paparazzi-devel
[Top][All Lists]
Advanced

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

Re: [Paparazzi-devel] Strange throttle behavior


From: Felix Ruess
Subject: Re: [Paparazzi-devel] Strange throttle behavior
Date: Tue, 20 Dec 2011 13:33:19 +0100

If you mean the bound_commands_step that was added by Piotr in May,
but unless SUPERVISION_MAX_NEGATIVE_MOTOR_STEP or
SUPERVISION_MAX_POSITIVE_MOTOR_STEP is explicitly defined in the
airframe file that is not used...

On Tue, Dec 20, 2011 at 1:20 PM, antoine drouin <address@hidden> wrote:
> I just looked at the file and noticed that someone added a sort of low
> pass filter in there. It would probably be nicer to handle the
> dynamics of the commands in the stabilization layer.
>
> On Tue, Dec 20, 2011 at 1:14 PM, antoine drouin <address@hidden> wrote:
>> take a look at
>>  paparazzi/sw/airborne/firmwares/rotorcraft/actuators/supervision.c
>> This was my first attempt at handling actuators saturation issues. You
>> are correct that the attitude stabilization is given priority over
>> vertical guidance. It's based on the assumption that the angular
>> dynamic is faster than the vertical one.
>>
>> Poine
>>
>> On Tue, Dec 20, 2011 at 1:06 PM, Simon Wilks <address@hidden> wrote:
>>> Hi Felix,
>>>
>>> On Tue, Dec 20, 2011 at 2:09 AM, Felix Ruess <address@hidden> wrote:
>>>>
>>>> Hi Simon,
>>>>
>>>> yes, when it's armed the motors will idle at low rpms with zero
>>>> throttle on the RC.
>>>> From your config it looks like you were testing in attitude_direct
>>>> mode?
>>>
>>> Correct.
>>>
>>>>
>>>> In that case the guidance_v p gain does not play any roll at
>>>> all, since throttle is directly controlled via RC.
>>>
>>>
>>> I see. I will give it a try seeing Eduardo seems to think it helps. It will
>>> at least not hurt if it isn't used.
>>>>
>>>>
>>>> I haven't heard or experienced your problem with throttle lag
>>>> before... maybe it's a problem with 8 motors and i2c? That is a shot
>>>> in the dark, but you could test with increasing the i2c queue to more
>>>> than 10 (that's just what others reported on the mailing list for 8
>>>> motors).
>>>>
>>>>
>>>> When you hold the quadcopter with zero throttle and tilt it a bit it
>>>> should immediately try to compensate for that.... sluggish response to
>>>> that might indicated an i2c queue problem...
>>>
>>>
>>> When I tip the copter in any direction it seems to respond correctly and
>>> very quickly. I will probably give the i2c queue adjustment a try as well.
>>>>
>>>>
>>>> With respect to switching off attitude control at zero RC throttle
>>>> input: I didn't quite get why you'd want to switch that off and IMHO
>>>> opinion that is not such a good idea (at least I can't think of a
>>>> reason right now).
>>>> E.g. when you want the copter to fall really quickly and you take the
>>>> throttle to zero, you really don't want stabilization completely
>>>> switched off.
>>>
>>>
>>> OK, I agree that it probably isn't a good idea then :) Didn't think of that.
>>>
>>> The throttle issue only happens when the copter is in flight and I have
>>> never noticed something like this when throttling up (and then down) while
>>> remaining on the ground. I wonder if this is an effect of it overdoing the
>>> attitude control as it should be sinking and it just puts too much power
>>> into the stabilization attempts, at least for an octocopter?
>>>
>>> Thanks,
>>>
>>> Simon
>>>
>>>>
>>>> If one wanted to do that at all, one could think about using something
>>>> like the autopilot_in_flight state to check if the copter is not in
>>>> flight AND the rc throttle at zero to switch it off... but that would
>>>> need some modifications for takeoff.
>>>>
>>>> Cheers, Felix
>>>>
>>>> On Mon, Dec 19, 2011 at 7:42 PM, Simon Wilks <address@hidden> wrote:
>>>> > On Mon, Dec 19, 2011 at 6:23 PM, Eduardo lavratti
>>>> > <address@hidden>
>>>> > wrote:
>>>> >>
>>>> >> As i tell you ... i have the same behavior when i using a HIGH valeu on
>>>> >> guidance_V P parameter.
>>>> >
>>>> > OK, got it.
>>>> >
>>>> >
>>>> >>
>>>> >>
>>>> >> This is the first think that i will modify in PPZ software.
>>>> >> When the throttle is around 0 the ATTITUDE control stop control the
>>>> >> motors.
>>>> >
>>>> > That would be great. This is pretty nasty (and extremely surprising)
>>>> > behavior.
>>>> >
>>>> > Thanks
>>>> >
>>>> > Simon
>>>> >>
>>>> >>
>>>> >>
>>>> >> ________________________________
>>>> >> Date: Mon, 19 Dec 2011 09:56:34 +0100
>>>> >> From: address@hidden
>>>> >> To: address@hidden
>>>> >> Subject: Re: [Paparazzi-devel] Strange throttle behavior
>>>> >>
>>>> >>
>>>> >> I just replayed the logs and the GCS shows the throttle input going to
>>>> >> 0%
