Quad setup feels not much sturdy. Version 4.0.2 [solved]

@xfacta Hello, i have done some autotune with the setup I am uploading the logs pls preview it and tell me if you find something suspicious.
The auto-tune took around 13-14min for all the axis.
What I should do next perform some more autotune on all three axes or should manually try to change the gains?
Are the vibrations looks fine?

https://drive.google.com/drive/folders/1iRysh9W1KzjrOUpKq9xhu0tyATFBmZjE?usp=sharing

@xfacta Hello, So till now i have performed autotune 3 times and just around off some value to simple digits. I am attaching the log file of the latest test flight done in loiter mode. The quad seems very stable and I am getting a good position hold and not done any aggressive maneuvers but it seems to center itself pretty well.
I have not done any testing with the payload bcz the vibrations on the Z-axis are still not much stable and sometimes cross 60. So should I decreased the gyro filter and acc filter value to 15hz?
I basically build a setup to carry a 303015cm payload box which will be around 800gm when fully loaded.

logs
https://drive.google.com/drive/folders/1diMGsvbYZEdBzZ8Jow7Fzfahhx3h7Y3s?usp=sharing

I am also attaching the flight video pls share some suggestions. What to do next
https://drive.google.com/drive/folders/110u6l-IhvizHMgMmJXvr_ss8XUv5F0E7?usp=sharing

Definitely work on fixing the Z vibrations before changing any more settings, I couldn’t really see anything else standing out.

@xfacta could you pls explain this? what the difference between these and other PID gains.

And you can now set these:
ATC_THR_MIX_MAN,0.5
PSC_ACCZ_I,0.3
PSC_ACCZ_P,0.15

There’s some key bits of information in the Tuning guide:
https://ardupilot.org/copter/docs/tuning-process-instructions.html
And some insight can be gained by looking through lots of parameters:
https://ardupilot.org/copter/docs/parameters.html

The tuning guide gives a safe starting point and initial values for some parameters that will minimise the risk of uncontrolled ascent or fly-away.
Then in some cases, a formula or method to set params to suit your aircraft after some tuning. Like all these parameters you can set them yourself to anything that suits your aircraft provided you test and examine logs.

So i have done a few test flight this morning and also managed to lower the vibrations on all axis significantly. The issue I was having was the gain generated after the auto-tune was too low and the copter doesn’t feel very stable I have now increased the gains where it feels better and more solid in the air but I am facing few issues like.

  1. The throttle to attitude response if not that great it still overshoots a little after centering the throttle stick.

  2. When trying to land it bounce back a little the regain altitude.

I am attaching the logs to the last flight.
https://drive.google.com/drive/folders/1pNBXpkfV5CrLupBERe05gKkGamzzvu6Y?usp=sharing

Hello, thank you for the link i have gone through those links and fine-tune a lot of things according to them. I have also fine-tuned a lot of setting according to my needs.

As i have mentioned above i am making this drone for lifting a payload of around 0.8kgs. So the dry weight of the drone without any payload is around 1.2kgs and the maximum lifting weight is about 2kgs

Now as per the advance tuning we need to set the PSC_ACCZ_I to 2 x MOT_THST_HOVER and PSC_ACCZ_P to MOT_THST_HOVER

At 1.2Kgs Dry weight

PSC_ACCZ_I value is 0.284
PSC_ACCZ_P value is 0.142

At 2Kgs maximum weight

PSC_ACCZ_I value is 0.484
PSC_ACCZ_P value is 0.242

Now as the throttle is set to hover learn this will not have any issue but what i should do about these values should i set them according to maximum take of weight or anywhere between them. Please provide some suggestions.

I am attaching logs of the recent test flight.
https://drive.google.com/drive/folders/1CMeDMs01XHfMl04mXLHp46fJJSQq3S1m?usp=sharing

Done a test flight after updating the PSC_ACCZ_I to 2 x MOT_THST_HOVER and PSC_ACCZ_P to MOT_THST_HOVER values according to the new throttle hover value and a max payload of 2kgs

The flight doesn’t go well as soon take off the copter i think went into fail safe and start climbing i tried to switch from loitering to Atti mode but it was still climbing and i only managed to land it in stabilize mode.

I observed the log i saw the mag compass test got failed so was it the cause for the failsafe or the change in PSC_ACCZ_I to 2 x MOT_THST_HOVER and PSC_ACCZ_P to MOT_THST_HOVER.

I am attaching the logs below.
https://drive.google.com/drive/folders/1LdYThKi8bNLERG__jyoXxH911IayP3eQ?usp=sharing

It started climbing because Vibration Compensation activated. This is the typical response to that Failsafe. Vibration levels are high with Accel clipping on 2 axis.

I have done a flight with a similar weight yesterday but that dint had any issue at that time but with
PSC_ACCZ_I value is 0.284
PSC_ACCZ_P value is 0.142

Here are logs to that flight.
https://drive.google.com/drive/folders/16y1fbPJpNHCCY32pUF_w35RuA0Sthbu0?usp=sharing

If it’s flying the Accel’s are clipping.

Set INS_LOG_BAT_MASK to 1 and make a short AltHold hover flight so we can see how bad the FFT result is.

BTW-.bin files are more useful than .log files. At least one Log Viewer doesn’t like .log files.

Dose PSC_ACCZ_I and PSC_ACCZ_P have anything to do with this? which values i should prefer the older one or newer?

here is the bin file for both the flights
https://drive.google.com/drive/folders/16y1fbPJpNHCCY32pUF_w35RuA0Sthbu0?usp=sharing

Perhaps not. The high vibe basic issue is probably just being exacerbated by the additional weight. Work on fixing the the vibration or you will chase your tail with PID’s.

should I try reducing the gyro filter value from 20 to 15Hz and will this change affect the PIDs?

No, you should address the hardware reasons why there are high vibes.

1 Like

I have done a test flight without any payload.

https://drive.google.com/drive/folders/1ScrllLxYWJsfiq8iz_WYTRSX32L7PJVk?usp=sharing

Much less clipping (single digit counts) so your frame is sensitive to the payload I suppose. You could try a Dynamic notch filter in an attempt to reduce it’s influence on the gyro and obtain a decent tune but it does nothing for the accelerometers. In order to do that you have to add the payload back on and make another AltHold hover flight to capture the FFT data specific to that condition. 30-60s of hover is all that is required.

This is what the FFT looks like now, it will be interesting to see the difference with added payload. Perhaps not much of a change in the frequency but magnitude will likely be different. And of course the MOT_THST_HOVER value will change which is a required parameter for the filter.

Hello, Sure today i will try to perform a hover with payload attached. Till now i was reviewing the logs after every flight and trying to keep the vibrations below 30. As per your experience what should be optimal vibration levels for this setup. I what cause z vibrations ?

Should I keep the vibration fail safe Disable ?
I am attaching a video of the copter with all same parameter but without any payload. Please observe and suggest of the gain were too high.
https://drive.google.com/file/d/13F-IaZyKZsRjMbMLbNusPTErbqLwbAJq/view?usp=sharing

Hey @dkemxr @xfacta I have done another test flight with the weight installed. I have disabled the vibration compensation and also change few loiter and attitude parameters like max lean angle etc.

The vibration lvl are very low this time.

I am attaching the logs pls review.
https://drive.google.com/drive/folders/1jCB8ky2MnbtHIFhHEfQcvdTqAq7vBcXT?usp=sharing

Yes, vibe’s are low. Interestingly the FFT signature is quite different with the payload. What is this payload? Is is fixed, can it move around?