Hi,
thanks for the questions and pointing out TrueWind.
TrueWind uses one vector with known IAS, true magnetic heading and GPS speed
and GPS direction.
The method in my mind used two different vectors with known IAS, and known
GPS speed and GPS direction. The two vectors would differ by some (30
degrees + ) angle (making up for the unknown true magnetic heading). This
way the pilot wouldn't have to follow one of the 12 predetermined headings.
But the calculation could only conclude after second steady glide, which is
sufficiently diverted from the first one.
Now I think the TrueWind method is better then "my" method.
TrueWind is also great because it's already implemented.... :-).
Would people like to have something like TrueWind in XCSoar?
Tibor
P.S.
There is still the open question of precision of magnetometers in Android
devices... Maybe they would be good enough to provide good data....
On Mon, May 23, 2011 at 16:17, Roman Stoklasa <rst...@gmail.com> wrote:
> Hi all,
>
> isn't it easier to implement (or merge) so called "TrueWind Calculation"
> from LK8000?
> If I understand the discussion correctly, TrueWind is doing pretty much the
> same, what is Tibor suggesting...
>
> For those who do not know this feature, more information could be found in
> the LK8000's manual (http://www.bware.it/LK8000/LK8000_MANUAL_122.pdf),
> page 98, chapter "TrueWind Calculation v1".
>
> Tibor, please have a look at the LK8000 manual and compare it to your
> ideas. Is it "the same"? Or what's the difference? If it differs, which
> method would be better and why?
>
> Regards,
>
> Roman.
>
>
>
>
>
>
> On Mon, 23 May 2011 15:27:41 +0200, Tibor Arpas <ti...@bidforfix.com>
> wrote:
>
> > Just to clarify after John's question.
> >
> > The pilot needs to do zig zag for the zig zag calculation to work. And to
> do
> > zig zag flight path you inevitably have to make a turn.
> >
> > I'm still working with the assumption that there is no airspeed sensor,
> or
> > vario with airspeed data output in the glider.
> >
> > But the pilot flies constant, known airspeed for extended periods of time
> so
> > the airspeed sensor would indicate the same known value for these
> extended
> > periods of time.
> > Except of course for the situations when the pilot is not maintaining the
> > speed.
> > - circling
> > - change of decision and a scrappy turn into a new direction
> > - other distractions
> >
> > So ideally, the known speed is maintained most of the time and only
> > interrupted by some maneuvers which are detectable from the GPS track
> (with
> > a little help of some user interaction).
> >
> > This style of flying and little user interaction should give XCSoar
> enough
> > data to get wind calculation as good as LX 7000 or similar.
> >
> > Tibor
> >
> >
> > On Mon, May 23, 2011 at 13:16, John Wharington <jwharing...@gmail.com
> >wrote:
> >
> >> Hi Tibor,
> >>
> >> Maybe you misunderstand, you don't need constant airspeed for the zigzag
> >> wind algorithm to work,
> >> assuming you have a vario that outputs airspeed to XCSoar.
> >>
> >>
> >> On Mon, May 23, 2011 at 9:15 PM, Tibor Arpas <ti...@bidforfix.com>
> wrote:
> >>
> >>> Hmm, no idea, I look outside, not into instruments. when searching for
> >>> lift or making ad hoc decisions during glide I wouldnt like to
> concentrate
> >>> on IAS or some steady turns. Maybe it would be most beneficial on final
> >>> glide...
> >>> On May 23, 2011 12:07 PM, "John Wharington" <jwharing...@gmail.com>
> >>> wrote:
> >>> > You also need to be turning for zigzag wind to work; i wonder how
> >>> accurately
> >>> > you can fly to a given true airspeed while turning?
> >>> >
> >>> > On Mon, May 23, 2011 at 4:03 PM, Tibor Arpas <ti...@bidforfix.com>
> >>> wrote:
> >>> >
> >>> >> Hi everybody,
> >>> >>
> >>> >> I just recently learned strong arguments for a very steady
> >>> inter-thermal
> >>> >> cruising speeds (ignoring the McReady speed to fly):
> >>> >> http://www.betsybyars.com/guy/soaring_symposia/72price.html
> >>> >>
> >>> >> Sure, you should consult your gliders polar and the estimated climb
> in
> >>> >> front of you.
> >>> >> But until you hit something clearly worth circling in, you can
> maintain
> >>> >> steady speed.
> >>> >>
> >>> >> I flew this way during my last competition and it really feels
> better.
> >>> And
> >>> >> I think it also makes the airspeed sensor input unnecessary.
> >>> >> For a given day I might decide I'll do 150 km/h glides.
> >>> >>
> >>> >> Feature request:
> >>> >> I would like to set-up my typical gliding (air) speed in the "fligh
> >>> >> set-up". Now whenever XCSoar would detect a steady (GPS) speed for
> 20+
> >>> >> seconds it would beep and I would be able to confirm with gesture
> that
> >>> I was
> >>> >> flying my typical airspeed during those 20 seconds.
> >>> >>
> >>> >> This way, I imagine, we would have more reliable wind computations
> even
> >>> for
> >>> >> those frequent cases when the wind changes and you don't circle.
> >>> >>
> >>> >> Tibor
> >>> >>
> >>> >>
> >>> >>
> >>>
> ------------------------------------------------------------------------------
> >>> >> What Every C/C++ and Fortran developer Should Know!
> >>> >> Read this article and learn how Intel has extended the reach of its
> >>> >> next-generation tools to help Windows* and Linux* C/C++ and Fortran
> >>> >> developers boost performance applications - including clusters.
> >>> >> http://p.sf.net/sfu/intel-dev2devmay
> >>> >> _______________________________________________
> >>> >> Xcsoar-user mailing list
> >>> >> Xcsoar-user@lists.sourceforge.net
> >>> >> https://lists.sourceforge.net/lists/listinfo/xcsoar-user
> >>> >>
> >>> >>
> >>>
> >>
> >>
>
>
> ------------------------------------------------------------------------------
> What Every C/C++ and Fortran developer Should Know!
> Read this article and learn how Intel has extended the reach of its
> next-generation tools to help Windows* and Linux* C/C++ and Fortran
> developers boost performance applications - including clusters.
> http://p.sf.net/sfu/intel-dev2devmay
> _______________________________________________
> Xcsoar-user mailing list
> Xcsoar-user@lists.sourceforge.net
> https://lists.sourceforge.net/lists/listinfo/xcsoar-user
>
------------------------------------------------------------------------------
What Every C/C++ and Fortran developer Should Know!
Read this article and learn how Intel has extended the reach of its
next-generation tools to help Windows* and Linux* C/C++ and Fortran
developers boost performance applications - including clusters.
http://p.sf.net/sfu/intel-dev2devmay
_______________________________________________
Xcsoar-user mailing list
Xcsoar-user@lists.sourceforge.net
https://lists.sourceforge.net/lists/listinfo/xcsoar-user
