We are trying to run a tradheli flightaxis simulation in RealFlight to collect data for a research project. We have it connected and set-up per the instructions on the documentation and the github RealFlight ready-made models page. We are using the Raptor720 (the only ready model on the github).
Our issue is this: everything connects fine and loads great – we start the SITL through the simulator tab on MissionPlanner and make sure that flightaxis is selected. When we arm and increase the collective, specifically in manual modes like stabilize, the helicopter pitches back significantly as it leaves the ground. When we decrease the collective, the helicopter pitches forward violently. This forward pitch is significantly larger than the stops should allow in stabilize mode and are much larger than full pitch-down with the right stick.
This Raptor720 helicopter works great and there is no uncommanded pitch when flying without SITL present – it only occurs when MissionPlanner and the flightaxis SITL are connected.
We have loaded the parameters into the sitl instance (a few errors occured when doing this), but this is the only thing that seems to be causing an issue currently.
@LinkRaymond try this model and use the param list that has E720 in the name.
Before you do that. Try the simulation using master rather than stable. There’s a change that Matt Incorporated recently for a problem with SITL and real flight.
I was able to test the dev flightaxis (master) as well as the files you gave me. This didn’t solve the issue, however. The models both flew great before connecting to the sitl instance.
Dev FlightAxis - no change was detected for this. I launched the sim from mission planner and it connected to RealFlight fine. I also removed the “sitl” folder from mission planner so that I could get a fresh download of the sim. Both Helo models flew perfectly before connecting, and then as soon as I’d connect, every time I increased the collective the helicopter would drastically pitch backwards. The opposite occurred when collective was lowered, usually resulting in a crash.
Files you gave me - I imported the E720 model, started the dev sim, opened the “config” tab on MP, then wrote the .param file for this model with “E720” in the name. Then I tried it, closed the instance and RealFlight to be safe, and then restarted and tried again. Same issue occurred with pitch.
The weird thing is that everything else seems to be acting great – I can do auto missions fine (at one altitude haha) and roll and yaw are fine. The only thing that seems to be coupled is pitch and collective increase. I checked MP radio inputs and everything looks fine, but the servo outputs confirm that pitch is being induced with a collective (Radio 3) increase.
Let me know if anything I’m doing is not the correct procedure or if there is something else I can try.
@LinkRaymond Ray, I think this is caused by your software radio settings on your real flight controller. Go into settings for the controller, then select edit. You should see two check boxes at the bottom of the screen for software radio. These should be checked. I think these are typically checked by default.
The reason that we use software radio for heli is because the swashplate type of the aircraft is an H1. The software radio modifies it to be an H3–120. It allows a more realistic set up in Arducopter.
Hopefully this fixes your problem.
@bnsgeyer Bill, thank you so much! That was the issue – when we created the custom controller profile for our Taranis controller, we didn’t turn those on. It flies great now.