so i finally got around to replacing the front left motor on my 5inch after i clipped a tree pretty hard last month. ordered the same exact motors, xing2 2207 1800kv, swapped it out and everything looked fine. props balanced, motor spins freely by hand, no visible damage to the frame anywhere else.

took it out yesterday and now theres this low frequency vibration i can feel through the entire frame that wasnt there before. its not terrible but its showing up on my blackbox logs as noise in the gyro data and my video is definitely softer than it used to be. ive tried rebalancing the props on that arm like three times, tried a different set of props entirely, same thing.

im wondering if maybe the replacement motor has a slightly bent shaft that i cant feel by hand? or could i have done something to the arm when i crashed and didnt notice? the arm looks straight to me but i guess it could be slightly off. has anyone dealt with this before where everything looks fine but something is just... off after a crash

bent shaft is almost definitely your problem. you cant always feel it by hand especially if its only like 0.1-0.2mm off center, but at 40k rpm that tiny wobble multiplies into a pretty nasty vibration. easiest way to check is to put a prop on and slowly spin it by hand while looking at the tip of the blade edge-on ? even a barely perceptible wobble will be obvious doing that. also check if the motor mount holes on that arm are still perfectly perpendicular, sometimes the arm takes a hit and the mount face gets slightly angled and thats enough to throw everything off. i had an arm on my hyperlite do this and the arm looked totally fine visually but with a straight edge you could see it was maybe 1-2 degrees off.

have you run rpm filters since the swap? sometimes after a crash the resonant frequencies of the frame shift a bit if something got cracked internally even hairline cracks can change how the frame behaves and your filters might need to be retuned to match. i know its a pain but running a quick rpm filter analysis in betaflight might at least tell you where the noise is coming from frequency wise