Interactive LED block wall

led_pixel_wall

[Dave Vondle] from IDEO Labs sent in the large LED pixel wall he built using BlinkM modules, an Arduino, and Flash to control it. The overall result is a blindingly bright, large, public display for people to interact with. The best part about the project is that [Dave Vondle] documents everything; from hardware to schematics to source code. Unfortunately, he was forced to remove the wall due to construction, but since every part of the project is open source, it lends itself to be easily recreated. I’m sure we’d all like to see a wireless controller hookup to play pong on the streets of Chicago.

12 thoughts on “Interactive LED block wall

  1. Ouuuchhh…126 BlinkMs? So expensive…and he’s just controlling them manually, not using the BlinkM internal scripting which is their main strength. I’m totally biased since I make and sell ShiftBrites, but they use the same LED and would have cost about $471 instead of $1300. The guys at metalab.at actually made a similar glass brick wall with ShiftBrites: http://www.vimeo.com/5714761

  2. Those shiftbrites are cool to know about. I wanted to have everything on a data bus so I could make the electrical grid the mechanical structure as well. Also since the wall is so big it is nice to have the ability to only update the pixels that needed to be updated.

  3. Not every part of the project is open source! The BlinkM firmware is closed. I do not have anything at all against ThingM on this point (A licensing issue, I understand), but I did want to mention that I recently got involved with the CYZ_RGB project that Matteo Caprari started a couple years ago to produce functionally equivalent open source firmware for the BlinkM board.

    CYZ_RGB has languished for a while, but I recently got involved and completed a port of CYZ_RGB to the ATTINY44 used in the BlinkM MaxM. I am currently in the process of enhancing the MaxM port to support 16 bit hardware PWM with logarithmic power output during dimming (to give brightness output that appears linear), and hopefully I’ll be able to add some other neat tricks too that are not present in the shipped firmware, such as light script enhancements.

    1. Ime a bit wazed off about the closed source firmware for blinkM
      they dont even make the blinks anymore,and the i2c led concept is locked away from us,we cant even develop ontop of it,i think it holds electronics back from further discoverys with i2c and led development to some degree,and what for? money so some one could mae a few bucks,i think it would of been better open source,i think they should try raise the money even put a donate to help the release let us people try team together and get it released.A few dollers each could easy raise the cash.i have the source code for the blinkM it easy to find when you know how.

      1. Sorry but did you look up CYZ_RGB? I (unofficially) host it now after google code closed. https://github.com/johnlaur/cyz_rgb The firmware is open source. A lot of others have extended it since as well. Compared to the last BlinkM firmware, the CYZ_RGB offers full 16bit PWM and a LUT for custom color calibration. It is tuned by default to be a pretty good match to the SRGB color space for most high brightness RGB LEDs, but the LUT can be adjusted . Also the BlinkM firmware has some kind of i2c bug which causes bus lockup in extremely large topologies. The bug was in the Atmel reference i2c implementation at the time; I had it too, so I am thinking this is the same bug.

        The thing that prevented ThingM from releasing their source code is that they proactively licensed a philips patent related to controlling RGB LEDs over i2c. The patent specifies that the i2c bus runs over “flexible” wire (like xmas lights and similar) so I can see their point. It is my understanding that the patent license prohibited the source release. In my own hardware I put many uCs on the same board and my bus ran only over PCB traces, and I drive long LED chains using high power mosfets.

        Whether it may have eventually found its way out or not is irrelevant, CYZ_RGB is better in nearly every respect. The only downside is that it does not offer as many preprogrammed sequences.

        In any case the shift-register based approaches that are de rigueur these days are considerably more cost effective than using attinys as led controllers. Though perhaps it would be fun to make a minimal port to attiny0

  4. Hi. I’m one of the guys who makes BlinkMs. I agree that shiftbrites are great for making large LED arrays, but what Dave is doing that IS much easier with BlinkMs is that he’s letting each individual unit do its own fading. That way, he has to send much less data to get each individual unit to fade from color to color. His hardware only has to send a color and a fade time and let the BlinkM do the hard work of interpolating between two colors and smoothly fading between them. That requires much less bandwidth and lets him address individual pixels in the array without keeping track of what’s in the rest of the array. Also, because he was buying such a large quantity, we gave him a discount, which we’ll offer to anyone else who’s buying that many. Send a note to blinkm@thingm.com if you’re interested.

    Also, fwiw, John’s doing great stuff with the cyz_rgb project. We will continue to ship BlinkMs with our own firmware, but if you’re comfortable with an AVR programmer and embedded systems programming, we encourage you to take a look at that project.

  5. holy crap, i live in chicago….id like to check this out sometime…

    is it always on? logan square aint far from me!

    ideo labs, do you guys have a hackerspace goin on up there? would be cool if there was one in chicago somewhere…

Leave a Reply

Fill in your details below or click an icon to log in:

WordPress.com Logo

You are commenting using your WordPress.com account. Log Out / Change )

Twitter picture

You are commenting using your Twitter account. Log Out / Change )

Facebook photo

You are commenting using your Facebook account. Log Out / Change )

Google+ photo

You are commenting using your Google+ account. Log Out / Change )

Connecting to %s