This is a video I created to demonstrate\promote these little lights I’ve been making.
I’ve got a fairly consistent process for making a few different styles so I’ve decided to call the main ones BLOOM, FLAME, and CLOUD. One-off experiments that work out enough to put in a light will be GUMBO. The larger fresnel lens tubes with little scenes will be SPACE.
I have no idea if anyone will think these are cool enough to spend actual money on, but at this point I pretty much have to start selling them just to justify continuing to make them.
I still have a little sorting to do but I plan to have the first Etsy listing of one of these lights up within a couple of weeks. Still not sure about a price, I’m thinking in the 60-80 range.
If you are an Arduino\electronics person and want to try writing some cooler FX loops or replacing the ATTINY85 with an ESP8266 for WiFI control let me know and I’ll see about sending you one for testing.
UV LEDs really make fluourescent materials pop but I’m way too lazy to be wiring up individual PWM dimmable LEDs, so I looked around for addressable UV options. The first searches came up with references to a few UV LED strip supplies, but everything I could find was out of stock. But in combing through Alibaba I found these PCBs made to drive a thru hole RGB with WS2812 signals.
I thought it might be possible to drive UV LEDs with them, so I ordered some.
I had a while to wait for the shipment and started searching again, and realized I’d somehow missed the most obvious supplier of specialty electronics had exactly what I was looking for- Adafruit FTW.. So I ordered a meter of them and got them about a week later.
They work really well, and I love that I can control 3 tiny UV LEDs on one chip and that does offer a little movement with the light when you rotate them, but I was looking for more definition. I’d already tried some little 30degree lenses for 5050 LEDs so I popped one on. It definitely tightens the beam, but the beams are so close it doesn’t make much of a difference in a small amount of space.
So I finally got the PCBs in and tested soldering up a few LEDs in a triangle.
Here’s a demo of the addressable 5050 bare, the 5050 with a 30degree lens, and the PCB with 3 individual UV LEDs (I think they have 20 degree lenses).
I really like the PCB version and I think I’m going to use them in these fresnel tube lights. This is a balsa plane one that is kind of boring because the plane is too level, but it’s good for testing.
I needed a phase-cage to trap a A tensor varmint outside of normal space-time, so I built this meta-harmonic open-closed cubical spheroid.
That, or this is the same as the hex tube from before, but in black PLA stand and with a cooler tube in the middle. I need to write some cooler FX loops for it, this one doesn’t even use the ‘area’ LEDS, the ones in between the tubes in the LED ring. Can’t really think of anything for the FX right now, but there have to be some cool combos for this configuration.
Plenty of other neat-o silicone\LED nonsense going on, but this is kind of a one-off I figured deserved a post so here is.
This is a tube of silicone with two bundles of optical fibers and some highlighter fluid. It’s lit from above with a UV LED. There’s a little 3d printer collar that holds the fiber bundles onto 5050 RGB LEDs. I did a few handheld tests with the fibers with a collar and I thought it transmitted enough light to be an interesting effect so I jammed some fiber into some silicone and let it cure.
The results are pretty messy and not really what I had in mind. I cut the ends at an angle to try to make them more visible, but as it turns out you can only really see them from above. I tried to get the bundles to fan out to the sides in the silicone but they broke at too great an angle so they mostly ended up semi-parallel with the tube so their lights are invisible. Also there’s no way to avoid big bubbles coming up through the bundles- silicone finds a way.
It was weird enough to justify documenting, but I don’t think I’ll be gung-ho with the optical fiber. There may still be some uses for it but this test was pretty unimpressive.
The rest of the video are various things in a temporary lighting rig with one RGB LED below and a UV LED above. Not sure any justify permanent builds because of various defects, but they’re part of the learning process.
This is the same balsa tube concept in the fresnel tube rig. I should have made the plane a bit bigger and rotated it a bit so it’s not just going straight. Also the powder smoke effect didn’t work out as cool as the other one but I think I know how that works better now. I sprayed the plane with a good bit of highlighter fluid and it pops pretty well with the UV but it gets washed out by the RGB in the video. I like the general balsa plane in a tube scene concept so I’ll keep working on that.
Same basic idea as the balsa plane but a paper plane also soaked in highlighter fluid. They’re surprisingly hard to fold symmetrically that small. I placed it too far off center so it’s too close to the edge. I also added some dandelion seeds but they squished so it looks pretty weird. I’ll probably work more with the paper plane but this one was kind of a bust.
This one was just stupid and gross. I gathered some little flowers and berries, sprayed them with highlighter fluid, and stuck them in silicone. I guess I was thinking silicone stops time but they continued to wilt but they aren’t drying out so it’s just a mess. The only interesting bit was I just squirted some elmers glue at the top for no reason and it make an interesting ‘concave cloud’ formation that I might try to work with later. Overall it’s just a nasty mess but I wanted to see what it looked like under the light so this is it.
