I believe that your projection screen is the culprit.  If you are bound to the type of screen for so long, you'll have to create a special pen for it.  A backwards facing pen, or a pen with LED in the front, but tilted at an angle so that your wiimote sees it directly would do the trick. Much cheaper than buying a new projection screen. 

I might make a pen with a LED facing backwards, that sounds like a good idea, as a last resort pen to get out of a pickle such as this one. As for buying a new screen, I don't have to, it wasn't mine I usually give presentations on site, so I use someone else's projector and screen. I just bring my Wiimotes and my pens.


Do you mean lower the sensitivity? Usually I leave it all the way to the max, because the few times I played around with that, it didn't do any good. If you mean the screen was scattering IR light in a way that confused the Wiimote camera (like, I suppose, seeing "ghost spots" or something), then presumably it would have registered clicks, even if they were detected at the wrong place.

The problem during my last presentation however was that it seemed to have a hard time detecting clicks at all. This is why I thought perhaps the screen was a little to directive, i.e. it didn't scatter the IR light in all direction like a plain white wall would, but rather reflected the light more like a mirror would: I figured the IR signal just didn't make it to the Wiimotes. This would explain why I had to have the Wiimotes facing almost perpendicular to the screen, and I had to hold the pen quite perpendicular too to get something out of the setup.

I'll give adjusting the sensitivity settings a shot.

I use 2 pens and 1 wand, all of which sport a Vishay TSAL 6400 LED. One of the pens is powered by 2 AA alkaline batteries through an appropriate resistor, the other by a single AA alkaline battery, and the wand by a LR1 alkaline battery. All batteries were fresh. Both pens worked sort of okay when held at the proper angle, but the wand performed really badly, presumably because the LED was shining to the side.

I think I'm going to try making a new pen with a modified LED: I'll drill a half-spherical hole at the top of the LED and glue a small glass bead in it to replace the dome of the LED. Hopefully it'll help reflect IRs backwards. I'll do some experiments and report back on the intensity of the reflected light with this mod.

UPDATE:

I made a hemispherical cutting tool to drill a 4.75mm spherical hole in the tip of a 5mm LED. I drilled an ordinary RED led and glued a 3/16" bearing ball in the tip, and results looked very promising, as the ball reflected the LED light in all direction and made it look bright even from the back.

So I drilled an IR LED and the result is disappointing: the ball doesn't shine at all. The only reason why I can think of is because the superglue (cyanoacrylate) I used to glue the ball to the LED may be IR-opaque. I know it's not the steel of the ball though, as I've tried to shine an IR LED on it and it looked really bright in my digital camera's screen.

I'll investigate some more. Another possibility is to stick the bearing ball in front of the LED (without any modification to the LED) and assemble them together with some kind of IR-transparent tube.