[ppnvedado]

Hi there I have a quad 16"55 props large and heavy motors. really flies well except that in altitude vario some specific roll moves end up in control oscillation and crashing. (only roll)
I suppose it has to do with PID config but I already try some modifications with no results. Can anyone give me some tips on large quads configurations to try?
Two serious crashes already.

[TheOtherCliff]

Altitude Vario controls only the throttle, so any crash caused by it would be too much power or too little power.  Is that what you are seeing?

Generally, you must put closed cell foam over the baro chip (and keep light off of it when flying) and also keep prop blast off of it.  The prop blast is probably one of your problems.  I bet you get problems if you yaw left and right repeatedly too.

If it is uneven in thrust at full power it may go off to one side or the other.  Are your motors and props well matched and ESCs calibrated and motor neutrals set?  It wouldn't hurt to do these things again.

[ppnvedado]

Maybe I try to put to much info in just few words.
I have no issues with vario control, runs perfect and the altitude response is quite well. My cased revo comes with the foam inside the case, first thing I check.
Revo is just at props level and far away, no turbulence, or blast.
the quad its very well balanced, at full power goes straight up and very fast. the quad can lift 1.5 its total weight, has no load now.
The problem I'm having its oscillation only in roll, where it needs less torque to level, the body is a rectangle long side front to rear, never have the problem on pitch. (cant send picture due to repairs  )
this oscillation, when start, grows like resonance in mechanical, structural systems, don't know the term in electronics or control, maybe the same.
Everything is calibrated and tested.
One thing I can do is to set neutral higher to reduce mechanical inertia and faster response but I need some advice if there is something else to do with pid (pid should manage this issue if it has a proper configuration)
Thanks

[TheOtherCliff]

So oscillation only happens in roll, and not in pitch?  Are you using stock PIDs or some downloaded PIDs or did you tune the PIDs yourself?

If all your ESCs have the same firmware (bought at same time or flashed with same version) and are well calibrated with neutrals set too, and all of your motor output channels being used (Output page: Bank Channels Mode:) use the same protocol (and speed for some) then...

If the quad is not symmetrical; that is front to back motor distance is different from left to right distance and / or weight distribution is very unsymmetrical and / or oscillation only happens in roll axis, not pitch axis, then... go to Stabilization and reduce the roll (roll only) PIDs (inner loop, Proportional and Integral).  Divide each number by 2.  Save.  Try that.

[ppnvedado]

thanks
Changes done on inner loop.
I"ll test it and post results. It will take some days due to frame repair.

[ppnvedado]

Hi there
modifications
programing
cut in 30% inner loop
hardware
+ mast and GPS antenna (heavier in roll)

no oscillation so far
If I have this problem again I'll post new modifications
 

[ppnvedado]

another post, another crash.
I´ll cut in half the original values (another 20%) and see.

[TheOtherCliff]

Maybe you should make videos of your flights?  So that if you crash we can help more.

Are you using GPS flight modes and a DJI/Naza GPS (authentic or clone)?  If so, you may find that this firmware fixes the issue.  If you try it, please post if it helps.
https://forum.librepilot.org/index.php?topic=3012.msg21154#msg21154

When using GPS flight modes it is a good idea to have a non-GPS flight mode (Attitude) on the switch, to use if anything goes wrong, until you know it works consistently.

[ppnvedado]

I'm the one who did the fix on gps problem and works, so I start flying around and have the roll stabilization issue in Attitude-Attitude-altitude vario flight mode. Always happens in this flight mode, the one I use most of the time.
I don't have a camera to film the issue before the 30% inner loop reduce. I try before to force the fail but had no results. I'll try again in my next flight.
There is a way to calibrate PID in the quad? a method to follow or something like that.
I'm asking because this will not be my final quad, this is only a test platform, I don´t want this issue on my final design.

[ppnvedado]

Problem solved with the reduction no crashes so far.