MPU-6050 angle readings under mechanical vibrations

Hello,
I'm in process of making a quadcopter. At this point I encountered a problem with MPU-6050 angle calculation.
When quadcopter is stationary with motors in off state and I'm tilting it, angle readings in all 3 axis is good, but when motors are in on state, readings starts to jump all over the place and can even differ from real value by 20 degrees.
I assume this effect is due to mechanical vibrations caused by motors. I include pictures of quadcopter and MPU-6050 board mounting also a graphs of angle readings of MPU-6050 in X axis using Kalman filter, complementary filter and MPU-6050 DMP (Kalman and complementary filter implementations as well as processing code for graphs is from Kristian Lauszus GitHub https://github.com/Lauszus , DMP used with Jeff Rowberg library from https://github.com/jrowberg).
Maybe someone had similar issues and could help me.

Martynas Jankūnas said:

I assume this effect is due to mechanical vibrations caused by motors.

Click to expand...

Did you try clamping the aircraft to a solid object, then take readings? That would quiet the mechanical vibrations.

Yes, but I didn't put much force or weight to hold quadcopter, because all its base is just a PCB and it's quite elastic, I don't want to damage the board.

Filtering could help. Possibly by adding capacitors. Either on the power supply, or on the incoming signal wire.

Also, for testing purposes, try powering your motors from a different supply.

I connected second MPU to different microcontroller, so it would not be connected electrically with a quadcopter and mounted MPU on board with two sponges from both sides of MPU. Results were pretty much the same. Then I put MPU between two copper clads and connected them to quadcopter ground, results was much better, max angle difference from real value was 7 degrees, not 20 like before. I will try to fully isolate MPU with tin foil next.