Also if you’re wondering the solder roll at the top is because the temporary rig isn’t secured to anything and the wires pull off the top so the roll is just as a weight.
I originally wanted to make a sort of ‘flame’ in the tubes but in playing with the fluorescent and silicone I stumbled on a nice floral aesthetic and made these stands to light them. These are lit from the top with one RGB LED and from the bottom with one UV LED.
The ‘stems’ are created by cooling the silicone and injecting warm gelatin\highlighter fluid with 16-18g dispensing needles. The cold silicone makes the gel set up quickly and gives the stems a bit of a texture. The ‘dirt\sand’ is coffee\grits. I thought I needed to leave a hole in the middle to allow the UV LED through so I used straws to leave a hole, but turns out it looks better with the UV on the tube edge so next time I won’t bother with the hole. The ‘flower’ tops didn’t turn out well this round- I let the silicone cure too much so I couldn’t get the ‘ribbons’ that I had in other tests but the next batch should turn out better.
Here are a the tests leading up to the first flower tubes. The bottom One of these tests was the inspiration for the tubes, it just started to look like a flower so I jammed some coffee into the bottom to look like dirt and I figured it looked good enough to try a few more.
And here are the first batch of intentional ‘flower’ tubes. Lots of mistakes but learned a lot and a few of them came out well enough to justify the LEDs.
Cubic portals are cool, but we all know tubes are the future so of course the next step in inter-dimensional balsa travel should be tubular.
I’m slowly learning some tricks to working with silicone but it’s a slow process partly because I’m slow and partly because silicone is a truly insane material to work with. There’s a gap in the top that formed slowly as it’s been curing over a month or so but now I’m leaving holes in both sides for curing so I can control the shrink better. That’s the thing with this stuff- you can’t really build the light until the silicone has settled and sometimes you have to keep adding silicone every few days so it meets where the LEDs will be. It’s kind of a nit picky process after you’ve laid in the silicone, you just pick at it once a week or so until it settles in to whatever it’s going to be.
The smokey trails are just baby powder dabbed in with a q-tip while the silicone is wet, that was a surprising success, it adds a lot.
This build is lit with an 8-LED RGB ring and an addressable UV LED I got from Adafruit. The UV is WS2812 but the RGB channels are all UV chips, pretty cool.
I spritzed the plane with highlighter fluid but I don’t think I used enough because it’s not really popping. The FX loop is just cycling colors and brightness on the RGB and fading between the UV chips to give the UV some movement. Unfortunately the UV\fluorescent didn’t really add much on this one, but I’m learning. I’ve got a few other plane tube type trials curing so I’ll try a different configuration next time.
I need to figure out how to get better pictures of these things because these look a lot cooler in person, but it is what it is for now.
This is an artifactually intelligenting quantumish singularitinity combining cartoon shnano-toob structures with nob-libnear crystalfinity waveforms.
Also it’s a 3d printed cube with fresnel plastic windows filled with clear silicone. And there’s a balsa plane too.
I screwed up in the process and let a big bubble form that blocks a good bit of the LED, but it was a fairly successful little test and I’m going to make another one with what I learned making this one.
This is the shotlight prototype I ‘fixed’ by replacing the dome top with a vinyl tub. I still haven’t made the version with the shotglass or last tube section yet, but this was one of my first fluorescent fluid tests and the lighting effect is coming along so I thought I post it here. There’s also some fluorescent drops in the smaller tubes- that was a strange test where I injected the fluid while I was pumping the silicone into the tube. I had to pump the caulk gun with my feet so I could handle the tube and syringe, it was nuts. Injecting it wet like that the fluid beads up into a perfect droplet from the pressure as the silicone flows down the tube. However if you look in the picture the droplets are distorted- after I cut a 10″ section and it cured a few days I found you could squish the tube really hard and the droplets would spread out in the mashed silicone. Interesting effect but I don’t think it’s cool enough for all the trouble.
Frame with LEDs filled with silicone and spastically squirted with fluorescent fluid and squished together with fresnel sheets. No pattern or anything, but even so it came out kind of interesting.
Just another iteration of the ring\tube light I posted before. This one uses slightly larger tubes on the perimeter and added an larger tube in the center.
One of the first tests of getting fluorescent fluid into the silicone, lots of fun potential there. Just made a stand because things need stands.
The tube to LED tube effect mastery tubes through many strange tubes. Some tubes are interesting concepts that become poorly thought out attempts, some tubes suffer from a lack of resources or patience in fabrication, some just don’t make any sense at all in retrospect but I have to assume they did at some point.
Here we have a few failed, but still educational, tubular journeys I’ve made recently.
