P4 Altitude Accuracy

[RémyHeli]

Did you notice that, before GO4 app existed, some previous version had slower RTH descend rate? I think since some 3.xx version or P4 firmware they increased RTH descend rate. I remember very well I did some RTH demo and was shocked seeing the craft having higher vertical speed descend - I add throttle and finally cxl RTH as safety. Imagine if it was auto RTH following RC disconnect: possible hard landing or worse..

I think at some point DJI must have increased RTH descend speed for wathever reason, but I have not seen any information about this. Just my experience. As per result, the craft might apply agressive breaking when VPS takes authority as seeing (if he can) the ground within last 10 meters or so.

I understand craft is only able to have linear vertical descend speed, until close to the ground were he realize it's time to apply breakes. Because he doesn't know exact altitute as already discussed in this forum (barometer).

Increasing RTH descend rate looks on my opinion more risky as possible hard landing or crash if VPS confused: less buffer when reaching the ground. So reducing speed for the last 20 meters instead 10 would be smooth but I doubt it's possible with current design.

Any though welcome.

 

[jsposito]

Yep, I don't trust the altitude readings at all. I always land manually because the auto landing is way to fast for me to be comfortable with it.
Jim
WA5TEF

ditto that, very much the needs the mavic landing system.

 

[Crack The Sky]

Well stated! I could not have said it better, myself.......

Loved your explanation/observation. When I am hovering at 4-5' agl right after take off, my P4 almost always reads higher, in the neighborhood of 10'-15' off, meaning about 15'-19' agl. It is really just a curiosity to me rather than a concern. Since the barometer/altimeter is indicating higher than it actually is, when setting the RTH, I add in the error plus a 100% safety factor. I am sure your math is spot on, (apx. 60 Pa higher pressure below the rotors than above them), and combined with the fact that the airflow around the machine is certainly not laminar, it's quite a software feat that the reported altimeter reading is anywhere close to the actual AGL. I guess the only way to record a very accurate reading consistently, (barring atmospheric barometric changes), would be to have a probe taking readings from 2-3 feet above the craft and then subtract the probe height in the final calculation.