Short version: Using an Android tablet, you can draw things on our Glass Block LED Matrix from the street, and it’s pretty awesome. Video here, photos here.
Things have progressed recently on the Glass Block LED Matrix which Chris Davis and Paul Vincent started. For a couple weeks, the code was already in place to let Processing talk to it via simple serial commands to the Arduino & ShiftBrite shield. We wanted to use the tools from Project Blinkenlights to control things over the network; while this didn’t entirely work as planned, the project offered a lot of ideas and inspiration.
The most recent addition I made was the inclusion of oscP5 to the Processing sketch to let it listen for OSC (Open Sound Control) messages. As it happens, a brilliant piece of free software already exists (Control from Charlie Roberts) which turns Android/iOS devices into control surfaces that send out OSC messages. On top of this, Control comes with a handful of example UIs, one of them being “Multibutton Demo” which provides a UI with an 8×8 button grid, sort of like a monome. (The tablet in all of the photos is running Control with that Multibutton Demo UI.)
As our LED matrix is 7×8, this UI was a good initial match. I set Control’s destination URL/port to the backend machine that was running Processing, set the sketch to parse the pretty simple OSC messages Control would send out at every button toggle, and then I was able to control what was on the LED matrix by drawing on that 8×8 grid on my tablet.
I finally got to show it off outside on Tuesday evening when it was dark, and it’s working pretty well, as the video shows.
- Making a Control UI that allows for color control. These are RGB LEDs, after all – we can control intensity and color, not just whether they’re on or off.
- Making this web-enabled. I think Control allows this?
- Fixing the glitchiness that I didn’t show in the video; something cryptic is going on on the Arduino side.
Check out the github project here and the project wiki page here.
At Hive13 we have a 5′ × 5′ glass block wall/window in our space and the first thing we thought of when we saw it was a 7 × 8 pixel grid.
We want to build a programmable full color, lo-res display using RGB LEDs.
To do this we obivously need LEDs – lots of LEDs. The brighter the better.
We could use these nice ShiftBrite modules, but we’d really rather go all out and get the ShiftBrite’s big brother – the MegaBrite.
ShiftBrites are about $3.50 each and MegaBrites are about $7.75 each. And we need 56 of them. We’ll also need some cabling to connect them all up. The cables are about $1.50 each too. Finally, we’ll need a a good power supply to power it all; that could be as much as $100.
($3.50 + $1.50) * 56 = $280
($7.75 + $1.50) * 56 = $518
All told, we’re going to need somewhere between $300 – $700 to do this. We’re shooting for $400 here. If we go over and get to $700, awesome. If we only get to $300, we’ll make do. Any funds raised here, but not used, will go directly to the Hive13 general fund.
Here’s the stuff on our shopping list:
Here’s the project page with our progress and prototypes so far:
CJDavis and I were at the hive until about 1:20 am this morning working on finishing up a second strand of LED lights for the Glass Block Matrix LED Display.
In this image we are diffusing the LED’s with a sheet of paper towel which causes a nice soft glow, however it greatly reduces the brightness. Therefore we are leaving the paper towels off even though they make for a great picture and viewing experience from inside the bathroom.
We have had some technical issues with both the RGB LEDs and the wiring system we are using. A good amount of the LED’s seem to have an issue where they will turn on even when the anode wire is not connected to anything. We are talking nothing, like it is just soldered into a proto-board. Current speculation is that either the LED’s are bad or the proto-boards are bad.
Another problem we have been having is that originally we were using Cat-5 to wire up each anode for the 8 LED’s in a column, however the current in the twisted pairs was enough to induce a current in the second pair causing issues with LEDs coming on when they are not supposed to. This is a fairly limited problem so far as the LEDs only glow faintly. In any case we are altering how we wire up future anodes.
I had some spare time down at the hive last night so I finished assembling the framework for the RGB Glass Block Matrix display. The next step is going to be to mark out where to drill the holes for the LED.
After that is done we can start soldering up the LEDs.
After our regular weekly meeting last night, Dave Menninger mentioned on the way out the door an idea he’d been kicking around: wiring up LEDs in the 5×7 glass block window above the entrance to the hive. Enthusiasm was expressed, and several of us came back in put together some tests with a couple of LEDs.
This was deemed a great success, and the project brainstorming quickly escalated from a simple 8×7 array of single color LEDs to a full color display capable of video. The video part we will hold for phase 2.
So, this Friday, we will be building phase 1 of the Hive13 windowmatrix. How capable it will be depends on just what parts we are able to procure by Friday.
update: The rainbowduino LED controller board has arrived – thanks to Great customer service from NKC Electronics – http://www.nkcelectronics.com/ I can definitely vouch for that!