AltHold does not respond to throttle after first flight

Hi everyone,

after replacing a faulty GPS with a new one my first flight was very amazing. AltHold and PosHold work as expected (with a high precision out of the box - wow).

Today I went to another fly with ground station (APM Planner) for RTL and mission tests. But for no reason AltHold did not work anymore.

If I (perfectly) hover the quad and switch from Stab to AltHold the quad stays at this altitude, but did not respond to any throttle input. For my understanding the quad should climb (2.5m/s) by rising throttle above 60% and descend (2.5m/s) by lowering throttle below 40%. This works for my first flight but not on second or third one.

If I climb/descend slowly (throttle within 40%-60%) and switch to AltHold the current altitude should be hold. This works for my first flight but not on second or third one. Instead of holding altitude it keeps climbing/descending for ever.

If I fly forward/backward/sideward with AltHold the quad starts ascending/descending after some time. Nothing can stop this except switch to Stab mode.

I did not change something on my quad or on APM configs and Im using an APM Kit, Turnigy 9X as radio, APM Copter 3.2.1 and 4s lips. My quad is hovering at 48-49% of throttle (nearly perfect at 50%). All config values are default. On first flight it was bit warmer and maybe a lower humidity, but could this be a reason?

After second and third flight I made an additional radio calibration. All seems to be fine, but one question: after calibration the GUI shows me for RC3 input (throttle) a center value of 1059 (min: 1058 max 1894). Is this a proper value? (seems to be a bit odd)

Whats wrong with my AltHold mode (or APM)?

Best Regards,
otbe

Observation on second flight in AltHold mode:
The ground station shows me an altitude of ~10m and a climb rate of -0.1 to +0.1 while using AltHold. But for any reason the quad climbs at high rate and the ground station did not recognized it.

Looking at your logs it hard to say as logging on the IMU has been disabled.

You issue is almost certainly related to vibrations. The autopilot is adjust DALT is increasing and the Alt is following.

[attachment=0]vibrationsII.png[/attachment]

Logging of IMU is disabled by default :frowning: (how to enable?)

But why this works for my first flight? With same props, same battery pack …

Correct me if I got it wrong, but in your picture DAlt (desired alt) starts decreasing at x~=5000 with no throttle change. This should not happen, shouldnt it?

Im using a vibration damping system like this.

Maybe vibrations where not as bad on the first flight, and only worsened in successive flights.

If you turn the logging on we can analyze it and see if that is the case.

copter.ardupilot.com/wiki/common … vibration/

Hope that helps :slight_smile:

[quote=“otbe0”]
Correct me if I got it wrong, but in your picture DAlt (desired alt) starts decreasing at x~=5000 with no throttle change. This should not happen, shouldnt it?[/quote]

Is this the truth?

Will give IMU logging a try tomorrow.

This is from my third flight. DAlt ist decreasing with no throttle change. In real life the quad was descending much faster than it looks like in this picture and much closer to the ground. Only switching to stab mode allowed me to safe my quad.

(sorry for double posting)

Z vibrations screw up the altitude estimate, so it thinks it’s actually rising, so it lowers the desired alt. (or vice versa)

So here is my log from today. Seems to be fine !?

Look at x=9000:
DAlt is increasing, BarAlt too. But z vibrations are “constant” and throttle input decreased to zero at x=9900 and no response from APM. Only switching modes stops infinite climbing of my quad.

The vibration seems high towards the end of the flight for what seems like you are hovering.

[attachment=0]vibrationIII.png[/attachment]

What was your assessment when you looked at the logs?

Your picture shows a successful hovering in AltHold mode. At this time my quad stays in the air with same altitude.

My assessment:
All axis are in range (+3 -3 for x and y; -5 -15 for z). Taken from copter.ardupilot.com/wiki/common … vibration/

Look at my picture:
DAlt is increasing, BarAlt too. But z vibrations are “constant” and throttle input decreased to zero at x=9900 and no response from APM. Only switching modes stops infinite climbing of my quad.

It seems that APM recognize my throttle input (~50%) as a request to climb :frowning:

In other words: the APM does not change ThrOut at all.

  1. If I hover perfectly and switch to AltHold it stays at current altitude. If I fly forward/backward/sideward it starts descending and no correction happens.
  2. If I switch to AltHold with a small rate of climbing, the APM does not tries to stop it. It will climb forever (ascending or descending).

But: If i rise to 50m, cutoff throttle (decrease to 20%) and switch from Stab to AltHold the APM tries to stop my quad falling. It does not work perfectly, but you can here props spin up a lot.

Your THR_ALT_P is 0. I believe it should be default 1.

THR_RATEs and _ACCELs are also 0.

This might cause your problem.

I’ve just posted a very similar problem here, which is under investigation. viewtopic.php?f=80&t=13883

These parameters are accessible via the Extended Tuning “page”, lower part about “Throttle Rate”, “Throttle Accel” and “Altitude Hold” as I understand it. My problem is that they are randomly set to 0 unintentionally and saved to the FC when saving another change I did.

That’s it. Or maybe this could be the solution. I will give it a try tomorrow. Thank you.

If I change some settings it causes in almost every case a invalid state of some other params.
After some time waiting it works as expected. GCS seems to be buggy :frowning:

I have a fix for the param = 0 issue. I’ll post a link to a test build later.

try this build and let me know how you get on, it fixes the param setting to zero issue.

firmware.diydrones.com/Tools/APM … 015-10-19/

If that helps resolve the issue, I’ll push 2.0.19-rc4 (this problem was introduced after 2.0.18)

Link is dead. Please correct :slight_smile:

done firmware.diydrones.com/Tools/APM … 015-10-19/

Thank you Flyhard and thank you billbonney. Quad feels amazing like the first day :slight_smile:

So far no param = 0 with new APM Planner version, but I will keep an eye on it :slight_smile:

Great!! Looking forward for som althold as soon as I get up in the Air again :slight_smile:

Sent from my iPhone using Tapatalk

2.0.19-rc4 is available for download from firmware.diydrones.com/Tools/APMPlanner/beta