Hi,
First af all I want to thank every contributor to the Arducopter project for their hard work and commitment.
A few months ago I got hooked to this project and first of all did a lot of reading.
Last month I started ordering the different parts and yesterday was D-day for me because I did my first test flight after some minor problem with bad contacts the IMU headers .
First of all, my setup
APM
IMU
Mikrokopter frame
KDA 20-22L motors
ESC's 30A
APC 10 x 4,7
DIY power distribution
Here are some photos of my quad.
Note, the ESC's will eventually be covered ;-)
Now for my problem: PID tuning!!!!!!
I calibrated everything according to the manuals, uploaded the alpha 1 firmware and used the standard PID values. I did a test flight in stable and + mode.
My observations:
-Front motor and rear motors tend to turn very late.
-Before lift of nose tends to go in the gound.
-When in the air quad tends to drift and yaw to the left
-At a certain height (1,5m) quad starts to oscilate in the role axes
Stable PID values are:
Roll: P=4; I=0,15, D=1,2
Pitch: P=4; I=0,15, D=1,2
Yaw: P=3; I=0,15, D=2,4
I've allready tried some minor changes in the PID values but this cost me some leafs of a plant in the living room. ;-)
I would try to change Roll P to a lower value (3) to minimize the oscilation
Any suggestions?
Replies
Still some wobble when descending fast.
Don't use the bullet connectors that comewith the KDA motors.
Mine were to loose (the male connector can still turn when connected to the female connector) and was the reason that sometimes the motors would get a nodge or even turn off resulting in a crash.
Now I can continue with the finetuning ;-)
And yes it flies very stable and its controlable even with the two motors not starting at the same time. I have a feeling that this has to do with the magneto not installed. Like you can see in one of my video's the yaws seems to drift. And I think that the APM wants to compensate this by only driving to motors left and right. I also did some further PID tuning, reinstalled APM with gyrotape and recalibrated everything.I'm happy with that now.
But that is not the big problem.
The problem is that sometimes the motors get a couple nodges and then fall out resulting in a crach.
I tested the motors seperatly directly with the RX ch3 and everything seems fine. Then testes it with one motor on APM and also fine. When I connect 3 the problems starts and its not allways the same motor. I disconnected the 5V from ESC thats suppling power to the system and hooked up external 5V. Same problem. The resoldered the header on the APM. Same problem. What now??????
Change ESC's?
Change motors?
In the meantime I completely redid the frame because in my opinion the arms would bend a little. Now they stay a lot stiffer.
As suggested I change the Min throttle value in the userconfig to the value were my motors start to turn with the motor command. This was 1250. I uploaded the firmware recalibrated TX.
Did the teset and the only thing that changed was that I had to push the throttle stick much further to start the motors.
Left and right engine start a little bit but then stop.
It only wobbles a little bitwhn coming down.
It seems that the wind has a lot of influance. The smallest gust takes the quad with it.
Whats the key to adjusting and maintaining hover?
Will the GPShold really keepit in place?
MIN_THROTTLE in the UserConfig.h to a Higher value (1100).
This seems to put on all motors at the same time and have a much smoother startup.
Of course I had to update the firmware of the ATMega328P because it had the brown-out bug before trying this. :(
Jani: could this be concerned to the type of motors we have (KDA20-22L) that nedd a higher value to start up?
Could this be modified in the code?
Best regards,
Emile
In the meantime I did some modifications.
I reinstalled the APM en IMU. Its a pitty that there are only 3 holes. If there were 4you kind a could calibrated mechanicly.
When you install the APM does it have to be exactly level with the arms and motors or as close as can be?
And do you do the rest with the offset in the configurator or with the trims of the TX.
My quad flies but still with a little bit of roll to the left. Less wobbly and the yaw is still an issue. I try changing some yaw PID values.
I my opinion the two motors that dont start when throttleing up isen't a big issue.
Eventually they start and turn fast enough to take off.
Meanwhile I lowered the D value to 1 and it started to fly much more stable.
But due to my fault an ESC got wet and blew. Is allready ordered.
More tests next week.
What is your arming procedure of escs?
-
1
-
2
of 2 Next