paparazzi-devel
[Top][All Lists]
Advanced

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

RE: [Paparazzi-devel] Throttle dies in auto2


From: Roman Krashanitsa
Subject: RE: [Paparazzi-devel] Throttle dies in auto2
Date: Thu, 31 Aug 2006 08:00:13 -0700

Anton, I used FAILSAFE value as a filler, you can alse use CLIMB_LEVEL_GAZ 
instead. Code snippet in the previous message gives just an illustration of the 
idea and the place in code where it can be used. As for the failsafe mode, 
nonzero roll should make your airplane come back, or at least sustain the 
flight long enough to correct the situation without flying off. This is how I 
understand failsafe mode :)

Roman

> -----Original Message-----
> From: 
> address@hidden 
> [mailto:address@hidden
> rg] On Behalf Of address@hidden
> Sent: Thursday, August 31, 2006 6:07 AM
> To: address@hidden
> Subject: RE: [Paparazzi-devel] Throttle dies in auto2
> 
> 
> Let me say having FAILSAFE_DEFAULT_GAZ as something other 
> than 0 is a gamble. If you lose GPS long enough for the 
> failsafe to work, and long enough for the plane to have 
> landed if you went throttle off, then its just not going to 
> come back.  Putting this to a number like your level 
> throttle, and using close to wings level can potentially 
> allow your AC to fly away if GPS is lost, 
> which can
> and will happen to everyone at some point.
> 
> Anton
> 
> 
> Quoting Roman Krashanitsa <address@hidden>:
> 
> > I found the problem. It seems that nav_desired_gaz should be
> > specified at least once taking into account logics of the flight 
> > plan. For example, if gaz is specified only once in the init block, 
> > the program may skip it due to (launch && z>ALT) already true and 
> > this variable will be left at 0, like in my case when it is being 
> > launched in manual.
> >
> > As a solution for silly people like me, we could modify 
> nav.c:359 from 
> > uint16_t nav_desired_gaz; To
> > uint16_t 
> nav_desired_gaz=TRIM_UPPRZ((FAILSAFE_DEFAULT_GAZ)*MAX_PPRZ);
> >
> > Or add nav_desired_gaz=TRIM_UPPRZ((FAILSAFE_DEFAULT_GAZ)*MAX_PPRZ);
> > To nav_init(void)
> >
> > Best regards,
> > Roman
> >
> >> -----Original Message-----
> >> From:
> >> address@hidden
> >> [mailto:address@hidden
> >> rg] On Behalf Of Pascal
> >> Sent: Wednesday, August 30, 2006 2:56 PM
> >> To: address@hidden
> >> Subject: Re: [Paparazzi-devel] Throttle dies in auto2
> >>
> >>
> >> Hi Roman, Hi Anton,
> >>
> >> the default is "no slew" (pid.c:131, CLIMB_MAX_DIFF_GAZ=1).
> >>
> >> What is expected (from what I reread in the code):
> >>   - MANUAL: no control on the throttle, you get what you 
> set on the 
> >> stick
> >>   - AUTO1, AUTO2: slew is active: you cannot immediately go from 
> >> throttle=0 to throttle=100%. The variation in time is limited by 
> >> CLIMB_MAX_DIFF_GAZ
> >>
> >> MANUAL -> AUTO1|AUTO2 :
> >>  Before the switch
> >>   - desired_gaz is set (main_ap.c:258)
> >>   - pid_slew() is run (main_ap.c:462)
> >>  so the controller follows (with a delay) your manual throttle 
> >> command (of course this desired_gaz value is unused by the fbw
> >> process in this mode)  After the switch: desired_gaz smoothly
> >> evolves from the current value
> >> to the one computed by the controller (AUTO2) or set by the
> >> stick (AUTO1)
> >>
> >>  So I can't explain the problem you observe ... And I 
> can't reproduce 
> >> it with the simulator :-(
> >>
> >>  About the CLIMB_MODE_PITCH: desired_gaz is set to nav_desired_gaz 
> >> which is set by the navigation task, i.e. the flight plan.
> >>
> >>  So we will have to boot a plane to figure out the 
> problem. I think 
> >> we almost never flew in MANUAL since the slew control was 
> implemented 
> >> ...
> >>
> >>  Last remark: do we really want the slew in AUTO1 ?
> >>
> >> --Pascalr
> >>
> >> address@hidden wrote:
> >>
> >> > Roman,
> >> > Ive seen this before.  What climb mode are you using?  
> If you arent 
> >> > specifying then Im pretty sure that part functions.  
> However There 
> >> > is
> >> a throttle
> >> > slewing
> >> > built in, Pascal would have to tell you whether its default
> >> on or not,
> >> > as I
> >> > dont have code infront of me.
> >> >
> >> > Anton
> >> >
> >> >
> >> > Quoting Roman Krashanitsa <address@hidden>:
> >> >
> >> >> Guys, I have this weird behavior while flying in 
> manual/auto1 and 
> >> >> then switching to auto2. First the throttle dies to 0, 
> then slowly 
> >> >> goes up to the required value. Im I doing something wrong?
> >> >>
> >> >> I suspect that this is related to the initial value of 
> >> >> nav_desired_gaz variable in nav.c when
> >> climb_mode==CLIMB_MODE_PITCH
> >> >> All other cases use controlled_gaz.
> >> >>
> >> >>
> >> >>
> >> >>
> >> >> _______________________________________________
> >> >> Paparazzi-devel mailing list
> >> >> address@hidden 
> >> >> http://lists.nongnu.org/mailman/listinfo/paparazzi-devel
> >> >
> >> >
> >> >
> >> >
> >> >
> >> >
> >> > _______________________________________________
> >> > Paparazzi-devel mailing list
> >> > address@hidden 
> >> > http://lists.nongnu.org/mailman/listinfo/paparazzi-devel
> >>
> >>
> >>
> >>
> >> _______________________________________________
> >> Paparazzi-devel mailing list
> >> address@hidden 
> >> http://lists.nongnu.org/mailman/listinfo/paparazzi-devel
> >>
> >
> >
> >
> >
> > _______________________________________________
> > Paparazzi-devel mailing list
> > address@hidden 
> > http://lists.nongnu.org/mailman/listinfo/paparazzi-devel
> 
> 
> 
> 
> 
> _______________________________________________
> Paparazzi-devel mailing list
> address@hidden 
> http://lists.nongnu.org/mailman/listinfo/paparazzi-devel
> 






reply via email to

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