Building the Frame (And Testing Some)

Well, I need something to hold the plexi glass right? If you see the post earlier in my blog about testing, you'll notice a 3d model. I scraped the little cut out in front, I thought it was a bad idea. I also added another 2 by 2 bored frame around the outside where the plexi glass isn't. This provides somewhat of a closed area for all my LEDs and wiring. I only needed 1 sheet of plywood to make both the top and bottom frame.

Measure twice.

Now that both frames were measured and drawn out, I propped the plywood up on top of some other pieces of wood to prepare for cutting. I just used a skill saw and got as close to the corners as I could. I finished cutting the corners with a hand saw.

I wanted the edges, and the top, to be fairly smooth, so I used an electric sander.

Here are the two finished frames. If I were to do this same thing in the future, I wouldn't make the bottom frame smaller. I would just make both frames the same size. You'll see why shortly.

The frame that is on the floor is the big one, and its the same one that will be on the very top of the table. I then glued my LED rails to the frame. I didn't get quite a good picture, but you'll see them all glued down in the background. I used spray glue, so I had to spray it in a cup first and use a cue tip to apply it.

If you look at that last photo, I also applied some to the plexi glass between the glass itself and the compliant surface. I wanted to make sure the vellum didn't slide around. I knew it wouldn't after the table had been set up, but I'm going to be moving the glass around a lot. This step isn't necessary.

It'll be hard to explain exactly how I'm building my frame. You'll catch on if you've seen the CAD 3D model and know I'm using 2" by 2" boards. Which, as you probably know is never exactly 2" by 2", its about 1.8" by 1.8". They sell it thinking that you're going to put drywall on top of it. Since one of the boards will be sitting on top of the acrylic glass, the LED rail, and the plywood, and one will only be sitting on the plywood. one board must be cut a bit thinner. Luckily I was able to get access to a nice table saw, and accurately cut 11/32 of an inch off the boards.

I ended up with thin strips at the end so it looked like this.

Then I had to cut these pieces into the size I needed.

When I got home I laid all the pieces out on the frame, which I already had the rails glued down and already had the acrylic on top of that in between the rails. Then I glued them together so they wouldn't move when I drilled.

The bottom frame, the one I glued the rails onto won't have anything screwed into it. At least for now. I set up both 2 inch frames and screwed the other piece of plywood into it.

You can't tell in the picture above, but where the (currently) top piece of plywood hangs over, is where I put the other 2" frame and screwed it in. My final product, all set up, and flipped over, looked like this.

I can now officially make some tests!!! I got my webcam set up perfectly underneath the frame, and was able to power up all of the LEDs perfectly. This is an extremely old laptop, so a lot of the multi-touch software didn't work on it. The graphics card just isn't powerful enough. So, the best thing I ended up finding, and since this is just a test its okay, was MTmini. It's actually a really great application. For the final table though, I expect to be running linux.

Seth (cerupcat) posted about this application on the NUI forums and thats how I discovered it. You can find his post here. Under the old archived downloads, I downloaded the v1 software package. You can download that directly here. Setup was EXTREMELY simple. Though, this software package was designed to sense black blobs instead of white ones. This was a simple fix. If you go into the folder 'Touchlib_Tracking_Software' and edit config.xml. All you do is delete the line that says "". Once you've done this simply go back to the root folder and double click on '1) Calibration'.

Once it looks good click the escape key. Keep in mind these aren't the only settings you can adjust. You can also adjust the camera settings directly. After you're finished calibrating, go into the demos folder, then the c++ demos, and run smoke demo. Mine turned out great!

Also, here are some images directly from the webcam so you know how much you can do with software! Anyway, I'll go more in depth about software and everything else in a separate post. This was mainly just for test purposes, and because I was really excited to try it out after all that hard work! ;)

Works Sited:

The Comunity. NUI Group - Natural User Interface Group. Web. 21 June 2010. <http://nuigroup.com/>.

Soldering the LEDs

This was actually a very tedious process, though I did get really good at soldering. I started off by bending all of the LEDs in the correct direction positive to negative and negative to positive.

I bought 4 1/16 inch blue strips of heat shrink, and 3 1/8 inch clear heat shrink to cover the resistors. I ended up having an entire strip of blue heat shrink left, and about half of my clear heat shrink. I measured and cut as many pieces of heat shrink as I needed for every LED.

After I cut the heat shrink, I put it on the LEDs. A few times I forgot to and had to un-solder everything, that sucked.

Then I put on the heat sinks, which was sometimes hard, because I still needed enough room to solder.

I simply melted some solder on the tip of the soldering gun and applied it to the wires. You have to make sure that you get the wires hot enough so that you don't have a cold solder.

Second I soldered the resistor.

I usually had to cut some of the LED wire off so that the resistor would fit between the two. I would use this spare wire to solder between 2 of the LEDs when they didn't reach each other.

I shrank all of the heat shrink with the tip of the soldering gun, plus, it looked cooler that way.

The final product looked something like this except with every single LED soldered.

Every single time I finished a set of 3, I tested it with a multimeter to see if the components still worked. When I was finished with every single rail, I soldered a wire along the positive sides, and a wire along the negative sides. I then tested that with a multimeter.

Through my regular camera, the final result looked something like this.

Through the IR camera, the results looked like this.

I am extremely pleased with the results I got. I wouldn't do anything different if I were to do it again. I might add more LEDs, but hey, I can still do that if things don't work out.

Works Sited:

Hewes, John. "Soldering Guide." Welcome to the Electronics Club. Web. 21 June 2010 <http://www.kpsec.freeuk.com/solder.htm>.

Resistor Specifications

Using Ohm's law, which states that voltage = current/resistance, and using Kirchoff's voltage law, which states that the sum of all the voltages in a loop must equal zero, I was able to find the desired resistance I wanted.

I wanted to use the computers power supply 5 volt output to power all the LEDs, so....

5V - IR(Ohm's law) - 3(1.6V) = 0 (1.6V is the primary operating range of the LEDs)

+----|>|----|>|----|>|---/\/\/----+

So the resistance of the resistor must equal 2.2 ohms. Thus, I searched digikey.com for some 2.2 ohm resistors. Since each resistor dissipates about 22 mW, a 1/4W resistor should do the trick. I ended up getting the OD22GJE-ND resistors.

Click here for the resistor specs.

Testing

This is probably a post I should have made a long time ago. Before everything I do, I test it. Through past experiences, I've realized that this is very important to do. You don't want to build something that takes you a few hours, then figure out that it doesn't work. I haven't taken many pictures of my testing, because, well, I'm not usually sure if I'ts going to work!

These are the LEDs that I used in my LED rails.

I also tested the IR rails, and if they would fit the plexi glass properly. I tested drilling the rails. I tested to see if the FTIR effect would work, and I tested the compliant surface. So, I've tested nearly everything so far.

I also designed a CAD model in Blender for my table, a basic guideline of what I needed to do. This showed me how much wood I needed and it what size.

To download this Blender file in .blend format, click here.