paparazzi-devel
[Top][All Lists]
Advanced

[Date Prev][Date Next][Thread Prev][Thread Next][Date Index][Thread Index]

Re: [Paparazzi-devel] rotorcraft / throttle sensitivity / rate mode


From: Oswald Berthold
Subject: Re: [Paparazzi-devel] rotorcraft / throttle sensitivity / rate mode
Date: Thu, 05 Jun 2014 13:08:12 +0200
User-agent: mu4e 0.9.9.5; emacs 24.3.1

Hey Felix,

sorry for the intermission. now i did check #737 and indeed it seems to
do well, no spinup on disarm any more. i didn't know about the
"in_flight" variable beforehand, but i think, as gautier comments, this
additional measure is useful.

bst, oswald

Felix Ruess writes:

> Hi Oswald,
>
> could you please check if https://github.com/paparazzi/paparazzi/pull/737
> solves the spinup on disarming via max yaw (if you don't wait for in_flight
> to be false) for you?
>
> In attitude mode all axes are controlled to an angle setpoint, including
> yaw of course... otherwise it could drift and would not be able to return
> to desired heading after disturbances.
> What you do with the RC yaw stick is simply advance the current yaw/heading
> setpoint...
>
> Cheers, Felix
>
>
> On Sun, Jun 1, 2014 at 11:18 PM, Oswald Berthold <
> address@hidden> wrote:
>
>>
>> hey,
>>
>> thx for all the feedback so far. briefly commenting inline ..
>>
>> Felix Ruess writes:
>>
>> > Hi again,
>> >
>> > Regarding motors spinning up when disarming via zero throttle full yaw:
>> > We could probably add something that doesn't pass through the yaw command
>> > if you have zero throttle in the read_rc functions...
>> > Added an issue/feature request for that:
>> > https://github.com/paparazzi/paparazzi/issues/736
>>
>> cool.
>>
>> > Regarding yaw behavior, are you talking about rate or attitude mode?
>>
>> ATT
>>
>> > In attitude mode, the yaw stick "simply" changes the yaw setpoint with a
>> > maximum rate of STABILIZATION_ATTITUDE_SP_MAX_R at full yaw stick.
>> > This yaw setpoint is then passed to the attitude reference model (roughly
>> > meaning it will "smooth" your attitude setpoint and calculate reference
>> > rates and accelerations to reach that attitude setpoint). The main
>> > parameter to adjust the "speed" of the yaw reference
>> > is STABILIZATION_ATTITUDE_REF_OMEGA_R.
>>
>> > You can also look at increasing STABILIZATION_ATTITUDE_REF_MAX_R so that
>> > this limit doesn't really come into effect.
>> > We could probably remove the REF_MAX_x parameters as in most cases they
>> > should not really be used (meaning it should be set such that they
>> actually
>> > don't limit it).
>> > Already discussed here: https://github.com/paparazzi/paparazzi/pull/663
>>
>> i briefly varied these settings this morning and it seems to go into the
>> right direction. so in short, you do a bit more work than simply rate
>> controlling yaw in ATT? (i'll need to do more experimenting, try RATE
>> etc.).
>>
>> cheers, opt
>>
>> >
>> >
>> > Cheers, Felix
>> >
>> >
>> > On Sun, Jun 1, 2014 at 12:44 PM, Felix Ruess <address@hidden>
>> wrote:
>> >
>> >> Hi Oswald,
>> >>
>> >> I'm not familiar with your ESCs, but if they expect a fixed 1000-2000us
>> it
>> >> should work just fine.
>> >> Just set it accordingly in your servos section.
>> >>
>> >> The most annoying ones are the ones that do some sort of
>> >> "auto-calibration" on startup, where you have to give full throttle
>> first
>> >> and than take this value as their internal maximum...
>> >>
>> >> Cheers, Felix
>> >>
>> >>
>> >> On Sun, Jun 1, 2014 at 12:46 AM, Oswald Berthold <
>> >> address@hidden> wrote:
>> >>
>> >>>
>> >>> hi felix,
>> >>>
>> >>> thanks for the nudge.
>> >>>
>> >>> turns out the multistars don't do calibration [1], so i swapped them
