OpenTX is open source firmware for RC radio transmitters

Tridge also happily uses an earlier version of OpenTX called er9x I believe. I've worked with him on a bit of the firmware you're talking about here, but it's not the same stuff. The most impressive thing about the SiK firmware you're talking about is the stupid range you can get from them on a good day; line of sight can be tens of kilometers (say 40+), and could potentially be used to communicate with satellites, all while using a max of 5W transmit power at 900MHz. There's plenty of more info on both the CanberraUAV and 3D Robotics blogs from him and others about both these topics. The following talk as LCA was about such satellites, and there was some talk about using these radios for these open source satellites.

There's also OpenLRSng which is open source 433MHz hardware and firmware for control links.

That site says you should look at http://plane.ardupilot.com/ for documentation.

>I know that's a neck beard comment but have things really got that complicated that computational power is required? I can understand mixing for helicopters but not much else.

Mixing is needed for pretty much anything and it sure as hell beats mechanical setup. For anything of even medium complexity it's a must. For example in a simple three channel glider (aileron, rudder, elevator) it's so much easier to just have 4 servos (1 on each aileron in each wing) and then do the aileron mixing in software. Otherwise you'll have to arrange the linkage geometry to get differential travel in the ailerons (for aerodynamic reasons you want more up travel than down travel) and then make sure the end points and central trim are all correct. Much easier to just stick the servos in the wing, set the linkage roughly and then trim it from a software interface. In the old days you'd have crazy stuff like left and right servos to reverse the direction of travel, now it's a software setting away.

And that's a simple example. Still in gliders a competition F3F glider will need much more mixing. You're basically doing laps between two pylons 100m apart. The world record is below 30s, so that's 120km/h of average speed including 9 full U turns. To actually be able to do those fast turns those gliders will actually move all 6 surfaces (2 ailerons, 2 flaps, 2 V-tails) at once, dropping the ailerons and flaps to increase the lift of the wing during the turn and using the V-tails as elevators. Then to land the plane you'll put it in "butterfly mode" moving the flaps way down and both ailerons up to get it nice and slow for landing.

R/C airplanes are actually very advanced stuff. The jets and scale stuff will have complex sequencing for landing gear retract, the gliders will do insane stuff like do over 800km/h dynamic soaring, so you need to have exponential mapping between controls and outputs to have precision in the center of the sticks. So radios have evolved to match those capabilities. Radios have been mixing for decades now, and have had replaceable firmware for a long time. It used to be that to get a really flexible radio you'd need to spend 1000$. Now you can get a cheap Chinese made radio for 100-200$ and put an extremely flexible open-source firmware on it that will do all the things the high-end radios do and more. Software is eating the world indeed.

For a toy, that would be enough. But most moderate to "prof" (geh) RC guys want stuff like: Gear up-down switches (for jets) - throttle cuts, channel mixers for flying wings and scale models, flap/butterfly switches, dual-rate and expo settings (for acrobatic flying) snaprolling, telemetry (for drones), RSSI levels, microtrims, etc.

And then there's the guys like Warthrox who are totally insane in reaction-speed: http://vimeo.com/92400774

Most RC transmitters are digital and use the 2.4 GHz spectrum. They do channel hopping, they do integrity checking (CRCs for the most part), have configurable channels and mixing. Some also do telemetry and support multiple profiles for the channel and mixing settings.