TouchFactors

We created a fully assembled multi-touch surface for the first time the other day. It looks rather nice, and produces good blobs! We believe there are a number of innovations in this process, which really simplified the construction time and cost:

• Use of infrared LED ribbons as our IR Light source
• Use of LED ribbon channel with neoprene spacers for the display frame
• Double-sided taped perimeter for applying Sulky Solvy and Rosco Grey

We captured the process on video and with photos. The photos are numbered and walk you through the whole process step-by-step, assuming you’ve ordered a few supplies first.

[Note: The idea of using IR light with plexiglass (FTIR) is from a paper by Jeff Han (http://www.cs.nyu.edu/~jhan/ftirsense/index.html).  If you are planning to sell something based on this you should probably check with him to make sure it is ok.]

Photos

View the photos with step-by-step instructions.

Video

High-Quality Video

Once we had made a small IR LED array and modified an X-Box Live camera to be an IR Cam, we found that testing for blobs was a pain in our backside.  We could get blobs, but not every time… we had too many moving pieces.  We figured that it would be easier once we built a frame to hold the LEDs against the acrylic at the right angle and had the screen in a consistent place with respect to the camera.  But every time we experienced a problem, we had to wonder:  Is it the surface, or the IR array, or the camera.  To make a long story short… It just wasn’t a reliable way to test compliant surfaces.

Our next thought was to build an array of visible light LEDs. The thought behind this was that with visible light, we should be able to see the blobs with our eye, and that took both the computer and the camera out of the equation… two fewer moving parts.

The thought of building another LED array was not a pleasant one though. Luckily, Jim had recently been doing experiments in Tele-presence and had a strip of white LEDs left over. Out of pure laziness, we decided to use that instead of building another array from scratch. As a bonus, the strips are a reliable product, so we no longer had to worry about whether we had a bad solder joint (one fewer moving part). We attached the strip to our test screen using scotch tape and we were rewarded with beautiful, and consistent, white blobs… (and that got us thinking why don’t they make these things in IR?)

Anyway, we highly recommend using the white strips for testing and experimenting. They are about $18 a foot, and they are reusable. You can find them at several places around the net, but please buy them from our friend Greg Thorson (since he was nice enough to take a chance on us and produce IR strips for us… which as far as we know had never been made before).

Here you can see a white light test using a sample piece of Rosco and Sulky Solvy as the compliant surface.

We did a quick test of the infrared LED ribbons to see how well they generated blobs. For this test, we just wrapped part of the LED ribbon reel around the acrylic. In reality, you would trim the ribbon to an L-shaped length. We’re still trying to see if a U-shaped wrap, or a total-perimeter wrap, will produce better results. But wow, the blobs are great with just an L-shaped wrap! Keep in mind this test is showing a direct touch of the acrylic with a person’s fingertips. We’re exploring different compliant surfaces, such as silicone, Lexel, and Sulky Solvy, and will post our findings with them later.

[Note: The idea of using IR light with plexiglass (FTIR) is from a paper by Jeff Han (http://www.cs.nyu.edu/~jhan/ftirsense/index.html).  If you are planning to sell something based on this you should probably check with him to make sure it is ok.]

High-Quality Video

Earlier this year, I was working with flexible LED ribbons for a research project investigating telepresence lighting and glare. I found that the ribbons, connectors and power supplies from EnvironmentalLights.com were pretty handy, and saved a lot of time. So when Jon and I began to build the infrared LED lighting circuit for our multi-touch display, we realized there had to be a simpler way. We looked all over the Internet, but did not find an easily ordered source of LED ribbons in the infrared wavelength.

So we contacted Greg Thorson at EnvironmentalLights, explained our situation, and convinced him there was an army of multi-touch prototypers that would love to simplify the LED lighting step. The result? EnvironmentalLights now has Infrared 940nm LED Ribbon sold by the foot on its web site. If purchased by the foot, it ranges from $20.10 to $14.63 a foot. You can also buy a 5-meter reel for $207. One meter contains 60 LEDs, and can be cut to any length. Don’t feel like soldering? Not a problem. You can purchase a 60-watt 6-12 VDC dimming power supply, a female mini power supply plug, a power line connector cable, ribbon-to-ribbon connectors, and connectors with a bit of wire between them. All of these parts can be found at the bottom of the product page for the reel.

The infrared LED reels we ordered from Greg have arrived, so our next step is to test them with a piece of acrylic, and see how well they produce FTIR blobs.

The first lesson we learned in our experiments is not to buy IR LEDs from a bin at a bulk electronics store unless they have the specs.  We bought 50 of these little black IR LEDs for 15¢ a piece.  We were hoping that since they came lined up in a row, it would make them easier to solder together.  Several months later, we still can’t figure out how to make them light up… we ended up running to radioshack and paying their outrageous price of $2 per LED for our first tests of the IR Cam.

Those little black specs are IR LEDs... or at least we think they are since we haven't actually gotten them to work yet.