BIG HONKING EDGELIT TUBE
While I was at the hardware store looking at tubes I couldn’t help but notice the largest diameter tube in stock was about right to try edge lighting with an LED ring instead of through the center. They’re 1-7/8″OD and a 12-LED ring lines up right the LEDs packages. It took a little thought to get a secureish connections with a 3d printed holder. Unfortunately the results were very underwhelming. The light mixes completely in the first two inches so you lose any color differentiation between the LEDs pretty fast. And it dims out after about eight inches. Worse the look of the light coming from the tube wall instead of inside the tube doesn’t look very cool to me. Just looks like a giant, crappy fluorescent tube with a color temperature that’s just wrong. The tube was about $10 and I wasted a lot of time getting the holder right, but my time is worthless and I can reuse the LED ring for other projects. So not a huge loss and I might try using the tube for something else later.
GOTTA DO SOMETHING WITH THESE FREAKISH 3W LED PROJECTORS
These modules are ridiculous, but I’m semi-committed to figuring out something to do with a lens with such an insanely tight beam that the RGB channels don’t even overlap. After giving up on bouncing the light off anything since literally any other LED is a better option- I decided to see if I could capitalize on the insanity of the module’s design by projecting the beam into a tube. That makes it sound like I really thought outside the box to find a use, but I think it’s pretty clear why my first thought was tubes.
The lens turned out to fit a 1″OD|3/4″ID tube perfectly so I just printed a little collar and pumped in a bead of silicone to make a watertight so it could hold water. I didn’t go with silicone fill in the tube because it’s just a lot of silicone for a test. The results are interesting, it lights up the tube and blends pretty well. I think two of these on opposite ends could light up a tube at least 2-3′. I’m not sure that’s worth buying any more of these but I might have found a way to at least put them in a one-off giant-tube build. Not sure why I’m calling this a failure because it kind of worked out, but these freaky little projector LEDs are pretty much a mistake to begin with. I’d like to ask whoever designed them what purpose they had in mind.
This is supposed to be an effect light with tubes that ‘fill’ a little bottle or shotglass at the top with light. After I had the basic build worked out I was somewhat shocked to discover I do not presently own a shotglass. I had the acrylic case from an hexbug toy so I chopped it down and filled it with silicone just to top off the prototype build. Once you put the silicone on its kind of a done deal, you can kind of unmess it after it cures on PLA, but it’s not worth it.
It took me a bit to see it, but I am no longer comfortable with the overall aesthetic of this build. I should have waited until I got a shotglass. But I’m going to add another section of tubing to complete the loop back to the base. so this prototype build just is what it is, which is a light that doesn’t not remind people of a dong enough to not be a considered failure.
THE ONLY TRUE FAILURE IS THE FAILURE TO OBSESS OVER FAILURE UNTIL YOU THINK OF ANOTHER INTERESTING WAY TO FAIL.
These tests actually gave me some other ideas that I’ve had some success with, but I’m going to save those for a post that isn’t about fun mistakes. So stay tuned, there might be a post about mistakes involving fresnel lenses and\or highlighter fluid to look forward to, and definitely more tubes.
I was searching for a high power addressable LED module and this was one of the few I found in a price range I could play with, and being 5V instead of 12V is convenient for me. They were $15 for 5 unit plus shipping.
Only found them on Ray Wu’s AliExpress store, can’t find them anywhere else. It took a few weeks for delivery but it was reasonable for international shipping.
They’re basically what I expected, they worked right out of the box with FastLED test code (WS2811\GRB). I’ve been using 2812’s for so long I was briefly confused by the 4th wire and was worried they were just common anode RGB. But they’re WS2811’s for sure.
I opened one up and should have taken a picture but forgot. The only things worth mentioning are that the lens is glued to the PCB and protrudes from the housing so you cannot remove the board from the aluminum block without breaking off the lens somehow, which is likely to damage the PCB. Also it’s notable to me that the aluminum housing won’t help cool the PCB at all, it’s held in a channel in the middle but doesn’t have any thermal connection to the aluminum. That seems like a missed opportunity.
As far as the light goes- they’re very, very bright. Incomparable to a standard LED strip LED, it’s a bitty freaking light cannon. I can see them being used for architectural displays that would be visible even with lots of ambient light. The video is with my apartment lights on and the FastLED brightness at 128, about 1/2 power.
The thing that I didn’t expect was the projection effect. It’s kind of bizarre and probably why these things aren’t as popular as they could be. If you watch the video you’ll see what I’m talking about.
It projects an almost perfect image of the LED chip and you can clearly see the different positions of the red\green\blue LEDs. It makes it pretty much unusable for direct illumination of anything, at least from close up. You have to bounce this off of something or it’ll be unbearable to look at. It might help to test 20-30ft away, but my apartment isn’t that big and my neighbors already think I’m nuts so I’m not doing outdoor testing right now.
Unfortunately that kind of defeats the purpose of a tight beam for me. Might as well just get a better diffused light if I have to bounce it anyway.
So I’m not sure where to go from here with there. Since I’m already going crazy with tubes I might see if I can use these for a giant silicone tube light or something. We’ll see what happens, but I was pretty excited to get them in and test them so I thought I’d share.