Replicating the fancy touch sensor that uses anything

[Sprite_tm], a name many of you will recognize from these pages, has wasted no time in replicating the latest cool thing in a much simpler fashion. En Garde is a touch sensor that can detect up to 32 different points of contact on… whatever you use as the surface.  He couldn’t sit idly by and let the Disney funded one from yesterday keep the spot light. As you can see in the video, it works pretty well. If he didn’t tell you that his can only detect up to 32 points as opposed to the 200 of the other, you probably wouldn’t even notice the difference.  Of course, [Sprite_tm] also shares how you could easily beef his up to be even more precise. You can also download his source code an schematics from his site and give it a try yourself.

22 thoughts on “Replicating the fancy touch sensor that uses anything

  1. Can we please have a post about how we can make our own touchsurface with position determining? This projects are cool, but I need to detect a position on a surface that is 20cm x 3m.

    1. If you want a single axis, you’ll want a taper of surface area on each of two different interwoven sensors. e.g. █▉▊▋▌▍▏ or ◣ where each of black and white are the two different sensors. Then you measure the capacitive proximity to to each sensor and linearly extrapolate.

      The other thing: The closest you can measure is approximately proportionate to the surface area of your sensor.

      1. I quite doint get it but am sure youre trying to give me a very useful info. Can you link to an image or an article so I can read more about this?

    2. I recall someone, somewhere had turned their wall into a touch surface.

      Was as simple as two piezo mics mounted on the wall. Tap the wall, measure the time until the tap is received at each mic, triangulate the position. This can only sense the initial tap.

      Surface acoustic wave touchscreens work similarly, but actively generate ultrasonic waves/pulses in the surface, and detect how pressure from a finger reflects them; so they can sense prolonged touch. Just like sonar.

      Given the size of your surface, that’s what first came to mind.

      1. Do you have a link? Because it sounds amazing to at least experiment with. I ant believe that this will be acurate enough.

      2. Sorry, I wasn’t able to find it again. If anyone else remembers this project and knows where to find it, please chime in – I’d like to see it again as well.

        Speaking of accuracy, what are your requirements? Transparent surface? Multitouch?

      3. Both. I need transparency and multitouch. For experiments it would be enough to have a dualtouch.
        Its really hard to find good sources with this conditions.

        I’ve read this entry with appetite. I have currently hooked up some small transparent Resistive touchscreens on an Arduino. I use those cheap Nintendo DS screens. But these doint support multitouch; But I can learn from them.
        I doint know how I can build a (ideality gap-less) surface in the dimension of 20x3000cm. I came to the point to look into custom made screens, but those are expensive, if even possible in extrem small quantities. Therefore I try to build it myself. As a beginner. Its hard! =(

      4. That’s a very limiting set of requirements.

        Capacitive and resistive touchscreens both have transparent conductors on them. While it wouldn’t be impossible to make a giant touch surface with an ITO (indium tin oxide) matrix on it, I think only the most hardcore of us would attempt it on such a scale (like Jeri Ellworth)! That rules out those two technologies.

        Multitouch also rules out both SAW, and infrared in the style of having a ring of emitters/detectors around the border.

        That really leaves only one practical option with decent accuracy. Project infrared light on the back of a pane of glass, so that when finger(s) touch it, a camera can detect the point of contact. That’s the “detection of touch via frustrated total internal reflection” approach in the MIT paper Stefano linked.

        Capacitive sensing as described in that paper has poor accuracy. Scanning laser rangefinders may be too difficult. Knock tracking is the approach I originally described (thanks for the link, Stefano).

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