Hmm. Not convinced.

My understanding was that keeping the PID loop active at low throttle is only part of it. Airmode gives the FC the ability to spin up the motors beyond the set throttle value in order to deliver the on the commands from the RX. That's where the infamous 'airmode bounce' comes from. The ground exerts an external force on the quad, which then attempts to correct using full authority.

JB, surprise surprise, has a video:

Imagine that, Pavel is not correct yet again.....

Without trawling through code and trying to work it out I don't know, but considering Pawel is one of the active developers for iNAV it means he knows how to program flight controllers so I'm sort of inclined to believe what he says over others who just go off theory based on stuff told to them from a third party. And I think his reference to losing his glasses and challenging "famous" people in the FPV world at the beginning of the video might actually be a little hidden dig at JB and the video you just linked to above 

I always have Airmode permanently switched on but I've never experienced the Airmode "bounce" when landing because I generally land in grass and because I don't want to spray chopped up grass clippings all over my camera and flight controller stack, to land I just disarm my quad when I'm a few centimetres off the floor.

It would be good if some of the experienced pilots on here could maybe do a real world test and do a flight with just Airmode enabled immediately followed by another flight with Airmode disabled and Idle Up enabled instead. Assuming that neither allow the quad to just drop out of the sky on a flip or roll, then I'm probably not a good enough pilot at the moment to be able to tell the difference, so someone else would need to do that.

From the Betaflight docs: https://github.com/betaflight/betaflight...-ctzsnooze

Quote:A quicky explanation from ctzsnooze:
Let's say you have PID_at_min_throttle enabled, pull throttle back to zero, so that all motors are at min_throttle, and pull the roll stick to the right quickly. The left two motors will speed up, but the right two can't go any slower. So only the left two, in speeding up, contribute to getting the roll going.
Contrast this to a right roll at hover throttle. The PID requested is the same as for the roll above, and the left two motors speed up to the same extent. But the right two can go slower than hover rpm, and they do.
That's why, before airmode, even if idle up was enabled, or PID_at_min_throttle enabled, a roll at hover throttle would always be quicker than a roll at zero throttle.
What airmode does is to compensate for the motors that can't slow down. With airmode enabled, as soon as the mixer determines that some motors have to go to min_throttle, the other motors are sped up accordingly, so that the differential required to generate the turn is the same as it would be hat hover throttle.
Hence airmode maintains roll rate when any of the contributing motors hit zero (or max) throttle. It ensures consistency of roll rate at extreme throttle values.
That's how airmode is different from pId-at-min-throttle.
The feature might still be optimized based on experiences of the Beta Testers, but is looking good already.

I guess one test would be to do snap rolls at 0 throttle input. But you'd need high enough rates for it to be conclusive.

Honestly the I switch it off to land and that's it (I use a three-position switch for disarmed/armed/armed+airmode). The only problem I have with it is in a crash it can try to correct using all the juice at its disposal and ping you off into the distance. That always makes me a little nervous.

(12-Apr-2019, 03:15 PM)Banelle Wrote: From the Betaflight docs: https://github.com/betaflight/betaflight...-ctzsnooze

I guess one test would be to do snap rolls at 0 throttle input. But you'd need high enough rates for it to be conclusive.

Honestly the I switch it off to land and that's it (I use a three-position switch for disarmed/armed/armed+airmode). The only problem I have with it is in a crash it can try to correct using all the juice at its disposal and ping you off into the distance. That always makes me a little nervous.

Hmmm. The more I read up on this and watch videos of Airmode in action, it definitely IS feeding back into the PID controller which is something Idle Up definitely can't do with it running on the transmitter because the transmitter will have no idea what the current attitude of the model is.

Like with everything, I guess it depends what is actually meant by the term "vs" in the author's eyes, so it follows the old adage of "it depends". If you mean every single thing then clearly not, but if you are comparing high level functionality then probably yes. There is now a small debate in the comments section under Pawel's video challenging him on his explanation.

I watched a few videos of Airmode in action but probably one of the best ones is by a "famous" person (LOL) below. It shows perfectly how Airmode is aiding the pilot and is using the PID controller to keep the model "in position" when inverted which is something that Idle Up on the transmitter clearly wouldn't be able to do (and can't). So there definitely is a difference if you refer to automatic pilot aid.

Snow I think you hit it on the head. Are both exactly the same, I don't think so. But on a basic level, they achieve the same inherent goal..which is your quad from dropping out of control at zero throttle. Here is another famous person demonstrating the slight difference. This person likes idle up better.