Just to check, which firmware version are you using?
I opted for the revo-mini-i2c fw 4.1.7.
Had my oscilloscope on it, but I am not an expert and I could not see any irregularity there …
You did see the yt video. Just swapped the servos and it looks ok now. There is still some movement when powering on, but not extreme. Strange.
It seems to me it’s different from my trouble. On my plane it’s normal, that during FC boot servos go once to extreme positions and then back and I don’t consider it a problem.
My problem was that after about a minute from boot of servos being calm when the GPS locked satelittes, servos started crazy dancing like in an epileptic fit and it kept moving on until I power off the plane or start RC transmitter.
Can you provide a log (set LOG_DISARMED = 1) and tell me what hardware you are using?
Who, rptacek or me?
Doesn’t it make more sense?. ‘First the radio on and then the rest, after flying, first the plane power off and then the radio’.
This topic I created when I was testing openHD with RC control transmission over computers and I needed often to have plane ON while the rest was booting/rebooting. Now I mostly keep RC on all the time the plane is on, but often when I’m just testing something, I don’t need RC at the time and have to keep it ON just because these movements.
Solution of this would be nice for some evaluation or testing, but it’s not critical.
My fault. Wiring error. 
On my test board I am powering motor and both servos with the speed controller w/BEC.
The other electronic components are supplied by a separate buck converter.
I measure the continuity of GND in both lines (from main switch over the speed controller to motor/servos and from main switch over the buck converter to fc and rest) and I lightheaded omitted a GND connection after controller and after converter.
Adding that GND connection and no extreme movement any more.
But that those different servo brands react so differently astonishes me a bit …
br KH
Just to clarify, you’re talking about connecting the output ground of the BEC in the ESC to the servo output ground of the flight controller?
As long as there is ground continuity between input and output ground of the BEC and input and servo output ground of the flight controller, they shouldn’t need to be connected together since they should all be essentially equivalent to the negative battery terminal.
It seems odd that there’s enough noise being injected that it would substantially mess with a PWM signal (which is essentially a full 5V signal).
It was amazing to see, that the HITEC servos showed a strong reaction and the HXT900 nearly did not show any …