[saleezy]

Hello!

Im almost done with my build but having some issues getting it to fly right. The quad arms fine but when i give it very very little throttle the motors go max speed. If i move the throttle to the top it doesnt change the speed sincw they are already maxed out. Been getting super frustrated towards the end of this and cant seem to find the answer anywhere online. Is there some setting im missing? All of the copter and transmitter setups appear to go smoothly so im not sure what is wrong at this moment. When I test the output and set neutral... it seems fine.

Ive got a QAV250ish clone with a CC3d and a Flysky FS-i6 controller. If any of yall have any idea id super appreciate it! Ahhh! I want to fly this thiiinnnnggg!!! 

[TheOtherCliff]

Look at the Configuration -> Output page.  Are your motor mins all 1000, neutrals between 1000 and 1100, maxs all 1900?  (2000 doesn't work with PWMSync, you have to reduce it to about 1900 when you calibrate ESCs).

With props off, do all of the motors go max, some of them, or do they work correctly?

Are you trying to use OneShot ESC protocol?

Props off, battery and USB plugged in, watch Configuration -> Input.  It does not need to be armed.  Does the Throttle channel act like you would expect?

Props off, battery and USB plugged in, armed, watch System -> DataObjects -> ActuatorCommand -> Channel.  Does zero throttle generate 1000's and small amounts of throttle generate numbers close to 1000 (like say 1060 or so)?  Medium throttle generate numbers around 1450?  High throttle generate 1900?

[nikhil12340]

Having the same issue...The motors rev up to full when throttle is minimum...

[Mateusz]

If the testing outputs works as expected in GCS, then I would suspect Flysky FS-i6 controller. Is it sending signal as it should ? What kind of input protocol are you using ? Posting exported UAV configuration file may give a better clue, or at least screenshot of Input and possibly Output tab ?

[Eel_dahc]

I am new here and may not be any help at all but my problem when I put my quad together was that I couldnt get anything until i was at half throttle and had to do the wizard a few times to calibrate escs and get the idle right.  I think that I also forgot to uncheck those link boxes afterwards from the first time I tried it. Not Sure if that link made a difference or not. I have a generic zmr frame with eachine racer cc3d, i6x and using Ibus. Using the second zmr settings from the wizard.

[ak74]

I do experience a similar problem. When setting up the rig the response seems very ok (low 1000ms - motor starts at 1062ms - max 1900ms). When trying to fly the props spinnes as expected from idle to takeoff, then something goes completely wrong: the props seems to go full throttle and the only option is to almost cut the power, and crash.

The setup is as follows:

• cc3d (rev 1 & rev 0 tried both, same issue)
• Frsky c8r, turnygy x8 - tried both, same problem
• Afro ESC 30Amp

The problem started around the time the Libre Pilot were introduced. I am setting up the rig through Ubuntu, Librepilot 16.9

[TheOtherCliff]

I bet that your quad is fairly powerful and hovers at much less than half throttle.

There is an issue that I call "invisible oscillation" where the oscillation is so quick and small that you don't see it.  The quad sees it though and the effect is that it gives full left roll then full right roll (and pitch) many times a second.  Our firmware sees that the throttle is very low, and to give full roll it must increase the throttle to 50% so that one motor can be 100% and the other one 0%, then it switches from 100, 0 to 0, 100 and back and forth, the result is that all motors are running about 50%.

The solution is to reduce the D gains in whichever PIDs are responsible.  When you do that, you need to reduce the PI gains a bit too to avoid oscillation.

If you are using AutoTune there is a setting that reduces D and slightly reduces PI.  DerivativeFactor (0.0 to 1.0 are reasonable values) default value is 1.0  Try setting it to 0.5 and "reexport PIDs".  Reexport happens whenever you move the SmoothQuick knob, or if you haven't set that up, while disarmed, just toggle the FMS quickly into and out of AutoTune mode, like 1,2,1,2,1,2,1 if position 2 is AutoTune.  If you do the FMS toggle with it armed, it rotates through a set of 5 SmoothQuick settings (read the AT wiki page).

