I develop helicopter with a propeller diameter of 1M and a weight of about 20KG. According to the documentation, I successfully completed the AUTOTUNE VFF RATE D/P. When I wanted to continue the YAW adjustment, I found that the ROLL was shaking when I hovered. YAW AUTOTUNE was operating. Not going well, the tail is DDFP
Can help understand the problem?
ROLL RATE D/P
PITCH RATE D/P
Thanks to Bill for assisting
In addition, I would like to ask the developer to help me adjust to optimization. How should I start?
I don’t understand what you are asking. Are you asking for help to optimize your tune? The autotune feature should determine the best gains for a given set of filters. However, a tune could be made better by adjusting filters and running the autotune again. I have been experimenting with this but haven’t come up with a way of automating this for all users to benefit.
Yes, I need support to help optimize
Are you sure this is right? 20kg for a 500 size helicopter.
I will do my best to help you get the autotune working properly but I don’t have the time to help with custom tuning. Sorry
Sorry, I said it wrong
This is a helicopter with a blade diameter of approximately 2.2 meters.
In the rate P and Rate D tuning for the pitch axis, it tuned the gains too high. One issue that I am finding is that the autotune doesn’t alert or decide on its own what the max response gain (AUTOTUNE_GN_MAX) to tune to. In your case, the shape of the gain plot does not peak very high.
It starts around 0.8 and then never peaks above it at higher frequencies. So first the FF gain probably needs to be raised to 0.165 so the initial response gain is closer to 1.0. so based on the tuning conducted as shown in the table, I think the Rate D gain is fine but the Rate P gain could be reduced to 0.05. I think it was possible that the high Rate p gain in pitch was driving the roll axis.
Looking at the roll axis gains, the FF gain looks good based on the low frequency response gain being close to 1.0. In this axis, you can see the response gain quickly grows to a max of 1.5.
I would suggest setting the Rate P and Rate D gains to zero and run the roll axis rate P and D tune again.
@bnsgeyer
thanks for your suggestion
I try my best to understand and continue performing autotune
But ROLL still has jitter after adjusting the ANG-P gain.
After reducing ANG-P, the situation was alleviated, but the problem still exists.
When I executed AUTO later, I found that the plane was flying very well, with some gusts. When the plane was affected by crosswinds, jitters would appear.
https://drive.google.com/drive/folders/1IZkDQDnogobs85v6OhpJaMToC6QlBUnk?usp=sharing
Specifications of this helicopter
The direct diameter of the cantilever is approximately: 2.2M
Direct drive motor 18040 30KV
Tail suspension: 1245
Tail suspension power: 3520 250KV
Weight:20.5KG
I think the angle P for the roll axis was too high. So I agree that it should be lowered. I would set it to 7.
There is a technique that I have been using to control the “jitter” that you have been seeing. I set the ATC_RAT_PIT_FLTT and ATC_RAT_RLL_FLTT to the equivalent cutoff frequency for 1/3 the ATC_INPUT_TC.
FLTT = 3 / (6.28 * ATC_INPUT_TC)
For ATC_INPUT_TC = 0.25, the FLTT for pitch and roll should be 1.9 hz. This should make the aircraft very smooth in its response.
For the yaw axis, the command model or the desired pilot response seem pretty high. You have PILOT_Y_RATE = 202.5 deg/s. You could set this a little slower. Set it to maybe 120 deg/s. Also you haven’t set the PILOT_Y_RATE_TC. This is the time constant or how fast the yaw axis achieved the steady state rate. you could start with 0.1 sec. That might help the yaw axis.
I am a little concerned with the yaw axis because the tuning is showing the yaw response to be very low. The autotune is not doing that good of a job because the response is so low. I will look more at this and get back to you
@bnsgeyer
Hello Bill,
I’m a diligent reader because I want to learn more.
Could you explain the term “steady-state rate” in this context in more detail. Of course, I can easily translate the term into German, but I don’t understand it.
Thank you very much.
Heri
@bnsgeyer
Hi Bill, long wait
I readjusted my drone due to structural issues
Also perform autotune again
Unfortunately, the last time I ran the ANG-P, the action was so intense that the rotor knocked into the tail.
After that, I repaired the drone and ran Autotune again.
This time I lowered the maximum frequency and GN_MAX
Not sure if I’m correct but the plane was completed in a short time and works well
https://drive.google.com/drive/folders/1ML_qnY2LOE5EVRHws1ZBbIm8zKbtftjy?usp=drive_link
Today I tested high-speed flight and high-speed flight missions. The aircraft seems to be in good condition, but I don’t know anything about IMAX’s adjustments. I’m still exploring.
@bnsgeyer
Hi Bill
I flew part of the mission and it was a little gusty but it was a good flight
During the landing process, I found that the descent rate was unstable. Is there any solution?
