a light source much further, yet always behind the light fixture. The illusion wouldn't be as strong as a real parabola, but it could work.

Turns out the film on TVs Matt mentions is for sale on Aliexpress when you search for "composite brightness enhancement film". The composite 3M BEF 5 stuff is multi-axis and works quite well on random LED panels. 🙂

@fhr I think the issue you would run into is the light would be reflected straight back into the fixture. You would probably have better success with a normal reflective surface rather than retroreflective tape. I have seen people use this approach with some success.