Airspeed sensor affects plane pitch when no GPS lock

Hello all,
I’m trying to find a reason why my plane crashes when flying without a GPS signal.
And I’ve found an issue similar to this topic
Shortly: if I blow on the airspeed sensor while the plane is on my table (without GPS lock) I see that the horizon on my GCS is moving like ±20°

You can see this in the picture below:

I see this behavior no matter if I use DCM, EK2 or EK3.

I suspect this causes my plane to crash.

Does anyone has any ideas on

  1. is this something expected?
  2. can this lead to a plane crash?
  3. how can this be fixed?

Thanks in advance!

TBH it looks like TECS compensating for increased (excessive) speed. Also your probe doesn’t look calibrated. From my experience simply blowing into the tube shouldn’t cause 50m/s spikes.

If you have GPS try flying circles with airspeed sensor enabled but not used and airspeed learning enabled.

Other than that check if your control surfaces move the correct way in Stabilize.

Thanks, I will check calibration and TECS behavior

TBH it looks like TECS compensating for increased (excessive) speed

Why would that cause ATT.Pitch to change? I could see that changing the “desired” pitch, but not the “actual” pitch.

Yeah, I’ve started checking TECS, and found the same stuff:
ATT.Pitch is the value from DCM and it is logged directly inside DCM, so TECS will not affect this

So, I’ve investigated the issue and found that the behavior is expected.

By default, wind is considered as sensor error, and it will decay overtime.