Thrust problems

[frank]

Hi everyone

just a brief information; I’m trying to calibrate a friend’s UAV. This quadcopter has always worked very well and now has about 110 flight hours (perfect motors with OK bearings) and in a moment of madness the friend got his hands on the settings without saving the originals.

I’m no expert but I will certainly perform an Autotune to adjust roll and pitch but the question is to understand why it always corrects the throttle value and how I can intervene. That Wa-Wa-Wa that you hear coming from the engines is the result of an imperfect compensation on the thrust but I don’t know how to set it (the payload is always the same)
Do you have any suggestions about it?

At the moment I have not downloaded the log but in case I publish it. Let’s say that in a general sense, the question is only to understand which instruction to intervene on

Thank you

(sorry for the video edited in 5 minutes with Filmora … just to hear the sound)

Little video: https://photos.app.goo.gl/gi3YhYMzHGfbW7nT7

[xfacta (Shawn)]

If you have the logs of flights you can easily extract the parameters from the log of a good flight - at least with Ardupilot anyway.

Let us know if you need help with that.

[frank]

Hi Shawn, you are always very kind and helpful.

Place the log file; it is important in size but I hope it is not a big problem

The aim of asking for this courtesy of analysis is to understand the reason for the Wi-Wa-Wa Motors in Hovering (only in Loiter … by heart, I seem to remember that in Althold it did not do it) and I know that the pitch is The roll will be recalibary via Autone

Where can I intervene? What instructions are involved to eliminate the described effect? Thank you

https://drive.google.com/file/d/1vdjN5fjIyU4dWtkdOW_TszHku3Blb7dx/view?usp=sharing

[xfacta (Shawn)]

Do you have a log of an old flight where everything was good?
Exactly what ESCs do you have?

A lot of the oscillations and noise will be coming from the D terms being way too high. You will have been lucky not to burn out motors or ESCs.
The harmonic notch filter is not configured correctly, and not doing anything - also probably not worsening any problems.
The rate P and I values look more suited to a small 5inch quad, so we might have a bit more work to do there.

First job - definitely update to latest stable firmware, there is important fixes that you need for Cube Orange (and plenty more)

Second - set all these:

ANGLE_MAX,20  // increase a bit for now, allow more movement
ARMING_CHECK,1
ATC_ANG_RLL_P,6.5
ATC_RAT_PIT_D,0.005
ATC_RAT_PIT_I,0.08
ATC_RAT_PIT_P,0.08
ATC_RAT_RLL_D,0.005
ATC_RAT_RLL_I,0.08
ATC_RAT_RLL_P,0.08
ATC_THR_MIX_MAN,0.5
AUTOTUNE_AGGR,0.1
BATT_ARM_VOLT,44.30
BATT_FS_LOW_ACT,2
BATT_LOW_VOLT,43.2
BRD_BOOT_DELAY,3000
FENCE_ENABLE,1
GPS_GNSS_MODE,65
INS_ACCEL_FILTER,10
INS_HNTCH_BW,17
INS_HNTCH_FM_RAT,0.7 //  you wont see this until after the firmware update
INS_HNTCH_FREQ,33
INS_HNTCH_REF,0.15
INS_LOG_BAT_MASK,1
INS_LOG_BAT_OPT,4
LOG_BITMASK,180222
PSC_ACCZ_I,0.4

Don’t race out and do another test flight yet - get back to me with the ESC details first, then we have more parameters to set.

[frank]

Dear Shawn, I can only admire how you spend with generosity and availability towards people who wander in the darkness in search of light … you are a really precious resource for everyone and thank you!

Unfortunately as I told you, the .param file has been completely tampered with people more inexperienced than me and therefore there is no valid trace at the PID level or other parameters so we start from scratch

I loaded 4.3.6, sized the batteries (50V), recalibrated compass, ESC and radio control as well as “Initial Tune Parameters” for 26 "propellers

The engines are U8II 100kv and the ESCs should be Alpha 60A HV

Thank you again

[xfacta (Shawn)]

OK, set these too

MOT_THST_EXPO,0.6
MOT_PWM_MIN,1100
MOT_PWM_MAX,1940

The expo may need adjustment but it should still work. It could go either way.
Now you need to use MissionPlanner motor test to check
MOT_SPIN_ARM,0.1
still works OK, and change it so that motors will all start smoothly and reliably.
Then take that value and add 0.03 and set it as MOT_SPIN_MIN
For example
MOT_SPIN_ARM = 0.11
0.11 + 0.03 = 0.14
MOT_SPIN_MIN = 0.14

Also set
EK3_PRIMARY,1
a recent change to the default IMU for more reliability.

Now just take off in Stabilise mode, if everything is going OK use AltHold mode and just hover, do some gentle pitch and roll, nothing radical. Maybe throw in some yaw.
Don’t use POSHOLD in future, just Loiter instead - but stay away from that for now, let’s just get some good old fashioned stability first.
Let’s see that log.

