I have spent days trying to sort out a severe servo jitter problem and am fairly confident it is an issue with the SKR2 board.
The problem is that the servo would jitter badly every time the heatbed switched on (during the switching event). This happened with PWM control and with BANG BANG using Marlin 2.0x bugfix.
I have a servo externally powered by a buck converter off the 24V supply to provide 5V for the servo. I have the servo signal from SERVO / PE6 / PE8 - I have tested each.
After eliminating many possibilities, I finally discovered that the problem ONLY occurs when using an external power supply. Investigating this, I discovered huge spikes on the SKR2 board's ground rail (with respect to the power_ground) when the heatbed switches on and these did NOT seem to occur when hotends heaters were switching.
This image shows the problem.
![20210612_081451 1](https://user-images.githubusercontent.com/20292325/121754016-e4222100-cb56-11eb-8c3d-cb3575894f9b.jpg)
The GND reference for the scope was taken from the ground pin on TH1 connector (for convenience).
The BLUE trace is the heatbed which can be seen turning on. (I am using a 24V PSU)
The YELLOW trace is the PWR ground at input to the SK2 board.
I can confirm that powering the board from either the USB or the PSU does NOT alter the situation.
Note that in this situation, the heatbed MOSFET is only switching the control signal for an external solid state relay. I originally was directly switching the heatbed power and I thought that using an SSR would alleviate the problem - but this was not the case.
The saving grace is that the onboard 5V rail seems to have enough capacity to supply the servo directly. When testing in this configuration, the symptoms are resolved.