Hi!
During flights with low throttle my pitch is not tuned very well. There are spikes which cause my copter to drop to one side (roll and pitch are affected). Is this a tuning (PID) problem or something else?
Any suggestions are appreciated.
First just do some basic tuning. The Rate Pitch/Roll P&I ratio you have probably isn’t great (set them equal). There is oscillation in the outputs perhaps because you have PSC_ACCEL_Z parameters backwards. Try 0.08 P and 0.16 I.
You don’t want this:
Vibration levels are too high also.
And is MOT_SPIN_MIN,.15 right? You can see the RCout’s hitting this floor.
And this isn’t doing you any good: INS_HNTCH_REF,0
Set this: INS_ACCEL_FILTER,10
Out of curiosity what ESC’s do you have?
Use the ardupilot methodic configurator software to get this vehicle correctly configured.
That would certainly normalize some things that are off.
Ok thanks four your inputs. I will try this out and get back to you with an update.
Regarding the ESC, I use a Fettec 4in1 ESC 8S - 65A.
Ah OK, was wondering where ESC telemetry was coming from. That explains it.
I did some tests with your recommendations. They worked pretty good for my initial problem. Now I don’t have any sudden drops on roll or pitch anymore.
What I don’t understand however, is why there is an oscillation on Rate.P (your picture). Especially since it only occurs if I fly in automission. Are there any separate parameters which need to be configured for that? Unfortunately PSC_ACCZ Parameters did not change that much for this problem.
You need to fully configure the vehicle.
Partially configured vehicle gets you suboptimal results. Is that hard to understand?
I also have two FETtec copters. Mine have no issues. But mine have been configured with the configurator.
No not at all why should it be?
But I think you missed my point. I am more curious how the autopilot works and therefore I am trying to figure out which parameters affect certain modes and behaviors. I do understand this is not the fastest way to configure a copter.
The best method to achieve that is to configure it properly and afterwards, play around with the parameters to see their effect.
Doing it on a un-configured vehicle, Param A can sometimes produce no effects and you will learn (wrongly) that a A parameter has no effect when in reality it does have effect if Param B had been configured correctly.
That is true. My initial thought was that someone might know this problem and could give me an answer.
However it does make a lot of sense to configure it properly and then tune it manually to see the effects.
Thanks for your swift responses and the methodic configurator.
There is a new version of it. Get that version before starting.