[frank]

Thanks Shawn,

As soon as it stops rain I will make some trial flights and I will send you the log.

There is a particular reason why do you prefer to use ek3_primary, 1? (by default on 4.3.6 is set as 0)

Thank you

[xfacta (Shawn)]

It is the new default for a Cube Orange, since it is more stable and reliable, as it turns out.

[frank]

Hi shawn

After a couple of days of rain, I finally made a couple of flights with your indications today

From 4.1.3 I went to 4.3.6 (just today the 4.3.7 but I’m calm), I calibrated RC, ESC and the compass

I made a first flight with your changes and everything is ok … you need to make some further tuning but the WA-WA of the engines remained very little

Let’s say we are on the right path

I wanted to make Autotune to better finish the flight but does not start

Not having vibrations on the axes (I checked), the suspicions fall on two things:

1. I did not recalibrated the accelerometers but I thought they shouldn’t be recalibrated every time I change release; Also because if I have a 300 kg manned drone it is quite uncomfortable turn it up, down right and left
2. I have the ESC excursion from 1102 to 1927 and Mid Point at 1515us … I know that 1515 is exactly half but I wanted to move the mid point to 1500; You never know that it was the problem … instead no

Conclusion: flies much better even if it still needs a fine tuning

If you have patience to take a look at pastime to log, here it is

Thanks again

https://drive.google.com/file/d/15vuqS0QoUzzAw3QHqYGj2dfPAXGEpSRb/view?usp=sharing

[xfacta (Shawn)]

Not much better than earlier - almost none of the changes I specified were done.
Work through every change I’ve listed and set them all, including

MOT_THST_EXPO,0.6
MOT_PWM_MIN,1100
MOT_PWM_MAX,1940

Setting those to something else and doing ESC calibrations doesnt work for T-Motor Alpha.
Be sure to also perform the test to correctly set MOT_SPIN_ARM and MOT_SPIN_MIN

[frank]

Hi Shawn … I’m applying the corrections indicated by you but I cannot fly due to bad weather

I have to understand what … in the MOT_SPIN_ARM procedure, after doing the test of the engines with “Motor Test” on Mission Planner to identify the percentage of Thrust in which the engine is rotating,

I noticed that 2 engines hook to 4% while two others hook to 3%

Can I consider all the engines at 4% ??
Therefore:
MOT_SPIN_ARM: 4%+2/100 = 0.06
MOT_SPIN_MIN: 0.06 + 0.03 = 0.09

Thank you

[xfacta (Shawn)]

Yes you’ve got those values correct!
That should allow for reliable startup.
MOT_SPIN_MIN might need to go up a little, but we can assess that later.

[frank]

Hi Shawn,

I hope I don’t take too long. I know it is all on a voluntary and passion basis but I don’t want to take advantage of your courtesy; However we are on the good road

I do an incorrect corrige on the first post … the correct values and which I inserted are:

Mot_spin_arm: 4%+2/100 = 0.08
Mot_spin_min: 0.06 + 0.03 = 0.11

Impressions after the first flight

→ tests carried out in Althold is then a pinch of Loiter before landing … if you wish I also have the video

1. Eliminated that annoying Wi-Wa-Wa of the engines … evidently I have never adopted this a little more scientific procedure, so leaving 1000/2000 the PWM limits and not recording MOT_SPIN_XXXX obviously did not be fine

2. I found the Thrust NOT very reactive … By providing more gases, it responds very slowly to the climb and also by removing gas, I have to lower almost all the stick to obtain a significant loss … a lot of excursion but it is normal at this moment

3. Pitch and roll respond well to the sticks … behavior a little oscillating and not very reactive to wind gusts (by the way: about 20 kn) but in the second phase we will work (I will work) on the tuning

How can I improve the reactivity of the gas and a dark resistance to the oscillations caused by the wind?
Could a Autotune be enough or adjust the bar in “basic tuning” to get him more “Twitchy”?

Thanks again for everything

https://drive.google.com/file/d/1ttJvp7r4QBLYYyR4U9tynb2cyfIJKmiH/view?usp=sharing

[xfacta (Shawn)]

OK now we are making progress.

Now that all the oscillations in the motor outputs are mostly gone, we can see there is a distinct difference between motors 1 + 3, and motors 2 + 4.

This means there is a significant weight shift to the rear making motors 2 and 4 work harder.
See if you can get this balanced better.
If there’s meant to be a payload like a camera, at the front and it is difficult to shift batteries, you may have to add some dummy payload there and make sure it is very secure.
It is detrimental to have a camera and gimbal fitted during tuning - we can compensate for that later.

For Arming Check, if you need to deselect something like GPS for indoor test flights (for example) make sure you have at least these others selected

Selecting ALL and unselcting everything else is normal - and if a feature is not used, like FFT or VisualOdometry then it doesnt contribute to the arming check and is ignored.

