It looks like a PID tuning issue to me. What I would play with: Acc / Gyro compenstation: 200 Your kopter will not fight with the wind and it will smooth the flight. Gyro-D: This also effects the flight stability. Default value is 10. Try 8. Gyro stability: Default value is 6. Try 4. And then 8. See the difference. If these doesn't help, you will need to start playing with P and I.
Hi there..I am also doing a similar build.. could you please tell me the FC mods for a 6S?? Many thanks..
Well it was the Acc/Gyro comp setting. I made a quick change right before a flight today and it's a night a day difference!!! I went from 32-40 and it's already so much better. I will be doing more tuning tomorrow morning. Makes me feel much better now. Josh
TFCM: I thought it was enough to change recoms and the capacitor for 6S. Recoms with Traco TSR1-2450 or Arch SR78 Capacitors with 35V But, on your pictures, there is a second capacitor installed and some bridges removed at the back of the controller. Can you give more info about that please?
Yes you're correct. Remove all capacitors > second image Replace all Recoms with TSR 1-2450 Ad 2x 35v capacitors > first image. And link ground from second cap to ground FC (second image)
Thanks for the info. Now it's clear. It turns out the official Flight Ctrl 2.1 ME is max 5s compatible normally without the removing that specific capacitors and installing the second elko. With this mod, I think it's possible to make the board even 8S compatible? About the capacitors, are you using 330uf or with bigger value?
Well no. Caps from second image (that you have to remove) are max 16v. Bigger Elco/Cap are already 25v so 5s capable. You only ad 1 Cap to compensate for the removed 16v caps Edit; it appears that indeed the SMD caps are already 5s. So only ad NEW recoms to get 5s
Bought a modded FC from http://plischka.at/index.html more then 2 years ago. I think the credit goes to them.
Btw before i forget and you start modding your FC. I did not test the FC above 26v so testing this is at your own risk. Modding your FC to 5s by only adding new recoms is not tested. I always removed the SMD Caps as shown in the image. Downside is that my current FC does not measure the correct voltage > diff 0.7v @ full voltage. No issues with 5s. Also re-check your work with a microscope or hi-res image. When using the Herkules boards there's no need to mod the FC > 8vbat output. But this has some disadvantages .
That's great tfcm. Now are the BL-ctrl can handle that power? or we have to change the elkos on the BL's also to 35v. Bill
Certainly, if the system can vibrate at that frequency then you can configure the PID controller to oscillate the system. To not oscillate at all the system (with controller) needs to be overdamped or critically damped. A well controlled system may be underdamped. (You can shorten the settling time over the critically damped system.) In this case, a disturbance to the system will cause the system to overshoot it's target and have decaying oscillation. This may be what is going on here. One thing I haven't seen discussed in forums where they are talking about PID settings is that changing your copter load, or propellers, or battery type, or many other things will change your systems model and so will require different PID settings to behave well. So if you tune your Cinestar for one camera and then swap out cameras you may find that the Cinestar no longer behaves right. Ideally, one would have a different set of PID parameters for each different build out.
William: I wonder whether having those different configurations built into MK Tool would be the way to change from one PID setting set to another? I suspect most people go with Setting 3 (Beginner), but I could see how those configurations could be used to select the parameters necessary for given payload systems. They can be set using the Tx control sticks so it would make it easy to change to preset values. Thanks for adding the voice of knowledge to the discussion. Andy.
Hey Guys, To be honest I really think the I2C->PWM converter may be the cause of these issues.... I think there is something else going on where it just can't process things quick enough for this many motors. I am going to attempt something to see if this is the case... I have a KK2.0 board that I am going to try and see what happens (dummy payload). Josh
I read many reports saying that BL Ctrl gets very hot under less than 20A load with heatsinks installed with 5s and 6s. If there is a work around for this heating problem, like adding extra caps, or water cooling, it would be nice. Even though, we will still have the I2C problem which is the weakest point of MK system.