Quick response with a few factual corrections: On Fri, 2007-03-02 at 00:56 -0500, John Denker wrote: > On 03/01/2007 08:02 PM, Dave Perry wrote: > > > .... there are only 3 options still being put forward; all having to > > do with how the kap140 gets the baro shift. These are: > > > > 1) computes it in kap140.nas using the pressureToHeight() > > function already there but not presently used (need to check the > > constants for consistency with John's model), > > > > 2) gets the kollsman shift value already computed by John's > > altimeter.cxx (but not presently saved to the property tree by > > John's altimeter.cxx) , or > > > > 3) gets it by fetching from the encoder propeties the indicated > > altitude and the pressure altitude, and then subtracting. > > I would add > 4) Use the shift() function from the Kollsman object in > http://www.av8n.com/fly/fgfs/atmo.nas
Option 4 is essentially the same as option 1 in functionality but not in clarity. As stated, in my note, the current kap140.nas will not compute (or get from the property tree) a new kollsman value unless the value of baro set has changed. And Roy's function includes the values of the constants used to compute the coefficient and exponent which has the advantage over 4) which just gives the two constants, coefficient and exponent. This noted, the computational efficiency is the same. If you won't share the computation of kollsman shift already computed in the encoder, I believe Roy will want to just use option 1) once the change to altimeter.cxx is in cvs. Have you asked to have atmo.diff applied to cvs?. > > The comparison as I see it goes like this: > > 1 2 3 4 > > verisimilitude + 0 0 + > > simplicity + 0 0 + > > accuracy + + + + > > computational 0 + + + > efficiency > > > It seems like option (4) is Pareto-superior. As I said previously, > I would be astonished if real autopilots didn't use this approach, > or something very similar. > > =================================== > > In the last couple of days, there have been bugfixes and > upgrades to atmo.nas, atmo.hxx and atmo.cxx ... the current > versions are, as usual, obtainable from: > http://www.av8n.com/fly/fgfs/atmo.nas > and > http://www.av8n.com/fly/fgfs/atmo.diff The only changes in the files in these links are not bug fixes but rather improvements of efficiency in the kollsman functions. The returned values are unchanged. However, I am glad to see that you did fix the bug in altimeter.cxx, so the PA low pass for tau > 0 now works correctly. -- Dave Perry ------------------------------------------------------------------------- Take Surveys. Earn Cash. Influence the Future of IT Join SourceForge.net's Techsay panel and you'll get the chance to share your opinions on IT & business topics through brief surveys-and earn cash http://www.techsay.com/default.php?page=join.php&p=sourceforge&CID=DEVDEV _______________________________________________ Flightgear-devel mailing list Flightgear-devel@lists.sourceforge.net https://lists.sourceforge.net/lists/listinfo/flightgear-devel