>>>> >> after about 3 secs but the octocopter was clearly still flying for
>>>> >> another
>>>> >> 3-4 secs after this I till I hit the kill switch.
>>>> >>
>>>> >> On Sun, Dec 18, 2011 at 10:13 PM, Simon Wilks <address@hidden>
>>>> >> wrote:
>>>> >>
>>>> >> On Sun, Dec 18, 2011 at 10:06 PM, Hector Garcia de Marina
>>>> >> <address@hidden> wrote:
>>>> >>
>>>> >> Yeah,
>>>> >>
>>>> >> regarding the safety, I agree with Eduardo, next time DO NOT run next
>>>> >> to
>>>> >> the copter if you have lost its control until it completely
>>>> >> is switched off.
>>>> >>
>>>> >>
>>>> >> Yes you are right. I should keep clear till it has settled.
>>>> >>
>>>> >>
>>>> >>
>>>> >>
>>>> >> Cheers
>>>> >>
>>>> >>
>>>> >> On Sun, Dec 18, 2011 at 9:31 PM, Eduardo lavratti
>>>> >> <address@hidden>
>>>> >> wrote:
>>>> >>
>>>> >> Hey man ... take care with your tests !
>>>> >> This props and motors are very POWER and DANGEROUS !
>>>> >> Take a look at this pictures
>>>> >> http://brquad.blogspot.com/2011/04/little-acident-with-my-big-hexa.html
>>>> >> This accident occurred when i make some tests with my big hexa whit
>>>> >> same
>>>> >> motors and props.
>>>> >> Never make this tests with children arround.
>>>> >> Well, for the JUMP problem reduce the "guidance_V / hover_KP" and make
>>>> >> a
>>>> >> test.
>>>> >>
>>>> >> Make the same for wooble. Reduce the P & D GAIN parameter for Theta and
>>>> >> Phi . You see this values in STABILIZATION_ATTITUDE section.
>>>> >> Reduce P for 300 and make a test.
>>>> >>
>>>> >> Stay sure that your props are very well balanced or your lisa board
>>>> >> have a
>>>> >> good vibration isolation.
>>>> >>
>>>> >> When i make my first fly using lisa/m i have similar attitude on me
>>>> >> small
>>>> >> octo (5"props) and on my quad (8" props)
>>>> >> after i balance the props and make a good vibration isolation all
>>>> >> problems
>>>> >> are solved.
>>>> >>
>>>> >> Regards and sorry for my poor english.
>>>> >>
>>>> >> Eduardo Lavratti
>>>> >> Brasilian UAV developer.
>>>> >>
>>>> >>
>>>> >> ________________________________
>>>> >> Date: Sun, 18 Dec 2011 20:38:21 +0100
>>>> >> From: address@hidden
>>>> >> To: address@hidden
>>>> >> Subject: [Paparazzi-devel] Strange throttle behavior
>>>> >>
>>>> >>
>>>> >> Hi
>>>> >>
>>>> >> I have been noticing what seemed like a lag in throttle control though
>>>> >> I
>>>> >> wasn't 100% sure till my first test flight today (octocopter, lisa/l,
>>>> >> spektrum satellite receiver). While on the ground I don't seem to
>>>> >> notice
>>>> >> this effect but when holding it above my head or lifting off strange
>>>> >> things
>>>> >> start to happen.
>>>> >>
>>>> >> When holding it above my head changing the throttle on the sender seems
>>>> >> to
>>>> >> take a second or two till it happens on the copter. I have also noticed
>>>> >> what
>>>> >> seems to be a general increase in throttle all of sudden, not just to
>>>> >> correct attitude but like a it wants to pull up and away.
>>>> >>
>>>> >> When I tried the first flight today I got a bit of a shock when I
>>>> >> wanted
>>>> >> to just lift off a bit then set down only to discover it did not want
>>>> >> to
>>>> >> respond to the fact that I had returned the throttle back to zero and
>>>> >> skipped off across the garden. In the end I had to flick the kill
>>>> >> switch.
>>>> >> Pretty funny after the event. I have a video if you want a laugh here.
>>>> >>
>>>> >> You will see the props rotating at minimum speed (zero throttle on the
>>>> >> sender) after arming. I assume this is correct? One of the props isn't
>>>> >> rotating properly and I suspect it might have a problem and need
>>>> >> swapping
>>>> >> with a spare (it has a bit more resistance to the hand when rotating
>>>> >> than
>>>> >> the others) but I thought it might be OK to try a small hover anyway.
>>>> >>
>>>> >> Any ideas what causes this kind of throttle behavior? Pitch and roll
>>>> >> seem
>>>> >> pretty responsive. Here is the aircraft config.
>>>> >>
>>>> >> Thanks
>>>> >>
>>>> >> Simon
>>>> >>
>>>> >>
>>>> >>
>>>> >>
>>>> >>
>>>> >> _______________________________________________ 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
>>>> >>
>>>> >>
>>>> >>
>>>> >>
>>>> >> --
>>>> >> Héctor
>>>> >>
>>>> >>
>>>> >>
>>>> >> _______________________________________________
>>>> >> 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
>>>> >>
>>>> >> _______________________________________________
>>>> >> 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
>>>> >
>>>>
>>>> _______________________________________________
>>>> 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
>>>
>
> _______________________________________________
> 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]