Drone behaving strangely ?

[spinzer]

Hi All

This is my first post and i hope someone could help.
I have been flying my old zmr250 for many years without problems but decided to upgrade the components as one motor decided to finally fry out.

But now i am having some problems ?

I am using cc3d and the latest librepilot and did the vehicle wizard setup first in librepilot

The first odd thing i realized is that librepilot put the first 2 esc's on 490hz and the last 2 on 50 hz.
I manually change the last 2 to 490hz ? was that the right thing to do ?

Then when i arm the heli the motors idle perfectly fine.
Holding the quad down i slowly advance thrust, it starts slowly then suddenly spikes to what seems like a peak, its as if the input i'm giving is not how can i put it linear or i don't feel a direct connection with my input and the motors, and i'm afraid if i finally let go to throttle up to let the drone fly for the first time it will shoot up like a rocket.

Is the odd behavior because i'm holding it down or is there a setting that needs to be changed ?

I have read that i need to adjust a sync setting in libre pilot but i can't find it, i used the sync motors to master in blhelisuite. and i put the motor timings to medium high.

What could be wrong or is it simply because i'm holding the quad down ?

these are the components:

CC3D FC

4s lipo

Hypetrain VORT3X 2207 2650kv Motor V1.0 × 4

Ethix S4 Lemon Lime Props

Flycolor Raptor BLS Slim 40A × 4




Reply

Report Edit

 

[Dugdog47]

Yeah you need to try flying it. Holding it down might be making the flight controller try to compensate.

 

[wafflejock]

Yah bench testing with props off or pinned down will result in some weird behavior would also suggest doing esc calibration:

Battery unplugged and props off set throttle through motors tab to 100% then power on with battery wait for esc tones to stop beeping, then throttle to 0, should hear some confirmation beeps then silence. This makes sure whatever low to high values the FC is using are "understood" by the ESCs.

After that would field test just use right stick only at first after arming to be sure roll and pitch work as expected if so should be good to give it some throttle.

 

[spinzer]

Ok i ticked the following option in the stabilization settings in librepilot:

"zero the integral when throttle is low"

The quad advances throttle much smoother now, however i seeming now have another problem.
(I'm not sure if it is directly related to the option i ticked above however i think it might somehow be ?) Or not ?

As i now advance the throttle i get oscillations/vibrations, more high frequency vibrations i would call it and i'm barely at half throttle. I can see this shredding my quad to pieces should i take off and fly.

Is there any settings i can change to make it stop doing this ?

(wafflejock, i calibrated the escs through BLHELI, however maybe its not the correct way, let me try the way you just mentioned and report back thanks)

 

[wafflejock]

If still just talking on the bench and pinned down the erratic behavior when using angle or acro (any sort of input from controller) will result in strange throttle changes on the motors. The problem is the quad expects the gyro on the FC to do a thing when it increases throttle on the back two motors it expects to tilt forward when that doesn't happen the FC gives it more throttle (it keeps trying to correct but has no control authority it can't apply force that has any effect on the gyro data). Also when on the bench strapped down or no props there's a bunch of vibrations happening from the motors and the frame that probably isn't being entirely filtered out (if at all) by the FC so it's seeing a bunch of little jerky motions but not able to correct for them. If controlling from just the motors tab or bleli you should expect normal behavior 50% throttle goes half speed and doesn't sound like it's ramping up or down at all but anytime you get the PID loop going the thing will start trying to self correct to reduce the error from some target/setpoint it's trying to achieve based on stick input etc.

 

[spinzer]

Wow Wafflejock that info makes loads of sense, however i'm just still a bit "wet footed" in pushing myself for the maiden.

I did the calibration as you suggested and when i armed the quad it seemed
like some parameters in the wizard setup changed as the quad at one stage only fired up 3 motors. (this after i did the motor calibration in the motor tab)

So what i did was after i did your calibration method (in the motor tab), i re-ran the whole setup wizard.

Now things seem ok at throttle just above idle, however i am afraid to advance throttle too much.

What would really help (even though what you say makes 100% sense) what would help a lot is if someone could just take their quad, arm it with props on and advance the throttle carefully for just a bit like i did while holding it down just to confirm if they also get high frequency vibration/oscillations.

(this will tell me 100% if the high frequency oscillation/vibration i'm getting is because i'm holding the quad stationary. (it really sounds wicked as if it would shred the quad to pieces in flight (also the quad is tamely set up, but the throttle response is insane, i am barely at 15% and it really ramps up fast) Maybe its just something i have to get used to coming from 3s and smaller motors on my previous setup.

