The way it is supposed to work is that if it needs more than 5% power for stabilization or control, then it uses what ever it needs (above 5%) to to do the job. That is true both for CruiseControl quad flying inverted with throttle stick anywhere but zero (meaning 5% power) ... or with ASWA set to 5%. Remember that CC will shut the motors off if you use zero throttle (safety feature).
10% for slowest motor speed (CC or ASWA) is probably better than 5% if you are going to do more than simple flips? I am more interested in waypoint flight and letting non-pilots fly my VelocityRoam than stunts and racing, so opinions of stunt and racing pilots are needed.
This is partly due to the fact that most people in my club fly airplanes, so no one to fly with.
There are many ways to test this, some of them that are in hand with props on are a bit dangerous because it will fight HARD to get to where it wants to be in any self leveling mode (Attitude, Rattitude). You can take props off and use Attitude mode on Roll/Pitch with Rate on Yaw to see that if you use low but non-zero throttle, flip it upside down by hand that the motors will run hard in either case (CC / ASWA) to try to level it out.
Assuming a normal quad, not one that is gone in one second of full throttle and assuming that you have a reasonable "Rate Mode Response" (like maybe 360, so full stick flips take a while second) so that a flip is slow enough for a human to reverse it exactly half way through... If you are OK with flips using CC with Rate/Rattitude, then the best way to test while flying using CC with Rate/Rattitude (not Acro) is to:
- start a flip a bit higher than you normally need,
- push the throttle to 100% (and leave it there),
- wait maybe one second so it now has some upward speed,
- use full roll stick to start the flip,
- when it is perfectly inverted, reverse the roll by very quickly applying and holding full roll stick in the opposite direction to come back out the way you went in instead of completing a normal flip,
- when level, reduce the throttle stick to hover power.
During a normal flip, the motors "coast" through the flip. In this case you should hear the motors spin up faster because they must reverse the direction, not just coast through.
One other thing: If doing this flight test using Rattitude with CC instead of Rate/Acro with CC it can come back out of the flip in a direction you don't expect if you pause at neutral stick when inverted. That is because it is inverted and acting like attitude mode and it will get out to level hover the quickest way it can. If you undershot or overshot inverted it will come back or continue the flip. If there is say some pitch forward from being otherwise perfectly level but inverted, then it will start coming out forward.
Another thing about CC: By default it increases the thrust when you are banked so you don't have to add power when banked in turns. If you don't like this part, you can just set MaxPowerFactor to 1.0. If you do that, basically all you have is "automatic motor power to idle when inverted".
Full disclosure or don't worry about something that happens when you do things outside of the flight envelope? Years ago there were one or two credible reports of quads not controllable in high speed falls or motors not responding as expected when falling upside down. The people who experienced this were not beginners. It was initially blamed on CC as I recall, but testing showed CC acting correctly. Trying to figure out what caused this we never got to an answer before life called us away. Possible reasons for this would include:
- prop stall issues when falling fast the prop never gets out of stall (I have had this one happen with slow RPM motors)
- prop stall issues when falling fast inverted and a slowly spinning prop has more drag than a stopped prop causing a reverse stabilization
- All it takes is one stalled ESC in an error state ... the ESC must see a stop motor signal before it will try to start again. An ESC spinning the motor slowly and then the motor stops because of reverse air blast through the prop.
- ESC issues that don't start correctly when props are spinning fast or backwards, etc.
- configuration issues there should be a high min motor speed (ESC neutral) so the motor never stops in a high speed fall.
- CC3D code (and Revo Basic AttiEstAlgo) is smaller and simpler and has some math singularities in self leveling modes.
If anyone sees an issue like this, get a telemetry log, FPV video, pilot description, preferably all the above
, and let us know.