Four months ago we were screwed.

But that's putting the cart before the horse - which as we all know is a no no. First, a brief background: I'm part of a volumetric imaging company, Looking Glass. We invented volumetric printing the previous year as a way to seed the ground for our big dream of dynamic volumetric display. You know, the Princess Leia technology of sci-fi dreams.

And four months ago, we'd just spent all of our personal money and meager company earnings developing the dream - a high-resolution, full color-depth true 3D volumetric display. The only problem was, this display of the future was the size of a refrigerator and made a volumetric image just a little bit bigger than an ice cube. It would be years before we'd be able to turn it into a commercial product.

Then, on the precipice, we had an idea.

This is the unfolding* story of the L3D Cube, as it happens. My hope is that by being a nearly current play-by-play update of how this story unfolds, the narrative fallacy that plagues all post-game product launch accounts can be mitigated if not altogether avoided. If additional details of any step along the way - technical, business or otherwise - would be useful for folks please let me know and I'll expand the section best I can.

*unfolding on Kickstarter now. I will be updating the later steps of this Instructable as the present catches up to the future.


Not interested in all of this strategic product development stuff and just want to build your own cube from scratch? The links below are how to make it happen!

All of the hardware & software described here is completely open source under a Creative Commons Attribution-Sharealike 2.5 license.

That being said, a word of warning: it will be hard to build a single high-quality cube from scratch, since it involves large even-temperature reflow ovens which are usually only available for large scale production. That's why we think there's value in us providing a kit, because we deal with that nonsense. But if you want to dive in, here are the detailed designs:

Cube hardware: https://github.com/enjrolas/L3D-Hardware

If you're looking for details on how to start creating on these volumetric LED cubes, more technical details on Alex's "How to Draw Sweet Graphics for LED Cubes" Instructable here.

Step 1: The Idea

Did you know the top Technology Instructable of all time is a volumetric display? It's called a 8x8x8 3D LED Cube.

Before discovering this, we'd felt pretty isolated. There were only a handful of hobbyists experimenting with high resolution volumetric display, using techniques like spinning LED arrays and highspeed projector systems. Work on volumetric display was generally considered so expensive and complex it was performed only in places like the MIT Media Lab or secret departments of Samsung.

But then one day we found that there were actually millions of people reading about how to make beautifully simple volumetric displays with arrays of LEDs. While these LED cube displays were of far lower resolution than the displays of our dreams, it didn't matter - they were true volumetric displays. That said, each of these 3D LED cubes would take legendary soldering skills and between 100 - 200 hours to assemble (and an equal amount of time to program) -- but a few people would do it and they would become heroes to millions of others.

Excited to find we weren't alone in the DIY volumetric display seas, we tried to buy an LED cube kit. Only it turned out there weren't any.

And that was the idea. We would make the first color 3D LED cube kit with easy to share volumetric programs that could run on any cube in the world. And we'd make sure they could be assembled by anyone in under an hour. To pull this off, we'd need to make some hardware improvements and create a way for cubes to connect to each other over WiFi, so we got to work.

The L3D Cube was born.

