Re: [Paparazzi-devel] Re: Bug that needs fixing

[Pascal]

Fixed (version 1.16 of sw/tools/gen_flight_plan.ml)

Thanks for you exhaustive tests !

--Pascal

address@hidden wrote:
> When someone gets a chance, could they try a mission in Paparazzi2 with 
> an xyz,
> a simulation that is.  When I run it the desired_pitch increases, and 
> continues
> to increase without bound.  I want to know if its Isolated to me or if 
> everyone
> has the problem with the new pitch code in the 1.54 update, thanks
>
> Anton
>
> Quoting Pascal Brisset <address@hidden>:
>
> > Hi Anton,
> >
> > address@hidden wrote:
> >
> > > Pascal,
> > > found a nasty bug in the code today.  If you use vmode="gaz" and give 
> > > a gaz
> > > setting gaz="?" you are unable to also specify a pitch angle.  This 
> > > may be a
> > > bug from adding autopitch I am unsure.  Good luck
> >
> >
> >   You are right.
> >   Before the auto_pitch mode was added, "desired_pitch" was computed in
> > navigation_task() (main.c) with
> >     desired_pitch = nav_pitch + pitch_of_vz;
> > and it was independent from the current vertical mode.
> >
> >   This line has been unfortunately moved into climb_pid_run() (pid.c) 
> > which is
> > called only with a vertical_mode greater than AUTO_CLIMB (i.e. CLIMB 
> > or ALT).
> >
> >   This is now fixed:
> >    - climb_pid_run() computes (if auto_pitch) or updates (else) 
> > nav_pitch ;
> >    - desired_pitch is always set to nav_pitch in navigation_task()
> >
> > Please check the source code before flying ...
> >
> > --Pascal
>
>
>
>
>
>
> _______________________________________________
> Paparazzi-devel mailing list
> address@hidden
> http://lists.nongnu.org/mailman/listinfo/paparazzi-devel