V. 2.0 Complete

Well the new circuit was a success, although it wound up costing much more than expected. It did however double the range off of v. 1 and it seems like it made it much less sensitive to positional variation of the emitter. (Eg, it works better for a wider range of angles)

Final schematic was 6 IR LED's wired in series with a 15 ohm resistor all on a 9V battery. I upped the voltage to keep current down and it also allowed me to leverage parts I already had.

Parts list
4 IR-LED's 1.99 (each)
1 pack 9v battery clips 1.99
1 set breadboard jumprers 7
1 pack 22 ohm resistors .99 (Picked these up only since a lot of the equations I'd been working out seemed to require this resistor, this was not an immediate need)
1 button, on-off, spst switch 1.99 (I think)
1 9v battery ?

I did however get to test the rig on the projection screen and it works great. We did uncover one major problem. The angle between the camera and the emitter has to be on a virtually horizontal plane both to each other and to the axes of motion. Now that would usually mean you'd have to hang the wii mote on the level at eye level, which is far from optimal (unles you've got a CRT that already stands at eye level in which it's moot) but for a projection environment, or in the case of your having to view it from a laptop the odds of getting the wii mote in such a favorable position are slim. This needs to be adapted for.

However, with funding being essentially out for this phase, I will attempt the finger tracking demo next, just as proof of concept and then we're going under the hood boys and going to see if we can't make these little babies useful.

Short Circuit??

Ok, well I've been out of physics for 2 years, and I've been out of high-school physics where I probably did this last for 4 years, so I can honestly say that my mind has slipped a lot when it comes to simple circuit design. I'm pretty sure the guy at Radio Shack last night sold me the wrong parts, but that's not too bad only a dollar or two wasted on this endeavor.

What it boils down to is this. Until I can get myself a pair of swanky built in glasses with IR Led's I'm going to need to find a way to get around it. Not owning a wii, I can't even just hold the sensor bar on top my head. So my solution is to build one. I've got all the components I think I need (with perhaps the proper current limiting resistor) and now it's time to crunch the numbers.

My IR LEDs are supposed to run on 1.2 v forward current, and 100mA of current. For the time being I'm looking at running the whole suite off of 3v. Possibly because I think that I can get most watch style batteries in 3 volts and anything less seems very difficult to work with.

So with a little number crunching on my part (and a lot on the internets part) I came up with that with a single strategically positioned resistor I should be able to cause a 1.5v drop across the LED's with the proper current. Somewhere along the way, the internet disagreed. I'm going to run my numbers again tomorrow and see where I went wrong, but I'm almost certain I got it correct. So at the moment with everything wired in series. 3v DC from 2 AA 1 15Ohm resistor and 2 IR LED's we're getting a signal that the Wii can read. We have component 3 in its infancy.

I want to decrease the dispersal on the LED's since we're not trying to light the world. I also want to optimize the intensity on the LED's because I strongly feel that we're running really low on the efficency scale.

There was an interesting idea suggested the other night by the guy at radioshack. Something that bears merit to investigate. He had mentioned that eyes reflect IR (hence Red eye and such) and that it would be possible to leverage that fact to build an eye tracking system. I'm calling in my buddy from Hopkin's who's kept up with this to see if there's any merit in it. After that come the health concerns.

We have ignition!

Ok folks,
The majority of the hardware has been acquired and some of the software testing has been done. I'd bet I can get a working copy of the head tracking software up by Friday evening at the latest.
Only hangups so far: Make sure you know what you need if you're going to build your own sensor-bar/clip on IRLed units. I went to Radio shack and realized I didn't know how to push the right amount of current through the device.

The salesman (while being very excited to my cause, handed me an equation that had been done, flat out wrong. I double checked it when I got home so all is well, no parts lost but some returns must be made.

The USB-Bluetooth interface works like a charm. Took a little wheedling to get the Wiimote to talk perfectly with my system but that was all over with in about 10 minutes. I've verified that it is indeed talking with a 3rd party library called WiimoteLib v1.2.1 . There sample application sheds some serious light on what the Wiimote can and is doing every second. It's remarkable really. Such a small, rugged, ergonmic device with so complex underpinnings.

I've gone with C#.NET as my language choice, not because I think it's the best, but because I know it best and all of my readily accessable testing gear will be the .NET framework. I do intend to develop for Linux and Mac X, yet that's a longer term goal. I need to get as much of this up and running as quickly as possible if only so I can maintain my own momentum.

There was another glimmer of hope. Supposedly RIT's robotics team is already working on similar projects. I'm hoping to set up a face to face with them at some point so we can combine ideas and hopefully learn from each other. As I've said, I've got an extreme deficit of manpower. So anyone who wants to help me on this little foray into the unkown I would love to hear from.