full throttle response slow when quad is descending

[squid10]

i have Crius AOIP V2.
Sometimes during fly i'm letting the quad to go lower from about 20, 30 meters to 10, but there is a point where i need to give lots of throttle back in order to avoid crashing in the ground, sometimes even full throttle doesn't help and the quad is scratching the ground.
it's very annoying cause i cant fly fast between altitudes...

maybe the gains are high? low?
anyone else encounter such case?

[PatrikE]

You could increase MAXTHROTTLE in config.
It's set to 1850 as default.

Vibrations in the frame can also reduce vertical performance A lot!

[crashlander]

Vibrations in the frame can also reduce vertical performance A lot!

Any vibration!
A LOT LOT!
On my quad I had faulty ESC that worked (partially died during sparks while connecting LiPo), but it was unable to give full power in short burst and the whole quad was unable to climb fast or stop from fast descend.

Regards
Andrej

[squid10]

ok, i'll try to increase max throttle to 1900.
thanks

[crashlander]

ok, i'll try to increase max throttle to 1900.
thanks

IMO most ESC will take 2000 without issues (especially if you calibrate endpoints that way).

Regards
Andrej

[scrat]

ok, i'll try to increase max throttle to 1900.
thanks

IMO most ESC will take 2000 without issues (especially if you calibrate endpoints that way).

Regards
Andrej

I think they can handle more then 2000.

[Christian W]

It probably wont help with increasing max throtte. The reason for loss of power on decent is usually to high pid which result in small ocilliation on decent. On mw 2.2 it's harder to see when calibrating, but you have to high P value.

[squid10]

yes i also think i have too high PIDs
I have a reptile frame (similar to F450 in the size) with SIMONK ESCs - my Roll/pitch P are 14, My Yaw P is around 10.
it looks very stable this way, but it's totally far from other people's PIDs that i saw in forums - P=5,6...

[Wil Chris]

Yep that's it. To high P. You can't see it before you descend and accelerate upwards. Had the same problem, tuned the PID like usual looked stable at P 15, except i got the problem you have. Turned the P down to 5 no problem. If you use a Bluetooth connection you can see the motors are overdoing the stabilization when you descend.

Edit: I just discovered I have two users.

[crashlander]

Edit: I just discovered I have two users.

That is not problematic, unless you also have two personalities!

[squid10]

tuned the PID like usual looked stable at P 15, except i got the problem you have. Turned the P down to 5 no problem .

You decreased your P from 15 to 5 just because of this? that's a lot!
actually when i think about it it's more related to I gain than P gain...

[Wil Chris]

I have tested it. "I" don't do really matter because it's not the effect of to little or to much drift compensation. I believe increasing D would work a little because it would dampen the oscillation, but in the end it's to high P. 14 it rather high.

[squid10]

ok, i decreased P from 14 to 11 , also i increased max throttle from 185 to 1950 , i did a small test flight and it looks much better... but still i have to test it more

[scrat]

Set throttle to 2000. It's much better.

[Christian W]

Have done some testing. It seems like stiff propellers like wood or carbon fiber helps. Wobble on descent can be caused by props bending a litle. I have not tested Simonk on ESC yet but it should also help.