Sooooo......how do I make one? What parts? What tools? Think this is an advert rather than an Instructable
<p>Hi!</p><p>I intended this instructable as a &quot;turning an idea into a product&quot; as-it-happens guide, and less technical, but I hear where some commenters are coming from, perhaps expecting a technical guide instead. </p><p>That said, some background on why I think a technical guide might be less useful than folks might think at first: Samtim, the guy on our team who did all the early heavy-lifting on the build, did handmake the first prototypes with a tiny reflow oven, hand stenciled solder paste, with carefully placed WS2812b LEDs on the custom PCB reed design you see in the instructable. WS2812b LEDs are super temperature sensitive, so in this homebrew initial setup, we would get 3-5% of the LEDs failing. To decrease the failure rate we needed to move to full pick-and-place + very large reflow oven production, for more even-temperature distribution. Then our failure rate dropped to nearly 0%. This means that homebrew versions of this cube will likely meet with similar barriers and I believe on the technical-front many more people will be interested in how to <a href="https://www.instructables.com/id/How-to-Draw-Sweet-3D-Graphics-for-LED-cubes/" rel="nofollow">create graphics for the cube</a>, rather than how to make their own cube from scratch. </p><p>That all being said, I'll add more details in this Instructable and might branch a separate one focused on technical assembly details anyhow, since it's nice to get that documented Instructables-style. </p><p>To start, everything in the cube is open source and is posted here. All schematics for the LED reed PCBs and base PCB are there and I'll post the lasercut profile for the housing soon!</p><p>Cube firmware: <a href="https://github.com/enjrolas/L3D-Software" rel="nofollow">https://github.com/enjrolas/L3D-Software</a><br>Cube hardware: <a href="https://github.com/enjrolas/L3D-Hardware" rel="nofollow">https://github.com/enjrolas/L3D-Hardware</a><br>Processing library: <a href="https://github.com/enjrolas/L3D-library" rel="nofollow">https://github.com/enjrolas/L3D-library</a></p>
<p>Hi</p><p>I had to RTFS to determine (hopefully correctly) how the cube is structured as I want to build one, but really huge (LED wider apart) but want at the same time to have it compatible with the L3D.</p>
<p>finally I don't get it. assuming pixel 0 is at 0,0,0 then the following are zigzaging throught the cube. But the code just uses x,y,z without any correction for y and z zigzag of the data line. Or is this hidden somewhere?</p>
<p>So, this is an 8*8*8 RGB LED cube but instead of using 5mm through hole type LEDs you've used 5050 SMD type LEDs facing two directions, baked onto basically PCB fingers. It seems to me that what you've created is a manufacturing process to mass produce these things rather than actually design a new device.</p><p>The new technique is cool but results in an obvious viewing angle limitation and an expandability limitation as those pcb fingers would block light too much at even a 16*16*16 cube (ie you couldn't see through it anymore)</p><p>I would also say that a youtube search will find tons of RGB cubes with this or much higher resolutions which are already being commercially sold (albeit for very high prices).</p><p>I appreciate the contribution in technique that you've made here but I think the additional limitations are too great and I say that having built a couple of the early monochrome cubes based partly on CHR's design. They take a lot of time, can't give the desired pixel density for a proper volumetric display and at a certain size the wire mesh blocks their transparency as well. That said, these are cool because only a few people have them; once they are mass produced they are no longer cool, just like the RGB, colour changing, individually addressable Christmas lights I see everywhere now.</p>
<p>Hi Razamatraz,<br><br>I failed to see what you called 'limitations' as limitations as all of the other cubes out there are based on hand-soldering big strips of LED's onto stripped wire, once they are done you get what you get. There's no 'expandability' as you imply.<br><br>So all cubes are equal in dimensions, if you get a 8x8x8 cube then you get an 8x8x8 cube, and that's that.<br><br>From all the cubes I've seen out there (and I've read through tenths of instructables and web pages and watched lots of YouTube videos). Most cubes out there are either low-res (4x4x4) and/or monochrome, they lack music support (apart from one, which I already wrote about and refuse to mention here due to the guy's awful manners with people and obvious intent to rip off those who are willing to entrust him with their money), &lt;none&gt; has embedded sensors (accelerometer) and &lt;none&gt; is wifi-ready. These cubes out there lack far too many features that this one has, than I would care to mention here.<br><br>And this cube is the only one I've found that's truly EXPANDABLE.</p><p>The LEDs being individually addressable means more than just power consumption, as you can spare lots of processing cycles avoiding having to code multiplexing for the LEDs (as some cubes use more complex bit-angle-modulation which is complex and takes lots of processing) and you can use these cycles instead, to build more elaborate animations (they even managed to chuck FFT analysis and display onto a single Spark core!). What you see as limitation I see as an enourmous advance. And the cube only draws 1.8 amps. I'd say not bad. Other cubes draw the same or greater current.<br><br>Honestly, Razamatraz, what's your point really? I fail to see it.</p>