Adjust these for the next tests (after you’ve adjusted that center of gravity/weight issue)

ARMING_CHECK,1
INS_HNTCH_BW,20
INS_HNTCH_FREQ,35
ATC_ANG_PIT_P,6.0
ATC_ANG_RLL_P,6.0
ATC_INPUT_TC,0.22
ATC_RAT_PIT_D,0.004
ATC_RAT_PIT_I,0.10
ATC_RAT_PIT_P,0.10
ATC_RAT_RLL_D,0.004
ATC_RAT_RLL_I,0.10
ATC_RAT_RLL_P,0.10

You’ll see I am asking for a slight decrease in the Angle P’s, and a slight increase in the Rate P & I s.
D terms are decreased a bit more just for good measure, since D was still over-reacting a little. At the moment these are just small changes, as bigger props and motors can be quite sensitive.
My reasoning is:
For larger movements the Angle P is over-reacting (too strong) and forcing the Rate P to work in the opposite direction to contain the movement. With Rate D damping that effect a little too much, this has limited the Rate P effect and allowed overshoot and some level of small instability.

Try another flight the same as before but take off in Stabilise mode first.
If there’s obvious oscillations then land immediately and reduce the Angle P’s even further

ATC_ANG_PIT_P,5.0
ATC_ANG_RLL_P,5.0

[frank]

Good morning Shawn

it seems incredible but here it has been raining for a month and a half and I’m afraid that today we will stay at home in front of the fireplace

But great time to make changes

I want to clarify a couple of things:

• as an aeromodeller of the 80s, I took particular care in checking the CG so before each flight I check that the gimbal (with sliding mount) is centered and it seemed to me that it was because the drone is perfectly symmetrical (with batteries installed) and lifting it in the middle of the frame, does not hang to one side or the other (right left and front and back)

• Originally it works with a Gremsy and an ADTI camera and not wanting to crush the expensive payload, I inserted a very bad 2-axis gimbal, made with an old Alexmos but absolutely functional and realistic in the movements (the ADTI camera is simulated by a block of wood from a pallet ) I thought the calibration was better with a payload of similar weight and size

• yesterday I didn’t want to risk flights in Stabilize because there was a sustained and annoying sell

Let’s hope that tomorrow will be good weather and as soon as it’s done I’ll update you

Thanks again sincerely

[xfacta (Shawn)]

I understand the bad weather.
Take a moveable gimbal off completely and replace it with some dummy payload, or have some system to lock it in place so there is no chance it can swivel and move around at all.

It’s possible the motor output imbalance is because of wind if you had the copter pointed in just one direction for the whole flight.

[xfacta (Shawn)]

Frank, full respect to aeromodellers of the 80’s Where would we all be now without them hey?
I can understand you are taking the correct actions regarding the gimbal and camera.

[frank]

Hi Shawn … but don’t think it’s old

Except that I started flying with aircraft in 1976 and I was 9 years old. I was early and I have an infinite collection of crash

With this post I must make you my apologies because the friend for whom we are optimizing his drone (with your very precious help) must urgently use it for some works that has in suspense and its client does not accept postponement (relief on crops they must be detected now)

They are standing from 6 in the morning to take advantage of the absence of wind and to make a calibration as reliable as possible.
However, I will return to the drone in a couple of weeks and if in the meantime I will not be crashed maybe, if I don’t disturb you further, we will try to finish these tests with which I am learning a lot. I limited myself to a rudeness via Autoton but such a fine regulation is very interesting

This morning, in anticipation of the return, I made the latest changes and I must say that it works much better … I made these changes (in addition to yours):

• Mot_thst_Expo = 0.75
• compass calibration (in flight)
• Autotune on Yaw (no roll and no pitch)

Now the engines work perfectly (it seems to me) or certainly better than it had arrived; However, the problem of the previous flight was precisely because of the front wind that made the engine 2 and 4 work more

Now it flies discreetly well even in the presence of a little wind risen during the end of the tests

The only note … after the laid he loses me some altitude … can I intervene without carrying out tests?

I am attaching the log

Still a dear greeting and a great thanks for everything

https://drive.google.com/file/d/1aY4wrBJzDMhVPVp0jm1lUU4UzFn/view?usp=sharing

[frank]

hi Shawn

a little update. The friend to whom we helped to set up the drone is really very satisfied with the behavior in flight considering that he messed up by putting his hands in the parameters without having any basis

He will bring it back to me in about ten days and if it’s not a problem for you to suggest any corrections, I will update you later

Did you get a chance to see the latest log? it flies much better but I don’t know if it could have problems in flight with the wind; it seemed to me that it did not oscillate particularly; I saw it rigid enough but I don’t have the skills to make an analytical diagnosis

Thank you

[xfacta (Shawn)]

That link was just the params file and not a log.
Once the copter is at a stage where it “looks OK” in the air, we still need careful examination of logs to make sure it really is OK and not on the edge of being bad. Usually Autotune sorts it out once everything is playing nice together.