I am experiencing an issue with my pitot tube calibration. I am flying a Quadplane equipped with a Mateksys 4526DO airspeed sensor. To calibrate it, I first performed a preflight calibration, and I received the ‘airspeed sensor calibration successful’ message.
Later during the flight, after completing the transition to fixed-wing mode, I set the ARSPD_AUTOCAL parameter to 1. I monitored the messages expecting to see the calibration process, but no related messages appeared. After the flight, I checked the parameters and noticed that while the offset value had changed, there was absolutely no change in the ratio value.
For context, my flight mission consisted of flying figure-8 patterns three times.
Why might the in-flight auto-calibration have failed to update the ratio? I know how critical accurate airspeed data is for VTOL vehicles, so how can I properly fully calibrate this sensor under these circumstances?
The calibration section of the wiki states " Note that when the calibration is active, an updated ARSPD_RATIO value is only saved in the autopilot’s non-volatile memory every two minutes, and only if the value has changed by more than 5% from the last saved value."
Was your flight long enough? If so, maybe the calibration was already within 5%.
Also, it’s possible the airborne calibration was continuously reset during the figure 8 maneuver when the turn direction reversed. Typically, a calibration like this cancels wind effects by making some assumptions while in a constant bank loiter or an oval circuit. I’d have to chase the source code to see how this particular routine works, but I think it’s likely to expect a same-direction circuit.
Instructions are here under “Automatic calibration”.
“Take-off and fly a repeated circuit or circular loiter for 5 minutes. This can be done in any mode, but if the autopilot is already tuned sufficiently well, this can be achieved by leaving it in loiter or RTL for 5 minutes.”
A few days later, I conducted another test flight. The aircraft completed 5 circles with a 70-meter radius, and the flight lasted for about 4 minutes. The ARSPD_AUTOCAL parameter was enabled while circling. I also performed a preflight calibration before takeoff. However, while the offset value changed, the ratio value remained exactly the same.
Can we consider the pitot tube calibration successful based on this, or should I perform another test flight? I can share the flight logs if needed.
You can check this yourself and make any necessary corrections by comparing the airspeed and ground speed while loitering. The airspeed (ARSP[0].Airspeed) should remain roughly the same and correspond to the average of the minimum and maximum ground speed values (GPS[0].Spd) .
Here is an example of a sensor that, in my opinion, is reading 0.5–1 m/s too high.