<p>I wrote this two months ago and you obviously failed to read my second comment where I followed up on the exact advantages you mention. The key disadvantage remains though; the columns are too fat so even in 16*16*16 it would be too opaque to see through.</p><p>1.8 amps? Maybe average; this cube at full white would pull 31 amps if they are fully utilizing max brightness while a multiplexed model would pull 1/8th of that max but obviously be less bright. It does make a difference as up to 3 amp PSUs are dirt cheap, bigger gets expensive quickly.</p>
31A? I wonder where you got these numbers from in the first place. The WS2812B datasheet states a maximum input current of 1uA per component, so multiplying that by 3 and again by 512 will yield a little above 1.5A. If you try to plug in a source capable of delivering half of your estimated total current (and assuming the circuit would actually draw that), I guarantee you'd fry the board before you even see a single LED light up.<br>These aren't regular LEDs, they're low power, high brightness components each having its own embedded driver with a constant current controller. That's precisely what drives the price of each of these units - I believe once the price of these components is softened a bit, we'll see it reflected on the units as well.<br>The problem when we start making assumptions is that we end up having a 50/50 chance of being wrong.<br>You also complained about the LEDs being soldered to a strip and not fully visible from all angles. I beg to disagree with that too, as these are diffused and not just flat as you implied. Anyway, if you're still not happy, you might as well get these instead (http://www.adafruit.com/products/1938) and provided you have connected them according to the schematic you can make your own strips to replace the stock ones.<br>As for your late post, I can't be blamed for the way this website reorders all the posts every time someone makes a new comment. I didn't come out with this. It seems to reorder them by some strange algorythm. I also didn't go through all of the posts fishing for a specific comment.
<p>I don't know what Ii is referring to on that datasheet but there isn't an LED on the planet that can create useful light at 1 microamp (your calculation would actually result in 1.5 miliamps total based on that number). Just like every other 5050 form factor RGB LED, each color will be 20mA rated but th eIC will limit total current to ~50mA. At 20mA per colour * 3 *512 you get the 31 amps I mentioned. I don't know why they opted not to include current ratings the same way they did in the WS2812 datasheet http://dlnmh9ip6v2uc.cloudfront.net/datasheets/Components/LED/WS2812.pdf . You can see the same thing when looking at the 30 LED strings of these; they require 1.5 Amp power supplies.</p>
<p>You're right - I made a boo-boo when converting units for calculating the total current (I ommitted one zero in my calculation). At least I didn't hand it to Stephen Hawking (&quot;- Another Fainter&quot;).</p><p>Either way, as you pointed out, those 5050 strips sold everywhere (I happen to own one that responds to audio as well) typically draw around 1.5A (that will depend on the length of the reel attached).</p>
<p>Figured it out; the 1 uA is the current needed on the signal line to tell the chip it has a 1 bit.</p>
<p>Hi Razamatraz,<br><br>The big idea with the L3D Cube is to have enough <br>of these out there that people start to generate content for the cubes. <br> 3D graphics, music visualizations, simple games - things like that. <br>Without a critical mass of displays, no one will develop neat volumetric <br> content.<br><br>And making the cubes easier to assemble by two orders <br>of magnitude in time is how we're attempting to do so. This is a story <br>that has repeated itself in several industries in the past, and the aim is <br>to get it to happen for the volumetric display industry.<br><br>The <br>second-order hope is that this then seeds interest in high resolution <br>displays. Arrays of LEDs is probably not the best way to achieve a <br>100x100x100 voxel volumetric display, but at least the dream for this <br>type of higher res display will expand and the volumetric display <br>industry can finally begin.</p>
<p>Well, having looked at that design in more detail it does contribute more than just manufacturing methods to the field; the use of individually addressable LEDs rather than typical multiplexing is relatively new (at least to me). I'd be concerned that it would take more power since every LED is on all the time rather than 1/8th duty but might get away with a slower controller (24 bits * 512 LEDS = 12288 bits per &quot;frame&quot; with a minimum 50 us reset signal per &quot;frame&quot;) So, 100 fps needs 12288 * 100 signals in a maximum of 995 ms = 1.235 million bps total transmission speed. That's certainly doable, whereas the 74XX595 shift register driver with software PWM takes ~ 512 * 3 * 7 bits of data per frame plus another 64 bits of shift clock signal and 8 bits of latch signal for a total of (512*3+64+8)*7*100 = 1.125 million to get the same 100 fps but only with 256 colors. The equivalent in 16.7 million colors (24 bit) would take 41 million bps of signal so this design cuts data flow rate from controller to array by a factor of 40....that's pretty cool. I haven't calculated it for TLC 5940 drivers with hardware PWM but I suspect they still require double the data flow of your method and they are kind of expensive (although that price has dropped 60% in the last two years).</p><p>I am concerned that a retail of $399 is too much to get many people designing content for these, but given the time mine took to build I realize you can't charge much less; these things take a lot of effort.</p>
