Re: [Flightgear-devel] Heads up: new autopilot subsystem implementation
Hi Syd, I just updated my fgdata and I noticed that the change you mentioned (in your message below) is not yet in de git repos. Should it not be in? Apart from that little remark, I can still not get the Aerostar700 AP to hold altitude. I have set the altitude on the dashbaord, armed it, activitated alt-hold with a altitude difference (between actual and desired) of less that 1000 ft (far less in fact, I switched to alt-hold on the desired altitude), but the plane keeps climbing anyway. The only way to keep a more or less stable altitude is by selecting pitch-hold and setting the pitch to a value close to 0.3. m Op 07-07-10 04:48, syd adams schreef: I can't update my aircraft at the moment , but to fix the ALT hold problem , change this starting at line 205 in aerostar-autopilot.xml ... input propinstrumentation/altimeter/indicated-altitude-ft/prop /input Must have been a left over experiment. My mistake , there is no VS hold with this system , but it holds pitch when the autopilot is turned on , so engage it when you've trimmed for a climb ... and make sure the Alerter is armed before your within +- 1000 feet of target altitude ... and it should work properly . At least it does for me . -- This SF.net email is sponsored by Sprint What will you do first with EVO, the first 4G phone? Visit sprint.com/first -- http://p.sf.net/sfu/sprint-com-first ___ Flightgear-devel mailing list Flightgear-devel@lists.sourceforge.net https://lists.sourceforge.net/lists/listinfo/flightgear-devel
Re: [Flightgear-devel] Heads up: new autopilot subsystem implementation
Just to chime in here with a me too, the altitude hold on the Rascal110 also seems to be broke. Something seems to be overriding the target altitude values and/or the mode selection, so the F11 autopilot dialog can't control these. Similar problem with target airspeed. The wing leveler and heading hold modes do seem to work. I haven't had a chance to investigate, but this is pretty recent breakage. Curt. On Sat, Jul 24, 2010 at 7:39 AM, wrote: Hi Syd, I just updated my fgdata and I noticed that the change you mentioned (in your message below) is not yet in de git repos. Should it not be in? Apart from that little remark, I can still not get the Aerostar700 AP to hold altitude. I have set the altitude on the dashbaord, armed it, activitated alt-hold with a altitude difference (between actual and desired) of less that 1000 ft (far less in fact, I switched to alt-hold on the desired altitude), but the plane keeps climbing anyway. The only way to keep a more or less stable altitude is by selecting pitch-hold and setting the pitch to a value close to 0.3. m Op 07-07-10 04:48, syd adams schreef: I can't update my aircraft at the moment , but to fix the ALT hold problem , change this starting at line 205 in aerostar-autopilot.xml ... input propinstrumentation/altimeter/indicated-altitude-ft/prop /input Must have been a left over experiment. My mistake , there is no VS hold with this system , but it holds pitch when the autopilot is turned on , so engage it when you've trimmed for a climb ... and make sure the Alerter is armed before your within +- 1000 feet of target altitude ... and it should work properly . At least it does for me . -- This SF.net email is sponsored by Sprint What will you do first with EVO, the first 4G phone? Visit sprint.com/first -- http://p.sf.net/sfu/sprint-com-first ___ Flightgear-devel mailing list Flightgear-devel@lists.sourceforge.net https://lists.sourceforge.net/lists/listinfo/flightgear-devel -- Curtis Olson: http://baron.flightgear.org/~curt/ -- This SF.net email is sponsored by Sprint What will you do first with EVO, the first 4G phone? Visit sprint.com/first -- http://p.sf.net/sfu/sprint-com-first___ Flightgear-devel mailing list Flightgear-devel@lists.sourceforge.net https://lists.sourceforge.net/lists/listinfo/flightgear-devel
Re: [Flightgear-devel] Heads up: new autopilot subsystem implementation
