ARDUPLANE - FLIGHTMODEI - STABILIZE
The mode - STABILIZE - does not work exactly as I am used to with a model airplane.
In the mode - stabilized - you tilt the F405 (for example aileron) by 30 degrees. Accordingly, the aileron servo deflects by a certain angle. This angle of the servo should also be maintained as long as you do not change the angle of the F405 (30 degrees).
Unfortunately, after a short time the servo slowly runs back to 0!
I have already searched intensively in the parameter list for a suitable parameter, but since there are so many parameters, the search is quite difficult.
Does anyone know the problem or does anyone know a suitable parameter (mode) to prevent the servos from running back?
Are you sure you are in Stabilize Mode and not Acro?
How you tested this? Do you have a logfile?
Sorry, I don’t have a LOGFILE
We have been flying with an autopilot with a pure P controller for a long time, which works very well. To check, the model is held at an angle, the rudders deflect and remain in the intended position. Well, AUDUPILOT always seems to use a P_I_D controller, which causes the servos to run back in STABILIZE mode. To what extent this is better for a glider model would have to be tested in flight. But you should also have the option of changing the controller values if there are problems. In the mode: Mission Planner - Config - Basic Tuning - the parameters for the function - Servo ROLL PID - have no noticeable influence on this phenomenon.
I have not found any other parameters so far.
It’s a little hard to follow your intent, but I think what you want is ACRO mode with attitude lock.
STABILIZE mode is self-leveling by design.
Neither of these features/modes behave better with P only control. You should tune the control loops appropriately and then use the flight mode best suited to your use case.
Here’s a place to start reading, but please continue searching the documentation for tuning procedures and flight mode features:
https://ardupilot.org/plane/docs/flight-modes.html
So you test this while the plane is not moving?
So describe exactly how you run the test like:
- hold plane leveled
- command roll left by Tx Stick
- aileron servos moved to roll the plane
- move the plane to a bank angle of 30° with TX stick not changed
- aileron servos going back to normal position
Thanks everyone for the support
OK, I’ve already tried that site. You can spend many hours reading documentation. Unfortunately, some documentaries are very superficial and some are even out of date.
The servo movements on a test board seem to cause irritation due to the PID control (which is always better), with pure p control this is easier (problem solved).
Today I flew a model with a V-tail with the autopilot to test it, which worked very well (aileron, rudder, elevator).
I would now like to rename the throttle stick for spoiler. I can’t find an RCMAP in the mission planner (according to the documentation). Servo parameters should be DifferentialSpoilerLeft1/-Right1.
For the spoiler function, both ailerons should extend upwards by a certain value (parameter?).
Is there a specific parameter for aileron differentiation or is this achieved with the deflection limitation? It would also be nice if part of the aileron could also affect the rudder.
If you could get all of that here, everything would be perfect!
Attempting to measure a PID controller’s output quality on the bench is almost certainly a flawed test. You need to tune the craft based on in-flight data logs and/or autotune routines. If you are mucking about with PID values on the bench, you are fundamentally misunderstanding the autopilot’s control loop.
As for mapping throttle to spoiler control, I’m afraid I don’t quite understand your intent.
FWIW do you mean the parameters called RCMAP?
RCMAP found Thanks
Sorry, was too fast at night, Flaperon problem solved.
https://ardupilot.org/plane/docs/flaperons-on-plane.html
The last problem, how do I get some of the aileron deflections to the rudder (compensation for negative momentum).
This article describes how to mix rudder for adverse yaw control:
https://ardupilot.org/plane/docs/yaw-in-plane.html
Tested further today (flew)
Thanks, now everything works as it should!