I held my quad from underneath and advance the throttle (cutting out table vibrations) but obviously can be done by pressing it down on a table as well, if anyone can do this please this will give me more confidence if to go
and do an actual test flight.

Why i insist is I don't have much money to spend on new motors and esc's should this go wrong on the maiden, that will be curtains for my fpv career, so want to make sure as best i can that it will be ok.

 

[spinzer]

Wow Wafflejock that info makes loads of sense, however i'm just still a bit "wet footed" in pushing myself for the maiden.

I did the calibration as you suggested and when i armed the quad it seemed
like some parameters in the wizard setup changed as the quad at one stage only fired up 3 motors. (this after i did the motor calibration in the motor tab)

So what i did was after i did your calibration method (in the motor tab), i re-ran the whole setup wizard.

Now things seem ok at throttle just above idle, however i am afraid to advance throttle too much.

What would really help (even though what you say makes 100% sense) what would help a lot is if someone could just take their quad, arm it with props on and advance the throttle carefully for just a bit like i did while holding it down just to confirm if they also get high frequency vibration/oscillations.

(this will tell me 100% if the high frequency oscillation/vibration i'm getting is because i'm holding the quad stationary. (it really sounds wicked as if it would shred the quad to pieces in flight (also the quad is tamely set up, but the throttle response is insane, i am barely at 15% and it really ramps up fast) Maybe its just something i have to get used to coming from 3s and smaller motors on my previous setup.

I held my quad from underneath and advance the throttle (cutting out table vibrations) but obviously can be done by pressing it down on a table as well, if anyone can do this please this will give me more confidence if to go
and do an actual test flight.

Why i insist is I don't have much money to spend on new motors and esc's should this go wrong on the maiden, that will be curtains for my fpv career, so want to make sure as best i can that it will be ok.






Reply

Report Edit Delete

 

[wafflejock]

Hey yup have seen it plenty of times when bench testing myself too (mostly with betaflight but same behavior) hence not too worried. Other thing is just use the right stick at first after arming no throttle at all and in angle mode in some short grass or propped up on a box or the like on the grass just use the right stick after arming and see if it pitches forward as expected and left/right, if right stick everything moved as expected and goes back to level when you let go then props are on right, motors are spinning the right way and everything is good. If anything is bad very likely it will either flip or go the wrong direction when using only the right stick. If everything goes good with that then can give it enough throttle to get into the air and start trying to hold a hover. Crashes are inevitable and just a part of learning so don't worry too much about that the quads are easily fixable just want to build so things are less likely to break and disarm before you crash into the ground or other objects to save your props and overall reduce damage done.

So long as your disarm is working reliably you should be good but can also double check what happens (props off) when you just shut off transmitter while throttle is up (make sure failsafe cuts motors to 0 and disarm stops motors as well so can recover it without fighting a Tasmanian quad devil )

Would say big tip is try to fly about 20 ft away from yourself and keep it below tree top level at first, learning to control throttle along with pitch and roll will take some practice. Flite test has some good walkthroughs as does Joshua Bardwell on YouTube for getting started and some things to practice for both line of sight and fpv flying. If it starts to get too high up just disarm you may not like to watch it fall but if over grass it will be fine minus some grass stains and better to cut off early than let it get away. Basically just build trust for flying higher/further bit by bit don't start off with how high and fast does it go as soon as you get in the air

 

[spinzer]

ok thanks i will go and give it a try, in the back of my mind i'm worried about the cc3d's capabilities ( i don't mind it not being as snappy as newer FC's) but, will it be able to handle the thrust these esc's, motors will demand, (the compensation) will there possibly be desync at full throttle ?, or is this mainly esc firmware related (demag compensation)

Or is the cc3d FC cpu not fast enough to handle high power ? ? Or is it ? (4 cells 2650kv motors)

i'm thinking about getting a newer FC but then i will have to fly with a heavy 2200 mah 4 cell 45c battery so i would then not be able to spend my last amount on a good light battery.

 

[spinzer]

Ok i just returned from the maiden flight !

The drone shakes violently but flies perfectly level and responds to inputs perfectly.

I just could not manage a good landing because of the shakes.

Voltage dropped Down from 16.80 volts to 16.37 in 10 seconds.
The motors were warm just for those 15 seconds of flight but not hot.

Motors are fully tightened to the frame, so is the props.

Which settings would any of you suggest i change ? (should i go into the advanced view ) ?

My configuration are as follows:

 

[spinzer]

Here are the other settings that can be changed ?

 

[spinzer]

ok i have adjusted the p and i values i went as low a 6 on (i) and 3 on (p) at one stage down from 65 and 30 and thought i noticed a slight difference however at these low settings control response was almost non existent.
I'm starting to lose hope in this controller.

