Hello everybody,
I'm thinking about some large FPV sailplane with APM for next summer, and I have a couple of concerns about the vertical speed measurements in RC world.
- It is possible to calculate total-energy-compensated vertical speed (true thermal speed) purely inertially, i.e. based on gyro and accelerometer inputs only. Will that be the best solution? How noisy are those sensors assuming the motor is not running (and the plane is heavy, say 2+ kg)?
- It is possible to calculate compensated vario speed based on pressure and speed sensors input. Will that be better or worse than the inertial approach? How fast and how precise those sensors are?
- Putting energy compensation issues aside, if I build some kind of "real" vario over the pressure sensor (something like measuring pressure in a chamber with precisely sized tiny hole connecting it to the static pitot tube), will it be any better than simply calculating the vertical speed from regular pressure sensor input? Is there any factory-made digital varios available?
- Getting back to the compensation topic, if I'll equip the plane with the "real" (sorry, this word again) compensated pressure probe, like on full-scale gliders, will it do better than any of the approaches described above? How hard do you think will it be to calibrate the thing?
What I'm looking for is the RC soaring experience which would be as close to the real thing as it is possible without feeling all the accelerations with my own butt, and realistic vario seems to be a vital part of the idea. Also, I suppose, that since it is the most important instrument in the full-size glider, it should be as important on it's RC model.
If I it will still be fun after some time, I'll probably get head-tracking goggles and pan-tilt camera, but the vario issue looks muck more important at the moment, and I don't want to start buying things without ensuring I can get at least the most important thing to perform good enough.
I have very little experience in RC flying, much less than in autopilot theory or flight dynamics, and even less than in full-scale soaring, so any input from you guys will be really appreciated. Please treat me like an ignorant child and don't hesitate to tell me if I'm thinking in totally wrong direction, or just throw a link to the appropriate resource if anything I'm asking about have already been detailed and discussed somewhere else. Maybe I just did my googling wrong.
Another topic I wanted to ask for you input on is not really related to soaring, but is still very important for me:
Have anyone tried to transfer fly-by-wire signals over X-Bee?
Even if not, what do you think about it? Will the communications latency be a concern? Are there any ways to reduce XBee latency, by sacrificing throughput for example, or something like that?
I mean real fly-by-wire here, like sticks positions, not high-level heading/speed/waypoints data. I imagine it would be perfect to pilot a plane in stabilized mode, but through the X-Bee and laptop (with normal stick, throttle, and maybe even real pedals) instead of conventional rx/tx system.
Interference around 2.4GHz band is quite high where I live, but it is almost quiet around 900MHz, which, with some amplification, provides much more coverage than any regular 2.4 tx, even when using simple half-wave antennas.
Or should I start a separate topic about fly-by-wire in the appropriate category? (which one?)
Replies
Hello Alexey,
I am one of the ArduPilot developers and a full scale glider pilot.
1 - A purely inertial approach is difficult with the current sensors. Several people have worked on it with limited results. The quad guys are still working on related ideas and getting closer to usable results.
2 - Doing a compensated with the pressure sensor and airspeed sensor is very do-able. It is polar dependent, and a niche feature, so it has not been pursued in the code.
3 - I doubt you can do better with a "mechanical" vario than differentiating the electronic pressure sensor value. Obviously any system based on a flask and calibrated bleed will still have to be measured by electronic means to relay the info back to you.
4 - Design and calibration of total energy probes is a bit of a black art. Expect LOTS of trial and error. In my full scale sailplane I use a mechanical vario with a total energy probe, and an electronic vario with electronic compensation. They give differing information; I think the superposition of their characteristics is useful.
The current ArduPlane code already supports flying in FBW with a joystick connected to the ground control station and communicating over the telemetry link. Latency is noticeable when flying in manual mode, but no problem when flying in FBW.