I am new to Flight Gear and have often thought it would be great to have an open source flight simulator. About three years ago, I began a project to model the Digitrak Autopilot in MSFS, which turned out okay, but not really like I hoped it could be. I was able to graphically imitate the AP and to connect it to the heading bug and existing internal autopilot and have it generally operate. I was able to model some of the special behaviors, such the initialization delay and display. However, it really could not go beyond that because you could not mod the internal autopilot (well, maybe some expert could by modding the code or doing all the calculations in the embedded scripting language, but that was beyond what I was willing to do). I was frustrated by the inability to model the unique characteristics with accuracy.
I discovered FlightGear supports XML panels and gauges, so I decided to give the project another chance. That's all very long winded, so I will get down to the matter at hand. The instrument I am trying to port is a model of the Digitrak digital autopilot for experimental aircraft to FlightGear. This is made by a company called TruTrak and you can download the PDF files on the Digitrak here: http://www.trutrakflightsystems.com/ttfsproducts.html#Digitrak I've read all the docs that came with the FG distributions and have made a start on the Digitrak for FlightGear, but the learning curve is steep. I've got the texture created, made a instrument XML file and loaded it onto the Cessna panel. The display works and the two main buttons are clickable. In addition, I have studied the documents on the PID controller and copied over the HDG one from the general autopilot for experimentation. So far, I can see that I will need some help pulling all this together and modeling the controller to get this actually working. Then there are a lot of unique features to model. The Digitrak uses GPS Nav or its own internal DG for directional information. I will probably need to use the nasal script to model the initialization delay at startup. This a/p has a variety of modes and nuanced behaviors that will likely be difficult to model. Although I do not intend to model all of the features, I hope to model the basic functionality. Even now I can see that some would make this a very complicated project given the limitations of FGFS and the amount of work to make many of the modes and display features work as in the manual is too much to consider at this time. It would require nearly every feature to be nasal scripted. This would probably be similar to the nasal scripted buttons/functions of the KAP140 autopilot. My attempts so far have left me with some questions. Is there a complete and comprehensive list of properties that are built-in to the FG system (not those created by a particular instrument)? Do I understand correctly that properties in the instrumentation/ hierarchy are built-in values, that come from the system and are not produced by a custom instrument or script? I do understand that you can bring a property into existence by naming it and it appears that the properties in autopilot/ hierarchy are somehow part of the standard autopilot. I believe this means I should create my own set of properties, such as autopilot/digitrak/settings etc. Is that correct? I tried adapting the controller, the cascading one for Heading Bug, but it will not track. It just spins around. I can see it hit a "bump" when it equals the setting. The original HDG controller uses the heading bug error as an input. But I need to input the difference between the orientation of the a/c and the GPS track into the controller. I tried making the orientation in roll axis the input and the GPS track the reference, but that does not seem to work. Also, I do not really understand the GPS system in FGFS. I need to take the heading value from the GPS ground track. The Digitrak in its NAV mode follows a GPS track and ground speed from the GPS. That is all it needs to function. It has various other modes as well, but that is the one I want to get working first. I appreciate your help, Steve _______________________________________________ Flightgear-devel mailing list Flightgear-devel@flightgear.org http://mail.flightgear.org/mailman/listinfo/flightgear-devel 2f585eeea02e2c79d7b1d8c4963bae2d
