APM 2.7.2 severe oscillations

BoFPV

New member
I just put a APM 2.7.2 (micro HKpilot Mega) on my 250 size quad to experiment with GPS modes like loiter.
When flying in stabilized and acro there are very severe oscillations, the arms move 2 inches up and down multiple times a second while I give no pitch or roll input.

I tried lowering the I value of my PID's with no succes.

Any suggestions on how to fix this problem are very much appreciated.

My apologies for any grammar mistakes, English isn't my foreign language.
 

Snarls

Gravity Tester
Mentor
APM is not initially designed for 250 sized quads so it is a little trickier to get working. First make sure that the APM is isolated from vibrations as best as possible. Unlike the miniquad flight controllers, APM and Pixhawk are quite sensitive to vibrations and need some form of isolation (could be as simple as foam tape). To check if vibrations could be an issue check out this page. Also make sure the APM is not near the ESCs or PDB as it is also sensitive to magnetic interference from power wires.

If you are getting severe oscillations try decreasing the P term as it has more to due with oscillations than the I term. Also make sure your ESCs are calibrated properly. Note that the APM runs older arducopter firmware so these pages may not be all correct.

Try these out and see if it gets better.
 

BoFPV

New member
Thanks a lot for the suggestion!

I will try vibration dampening the flight controller, now it's just sitting right on top of the pdb with 6mm standoffs so magnetic interference could have been the cause to.
 

BoFPV

New member
I tried vibration dampening the APM with no succes. The vibrations where a little bit less severe but still pretty bad. I also tried lowering the P value untill the point the quadcopter was very sluggisch and uncontrollable. I added a video of the result.

It just came to my mind that I had a similar problem with a CC3D, the problem was that the ESC frequency was set too low, the input frequency suggested for my Afro esc's is 1kHz. I don't know what frequency they are getting because all the esc settings are grayed out in Mission Planner.
ESC type grayed out.png
 

Snarls

Gravity Tester
Mentor
What firmware version are you using? Is it possible for you to post a flight log (im not sure APM has logging). Can you also take a picture of how the APM is mounted and wired on your quadcopter. And you are sure you calibrated the ESCs though mission planner?
 

BoFPV

New member
My firmware version: copter AC 3.2.1/AP 3.4.0
I don't think I can do a flight log, I don't have a wireless downlink or telemetry.
I did calibrate all the esc's but not through mission planner, I armed the quad set full throttle on my radio and plugged it in, then lowered the throttle until zero. The esc's did the usual calibration beeps.
When I try calibrating through mission planner I get an error that says that I need certain firmware on my board that my board doesn't support.

setup.jpg

Thanks a lot! I really appreciate your help!
 

Snarls

Gravity Tester
Mentor
Thanks for the wiring picture, everything looks good to me. I guess the automatic ESC calibration in MP is relatively new. Calibrating one by one with your transmitter should be ok.

When you're flying what mode are you in (acro/stabilize) and what PIDs are you adjusting. There are PID for rate (acro) and P for stabilize. Both P in acro and stabilize could cause oscillations if you are flying in stabilize.
 

BoFPV

New member
My oscillations are gone! All this time I was adjusting the wrong value, I was adjusting the P value (error to rate).

On another page in Mission Planner I found a P gain that represented Rate Roll in the pid's.

Lowering this value solved all of my problems.

Now I'm having another problem when the quad is in AltHold, the quad jumps up and slowly lowers throttle until it descends lower then it was before to jump up again and start everything over. (hope my explanation is clear :D ).
 

Snarls

Gravity Tester
Mentor
Nice glad you sorted the oscillations out.

One thing I notice in your photo above is that the barometer is uncovered. It is advised to put a peace of soft foam over the barometer to protect it from wind interference, and in this case with such a small frame, from prop wash as well. Then if that does not solve everything you can start playing with altitude hold P gain (in extended tuning).
 

BoFPV

New member
I have covered the baro already, I put the "fuzzy" side of a piece of velcro on it. not sure if that's sufficient but that's the only thing I have by hand now as I'm not at home.

I'll have a crack at tuning the P gain tomorrow and post a video of the results.

Thanks a lot for the help!
 

BoFPV

New member
I tried lowering the Altitude hold P gain, didn't help.
Here's a video of the result of multiple hours of fiddling around with the P gain.

If you listen closely you can hear the switch clicking when Iput the quad in Alt hold. When the quad is jumping up and down it's in Alt hold. My throttle management is a bit better than that :D
 

Snarls

Gravity Tester
Mentor
If you haven't I would try some foam over the baro instead of velcro. Also try to get the board as isolated from vibration as possible, maybe with some soft material below and above the board on the standoffs. I have read that vibrations can freak out the sensors for alt hold. Without logging it is hard to tell exactly what is going on, but with some experimentation you should be able to find it soon enough.
 

BoFPV

New member
I'll put the APM on a vibration mount "sponge :D" and cover it with some open cell foam as soon as I'm home again.
Thanks for the help!