We built a custom IR transmitter that uses a single high power IR diode and a collimating lens like described in
http://lasertagparts.com/mtoptics.htm . The idea for this was not to use this to send commands to the IR receiver but to detect uniquely modulated reflected IR energy off a track ball and seek the reflected energy. It has promise but it is not there yet mostly because we tried to design our own differential receive signal processing and used a standard set of IR phototransistors with a 5 degree receive pattern and it was difficult to acquire the target. We might throw in the towel on the target designator idea and change the modulation to a standard carrier frequency with a unique bit sequence so it can be used has a high power IR transmitter. If we do that, it will be very directional (at least until it is reflected). It will be interesting to see how much jamming is going on at the competitions. I’m assuming that teams will swap out the IR receiver module for different carrier frequencies but that still may not be enough. We attempted to do some IR spectral analysis by just looking at the frequency domain from our IR photo transistor with a bunch of different remotes including a TV-B-GONE (4 high power IR LEDs) and there are harmonics and spurs all over the spectrum. There do seem to be some nice quiet regions, though and that’s where we have been concentrating.