Agressive pitch up when switching from FBWA, Loiter and back. Log attached


I have a flying wing equipped with a APM Mini running ArduPlane 3.4.0. Trimmed and tuned via autotune following the steps in the wiki. Flies very well in Manual and FBWA but I am experiencing a problem that I cannot find the cause for.

Flew a short flight with LOG_BITMASK set to 65535 demonstrating the problem.

When I fly in FBWA and switch to Loiter the plane very aggressively pitches up for a few seconds. This is totally uncommanded by me from the ground. It then levels out and flies as expected. This is also present in RTL.

I have tried retuning and checking over the tune of the TECS system.

I have poked through the logs but have yet to find something to point me in the right direction.

As you can see in the attached screenshot: elevon channels (RCOU.Ch1&2) are deflected full up and pitch sharply turns up (ATT.Pitch). This is uncommanded, (ATT.DesPitch rather tries to compensate and level out). Maybe someone can find more clues from the attached bin log?

Does anybody have any ideas of what could be at play here? Let me know if more info is needed in any way.

Views: 866


Replies to This Discussion

I seem to remember something about the auto tune trim vs the manual trim.

I don't know the answer, but I would try setting the manual trims to stop the pitch up.

Manual trim as in trim clicks on the transmitter? Or are you referring to some specific parameters? Could you clarify? I am not sure that I follow. The problem occurs when switching to from FBWA to Loiter, not sure that manual trim has anything to do with it.

Very happy for the input!

I really haven't been through the process myself, I just remember an email conversation about this exact problem and it has something to do with the auto tune inserting a trim value that is not used n all modes.

I can't remember for sure if the manual trim had to be inserted in code, or just transmitter clicks, I think it was transmitter clicks

Is your plane properly leveled? Your artificial horizon should show 0 degrees pitch in level flying. If it's not level, the APM wants to level out first (pitch or dive) but then, in auto mode, it will try to maintain altitude and you'll have two conditions fighting each other.

My 2c

I have now double checked. The artificial horizon stays the same between the modes in level flight. So I do not think this is the problem...

I would be interested in reading the conversation you are referring to, if you can find it.

I should be able to simulate transmitter trim by switching to the Loiter mode with the pitch stick forward. But something tells me that something else is at play here. The pitch up is VERY aggressive. It is like both servos go to full deflection very fast and locks in place.

I meant to say that your pitch according to your OSD / Telemetry should be 0 degrees while in level flight. Although you say the pitch is aggressive, so I'm assuming you'd notice it. I'll check the log, if no-one else has better advise.

If you could have a look I would be eternally thankful! I hope that the log level is sufficient. I chose the pixhawk standard I think. Let me know if I need to change the bitmask for more data and do another flight. 

Sorry tried to post this yesterday but I had connectivity issues!

After looking at your log, these are NOT the problem:

It's not your level - the Y Gyro seems level when it's supposed to be.

I thought it might be your gains that's too high, but it does seem to be under control after the initial pitch.

It's not your RC trim - elevator seems almost center on trim: PARM, 9591334, RC2_TRIM, 1488

If I have time later I'll go into more detail.

Great to confirm these! I have noticed a spike in NTUN.TargBrg around the switch of modes. But I am unsure of what this might mean. I cannot find any documentation regarding what the more obscure data points stand for. If it is a metric reflecting error in navigation is it bound to spike right after engaging loiter, since it has not had the time to correct the heading etc. Do you have any thoughts on this?

Thank you so much for your time and input!

Does anybody else have any pointers? I am starting to lose my hope here :( Would be very sad to scrap this project... I see so much potential in the platform

Here is a video demonstrating the problem 


© 2020   Created by Chris Anderson.   Powered by

Badges  |  Report an Issue  |  Terms of Service