Flipping quad help?

[Mi96ke]

I've built a few quads so far so know the basics,
I've seen a few and now I've come across the same issue
Everything seems set up and working fine however as it takes off it just flips over.
Seems like the pids but I'm a bit out of practice with them and could use some help.
Using pwm on an spf3 acro board, dshot 300 escs.
Holding the quad I can feel that as I push left and right under higher throttle, that side just keeps pulling harder, doesn't correct.
I was thinking either pids or could pwm and dshot cause any issue? Not sure on that part
Thanks

[hoverfly]

First thing to do is check motor  order and f/c orientation..

[ched]

Have you run a calibration on gyros? I assume you are using Betaflight?

[Bajadre]

Prob already checked but make sure props on right way 

[Mi96ke]

Motor order and orientation are fine, it all acts as it should with low throttle, but as it starts to lift it oscillates like crazy. Was always gonna get the prop suggestion on this post 😂 is calibrating the gyros different to the accelerometer?    Thanks

[ched]

I may be wrong but I think gyros are to do with level and accelerometer are to do with speed of movement.
i.e. The quad needs to know when it is level.

Another thing to check is that the FC is in the correct orientation i.e. does the model in BF move the same way as your quad when you move it?

I assume all your tx/rx sticks are correct direction and dont need reversing?

[Mi96ke]

Well I've put it on a level surface and calibrated the Accel. All orientations are good, sticks are fine, as I said it all seems fine and normal until take off, seems like the motors keep pushing when they should back off

[ched]

I know it may sound silly but do a cli diff dump and save the config. Then reflash the fc, and reload the config.

I assume all motors feel smooth, no continuity between wires and motor bolts?

You are running default PIDs are you? Any way to double check the current PIDs with others that have a similar quad?

[Cheredanine]

Motor order and orientation are fine, it all acts as it should with low throttle, but as it starts to lift it oscillates like crazy. Was always gonna get the prop suggestion on this post 😂 is calibrating the gyros different to the accelerometer?    Thanks

So it lifts off but shakes like mad?

Unlikely to be PIDs, stock pids are not that far out.
It may be useful to list all parts if you can.

The behaviour you are describing is typical of something lose physically, either the motors, props or the flight control stack (can you take a photo of the stack?)
Other stuff that could potentially cause:
Wires carrying power too close to the gyro on the fc (it would have to be pretty much touching).
Motor screws too long - touching coil - isnthere any heat in the motors?
Potentially unbalanced motors or props (or damaged) but really modern fc software should handle that
Damaged gyro

May be worth reflashing the fc

Gyro and acc are on the same chip but are calibrated separately, however both are calibrated on boot
Gyro measures change in aspect, acc measures force and is only used in Self levelling modes (angle or horizon)

Some questions/stuff-
1.please provide a screen shot of your filter settings on the pids tab and confirm if you are using dynamic filter (config tab)
Blackbox log might be useful
Are you flying in acro, angle or horizon mode?
As above please confirm if you are using air mode and what the build is

[aburn09]

Might help ?
https://youtu.be/X4746_KLVEI

[Andy7]

How about a damaged sensor? I know they can be damaged from too much of a knock.
Wonder if it's possible to see sensor damage with the configurator trace if you tilt and shake the quad?

[Cheredanine]

How about a damaged sensor? I know they can be damaged from too much of a knock.
Wonder if it's possible to see sensor damage with the configurator trace if you tilt and shake the quad?

It is pretty obvious from gyro traces in a bb log, had a damaged gyro once, clean takeoff, mid flight one axis swings without rc inout and leads motor traces, axial movement is opposite of what quad actually did (because the fc commands the motors to fight the non existent rotation

[Mi96ke]

Thanks for your help guys, I think I got it now, reflashed the FC this time the pitch and roll were showing inverted, so I offset them by 180, just going out to test it now 👍

[Saleem]

Be sure to set and test failsafe if your tx/rx has it.