<p>I think some are too quick to jump on these guys when they are, in fact, offering a free, open source, project for us techies. We can discuss whether or not these types of Instructables belong here, but let's not be too pious since it may stifle future contributions. I'm okay with defining a new category if that makes some happy.</p><p>As for me, I was building &quot;color organs&quot; 40 plus years ago. This is the best improvement I've seen so far, a thousand times so. I might be willing to buy a kit, but here's another idea - use pre-wired LED strips which are actually pretty cheap. You cut them to the length you need. I've been using them for XMAS displays at home. Lately, I've been using the ones that don't have the external vinyl protection sheathing needed for outdoor use. That's been okay since the LED strips are not directly exposed to weather. Here, the strips are inside so not an issue. These strips have a self-sticking adhesive tape on the back.</p><p>My thought is to create the &quot;sticks&quot; using my new 3D printer (don't have any ideas yet as to a connector). Would have to run a wire back to the MotherBoard with the &quot;goesout&quot; signal if necessary. Adhere the strips to these sticks. There's no way I would try to solder SMD parts to anything. As these guys point out, I think, the least fun and hardest work of any electronic project is the packaging which they have worked on.</p><p>I haven't done this yet and I don't have any idea as to what challenges will be with the software. But, having some skills with C++ and Arduino, I think I'll try to do something with their ideas.</p><p>Thanks, guys, for sharing and being prepared for whatever criticism ensues. The feedback this community provides should be useful in determining what to do in the future.</p>
<p>Couldn't agree more.</p>
<p>I came to know this cube after having a less-than-satisfying experience with another cube maker (let's not name any names - sufficient is to say this one has been mentioned in here already). His design is actually a board that's hitch-hiking on somebody else's design (namely, Kevin Darrah and Nick Schulze's), is clumsy (he keeps launching newer releases, each one adds to the stack of fixes one has to perform to get it to actually work), requires countless hours of soldering LEDs to stripped wire (to make the panels), his boards are &lt;not&gt; open source (you have no choice other to buy them from him at &lt;his&gt; prices) - each new board comes out more expensive and requires purchasing of extra boards ('bridge' boards) because of the way this guy designed them.<br><br>The only feature similarity this guy's cube has with the L3D is the music support - with the added cost and burden of stacking yet &lt;another&gt; board to the pile to support audio, I shall add.<br><br>Nevertheless, yet unaware of the L3D's existance, I decided to go for it: I ordered a set of boards (baseboard and 2 'eliminators' - these are said to be optional, for those who want to build their own 'arduino' and 'chipkit' to stack onto the baseboard - for the 'bargain' (his words) price of $95.00 - ONE baseboard!) and waited, and waited, and waited. After 35 days, without getting any evidence of anything being shipped to me, I confronted him, just to get a stream of insults directed at me. I forgot to mention he actually expected me to pay $50.00 for a music module. Talk about weighing gold.<br><br>That was it - I decided then to part ways, and fortunately, I got my money refunded without having to go through PayPal's channels. The refund went straight to help funding an L3D order, which I just placed today. ;-D<br><br>Undoubtedly, the L3D is far more elegant, full-featured and easier on the eye and I can't wait to start experimenting in porting Nick's and Kevin's animations to this one.<br>Should be challenging, but also fun.<br><br>I am more than happy with my decision. Hope this helps others who may be in doubt whether or not to buy the L3D or 'some other cube' that's someone's excuse to make a quick buck out of people who are just on the lookout for a fun and educational project.<br><br>And as a bonus, I no longer have to cope with that other guy's 'charming' personality and hightened 'God' complex - as he certainly fails to understand the basic laws of demand vs. supply. ;-D</p>
<p>This page don't show clearly how to assemble. Its a shame because I'm an engineer and fully capable to recreate this marvelous ideia. It's a very big shame! You guys are thinking just in money and not to disseminate knowledge.</p>
<p>I think you're mistaken.</p><p>In case you missed it in the description, the L3D is a fully open source project. You can get all the schematics, PCB layouts, panel cutouts, everything is freely downloadable.</p><p>Cube hardware: <a href="https://github.com/enjrolas/L3D-Hardware" rel="nofollow"> https://github.com/enjrolas/L3D-Hardware</a> (in case you missed it).</p>
