[greggles77]

Hello, I am suffering a similar problem to what was posted in last years thread :
https://forum.librepilot.org/index.php?topic=2475.msg17357#msg17357

I have a King Hex 300 Drone, with a Turnigy i6 (mode-1) and when I give it some throttle, it lifts off the ground an inch or two but does not hover, the throttle increases and the thing goes straight up. I then try to reduce throttle and it does not seem to have any effect until I get to zero throttle.. which then the props stop (as expected) and it drops like a brick. It has the CC3d Atom although it detects as the normal CC3d board in librepilot.

I asked Hobbyking, and they suggested reduce my PIDS by 20%. I then tried to do that in Librepilot last night but had no luck, I was not sure what to do.

I read the linked post above that suggested use EasyTune. However I'm not sure how I do that to reduce the correct pid to affect the throttle. I have channel 5 set to flight mode and channel 6 set to accessory, that I'm using for arming.  Not sure if that's relevant, and I'm still trying to grasp the concept of having flight modes on a quad (why, what good are they, scenarios) and what effect they have on my problem now. I've not setup any secondary modes as far as I know. I also read about having to set a 'hover point' which seems a bit odd to me, as I would think that at X percent throttle, it should just hover. I'm guessing the idle rate (1046 for me) is just literally props spinning a little bit.

If somebody smarter than me (most people) could have a read and think, and point in the right direction as to adjusting the pid that will affect my throttle that would be great.

Thanks!

Greg

[TheOtherCliff]

Are you using stock PIDs?

Take props off for a test.  Can you control the motors through the whole RPM range then?  If so, you should try zeroing all D terms, and / or cutting PIDs in half for a test.  An "invisible oscillation" can be so fast that it doesn't move enough to see, yet the oscillation is bad enough that it causes the motors to go L=0,R=100 then L=100,R=0 with the result that as soon as you give a small amount of throttle, instead of L=1,R=1 you effectively get it jumping back and forth between 0 and 100 power with an average of 50% or more.  So as soon as you give any throttle, you get 50% throttle.

[greggles77]

Thank you @TheOtherCliff

Yes I am using the stock PIDS however I'm not sure exactly where I should be checking and changing these, but assume txPID section?

With the props off, it either jumps straight to 100% as soon as I move up round 10% stick, OR if I don't move it the throttle up past that point, the motors accelerate up to 100% on their own. Either way I need to back it off to 100%. This was after I did a re-configure with the hardware wizard last night. When I had the props on and flying it, literally it got 2 inches off the ground then just gained height and motor speed all on it's own - so I had to cut power and try to minimise damage when it bounced.

Are you able to explain how to zero D-terms, or at least cut the PIDS in half, which is inline with the Hobbyking suggestion. I 'm assuming it is all done on the txPID page as mentioned above - https://librepilot.atlassian.net/wiki/display/LPDOC/TxPID

I'm hoping the Turnigy-i6 (flysky-i6) is not the problem, there seems to be a few of them around reporting similar sort of issues...

[TheOtherCliff]

If motor speed jumps up at 10% stick with props off, then it is more likely that you just need to redo ESC calibration and neutral setting (and less likely that it is a PID issue).
https://librepilot.atlassian.net/wiki/display/LPDOC/ESC+Calibration

[greggles77]

I'll give that a try tonight and see how I go. What should I be seeing with the rotors off? I'm guessing a smooth up and down control of the acceleration in line with the amount of throttle? Will the FC try to compensate by thinking it is not taking off or anything like that?

Cheers

[greggles77]

Good morning. I have done the calibration steps and now the motors just rev themselves up to 100% initially (the first time I move the stick), then after that it is either 0 or 100%. Going a bit crazy now! Settings attached if anybody has suggestions.

[TheOtherCliff]

You mention "initially".  Is initially different or does zero throttle stick always make zero motor speed and anything above zero stick always makes full motor speed?

I didn't see anything wrong with the setup that would cause this, and that makes it sound like the ESC calibration did not work.  It is best to use the Home and End keyboard keys, not the mouse moving the slider.

Take prop off a motor and connect that ESC directly to the RC receiver throttle channel and see if you get normal response (or the bad 0 or 100 with nothing in between) that way.  If you get only 0 or 100, then the ESC calibration did not work correctly.

[greggles77]

Ok I will try that and have another go at the calibrations.

Here is a video of my woes, that show it misbehaving.

[TheOtherCliff]

So it does act differently the first time.  That partly looks (sounds) like I-term wind-up that is not cleared by the throttle being zero.

"Not being cleared" would happen if you disabled "Zero the integral when throttle is low" at the bottom of the Stabilization page.  You need this enabled.

After fixing that, redo your Attitude -> BoardLevelCalibration.  Your board is not leveled correctly, so it immediately tries to make it level, which it can't do because the props are off.

After fixing that, another test is to set your flight mode switch to "Rate mode" instead of "Attitude mode".  It will not wind up as much in Rate mode.

Another way to test if the ESCs are calibrated correctly is to (props off) Output page, click "Test Outputs", test each slider to make sure each motor works correctly with slider.  This should work as expected as it only relies on Output min/neutral/max correct and ESC calibration correct.  Then turn test outputs off.  Do NOT save after this since your neutral settings are now probably messed up.

[greggles77]

ok great, I will run through those as well. That all makes a lot of sense.
Thank you for taking the time to review my problem TheOtherCliff

[keiphur]

I've recently purchased the same red King Kong hex 300.  I am new to RC and this is the first not Ready To Fly unit I have.  My transmitter receiver is on order so I cannot attempt to fly it yet.  I can link to the hex with librepilot and run motors and view settings.  If I can help you with any settings or information let me know.   Please be specific, and possibly with steps.  As I said, I am new to hobby RC and librepilot. 

[greggles77]

I think my throttle problem is fixed now. 
It appears that " "Zero the integral when throttle is low"" was indeed the problem, and also changing to 'Rate Mode'. FLight modes are still a bit of mystery to me, I need to research more. It was still taking off unevenly, (ie front right corner would try and lift first), and it moved it slef around when hovering,  however I'm suspect I now need to re-calibrate. I still manage to snap an arm off but either way, it did not rocket up to the moon which is really good. I manage to fly it around at about 2-3 ft in height and then bring itback down again gently. I'm going to give it another run during the week and will see how I go.
TheOtherCliff has been super helpful. I'm really keen to learn about the intricacies of configuring this drone to get it perfect.

[greggles77]

Thank you, yes this is resolved. The drone still drifts around as it rises, and it is still very sensitive to throttle however it is behaving as expected. The attitude mode is useless, it just makes the motors rev up to 100% (at least with props off).
Thank you to theOtherCliff!!
 

[TheOtherCliff]

The normal windup that I discussed doesn't happen when it is in the air because a tiny bit of windup makes it tilt, which makes it stop winding up.