There is nothing wrong with it it worked perfectly on my previous setup with 3s batteries.

I guess modern motors and escs are just no meant to run on librepilot and cc3d

 

[wafflejock]

I think @Dugdog47 has more experience with cc3d on some modern stuff maybe but might want to wait for him to circle back around and give his 2 cents. Only CC3D librepilot stuff I ran was on my 3D printed frame (first "real" 5" quad) and was a wobbly mess too but I didn't know any better at the time so just learned to deal with it until I upgraded eventually. Now I basically run either the omnibus F4 based FCs or Kakute F7 is nice if need more UARTs or just have spare cash and using Betaflight 4+ on them (generally flash them with whatever is stable before configuration).

 

[wafflejock]

Newer versions of betaflight can use "bidirectional" d-shot if you have blheli32 ESCs and can flash them with newish firmware then it uses the one signal line from FC to ESC to both send throttle control and for the esc to report back real time rpm that it is achieving at a given moment. With that info the FC can be setup to filter out the motor shakes better by ignoring stuff around the frequency of the motor rpm in the gyro data (really want the gyro to just experience the overall motion of the quad not every little vibration from things on the quad but rather react to the wind and external forces).

Basically just saying I think it's worth making the switch even if it is a fair amount of soldering (get a good audio book or podcast and can get it done )

Also also also if you haven't already looked at soft mounting the FC or motors that can help physically take the vibrations out of the gyro on the FC

 

[spinzer]

Ok thanks wow that bidirectional setup sounds good. My esc's although quite newish only supports dshot 600 though and blhelS not '32"

I guess my only option then is to ditch my trusty old cc3d, never gave me 1 hint of a problem on my old setup.

I will have to use my heavier 2200 mah 4 cell packs as the FC is going to be expensive.
These are the 2 FC's in stock at my local hobbyshop (its the actual shop i will go to), which one would you suggest i get ?

Strix Binary F10 FC

RC EDGE - Strix Binary F10 FC

rcedge.co

Helio RC Spring AIO FC

RC EDGE - Helio RC Spring AIO FC

rcedge.co

It would still be great if Mr @Dugdog47 could give an opinion i will try send him a pm to see if he could perhaps give me last minute hope for the cc3d or i would need to go with one of the fc's above in the links ?

 

[Dugdog47]

Nah I've never messed with the cc3d boards, before my time. Now the naze f3 boards with afro 12 amp escs is probably what @waffljock is thinking of.

 

[wafflejock]

They both look pretty darn good from the mcu and imu perspectives but they both only list butterflight support which I'd be a little bit worried about (I expect they will get betaflight targets/configs setup eventually but would be good to find out if that's a thing yet or talk to support about when they plan to implement). I've heard good things about butterflight think mostly Alex vanover on rotor riot talking about it.

Ah that was probably it just mixed up my fpv people

 

[spinzer]

Ok thanks, i think those controllers support cleanflight too ? its all i can get so if it supports butterflight, should be much better than the cc3d and libre pilot problems ?

i know it closed proprierty software and i don't think betaflight will ever be supported because there was a fallout between them and betaflight from what i understand. Anyway these controllers are expensive so i would need to make the right choice.

 

[spinzer]

Wafflejock could you perhaps tell me if the new controllers support a pwm receiver ? i have a frsky V8FRII RX:

V8FR-II - FrSky - Lets you set the limits

www.frsky-rc.com

If not is there a way to wire it up to make it work ?

 

[wafflejock]

Sorry kinda useless here I haven't used those PWM output receivers with an FC so honestly could not tell you clearly on that one either, in general I've got Spektrum or FrSky receivers that send all the control signals on a single wire (SBUS or otherwise).

I have only used ones that output PWM directly to control servos and a motor on a fixed wing directly but if feeding the PWM signals into the FC would just need enough input lines that support the channels being sent over as PWM, from some googling it's looking like it's kind of FC specific:

https://www.reddit.com/r/Multicopter/comments/5wm079

I know for a fact with Arduino/AVR chips there are limited output pins that can generate a PWM signal (I believe all pins support analogRead and so can read the average voltage from a PWM signal but only certain lines can output PWM on Arduino/AVR processors). Didn't really pay much attention to price there but those do look like premium boards given the dual processor and newer IMU etc. I would be worried about the fall-out between the two (hadn't heard about that before this), but who knows which of the projects will continue to get the most support in the future.

TLDR, I think if it has enough RX pins (4) it can take in 4 channels of input from the receiver but personally like to go with the 1 wire protocols/receivers where all the data is lumped up onto a single line just makes wiring and configuration simple.