• This site uses cookies. By continuing to use this site, you are agreeing to our use of cookies. Learn more.

APM 2.8 destabilizes on descent. PID help?

Yags

New member
#1
I have an F450 frame with an APM 2.8. It flies well, but especially in autonomous flight, it gets a growing oscillation on descent that builds until the quad is tipping 90 degrees in either direction and it falls out of the sky! I'm trying to adjust the PIDs to fix the problem, but am having a hard time.

Raising the P seems to help, but I quickly get a fast oscillation with throttle. It flies crisply at P=0.17 or so, but with that oscillation. If I lower it to 0.14 the oscillation goes away, but it feels sluggish and unstable.

My main confusion seems to be with the I. I tried raising the I as high as 0.30, hoping to see some slow oscillation on throttle or in hover to tell me that I went too far, but I'm not getting it. So far it is hard to tell if the high I is making my descent problem better or worse.

Anyone have any thoughts?
 

LitterBug

Troll Spammer
#2
Sounds like prop wash. Is it possible to slow the rate of descent or add lateral motion so it is not flying through it's own turbulent air? I believe D has a bigger affect on things like prop wash. Still flying stock PIDs on a similar setup. A450 (it's a clone) with APM 2.6. Have a real F450 too, but it is going to be running dRonin on a more modern F3 or F4 FC.

Cheers!
LitterBug
 

Snarls

Gravity Tester
Mentor
#3
Usually I gain is what you tune to help keep the craft stable in prop wash. Of course too high P gain can cause instabilities. I fly a TBS Discovery with P and I gains around 0.1 for both roll and pitch so it is possible your P gain is too high and an I gain of 0.3 is also likely too high. It is not uncommon that you will not see slow oscillations from too high I gain. I would put the I gain on one of your transmitter knobs if you have one and adjust it in flight. Start low at like 0.02 and do some climbs and descents while increasing it a little at a time until it becomes stable on descent. If it is getting into an uncontrollable oscillation at low I gains you know the issue is with P. Note that descending directly in prop wash will always have some sort of wobble.

Also have you tried autotune? That may be useful to get you in the ballpark of some PIDs that are generally stable all around.

If all else fails you can decrease your auto mode descent speed so that the encountered prop was is lessened.