Dear All I am a new comer to your devel. forum I am using flight gear simulator to validate some of my theoritical design on the airport runway. Could anybody tell me what is reliablity of the Flight gear simulator to peform my research study? I found it is only 75 percent. Am I correct? Greets all Prof. Rao From: Curtis Olson curtol...@gmail.com To: FlightGear developers discussions flightgear-devel@lists.sourceforge.net Sent: Sat, July 24, 2010 6:28:57 PM Subject: Re: [Flightgear-devel] Heads up: new autopilot subsystem implementation Just to chime in here with a me too, the altitude hold on the Rascal110 also seems to be broke. Something seems to be overriding the target altitude values and/or the mode selection, so the F11 autopilot dialog can't control these. Similar problem with target airspeed. The wing leveler and heading hold modes do seem to work. I haven't had a chance to investigate, but this is pretty recent breakage. Curt. On Sat, Jul 24, 2010 at 7:39 AM, wrote: Hi Syd, I just updated my fgdata and I noticed that the change you mentioned (in your message below) is not yet in de git repos. Should it not be in? Apart from that little remark, I can still not get the Aerostar700 AP to hold altitude. I have set the altitude on the dashbaord, armed it, activitated alt-hold with a altitude difference (between actual and desired) of less that 1000 ft (far less in fact, I switched to alt-hold on the desired altitude), but the plane keeps climbing anyway. The only way to keep a more or less stable altitude is by selecting pitch-hold and setting the pitch to a value close to 0.3. m Op 07-07-10 04:48, syd adams schreef: I can't update my aircraft at the moment , but to fix the ALT hold problem , change this starting at line 205 in aerostar-autopilot.xml ... input propinstrumentation/altimeter/indicated-altitude-ft/prop /input Must have been a left over experiment. My mistake , there is no VS hold with this system , but it holds pitch when the autopilot is turned on , so engage it when you've trimmed for a climb ... and make sure the Alerter is armed before your within +- 1000 feet of target altitude ... and it should work properly . At least it does for me . -- This SF.net email is sponsored by Sprint What will you do first with EVO, the first 4G phone? Visit sprint.com/first -- http://p.sf.net/sfu/sprint-com-first ___ Flightgear-devel mailing list Flightgear-devel@lists.sourceforge.net https://lists.sourceforge.net/lists/listinfo/flightgear-devel -- Curtis Olson: http://baron.flightgear.org/~curt/ -- This SF.net email is sponsored by Sprint What will you do first with EVO, the first 4G phone? Visit sprint.com/first -- http://p.sf.net/sfu/sprint-com-first___ Flightgear-devel mailing list Flightgear-devel@lists.sourceforge.net https://lists.sourceforge.net/lists/listinfo/flightgear-devel
Re: [Flightgear-devel] Heads up: new autopilot subsystem implementation
I can't update my aircraft at the moment , but to fix the ALT hold problem , change this starting at line 205 in aerostar-autopilot.xml ... input propinstrumentation/altimeter/indicated-altitude-ft/prop /input Must have been a left over experiment. My mistake , there is no VS hold with this system , but it holds pitch when the autopilot is turned on , so engage it when you've trimmed for a climb ... and make sure the Alerter is armed before your within +- 1000 feet of target altitude ... and it should work properly . At least it does for me . Cheers On Mon, Jul 5, 2010 at 10:54 PM, syd adams adams@gmail.com wrote: I'll take a look at it , could be broken again ... but you need to set the altitude in the Alerter and arm it, then climb to that altitude (with a VS mode, or manually) ... it wont actually ALT hold until your within +- 1000 ft of the target. Pressing the ALT button simply holds the altitude your were at when the button was pressed . That said , it still sounds like something's not quite right with it. Cheers On Mon, Jul 5, 2010 at 10:22 PM, fiers...@zonnet.nl wrote: Yes the AP worked fine before. First select the desired altitude on the dashboard (top). Then select the desired heading on the dial on the dashboard. Next switch on the AP in the dialog and select ALT and HDG. On the dashboard click arm next to the altitude setting. The plane is now supposed to climb or decend to the set altitude, but it keeps climbing and never decends. m Op 05-07-10 21:20, Torsten Dreyer schreef: Hi Torsten (and others), I noticed that the autopilot on the Aerostar700 keeps climbing, despite what I set the AP to. I was trying to level out at FL120, but the AP was happily climbing to almost FL200 when I disabled it. Is this in the new code or in the plane? I can't tell for sure because I am not familiar with the aerostar. Just did a small checkflight and noticed that the autopilot did not act as I had expected. The altitude select at the panel seems to have no effect and using the autopilot dialog did not produce the expected results. Did the aerostar A/P work correctly before the