Servo Auto Trim Mechanism?


I believe that servo autotrim does not affect the flight performance in auto mode or FBWB flight mode.

It simply saves trim tendency in auto mode and uses them in manual mode.

However, my friend said that according to the discussion in the following link:

They are saying “Servo auto trim” works in auto mode.

So, my friend is suggesting that auto trim will indeed affect flight performance in auto mode.

Is that correct?

Do anyone have idea on this?

Don’t anyone have idea on this?

Looking at the post you reference, if Tridge says it’s so then it’s hard to argue.

You could pull a flight log in auto mode and check for parameter changes on the trims.

1 Like


I think there is some misunderstanding.

Tridge is saying that “Autotrim works in Auto mode,” but he didn’t clearly specify whether “works” means “saves” or “changes the servo movement and affect flight performance even in auto mode”

and, there is no need to check for changes in trims in parameter.
What I wonder is if the servo trim affects flight performance even in auto mode(not manual mode). parameter change mean that the trims are being saved to the aircraft. That is not related to my question.

Okay, if you’re having difficulty understanding me, let’s simplify. Sorry.

In auto mode, the aircraft behaves strangely.

My friend said it’s because autotrim hasn’t been learned yet.

I said the autotrim function is for manual mode flyers, not auto mode flyers.

He didn’t fly aircraft at manual mode, so I think he misunderstands something.

Then my friend shared this link showing that servo autotrim works in auto mode:

Servo Auto Trim will adjust the trim in FBW or Auto mode. You’ll see the trim values for pitch and roll adjusted in the flight logs. It will not make adjustments while flying in manual mode, but changes made in auto/FBW will be there for flight in manual.

If your plane is flying strange, that could be another issue.

Servo Auto Trim will adjust the trim in FBW or Auto mode.

← So, Does “Servo Trim” affect flight performance in FBW or Auto mode?

Will the aircraft use the servo trim value even in Auto mode, or does it fly solely with the PID loop?"

The plane will fly with the pid loop. But, if the plane is out of trim then the pid loop will always be in an error state. For example, if the plane is trying to fly straight and level, when the plane reaches it’s desired angle the controller will try to return the controls to their trim position. If the trim is out when then the plane will roll or pitch. In reality, how much impact this will have totally depends on how much out of trim the plane is. If we’re only talking 10-20 PWM, I expect it probably wouldn’t matter. If it’s 100 or 200 PWM, then that’s a problem. I’d also want to know why the plane isn’t in trim. Flight controllers can’t fix mechanical and aerodynamic problems.

I always leave SERVO_AUTO_TRIM,1. It’s normal to see minor shifts in trim on any given flight. You know it’s close when those changes are few and far between.

1 Like

No. It gathers data in these modes, so it can apply auto trim in manual modes.


I think there might be a misunderstanding.

The plane was already flying well, so I am not concerned about the solution for an out-of-trim situation. The issue I mentioned was not related to trim.

Oli1 is addressing my concern correctly.

I respect your wisdom and experience and want to avoid any misunderstandings. It seems that I received a response that was not related to my question, and I believe this might be my fault.

My question is simple : Will “Turning on and off autotrim” affect auto mode’s flight performance?

Thank you for your answer.

I received messages from my colleague about this matter several minutes ago. Could you review the following and share your opinion?

[Evidence of Auto Trim applied in FBWA mode]

  1. My colleague is saying that if auto trim is applied, the plane’s transition becomes smoother in “FBWA” mode. He firmly believes that FBWA mode uses auto trim, not just saving auto trim settings.

  2. He is also saying that in FBWA mode, even if the pitch and roll are zero and the gyro is parallel to the horizon, the servo PWM is not 1500 if auto trim is applied.

Can any expert developer confirm this matter?

My question is simple. : Will “turning on and off autotrim” affects auto mode’s flight performance?

First of all, the question is about as “simple” as I am an expert developer. So it is not simple.

Second, as far as I understand the question, Allister gave you a pretty conclusive answer. But then again I might not understand the question correctly, if you rejected his answer already.

Regarding scenario 1:
Let’s set the reference frame right: was this a transition where auto trim had been learned earlier or was the transition with default trim? As a stabilized mode, it uses a closed control loop to control the attitude. For a closed control loop you need an input, some transfer function, an output and a feedback to calculate error and apply correction. When you transition into a stabilized mode from a non-stabilized mode, you have an input and a feedback to calculate error, but you don’t have a previous control loop output that you can apply that error-correction to. So naturally the starting point that makes the most sense is to use the trim value. Now, if it was a transition with a previously learned auto trim, of course the starting point is better than with default trim. If it was with default trim though, I’d say it was a fluke and/or placebo, because auto trim takes time to learn, so there’s no way it can already have an effect for the transition.

Regarding scenario 2:
That is expected behaviour, regardless of auto trim being enabled or not. The question is, what is the difference between these values in level flight when auto trim is disabled vs when auto trim is enabled but hasn’t learned yet vs when auto trim is enabled and has learned values? According to the PID-loop, they should all be about the same. Also what are these values and what is the learned auto trim value after this?

1 Like

Scenario 1:

Auto trim had been learned earlier when the transition started.

Regarding Allister’s and your answer, the starting point of the servo is the trim point, even in FBWA or auto mode, not only manual mode.

when started transition with auto mode, the servo starting point is not PWM 1500, right? That could be the answer to my question. Thank you.

But I wonder, in QStabilize or QHover, QLoiter mode, there is also PID control on the wing. Does the wing servo position suddenly return to the servo trim position when starting the transition?

If so, I think the Wiki should be modified to clarify this. The auto trim value is also used in FBWB or auto mode as its starting point during the transition, So enabling Auto trim affect transition, and make it smoother If there is CoG issues(but we have to correct CoG first.)

Scenario 2:

You are right. The three values you mentioned are the same, except during the transition time.

My question should be clarified as follows: Will “turning on and off servo auto trim” affect the early stage of transition in auto or guided, FBWA, FBWB mode?

If it takes time to learn auto trim value, transition will be affected by autotrim much. and your forementioned three value will be different for some time.


Depending on the trim. If the trim is exactly 1500, then it is 1500. But most times it will not be exactly 1500.

I didn’t think about transition from Q-modes. When I wrote the post I had (non-Q) transitions between manual and stabilized modes in mind.
When you transition from Q modes the angle-controller is already active, as you mentioned. I don’t know enough about that to say if those transitions are effected by auto trim. My intuition would say it shouldn’t have an effect from one stabilized mode to another stabilized mode, but I also have no idea how exactly the code handles Q to non-Q transitions, if control loops get reset or something.

Keep in mind 1) I don’t know how much time it takes to learn and 2) that depends on the trim and how much out of trim the starting values are. And after the trim was learned once every starting point will be much closer to the later learned values, so it will have less effect on every transition after the very first one.

Okay, I understand what you’re saying.

Based on what my friend said, when we transition from Q-mode to FBWA mode, the servo starting point is not determined by the Q-mode PID loop but by the learned trim.

Thank you. I believe auto trim affects the early stages of the transition, when we enter FBWA from Q-mode.

1 Like

It would be better to add “auto trim can affect the early stages of the transition” in the wiki.

Those who are already familiar with this system can read it without a problem, but for beginners, it would be better to be modified.