Crashing after Autotune and Temperature problems

Hi to all,
Today I crashed it and I need some help to understand why.

I have a quadcopter with Cube Black 2.1 and HERE gps.
In some occasions I noticed that after perfoming Autotune (with relaxed setup) the quad became unstable, due to small oscillation.
So I lowered some PID and it was ok.
The quad today was flying well until oscillations killed him at 14 mt, but I didn’t understand if the problems was oscillations due to high PID gain or something else.

Can you check log file and tell me if you can find somenthings wrong?

https://drive.google.com/drive/folders/1CJKO38y7gITMxf11AG4BsBRur8Rn09Hc?usp=sharing

Another problem was that the Cube couldn’t reach target temperature.
Outdoor temp: 8°C
Target temp : 55°C
Sensor temp 1: stuck at 28°C
Sensor temp 2: stuck at 35°C

Have you got any idea?

Thank you.
Romeo

In the log, it looks like there may be an oscillation that is confusing the rate controller, but it’s hard to tell because the logging rate is too slow to see it. Do you have a log of the autotune flight? Autotune logs controller outputs at full resolution.

I have to search it. When I find it I will upload.

And about temperature, have you got any ideas?

Thank you Rick.

Romeo

If the heater can’t reach the target temperature, that just means the Cube is dissipating heat faster than the heater can generate it. If your Cube is exposed to moving air and the ambient temperature is only 8°, I’m not surprised that it doesn’t reach the target. If your Cube is exposed, try covering it with something.

1 Like

Hey, Can you provide me more information about your setup please ?

What motors, what propellors, battery, AUW, ESC?

So, @philip @Romeo_V

Please check the start similarity in RCOUT of this log as well as the log in the post : Potential Thrust loss: Reason Unknown

One motor has always been commanded a PWM value so low! While another peaks out and results in a potential thrust loss warning/crash

Hi @Vandan_Rao ,

my setup is 5005 280kv, 18 inch props, 10000mAh 6S Tattu, 25A esc (it looks undersized, but actually it works well, and is reliable), and the auw is about 3.5kg.
Hovering around 1650 of PWM.

I see the post you linked about potential thrust loss… I have done esc calibration when I built it. Yes, in the log you can see that 2 motors spin slowly cause cog is slightly off center.

That quad was fliyng very well until the crash, and I didn’t understand where there were innovations during flight.

Maybe I have to raise PWM_MOT_MIN?

Romeo.

There’s a fine line between getting something airborne and doing some actual flying. That’s why Wright Brothers aeroplane is museum stuff and not seen wandering around the sky today.
Your multicopter is over-propped and under-powered. It may look efficient on paper, and may be able to take off and hover in standstill weather, but there’s where it’s capabilities end. It isn’t designed to survive the relative harshness of an autotune process, much less what you encounter in everyday flying.

Sorry,
Para.

So, We flew again the next day with AP4.0.0 but with PIDs that we saved after an autotune in 3.6.x (before we upgraded to AP.4.0)

I see that you were also auto tuning when you suffered a crash. @rmackay9 has posted a warning on ardupilot forums - Check : https://discuss.ardupilot.org/t/warning-autotune-bug-in-copter-4-0-1-4-0-0/51709 and that reminded me to update the thread here.

Sorry for the delayed follow-up.

Hi @Vandan_Rao ,

Thank you for update me.

After few flight we decided to swap autopilot with a new one, and it works well (temperature is correct).

Yesterday I did an Autotune on Arducopter 4.0.0 and fortunately it didn’t crash like bug you linked.

But although I used autotune aggressivness at 0.05 (Weak) it isn’t really stabilized yet.

I will proceed with manual tuning.

Thank you, Romeo.