Home navigation question

[Jippieee]

@Jippieee: Thanks for the KML. Could you better describe the problem and also attach the datalog .xml file? It looks like a weird flight, so could you also include the navigation commands?
For a basic test flight I usually fly a circle with a radius of about 120m.

Be aware that the SVN servers contain development code. The good & tested versions are on the download page

Thanks for having a look. The problem was that the plane sometimes went into a steep dive in aut mode and showed some gliching moves that are beyond the configured allowed roll/climb angles (which is only 30deg for roll and 20 deg for climb). Also it did not navigate home but kind of zig/zagged with some breakaways.

The navigation command is
1: CIRCLE_REL offset 0,0, 40 dia, 50 height
2: GOTO 1

It looks like some toggeling between manual mode (yellow) and auto mode (red)

[Tom]

Indeed a strange flight.

How did you configure altitude control? (in the control tab)

You did set the altitude to 50 meters, and the altitude when the plane started diving in autonomous mode the altitude was 150 meters, so not unusual it started diving. Please next time use a circle radius of 100 meter and an altitude of about 100 meter.

The strange thing is that it didn't dive to the correct altitude during the 1st autonomous part. This problem was due to the incorrect calculated pitch angle. Possible reasons for this are:
- incorrect calibration (has it been a while since you calibrated the sensors? did you calibate with the real power source connected and the red wire of your ftdi cable disconnected?)
- incorrect mouting of the module (servo connectors should be facing forward)
- vibrations (is your propellor balanced?)


Were the flights after this one better?

[Tom]

Had any better lock later on? Lukasz had a similar problem and it turned out his module was undergoing too much vibration due to a non-balanced propellor. Maybe this could be the problem?

[lukasz]

Right, I had similar problem where the plane in AP mode was flying out in not specified direction or did aggressive, uncontrolled maneuvers.
I was doing a lot of tests and it came out that this was caused by vibrations coming from the motor.

On ground tests (having the plane perfectly still and calibrated), when switched to stabilized mode I could see that at some throttle positions (50-100%) servos started to oscillate (up/down) or even started to move to either roll or pitch position.

I solved the problem by:
1. Balancing the propeller
2. Putting dampers on the motor to absorb vibrations
3. Put dampers on the gluonpilot.

Hope that above helps somehow

[Jippieee]

Thanks for the follow up.

Ok, I have already carefully balanced the propellor but that did not resolve the aggessive spikes. Perhaps the motor axle is bent or the prop does not centre on my axle, which is a bit too small. I will resolve this first and mount the Gluonpilot on a foam pad to see if it helps. I also fly helis and had some issues with a head lock gyro in the past so I am already keen on vibrations.

@Lucasz, what dampers do you use?

[lukasz]

Here is a simple drawing that I made to show my motor mount:

motor_dampers.jpg (42.38 KiB) Viewed 6910 times

And here is how this looks on the plane:

IMAG0054.jpg (435 KiB) Viewed 6910 times

Additionally I have added dampers on the gluonpilot module.
I have pulled those out from an old CD-ROM:

IMAG0053.jpg (406.8 KiB) Viewed 6910 times