a better autopilot for airliners

[Guest tralluccio]

hi everyoneI write to ask your help regarding how to get a more realistic autopilot for the boing737 in fs9. I refer particularly to the ias mode, it presently only works with autotrottle, while it should also work without autotrottle, indeed during take off and climb thrust is set at a fixed value and flight director or autopilot should mantain the proper speed (i.e. v2+10) with the proper attitude. This mode is absolutely needed for a realistic airline flight. Do you know if there if a freeware product that can give me this autopilot improvement or there is a way for modifying the autopilot in fs9?thank you

[KevinAu 441]

I haven't tried it in FS9, but the AP from the FS2002 ATR made by Sanchez-Castaner included an IAS pitch mode. I had added it successfully into the Falcon50 also back then. Espen's Fokker 50 panel also includes an autopilot with the proper IAS pitch mode and so does the Grabowski ERJ panel.

[phjvh 0]

Hi,As far as i know these panels have big C++ gauges.In xml it is possible to control speed with pitch after setting a certain amount of power (TOGA, CL, CL I, CL II, CON and CRZ).It is some code and calc, but fun to figure out!Jan"Procul Negotiis"

[Guest Ian_K]

Hello,a working AP would be a useful project. However it would not be easy to get right immediately, given it would have to cover the vast differences in the aerodynamics of the different aircraft that can be built in MSFS.Even Burt Rutan can't get a working AP for his GlobalFlyer as it has a massive weight change in it's round the world trip.The AP needs to be able to predict the thrust and flying control settings that are to be output to control the aircraft in a smooth and predictable way. This means having an internal aerodynamic model to compute the future track and orientation of the vehicle for the next 10 mins of so. This needs to be iterated until the predicted and required paths coincide. Some sort of Runge-Kutta intergrating extrapolator would be used.Have lots of fun :-)Ian

[phjvh 0]

Hi,In my case, using xml, i set N1 and let IAS or MACH regulate VS.Something like: Setting the throttle axis dependant of parameters as weight, temperature, altitude, ias, cn1 etc.Going to fast then climb rate increases etc.Quite some code, but the result is a stable toga, climb and cruise.Also possible with this system to calculate the "plafond" and make step climbs automatically.Still very busy with it.May be not real world, but fun.Jan"Procul Negotiis"

[Guest tralluccio]

thanks kevincan you explain briefly how could you add the ias mode from the atr panel into a different aircraft's panel? were those xml panels?thanks

[rcbarend 37]

Hi,Interesting you should ask...:-)I made such a gauge, and have just uploaded it to flightsim.comFile: rcbfl-10.zip.I can't publish here at avsim now, because of the library problem (no upload possible). But I've attached it to this post.It might not do exactly what you want, since it was intended only for flight level changes, not takeoff. But if you try it out, and specify some other requirement, I might extend it for take-off :-)What it is:""FS2004 Gauges: Flight Level Change gauge.This freeware gauge is an extension to the standard FS Autopilot functions, and controls the climb/descent rate during an AP-controlled flight level change, whilst maintaining a user-set thrust and airspeed.It can be added to any panel that uses an AP with a Altitude Set/Holdand V/S Set/Hold function, and can be used in both FS2002 and FS2004.By Rob Barendregt.""Cheers, Rob

[phjvh 0]

Nice work Rob!Isn't that a coincidence?I tested it, but it interferes with the rest of my code.Do you use somewhere VS Hold or Set?I never did.What is your experience with those events?Jan"Procul Negotiis"

[n4gix 4,947]

>It might not do exactly what you want, since it was intended>only for flight level changes, not takeoff. But if you try it>out, and specify some other requirement, I might extend it for>take-off :-)Hi Rob!I just wanted to let you know that your FLC gauge seems to work a real treat! I plopped it into a project I'm currently working on (Citation II SP/2), and it functions splendidly.I've been trying for ages to write an adaptive algorithim that will enable a true VS mode on the CitII's AP, but with no real success.Now, if I can manage to parse the XML and translate it to C-speak, I may have found a solution... :)Thanks for the gift! :)

[rcbarend 37]

Hi Jan,""Isn't that a coincidence?""I had that several times when doing a project :-)...I don't (or have used) the VS Hold events yet, in my gauge I just test/assume it's one (and periodically set the VS value)Cheers, Rob

[rcbarend 37]

Hi Bill,Glad you like it."Now, if I can manage to parse the XML and translate it to C-speak"Well, to be honest, for complex algoritms, I write them in "C" first and them "translate" it to XML (Revers Polish Notation).If you open the .xml file with Notepad: in the Comment section I described how the algoritm works. Maybe that helps :-)On the other hand: if you put some effort in trying to grasp how XML is coded, you can learn of a lot from it; especially how-to use the various FS variables. There are so many good examples out there...Cheers, Rob

[n4gix 4,947]

>Hi Bill,>>Glad you like it.>>"Now, if I can manage to parse the XML and translate it to>C-speak">Well, to be honest, for complex algoritms, I write them in "C">first and them "translate" it to XML (Revers Polish>Notation).>>If you open the .xml file with Notepad: in the Comment section>I described how the algoritm works. Maybe that helps :-)>On the other hand: if you put some effort in trying to grasp>how XML is coded, you can learn of a lot from it; especially>how-to use the various FS variables. There are so many good>examples out there...Actually, I've already done that. XML is no stranger to me, but parsing a complex statement always gives me a headache! I much prefer the more structured and linear approach of a language like COBOL, Fortran, or even... C... :)Having to sandwich an entire algorithim into ONE statement is deucedly more complicated... :)For my purposes however, I'll have to make some concessions to the authentic visual cockpit, by having a pre-set target throttle %'age for climb, and another for descent. I'll also have to grab the "target" FL directly from the preset in the ALT SET gauge. IOW, the operator will dial in the new FL directly and allow the 'normal' AP ALT to begin the climb, then immediately select the VS button to engage the (hidden) FLC function.It's a bit of fun watching the CitII climb to FL490 with 20% fuel load (which is well above its celing, BTW)... :)

[phjvh 0]

Rob,I managed to incorporate your FLCH Mode and Warning Sound in my Autopilot and Thrustmode Computer.The big deal was the interference with your and my (G:Var's)...I'm still testing, but i can say it works very good and even better than my original code, see above.Thanks, hope you don't mind some renaming!Have a nice Sunday,Jan"Procul Negotiis"

[Guest tralluccio]

robit's a great job! Ive no time to test it now but I will do next week and report to you in this thread. May be better if one can set the thrust with trottle.bye

[rcbarend 37]

Hi Jan,Do you mean to say that G:Var* variables used in 2 .xml gauges interfere with each other (i.e. the visibility of a G:Var is NOT limited to one .xml gauge) ???Rob