Throttle issues in maiden flight on 400 frame

[axi]

Hello,
after migrating from a ladybird frame to a QAV400 frame with N4 and M4 v2 , I just did a quick test on my backyard and I found some issues. I had a lot of troubles controlling the altitude.

I only flew it on manual mode. First of all, the throttle is extremely sensitive, very hard to maintain hover. Besides that, when it takes off, it does at very high speed, it's not an smooth usuall take off, it's like it's holding and suddenly goes like a rocket. The little time I had on the air, I saw small high freq oscillations on roll axis mostly.

After carefully reading some forum posts, I think, I can have a vibration problems or some high PID values. Here are some logs. I'm curious why mot8 output isn't logged.
I can fix the throttle sensitive thing with a proper curve on the tx but I think it can be a sympton of some major problems. My PID values are the default ones on beta A4 firmware and I'm using SimonK flashed ESCs, the quad flies great with CC3D and NAZA, I'm surprised vibrations can affect this much.


Here is the complete log link: https://www.dropbox.com/s/sww27jszbc510 ... L.LOG?dl=0
Any help would be appreciated.

[aBUGSworstnightmare]

[axi]

Thank you very much for your input!!! I'll report back my changes! You're the man!

AFAIK the log is the same, I just have outlined the part between 72 to 82 when I was "hovering" or trying to hover.

[axi]

Seeing your param file, I noticed that:


CTRL_TLT_ANG_D 4100
CTRL_TLT_ANG_DM 900

and

CTRL_TLT_RTE_D 12550
CTRL_TLT_RTE_DM 9999

¿Having the max lower than the current value makes sense?

[chschmid]

Hi Axi

Have a look here to get familiar with the logic behind these values.


Cheers
Christof

[axi]

aBUGSworstnightmare, big thanks for your parameters! they worked great and now it's flying much more nicely, the throttle sensitivy is still there, but I cured it with a proper throttle curve.

I forgot to put the sd card back on the quad, so I don't have logs, but it makes some small unwanted yaw movements maybe due to the PIDs and DVH engaging without any apparent reason ( I think I didn't surpass the 37 degree angles ). Anyway, there is a way to establish independent yaw deadband on the fc ? on the ground control sw I only see it for roll/pich and throttle.

Again, thanks for your support! I hope I can improve my knowledge and add something of value programming for the project at some point.

[aBUGSworstnightmare]

[axi]

How are supposed to know default PIDs new users to Autoquad with M4 for a more bigger frames?
It would help a lot to know them beforehand.

M4 v2
1100kv T-motor 2216 v2
30A SimonK ESCs
8x5 Graupner props.

AUW aprox 1,1kg. + 3300mah 4S, 1.3-1.4kg
Here you have http://www.rctigermotor.com/uploadfile/ ... 558782.jpg the motors data.
It seems the motor are pushing 1,9+kg power to hover aprox at 65% ( 2500 ). The maxium output per motor is 4000? I don't get that, It shouldn't be 2000 max?
Problems I have found:
- Throttle sensitive, throttle curve cured most of it.
- In manual mode, you can hear, sometimes it speed ups and goes up with very slight or none throttle touch.
- On the first flight, the position hold was very good, after decreasing a bit I and changing the battery,
suddenly it started with small high freq oscillations. The battery change affected also the manual mode,
so maybe the CoG was altered with the battery, but the position of the battery was similar...
- Sometimes I have a yaw drift, you can see it on Youtube video.
- MAG_Z on the logs, makes some strange spikes, this can be a problem? AFAIK the magnetometer is used mostly on the takeoff.

MAG_Z strange behavior

First "good" log:
https://www.dropbox.com/s/gmc1cpfukpnte ... L.LOG?dl=0

After battery change some issues started to show again:

https://www.dropbox.com/s/c460mrcf0mfzy ... L.LOG?dl=0

YouTube video showing some behaviors:

https://www.youtube.com/watch?v=MvFke5gNFtA

[aBUGSworstnightmare]

[axi]

Hi Abug,
thanks for the video, but, that is what I used as guide to tune position hold, if it wasn't for that video, only with the "throubleshooting" section in the wiki, it would be very difficult to have a practical clues.
In the video I posted I show the problem on manual mode, I'm not refering to the high freq oscillations at the end.
¿Altitude PIDs affect also on manual mode?

My frame is difficult to put "on the ropes" for angle and tilt tuning because it has four bobbins holding the upper part of the quad, and I'm afraid they could tear ... I usually tune my minis from scratch with that method because is the easiest way to get good results.