True airspeed offset in flight

PX4 Autopilot
1

Hi,

After searching quite a bit everywhere, I can’t find a way to correct the offset of my Pitot tube in flight.

I did perform the ground calibration of the Pitot tube, which went well, but it cannot account for the dynamic offsets that can occur in flight.

image
image722×325 50.3 KB

PXL_20250514_161155608
PXL_20250514_161155608302×403 26.1 KB

Can someone help me to find the way to “calibrate” it arround the flight speed ?

Thanks in advance

1 Like
2

I don’t think you can easily calibrate in flight, because you’re measuring wind, which is sorta the point. Are you sure you have an offset in all directions consistently?

3

Yes, it was a small flight without wind…

4

What’s about the ASPD_SCALE_1 parameter ?

5

Ok, would be interesting to see your flight log. I can’t really comment without more information. I also don’t even know which airspeed sensor you’re using.

6

Hi everyone,

I conducted more flight tests:

  • First flight with the UAV using the same configuration (same issue):

image
image945×410 72.3 KB

  • Second and third flight with ASPD_SCALE_1 set to 0.7 (instead of 1):

image
image945×395 71.3 KB

=> This time, the ground speed closely matches the airspeed.

Note that ASPD_SCALE_1 can be automatically updated using the ASPD_SCALE_APPLY parameter. In my case, I manually set it to 0.7, but the system is now adjusting it automatically (currently around 0.74).

I’m not sure why I have such a large offset with my sensor:
Holybro High Precision DroneCAN Airspeed Sensor (DLVR)

I did a test flight with the Holybro Digital Airspeed Sensor MS4525DO, and only observed a small offset.

I’ll be receiving two more DLVR sensors soon and will do some cross-checking in flight to see if they exhibit the same offset.

Stay tuned!

7

Interesting, thanks for sharing the data. Anything out of the ordinary in terms of pitot tube, pressure, temperature?

8

You can find the log of the last flight here : https://review.px4.io/plot_app?log=680589ab-fdd6-4df4-b6c0-1c40d9c9f2e2

9

Thanks. I can see how it’s definitely not wind since you’re flying all over the place in all directions, so your finding is valid. I don’t have other data on this sensor, so I’m not sure if it’s specific to your specific sensor or a general issue with this one.

@ryanjAA do you have experience with the DLVR-L10D sensor having a large offset like shown above?

@Vincentpoont2 FYI

10

I made a new flight with a third airspeed sensor : Matek DLVR https://www.mateksys.com/?portfolio=aspd-dlvr

Here you can find the logs of the flight test : https://review.px4.io/plot_app?log=214dbc8e-4eab-456c-bdbc-1900188f237b

This one seem to be ok (with arround 7% of offset)
image

==> I’m already waiting for 2 more Holybro High Precision DroneCAN Airspeed Sensor (DLVR) to replace the first one with arround 30% of offset !