fluctuation of altitude reading

[mazi14]

During some longer flights (typically more than 40 minutes) with my plane equiped with FC SpeedyBee F405 wing mini (iNav 7.0.1) at the end of the flights in descend phase altitude readings (both OSD and logged in transmitter) fluctuate up and down randomly-see attached graph. I am afraid it is related with HW as I have not met this issue with Matek F411 wing mini FC in another plane. Have you ever experienced anything like this?

[sensei-hacker]

Estimated altitude comes from three sources - barometer, GPS, and accelerometer. You can look at the black box logs to compare the three sensors and see which one might be off, if any. Particularly you might want to look at the accelerometer Z reading.

Baro will drift slowly over time, and can be affected by airflow. But it tends to be good at distinguishing altitude changes over the course of a few seconds. It can have problems if the prop or ram air is pushing air into the fuselage, increasing the pressure in the fuselage.

The GPS does not drift over time, but isn't precise or quick to react to sudden changes. GPS vertical accuracy is best when at least three constellations are enabled.

The accelerometer quickly detects whether the aircraft is falling or rising, but is subject to noise from vibration. Vibration can therefore cause a sudden jump in the reported altitude. That jump will be corrected after a few seconds as the baro and GPS readings overcome the "ascent" reported by the accelerometer.

[mazi14]

Thank you for explanation, but provided graph shows poor barometr data. I am attaching another graphs which show that fluctuation is not related to GPS altitude neither Z acceleration. I can not see any corelation between altitude spikes and GPS speed too.

[sensei-hacker]

Excellent. If you've narrowed it down to barometer data, we'd want to look at air flow and pressure in the vicinity of the baro. The baro is of course measuring air pressure.

Can we see a couple pics of the plane and any air openings into the fuselage?

Also, this board isn't conformal coated, is it?

[mazi14]

Unfortunatelly, the air pressure has not been logged. This FC does not send air pressure data as a standard telemetry and I have not logged data onboard (black box) till now. The plane is Ranger 2000 from Volantex, so there are 2 side air inlets in the nose and a hole for camera. My camera is placed on the top of the canopy. FC is placed on the wooden plate between wings and it is not coated. Cooling air is definitely flowing through the fuselage, but I would expect altitude fluctuation during the whole flight in this case, not only at few last minutes. For majority of the flights, there is no altitude fluctuation at all. Once I had to speed up the plane up to 100 km/h to lose altitude (I usually just cruise at 40 km/h) and there was not any fluctuation in the logged data. Maybe something is overheating near the pressure sensor I guess.

[sensei-hacker]

Ah yes, the baro will be affected by temperature.

I'm not sure what would cause a spike like that, though.

[mazi14]

I am going to start logging to black box and next time when I experience it, I will analyze that log.