This is a discussion re the ArduCopter-3.1 AutoTune feature.  A place for people to report successes and failures, ask questions and get answers.

The AutoTune wiki page is here.

 

Views: 74326

Reply to This

Replies to This Discussion

A few pages back I wrote about doing Auto Tune on my quad and finding a strange oscillation when flying fast forward with FPV goggles on. I don't notice it when flying visual but it is quite annoying when doing FPV. I tried to tune it out but couldn't get there. Putting 3.0.1 back on made the problem disappear.  I may put 3.1 back on and try lowering the I value on roll and pitch as it never occurred to me to try that.

It's really great to have sharp minds on this forum to learn from. :-)

Hi Leonardt, Hi Randy,

today autotune worked out fine for me. I had loaded the firmware again, did a safe trim procedure and then started autotune. It worked completely. I did two runs, one for a 3S battery, one for a 4S. The log for the 4S is attached below (edit: where can I attach files to a reply?).

I wasn't aware before, that if you didn't put your radio trims to zero, autotune would not start. Now I know, read a hint in a post here.

Thanks for your help. Now the only thing left to do is to get rid of the yaw oscillation. Maybe it is a question of the tail servo.

I think that there was a hint somewhere in the forum, that yaw control has been changed between the version 2.9.1 and one of the following versions. Do I remeber that right?

John

Just a quick note to the devs to express my appreciation for the work you have done.

Yesterday, I disabled the GPS and fence failsafes on my Stock Arucopter Y6, and at our indoor arena did 3 autotunes in a row (just for kicks. LOL)

The autotune completed without a problem , and the copter flew ever so smooth afterwards. Unfortunately I could not remember how to save the new values, so I was not sure if it had or not, so this morning I re-did the AT outside in the parking lot, this time saving the values as per the Wiki.

Flew around a bit, and even now that the wind has come up, I cannot tell you how much confidence the handling instill now.

Gret Job!

Arducopter Y-6 post-AutoTune Standard Pids :

Going out now to loiter in the wind... will keep you updated!

Ok 20-30km/h gust wind outside, ROCK SOLID loiter. WOW, One happy chappy over here!.

Question, the controls seem very "connected" and "direct". I assume to smooth things out for camera work, I presume turning the top row P values down a touch will do? Or is it the Rate Loiter P?

Is there a description somewhere that links the above picture to a particular aspect of flying?

EG. This set of values influence flying around in STAB mode, while that set of values affect how well it loiters, and those ones have to do with flying missions. (In laymans terms I mean.)

Wessie,

      Yes, i think reducing the Stabilize Roll/Pitch P values a bit (20%?) will make it less "sharp" feeling.  I think leaving the Rate P values as they are is a good idea.

Hi Wessie,

I am happy you are enjoying it!!

So Autotune chooses the highest value of Stab P you should use but you can soften the response by reducing it. Don't adjust the Rate P or D terms. You can play with Rate I as you may be able to tune out decent wobbles by dropping it to P/2 according to one user.

Any value of Stab P that is comfortable to fly should also loiter well.

Now the good news. Because of the performance and sharpness you are seeing, and the general quality of control we are seeing now we can find the optimal gain values. We are looking at adding some parameters the user can adjust to specifically control the feel of the copter. So you will be able to have a very sharp tune but still have gentle camera ship performance or razor sharp reactions (see the heart beat in your thumb), all with the same tune. 3.2 should be fun :)

Cool thanx guys, good news indeed. Im going to put Stab P on the rotary knob of my new DX9 (brilliant radio for copters by the way.), with about 30% play for some inflight responsiveness setting. Once I have the correct values, I may assign it to a switch for "Fly around" mode and "Camera stability" mode.

Any ideas on my 2nd question? Would like to have a clear understanding too...  ;-)

Sorry mate, I don't understand your second question.

Is there a description somewhere that links the above picture to a particular aspect of flying?

EG. This set of values influence flying around in STAB mode, while that set of values affect how well it loiters, and those ones have to do with flying missions. (In laymans terms I mean.)

IOW

Without going into detail of every PID option, it would be nice to be clear on how changing a SET of values will influence the copter.

For example, what does the STAB Pids do to how the copter loiters, or flies, or navigates, compared to the RATE Pids. Call me obsessive but I want to know WHY I shouldnt mess with my Rate P or D terms (for instance) but I know nothing of how to interpret the control loops, or anything of programming.

Wessie

Ok, I see what you are asking.

Unfortunately there isn't a simple answer for most of these values. The two aren't too bad to explain are the Stab P and Alt P. Both of these parameters determine how aggressive the copter is at correcting the angle/altitude. If you want a softer response, reduce them and visa verse.

The lower parameters have a similar relationship however they are much harder to understand because it is harder to see velocity compared to position. Also these loops work much faster again, making it harder for the pilot to see what they are doing.

Sorry but it is hard to be any clearer without understanding how PID loops work.

Hi guys,

A friend of mine tried the APM for the first time and used autotune. During the tuning process he crashed. I was not there, but by looking at the logs I notice a few things.

The first axes (roll) went fine, then it switched to the second axis (pitch) and at the end of the first step something happens. I notice a large throttle-out and a peak in the yaw. I also notice the pitch-in is only in 1 direction but that could be a balance issue. Here are some of the graphs ive put together. The last chart are the autotune steps and axis. The logs are also in the attachment.

Hopefully you guys know what happened.

Regards,

BasTijs

Attachments:

I would like to request a fix: If ArduCopter senses a motor stall, then the AutoTune should automatically abort.  Why would I request such a feature?  I'm glad you asked.....

Reply to Discussion

RSS

© 2019   Created by Chris Anderson.   Powered by

Badges  |  Report an Issue  |  Terms of Service