When switched into loiter mode my T-Rex 450 does 2 or 3 small rotation like a toilet bowl before settling into the hover. It will also do the same thing when breaking from forward flight.
Could this behavior be caused by the compass being 250mm from the CG
David Ardis
What type of controller do you have ?
Do you have 4 compasses and baros in your system ?
it looks like the commanded yaw goes bizerk about 5 seconds after entering the loiter mode
Check the errors you are getting …
@ZvikaF Its a matek f405 wing flight controller using a 3DR compass and GPS. I only have on baro the one on the flight controller.
According to the documents, toilet bowl behavior is traced to compass or excessive vibration problems.
I have 2 Trex 450, both have toilet bowl effect (TBE) problem during loiter for couple of years.
Heli 1
Trex450 flybar, APM2.5, Compass HMC5883L mounted underneath the bottom mounting plate (pencil pointing), GPS mounted on the anti-rotation bracket. See photo Heli 1.
Heli 2
Trex 450 flybarless, Revolution full size FC (with integrated compass) mounted ON the bottom mounting plate., GPS also on the anti-rotation bracket. See photo Heli 2.
I have added 1 more bering to the torque tube to suppress vibration. But this did not solve the problem.
In fact, in stabilize mode, both heli are rather stable without much vibration.
Then I turn my attention to the rotor speed by doing 2 things :–
a) When using 3 cells, I increased the rotor speed by turning the dial knob from the usual 1535pwm to 1650pwm, it seems the heli is more stable in Stabilize mode. Will test more.
b) Instead of using 3 cells I changed to 4 cells. Now testing.
Let me know if you have any progress.
https://drive.google.com/file/d/11d88uLQT7OEXj7qMGW43MiNor8tmOvh2/view?usp=sharing
https://drive.google.com/file/d/1Xg7VNtLFIsyDbXULC_QqY8D8iAErbIn0/view?usp=sharing
David,
Everything looks pretty good as far as tuning. you can see that the actual velocities are following the requested velocities pretty well. So I don’t think this is a tuning problem
You could try increasing PSC_VELXY_D a little to see if that helps any but it should just make the actual and desired velocities match better. I would try increasing to maybe 0.2. Be ready to flip out of Loiter if you happen to see any bad oscillations.
@bnsgeyer
That made a big difference had to drop the value to 1.5 as there was a slight oscillation when letting the coper stop from forward flight in loiter.
I have another problem with the take off sometimes it will jump into the air and be really unstable for a few second then settle down. Is this a problem with the setup or perhaps I am over controlling. Enclosed a link to a log file.
David Ardis
What parameter did you change and drop to 1.5?
Sorry should read 0.15 for PSC_VELXY_D