<p>What I would love to see is a way to take people who have made other cubes (like SuperTech-IT's board) and make it compatible with the L3D. The Fact is that they found a way to bring LED cubes to the masses, and that can bring a lot of change to the cube community.</p>
<p>Thanks! Agreed, hopefully mods for other cube designs to enable shareable content between different cubes are coming soon -- neat stuff is starting to happen already....check out the new all-things-led-cube forum here: </p><p>http://cubetube.org/forum/</p>
<p>Awesome, thanks! I will be checking up and seeing what I can do to make mine compatible. I'm using the SuperTech board, which seems to be the most popular Instructable on the subject.</p>
<p>Now if only someone would make an instructable on how to sue someone, everyone will be happy.</p>
Great idea and it looks awesome but correct me if I am wrong instructables is a place to come to learn how to build things or make things... When did it turn into an advertising post for kickstarted
Hi Frogz79 - I agree, and was actually hesitant to post this Instructable. However, there were enough people asking me about how to run a Kickstarter project of their own that I figured a post about the detailed day-to-day running of a Kickstarter would be useful to folks. <br><br>I was inspired by these guys' post for their product, which I found to be really useful: http://therussiansusedapencil.com/post/2794775825/idea-to-market-in-5-months-making-the-glif<br><br>But I understand where you're coming from. I'll be working on growing this instructable as we take this product forward and hopefully it gets more useful to folks.<br><br>-Shawn
<p>Personally I found it interesting. As a small business owner myself, there's a lot of things that go on &quot;behind the scenes&quot; that the average consumer doesn't realize, understand or ever get to see. I can relate to how you need to maximize spending on advertising, etc. and the constant struggle to balance it all. Everyone things success happens overnight. This shows the hard work and determination behind any good project. Kudos! I like the product, and when I get some extra folding money I might come see you for one. Right now it's a little out of my price range (unless I turn it into advertising somehow) ..</p>
<p>I did not thought about this, but it is a valid argument. Good thing you can put other people experiences to good use.</p><p>Fun side fact: Instructables shows just the matching advertisment for the L3D campain on kickstarter. Also, congratutalions on reaching your goal that early!</p>
<p>Yeah, while it was an interesting read, there is no how-to here. Perhaps instructables.com should classify those type of posts differently (easier said than done, where to draw a line?) and not make them 'featured'. </p>
<p>I think with the title saying something more along the lines of &quot;Development of ...&quot; it would check out as an 'ible for business development or something similar.</p><p>I like the concept and the end result and hope that you will update this 'ible eventually.</p>
<p>Even thou this project is awesome, I've to agree with you. This <br>instructable is basicly just a rather long diary entry, it does barely <br>contain any additional information valuable for an own build.</p><p>True, <br> the title clearly says &quot;Making of&quot; so I don't blame you. However I <br>expacted you to share at least *some* personal experience (general stuff <br> that could be done better, if you knew about that before) you gathered while working on your project. <br>You don't need to reveal all your secrets, but some tips and tricks <br>would be nice, especially because you got inspired by our community.</p><p>After all it's all about sharing. If you're just searching for a place to advertise with, I'd suggest to keep looking.</p>
<p>The primary benefit of using traditional LEDs in LED cubes is that they are visible from all directions, hence a cube will look pretty much the same from each face. The SMD 5050 LEDs have the limitation of projecting most light in one direction only, the direction they are facing. Did you put these SMD LEDs on both sides of the strips? Is there a visible difference in looking from different sides of the cube, especially top?</p><p>Cool build nonetheless and best of luck with your kickstarter :)</p>
<p>The picture in step 2 shows LEDs on one side only. That's quite limiting but it's still a clever shortcut, just put the cube in the corner of your room</p>
Try indigogo
This would be a awesome idea to make 3D holograms. if we could add about 720x480 small low level led's you could make a HD 3D model that talks or something. maybe in the future we will have led's that act like colored displays for one pixel on a HD TV now that would be amazing I would love to make a game using something like that.
<p>Using qdleds this will be done. Also the first LED 3d printer has been showcased, so there is not much time until amazing new lighting solutions appear</p>
<p>This is the third instructable I saw today that was just a useless add. This used to be a fun site, now its being taken advantage and ruined by people posting promo garbage. I usually dont complain but its getting outrageous. </p>
<p>I second that....</p>
Best idea I have ever heard
<p>Very impressive.. and good idea...</p><p>and Good luck..</p>

About This Instructable




More by sfrayne:Making of the L3D Cube: Idea to Product in 4 Months Into the Looking Glass: 3D Scanning Scenes 101 Free Range Color 3D Scanning with an iPad 
Add instructable to: