I’m conducting a research project at my university using a fixed-wing aircraft powered by a jet engine. During a recent flight test, I encountered an issue where the aircraft developed growing pitch oscillations. At some point, these oscillations may have contributed to the engine shutting down, as the airspeed dropped significantly shortly after.
Fortunately, the plane managed to glide safely and landed without further incident.
I’d greatly appreciate any insights into what might have caused these oscillations.
There appears to be a significant issue with the EKF3’s primary altitude source (BARO). I recommend addressing the barometer oscillations first, as they could be the root cause of the observed behavior. It’s best to resolve this before proceeding with further tuning.
As a side note, could you describe how the barometer is mounted? Improper placement—particularly if it’s exposed directly to airflow or prop wash—can lead to unstable readings
@randyV
Thanks for the feedback! Just to confirm, the barometer is the internal one within the Cube Orange, mounted inside the fuselage. It’s not exposed directly to airflow or prop wash.
The Cube (mounted inside the fuselage) is placed at ~ 30 cm in front of CG. Could that affect the baro’s measurements? I mean since it’s now more subject to attitude/orientation angular disturbances?
Internat placement matters for baro though I wouldn’t expect intuitive relation between attitude (more specifically angle of attack and sideslip) and indicated baro pressure like there is for accelerometers.
There definitely is some serious coupling between AoA and and barometric pressure.
Top three lines with squiggles are barometric altitudes.
The brown line on top of them is canonic vehicle altitude offset to match baros before takeoff.
Blue line is vehicle measured Z-acceleration not compensated for gravity.
Yellow line is “expected” Z-accleration due to IMU offset with a bit of temporal offset due to filtering.
Bottom brown line is AoA. As can be seen high baro alt variance happens when there are rapid AoA changes.
I would strongly recommend equipping the plane to fly with RC control as it should allow safe recovery of control in a case of such oscillations provided the vehicle is aerodynamically stable or oscillations are caused by outer loops (outside of rate and attitude loops).