code change? Torsten -- This SF.net email is sponsored by Sprint What will you do first with EVO, the first 4G phone? Visit sprint.com/first -- http://p.sf.net/sfu/sprint-com-first ___ Flightgear-devel mailing list Flightgear-devel@lists.sourceforge.net https://lists.sourceforge.net/lists/listinfo/flightgear-devel -- This SF.net email is sponsored by Sprint What will you do first with EVO, the first 4G phone? Visit sprint.com/first -- http://p.sf.net/sfu/sprint-com-first ___ Flightgear-devel mailing list Flightgear-devel@lists.sourceforge.net https://lists.sourceforge.net/lists/listinfo/flightgear-devel -- This SF.net email is sponsored by Sprint What will you do first with EVO, the first 4G phone? Visit sprint.com/first -- http://p.sf.net/sfu/sprint-com-first___ Flightgear-devel mailing list Flightgear-devel@lists.sourceforge.net https://lists.sourceforge.net/lists/listinfo/flightgear-devel
Re: [Flightgear-devel] Heads up: new autopilot subsystem implementation
Hi Torsten (and others), I noticed that the autopilot on the Aerostar700 keeps climbing, despite what I set the AP to. I was trying to level out at FL120, but the AP was happily climbing to almost FL200 when I disabled it. Is this in the new code or in the plane? m -- This SF.net email is sponsored by Sprint What will you do first with EVO, the first 4G phone? Visit sprint.com/first -- http://p.sf.net/sfu/sprint-com-first ___ Flightgear-devel mailing list Flightgear-devel@lists.sourceforge.net https://lists.sourceforge.net/lists/listinfo/flightgear-devel
Re: [Flightgear-devel] Heads up: new autopilot subsystem implementation
On Mon, 2010-07-05 at 10:52 +0200, fiers...@zonnet.nl wrote: Hi Torsten (and others), I noticed that the autopilot on the Aerostar700 keeps climbing, despite what I set the AP to. I was trying to level out at FL120, but the AP was happily climbing to almost FL200 when I disabled it. Is this in the new code or in the plane? m This sounds very similar to the problem I encountered 3 days ago with a number of aircraft. If you are rebuilding from the (fairly) latest git code base for simgear and flightgear, then you need to also update your data directory from fgdata. I've copied Torsten and James reply below; Greetings, I did a GIT update and rebuild recently and have noticed that no properties under /autopilot/internal/ are being set or updated. Quite a few autopilot XML files use the lookahead speed and fdm-heading-bug-error amongst some others. Does anyone know where they went and if they are coming back. The calculation of these properties is no longer performed hardcoded but using a separate and additional autopilot configuration file. This should be included by your preferences.xml. Please check if you have something like autopilot pathAircraft/Generic/generic-autopilot-helper.xml/path /autopilot near line 226 of preferences.xml If not: do a git pull on the fgdata. If yes: You found is is a bug. Greetings, Torsten -- On 1 Jul 2010, at 19:53, Torsten Dreyer wrote: Please check if you have something like autopilot pathAircraft/Generic/generic-autopilot-helper.xml/path /autopilot near line 226 of preferences.xml If not: do a git pull on the fgdata. If yes: You found is is a bug. And if you're using the 777, it's a known bug, which has a short-term work-around; longer term, I have a rewrite of the AP pending, but it's a couple of weeks from being committed. James -- This SF.net email is sponsored by Sprint What will you do first with EVO, the first 4G phone? Visit sprint.com/first -- http://p.sf.net/sfu/sprint-com-first ___ Flightgear-devel mailing list Flightgear-devel@lists.sourceforge.net https://lists.sourceforge.net/lists/listinfo/flightgear-devel
Re: [Flightgear-devel] Heads up: new autopilot subsystem implementation
I already had updated my data dir. Just did it again, but it is up to date, it says. m Op 05-07-10 12:55, Scott Hamilton schreef: This sounds very similar to the problem I encountered 3 days ago with a number of aircraft. If you are rebuilding from the (fairly) latest git code base for simgear and flightgear, then you need to also update your data directory from fgdata. I've copied Torsten and James reply below; -- This SF.net email is sponsored by Sprint What will you do first with EVO, the first 4G phone? Visit sprint.com/first -- http://p.sf.net/sfu/sprint-com-first ___ Flightgear-devel mailing list Flightgear-devel@lists.sourceforge.net https://lists.sourceforge.net/lists/listinfo/flightgear-devel
Re: [Flightgear-devel] Heads up: new autopilot subsystem implementation