>> for
>> >>> some other ESCs, now throttle is fine.
>> >>>
>> >>> spent some time playing with the attitude params. horizontal is fine
>> but
>> >>> yaw behaviour irritates me.
>> >>>  - very weak response build-up to yaw stick and equally slow decay of
>> >>>    yaw response (still yawing on when stick is back to zero).
>> >>>  - thrust is diminished strongly during yaw
>> >>>  - during disarm (full yaw, throttle neutral) motors spin up and frame
>> >>>    tends to jump
>> >>>
>> >>> i couldn't find any setting for PSI_[PID]GAIN where this would really
>> go
>> >>> away.
>> >>>
>> >>> i'm on git master and using essentially
>> >>> examples/quadrotor_lisa_m_2_pwm_spektrum.xml
>> >>>
>> >>> bst, oswald
>> >>>
>> >>>
>> >>> [1] they seem to have worked quite well with a multiwii type system
>> (for
>> >>> which the tx travel is adjusted to precisely 1000-2000us).
>> >>>
>> >>>
>> >>> Felix Ruess writes:
>> >>>
>> >>> > Hi Oswald,
>> >>> >
>> >>> > this very much sounds like you either need to calibrate your ESCs
>> and/or
>> >>> > set up the min,neutral,max properly in your servo section.
>> >>> > The neutral value should be the pulse length in us at which your ESCs
>> >>> turn
>> >>> > the motors on, but at an acceptable idle speed. Min is off in that
>> case
>> >>> and
>> >>> > max is well, max...
>> >>> >
>> >>> > This has nothing to do with the gains or anything like that at this
>> >>> > point... first you need to make sure that if you send the "neutral"
>> >>> value
>> >>> > all motors will rotate at idle speed and reach their maximum speed at
>> >>> max
>> >>> > and not before.
>> >>> >
>> >>> > I never use rate mode, but the ahrs (attitude estimation) should run
>> as
>> >>> > normal... and works as expected here...
>> >>> > Can you provide some more details?
>> >>> >
>> >>> > Cheers, Felix
>> >>> >
>> >>> >
>> >>> > On Tue, May 27, 2014 at 10:45 AM, Oswald Berthold <
>> >>> > address@hidden> wrote:
>> >>> >
>> >>> >>
>> >>> >> dear list,
>> >>> >>
>> >>> >> i am very glad finally having a real paparazzi system for work and
>> >>> >> play. as you're already suspecting, i have some questions.
>> >>> >>
>> >>> >> system:
>> >>> >>  - quadrotor, X configuration
>> >>> >>  - lisa/m v2
>> >>> >>  - orange rx satellite
>> >>> >>  - turnigy multistar 20A ESCs
>> >>> >>  - git 82d414d8 .. (20140520)
>> >>> >>
>> >>> >> throttle issue:
>> >>> >>  1. arming starts the motors but i have to keep them alive by giving
>> >>> >>    some more throttle, otherwise some of them starve. igains are
>> zero
>> >>> >>    for the attitude controller. i cannot raise "neutral" because of
>> the
>> >>> >>    next subissue
>> >>> >>  2. i have been able to make a first flight with my provisional PID
>> >>> >>    settings and throttle seems oversensitive. quad takes off with
>> maybe
>> >>> >>    20% of stick and it was quite a challenge to hover manually.
>> >>> >>
>> >>> >> rate mode issue:
>> >>> >>  - when i alternatively configure rate mode for switch 1 manual, the
>> >>> PFD
>> >>> >>    doesn't work. is attitude estimation disabled in rate mode?
>> also, in
>> >>> >>    messages i see that the phi/theta/psi ests are not updated.
>> >>> >>
>> >>> >> regards and thanks for any feedback,
>> >>> >> opt
>>
>> _______________________________________________
>> Paparazzi-devel mailing list
>> address@hidden
>> https://lists.nongnu.org/mailman/listinfo/paparazzi-devel
>>
> _______________________________________________
> Paparazzi-devel mailing list
> address@hidden
> https://lists.nongnu.org/mailman/listinfo/paparazzi-devel



reply via email to

[Prev in Thread] Current Thread [Next in Thread]