Main P gain for ail / ele. My ail is at 55 and elevator is at 85. Seems pretty good except I still haven't seen any high gain wobbles yet.
I noticed on elevator or aileron giving a sharp input and quick stop that it moves quickly at first but coming to the stop it slows down and drifts into the final position instead of stopping cleanly and sharply.
I'm assuming my P gains still need to go up. Just seemingly like the elevator gain is getting pretty high up there?
My setup: black thunder, theta C1 @560hz, headspeed 2000rpm, Rail 716mm..set 10deg in cyclic throw setup and values are in proper range.
Thanks.
When you say that aileron and elevator move quickly at first, but when they stop they slow down and drift to the final position instead of stopping sharply and decisively, are you referring to the model in flight or to the movement of the swashplate with the model stationary on the workbench?
With the model stationary on the workbench it is very normal for the cyclic controls to “insist” for a while. This is because with the model stationary, the gyroscopes do not read any rotational speed and therefore “prolong” the duration of the initial commands.
By increasing the gains you do not change this phenomenon in any way but instead risk that in flight, the rotor may enter an uncontrollable self-oscillation.
If you had done what is written in the sticky post IF YOU ARE ASKING FOR SUPPORT... we might have been able to give you a more precise answer and less general and vague.
Didn't think it was a support question because it's not a problem with the unit as it functions great. It is just a general fine tuning behavior question to get that last bit of performance. Yes, the behavior is in flight, not on the bench. I have no errors in the logs.
I only change values by 2 to 3 points at a time as described in the instructions and set it on a separate bank to be able to get out of quickly in case I went to far.
Doing pitch pumps the rotor disc stays flat and level at the settings I posted.
I am tempted to go to D gain to affect the stop behavior, but that is my main question. Should I just keep to the instruction and increase P gain till I get oscillations and back off a few points before messing with other parameters?
At the current gain settings doing chaos, tic tocs, piro ticks tocs, snakes, hurricanes, funnels piro funnels ECT, there is no concerning oscillations at the current gain values. Coming from another brand of flybarless unit, I love how the brain/ikon feels.
Also forgot to include in my parameters that my agility is at 65
If the difference in decelerations of rotational velocities versus accelerations of rotational velocities occurs in flight, then it cannot depend on PID parameters because the same PID parameters are used in both accelerations and decelerations.
Instead, it may depend solely on the rates at which the pilot moves the sticks to increase the rotation of the model on one of the axes and the rate at which the pilot returns the stick to zero.
However, without data it is only possible to make assumptions, hypotheses and theories.
It is highly probable that the moment you give commands to rotate the model you do it very fast by also making a slight overshoot of the stick movement while when you return the stick to zero you do it more slowly and without giving any overshoot (command to rotate in the opposite direction for a brief moment).
However, without any flight logs it is impossible to analyze the trend of the curves of the signals received from the transmitter on the input of the flight control unit in order to be able to check if overshoots occur and to measure the rising and falling rates of the signals with respect to time and also to be able to compare them with the signals coming out of the flight control unit to the swashplate.
This thread is of no use to other forum users because it is based solely on pilot's feelings and perceptions that are not measurable and cannot be quantified instead of being based on accurate and measurable data from a flight log that other users can compare with their own.
Therefore, we consider this thread concluded.