In the Citation II, the AP seems to level off at around 24000ft, while it is set for 3ft. Pressure set is 2992. Also, initially I set the AP for an ALT of 35000. While climbing at around 22000 I changed the setting to 3ft. The plane went into a dive and decended to below 2ft. I took control to stabilize the altitude and reengaged the AP, set to 3ft, but it never reached that ALT. I compiled this morning and the data dir is up to date. m -- This SF.net email is sponsored by Sprint What will you do first with EVO, the first 4G phone? Visit sprint.com/first -- http://p.sf.net/sfu/sprint-com-first ___ Flightgear-devel mailing list Flightgear-devel@lists.sourceforge.net https://lists.sourceforge.net/lists/listinfo/flightgear-devel
Re: [Flightgear-devel] Heads up: new autopilot subsystem implementation
Hi Torsten (and others), I noticed that the autopilot on the Aerostar700 keeps climbing, despite what I set the AP to. I was trying to level out at FL120, but the AP was happily climbing to almost FL200 when I disabled it. Is this in the new code or in the plane? I can't tell for sure because I am not familiar with the aerostar. Just did a small checkflight and noticed that the autopilot did not act as I had expected. The altitude select at the panel seems to have no effect and using the autopilot dialog did not produce the expected results. Did the aerostar A/P work correctly before the code change? Torsten -- This SF.net email is sponsored by Sprint What will you do first with EVO, the first 4G phone? Visit sprint.com/first -- http://p.sf.net/sfu/sprint-com-first ___ Flightgear-devel mailing list Flightgear-devel@lists.sourceforge.net https://lists.sourceforge.net/lists/listinfo/flightgear-devel
Re: [Flightgear-devel] Heads up: new autopilot subsystem implementation
--
From: Torsten Dreyer tors...@t3r.de
Sent: Thursday, June 24, 2010 4:24 PM
To: flightgear-devel@lists.sourceforge.net
Subject: [Flightgear-devel] Heads up: new autopilot subsystem implementation
Hi all,
I have just commited a pretty large commit for a new implementation of the
autopilot subsystem(group). The reason for that was that the old way of
having
everything stuffed into one single source file of 1360 lines and a header
file
of 514 lines made working on this piece of code allways a little
challenging.
While at it, I replaced some design patterns to look a little more OO-ish
and
added some doxygen compatible comments to the class definition.
snip
Because I touched at least 90% of all code lines, it is very likely that I
introduced a new bug. Especially the build process was tested on linux
only. I
have no chance to test the build on Windows on OS/X.
snip
Torsten
Now that the config.h file is sorted (many thanks for the commit) I have
again been able to make an MSCV09 from the next branch.
However it was not without problems ;-)
In the autopilot logic.cx was failing due to NOMINMAX not being set. There
were also several problems due to SGvec3 and SGvec4 not being defined.
These I have sorted by re-arranging the order of include file calls. There
may be a better solution, and perhaps my solution will not work on other
platforms.
The changes are all at the beginning of files:-
in analogcomponenet.hxx...
#ifndef __ANALOGCOMPONENT_HXX
#define __ANALOGCOMPONENT_HXX 1
#include component.hxx
#include inputvalue.hxx
namespace FGXMLAutopilot {
in inputvalue.cxx
// Foundation, Inc., 51 Franklin Street, Fifth Floor, Boston, MA
02110-1301, USA.
//
#include Main/fg_props.hxx
#include inputvalue.hxx
using namespace FGXMLAutopilot;
in logic.hxx...
#ifndef __LOGICCOMPONENT_HXX
#define __LOGICCOMPONENT_HXX 1
#include analogcomponent.hxx
#include digitalcomponent.hxx
namespace FGXMLAutopilot {
Once it all compiled there were some undefined variables.
To solve these fdm_shell.cxx/.hxx and flightProperties.cxx/.hxx needed to be
added to Lib_Flight.
This produced another compilation error:-
The correction in flightProperties.cxx is
#include Main/globals.hxx
#include FDM/flightProperties.hxx
#include simgear/props/props.hxx
#include simgear/math/SGMath.hxx
#include Main/globals.hxx
FlightProperties::FlightProperties(SGPropertyNode* root) :
Also note that fdm_shell.cxx/.hxx and flightProperties.cxx/.hxx do not have
the usual copyright header.
All compiles and links now. I have not tested to see if it runs OK.
Thanks for all the continuing work by all concerned.
Alan
--
This SF.net email is sponsored by Sprint
What will you do first with EVO, the first 4G phone?
