Help with yaw drift

ArduCopter Copter 3.5
21

The copter looks to a bit unbalanced and underpowered.
The perfect copter hovers at PWM1500 output.
Consider PWM1800 as your maximum output as the FC needs the extra to maintain stability.
Your motors are having tops to PWM2000 to maintain stability, which will cause slight yaw changes until it is stable and can bring the yaw back.
Stability takes precedence over yaw.

The induced yaw you see is from the motors maxing out, in particular motor 2.

13%20am
Screen Shot 2018-04-10 at 6.22.13 am.png2283×951 320 KB

To improve what you have, balance the copter.
Use the external compass only.
Redo your compass calibration.
Do a motor/compass calibration.
Do an autotune on yaw.
This should improve things.
Getting more power or lightening the load will improve things greatly.

1 Like
22

Thank you very much for your analysis. By balance I assume you mean lat and long balance, will do.
I’ve been unable to get the copter to enter autotune. I get a fm change failure audible on my Taranis. The bar graphs is MP setup show channel 7 with the proper PWM change to invoke the change, but nothing happens. The very first time I ever tried I saw a twitch and heard a beep in flight, but nothing happened, and I don’t even see that now.
As far as more power I assume you mean a different motor/esc/prop combo?
I appreciate everyone’s input greatly.

23

I don’t think you have many choices with that frame at that weight. As I mentioned I used 2216-1100kv motors on 9" props. Performance was good flight time was poor.

24

I was afraid of that. Thanks for the input. I bought the airframe cheap on EBay and put on the gimbal, fc etc on hoping to cobble together a good machine.

25

I agree that at first glance aircraft appears to be underpowered, but after looking at the log, there is no amount of balancing or tuning that is going to fix this.

If you use the log browser in Mission Planner and look at RCOU > C1,C2,C3,C4, you will see the motor outputs clipping at 2006. The motors are being maxed out, and there ain’t no way that should ever happen on a camera platform…

As for the maximum output PWM number of 1800, that leaves just 200 for the FC to maintain stability. A PWM of 200 is no where near being enough head room.

A “typical” throttle range is from 1000 to 2000. With 1500 being 50%, 1600 is 60% and 1650 is 65%. 1800 is 90% throttle so your useful headroom is only 5% because the thrust change from 95% to 100% throttle is insignificant. http://ardupilot.org/copter/docs/motor-thrust-scaling.html. Read MOT_THST_MAX…

When I’m number crunching for a new aircraft I shoot for a hover throttle of 50% or slightly lower. I use 60% throttle as a warning, and 65% throttle as a hard limit. This method puts the Hover PWM between 1500 and 1600, with a hard limit of 1650.

If the projected Hover Throttle for a given configuration comes out above 65% I start looking for ways to cut some weight, or I start looking for a better power system configuration.

From where I sit the problem is a motor/prop/battery miss-match.

First off, these motors are not rated for 4S.
Second, the reason the motors are being maxed out is the props are being over sped which actually lowers thrust.

Here is a link to the motor test data:

https://www.foxtechfpv.com/sunnysky-221213-kv980-p-384.html

With an AUW of 1800 grams Mike should be swinging 9047 props on 3S, which sounds a lot like what a DJI Phantom runs…

26

Thanks for the analysis OG. I overlooked the basics I should have looked closer at the motor specs. There’s a reason it was for sale cheap on Ebay I guess.
Can you tell me the risks of using the 4s with these? I assume it would be heat and early wearing bearings?
I’m pretty disappointed.

27

Oldgazer is right that they are rated for up to 3S but you can run 4S with those motors. I had the same motors on a F450 frame on 4S and with larger props than you are probably running with no problem at all. On 3S that quad would barely get off the ground. What props are you using?

28

They are 8x5s. i think I could squeeze a 9 on there. Maybe a change of prop might help? todays logs look marginally better after balancing and shedding the 900mhz telemetry and external led. the weight wasnt much but the balance probably made the biggest difference, as it was quite a bit off. this time i balanced it on a small screwdriver tip. I really appreciate all the help guys.Next up is all the recaliberation. this time Im going to sit out in a field.

29

Put the largest props on that will fit. Sure, they will tax the motors and ESC’s further but for a sedate camera craft no big deal.

30

So I went to 9 inch props and turned off the internal compass. The yaw drift is greatly reduced, the motors and speed controllers are barely warm to the touch. The PWM avg at hover is right around 1500. I still have balance work and calibiration to do, but the difference is significant. endurance is also up around 20 percent. logs here: https://drive.google.com/file/d/1hiDe3d3uBxSdydowXbY81ne9gilxo036/view?usp=sharing

Thanks again for the help.

31

And make sure the center-of-gravity is in the geometrical center of the copter.