My two largest quads, a 960 class Tarot X4 and a 690 class (modified) Holybro X500-V2, both exhibit significant “wobble” before becoming airborne.
It’s the sort of wobble that makes you want to shut down because it seems something is physically very wrong.
An interesting note - On the X4, when I take off in Altitude Hold, the wobble is less than when I take off in Loiter.
A fellow in the community suggested that the root cause of this behavior stems with the PID’s - and how they’re operating. That was the first time I’d heard that suggestion.
I’ve read the other related “wobble” related posts - and can’t seem to find information that would be helpful.
I don’t have any suggestions, but I am curious about what motors and props you’re using on the X500V2? We use the X500 for student projects and I have seen a few of them get some crazy shakes just before take off. It’s not consistent though. Some get it, some don’t. And those that get it, don’t get it all the time. I have theories, but nothing definitive.
I’ve seen this, also, and I don’t have a root cause, though I suspect it’s the control loop attempting to correct attitude on a slightly flexible frame that isn’t airborne.
My remedy is not to nurse the throttle to get airborne. Stab it quickly and smoothly and limit time on the ground with the props spinning above MOT_SPIN_ARM.
But it does support one of my ideas, that goes along with what @Yuri_Rage mentioned. I wonder if the landing gear on these allows some sway or oscillation that messes with the PID loop. That would also explain why some machines get it, and others don’t. Slight differences in assembly may cancel or accentuate the issue. Those big arms on that frame would only amplify the situation. I’m seeing the issue with 10" props.
BTW - I extended the two horizontal pieces (where the foam tubes reside) a few inches - to help insure the copter is less likely to tip over on landing.
If it breaks - I’ll see about fabricating something better.
In my opinion, the thing that makes this such a good quad is the thickness of the top and bottom plates. And a bit to my surprise, the plastic arm joints handle the extra load OK.
I’m also using a 4-in-1 esc instead of a PDB - running BLHeli_32 and bi-directional Dshot. Due to space constraints I’m using a CubePilot Mini Power Brick instead of my preferred MAUCH current sensor and BEC. But it does the job OK.
I don’t know if this would help - but I almost always swap out those plastic T-Joints with an aluminum equivalent for better resilience on landing (have had them crack a lot with newer pilots).
If the holes line up - I would also recommend these instead:
I think that one just needs more tuning before go further down the wobbling/oscillations rabbit hole.
What I mean is those issues could be solved by following the normal processes, since it’s a fresh build and you are on defaults. There are suggestions in one of your other threads.