Visit sprint.com/first -- http://p.sf.net/sfu/sprint-com-first
___
Flightgear-devel mailing list
Flightgear-devel@lists.sourceforge.net
https://lists.sourceforge.net/lists/listinfo/flightgear-devel
Re: [Flightgear-devel] Heads up: new autopilot subsystem implementation
These I have sorted by re-arranging the order of include file calls. There may be a better solution, and perhaps my solution will not work on other platforms. This was due to the missing include of config.h. A very common pitfall for linux developers. Should be fixed now. Once it all compiled there were some undefined variables. To solve these fdm_shell.cxx/.hxx and flightProperties.cxx/.hxx needed to be added to Lib_Flight. They are added to the vcproj file. Also note that fdm_shell.cxx/.hxx and flightProperties.cxx/.hxx do not have the usual copyright header. I'll leave it to James who is the author of these files to add a copyright and license statement. All compiles and links now. I have not tested to see if it runs OK. Please check, if this is still the case after my commits or if I broke anything new. Thanks for all the continuing work by all concerned. Thanks for testing and reporting! Torsten -- This SF.net email is sponsored by Sprint What will you do first with EVO, the first 4G phone? Visit sprint.com/first -- http://p.sf.net/sfu/sprint-com-first ___ Flightgear-devel mailing list Flightgear-devel@lists.sourceforge.net https://lists.sourceforge.net/lists/listinfo/flightgear-devel
Re: [Flightgear-devel] Heads up: new autopilot subsystem implementation
On 27 Jun 2010, at 14:12, Torsten Dreyer wrote: Also note that fdm_shell.cxx/.hxx and flightProperties.cxx/.hxx do not have the usual copyright header. I'll leave it to James who is the author of these files to add a copyright and license statement. Good point, I'll add these. James -- This SF.net email is sponsored by Sprint What will you do first with EVO, the first 4G phone? Visit sprint.com/first -- http://p.sf.net/sfu/sprint-com-first ___ Flightgear-devel mailing list Flightgear-devel@lists.sourceforge.net https://lists.sourceforge.net/lists/listinfo/flightgear-devel
[Flightgear-devel] Heads up: new autopilot subsystem implementation
Hi all, I have just commited a pretty large commit for a new implementation of the autopilot subsystem(group). The reason for that was that the old way of having everything stuffed into one single source file of 1360 lines and a header file of 514 lines made working on this piece of code allways a little challenging. While at it, I replaced some design patterns to look a little more OO-ish and added some doxygen compatible comments to the class definition. Every single algorithm should work as before and the xml syntax should be 100% compatible with existing configurations. Please note, that the former hard coded autopilot helper functions are gone now. These functions are now implemented as xml components within $FGDATA/Aircraft/Generic/generic-autopilot-helper.xml which is included in preferences.xml by default. Every property that existed before this change should be in place after the update. Because I touched at least 90% of all code lines, it is very likely that I introduced a new bug. Especially the build process was tested on linux only. I have no chance to test the build on Windows on OS/X. I have flown some aircraft to test the new implementation but certainly not all of them. Please give your favourite aircraft a test flight with the autopilot and report any failures. Torsten -- ThinkGeek and WIRED's GeekDad team up for the Ultimate GeekDad Father's Day Giveaway. ONE MASSIVE PRIZE to the lucky parental unit. See the prize list and enter to win: http://p.sf.net/sfu/thinkgeek-promo ___ Flightgear-devel mailing list Flightgear-devel@lists.sourceforge.net https://lists.sourceforge.net/lists/listinfo/flightgear-devel
Re: [Flightgear-devel] Heads up: new autopilot subsystem implementation
On 24 Jun 2010, at 16:24, Torsten Dreyer wrote: Every single algorithm should work as before and the xml syntax should be 100% compatible with existing configurations. Please note, that the former hard coded autopilot helper functions are gone now. These functions are now implemented as xml components within $FGDATA/Aircraft/Generic/generic-autopilot-helper.xml which is included in preferences.xml by default. Every property that existed before this change should be in place after the update. Great work, Torsten! James -- ThinkGeek and WIRED's GeekDad team up for the Ultimate GeekDad Father's Day Giveaway. ONE MASSIVE PRIZE to the lucky parental unit. See the prize list and enter to win: http://p.sf.net/sfu/thinkgeek-promo ___ Flightgear-devel mailing list Flightgear-devel@lists.sourceforge.net https://lists.sourceforge.net/lists/listinfo/flightgear-devel