Vibration from motor/shaft/adapter/prop being out of balance or out of alignment makes this a lot worse.  Balance and track those props! 

[ak74]

Thank you TheOtherCliff,
that sound pretty plausible. Having SUNNYSKY X2208-13 1500KV with 8x4.5 props may be over-powering a 1 kg rig.

 I am still a noob on the PID settings, so i need to learn some more. I will test the input you are providing and see if i kan get it better... Haven't figured out if i can get autotuning to run, will try.

Yesterday I just realized that the throttle curve in the vehicle setup could be set as non propotional (lowering 25, 50 and 75% points) and the experience was ok. Still a lot of occilation but now the rig doesn't just take off.

So that in mind tells that the rig is overpowered.

[TheOtherCliff]

Nothing wrong with being overpowered. 

Just make sure your props are balanced; especially when you run into these issues.

[ak74]

Hi again,
Well got it off the ground without tweaking the PID settings (i am a noob in libre pilot). Exponential on Thr, Pitch & roll did make it a bit dull and controllable. I am sure that you are right (TheOtherCliff) in the micro occillations. Now for the major issue: Which screen to go to first? Any good tutorial on autotune (How to start?)

Do i make the FMS-page like:
* Pos 1 = Stabilised1 + Bank1 + None
* Pos 2 = Autotune + Bank1 + None
* Pos 3 = whatever

Stabilisation bank 1 = Moderate (default)
What else to tweak?

[TheOtherCliff]

Do you get a lot of power with very little throttle stick and yet don't get much more power when you give full throttle stick?  If so, then you definitely have these invisible oscillations.  Fixing your vibrations and reducing the D terms is what will fix this problems, but there is no way to tune it if there is too much vibration.

I would suggest Pos3 = Stab1 + Bank3 just because Stab1 is the easiest to fly if it does something strange.

You must get your props balanced well to be able to fly it.  I spent a long time fighting vibrations on a cheap plastic quad.  Arms were not really stiff enough.  I static balanced the props, then used a "servo tester" connected to one motor at a time and dynamically balanced each prop by finding the best place to put a piece of black electrical tape to minimize vibration (as felt) (measured would be more accurate).  Even that was not good enough.

I noticed that the vibration only happened at one frequency, and that a small amount of holding it with my hand would stop it.

Each arm has a small landing gear extension at the bottom.  I finally got the vibration manageable by running a string around from gear to gear with a couple loops at each one so that each string was tight.  The string damps the vibrations enough.

[ak74]

Thank You so much The other cliff....
Balancing the rotors (which i have neglected for many rotor changes) did do the trick. Its stable as a rock and flies completely as expected... I think i have got a poor batch (chinese mailorder) as they did require a fair amount of hot glue at the tips.... (my version of black electric tape)

Now my new Taranis will come into action... And maybe i will find out how to set up auto tune...

[f5soh]

And maybe i will find out how to set up auto tune...

Autotune only works using a Revolution board. Seems you are using a basic CC3D.

[ak74]

Thanks f5soh
Yes, its the basic board. (rev 2)

[Visesh Lal]

Hey all, so I am having a similar pblm with my motors misbehaving.

My build, a 550 hex frame, 1500kva emax motors, 30amp simonk esc w/bec, frsky x4r receiver in sbus config and taranis 9d plus radio, revo board with updated firmware.

I set everything up correctly, ran the wizard as I did in a previous build, everything went as planned, the motors tested as they should, using the gcs, the motors increased/decreased, and all worked together.

The problem came when I tied to fly. Some motors will come on, some wont, the ones that came on went straight to full with a mm of throttle, the others might come on at about half throttle. I swapped around escs, I swapped around the motor inputs on the revo board just to see, then I changed out my revo board for another. The problem never followed a component, nor occurs with any pattern, it is completely random. And if I connect it to the gcs, I link all the motors, move the slider side to side, and all the motors respond perfectly.

I saw on another similar post, on pg 1 say something abut using settings to configure the motors as opposed to the output pg. Any more info on that.

Well thanks in advance for any info, I will be sure to post updates as I go..