Hello Repetier
Thanks for the fast response.
I use 24V for the motors.
I did some further investigation with this printed octagon:
Octagon
X-Side
Y-Side
X-Y-Side
X-Side with measurement
X-Y-Side with measurement
This part is 10 mm high, print settings are 0.2 mm layer height, 3 perimeters and 25 % infill.
Every 3 mm, the printing speed has changed:
0-3 mm: Default 66 mm/s, Outline Underspeed 40 % (26,6 mm/s)
3-6 mm: Default 100 mm/s, Outline Underspeed 40% (40 mm/s)
6-10 mm: Default 133 mm/s, Outline Underspeed 40% (53,3 mm/s)
(I use Simplify3d for slicing, so the "outline underspeed" means that the outermost perimeter is printed with 40% of the default speed, the middle perimeter is printed at 60% and the innermost perimeter is printed at 80%. )
As you can see, the distance between the waves is longer at the higher printing speed. To be exact, a double printspeed results in a double wave length. (maybe this isn´t clearly visible at the photos, but I can definitely see it here)
In my opinion, this shows that the waves are generated at a fixed time interval, so only the Due or the firmware can be responsible for it.
The linear bearings I use are running absolutely free nearly without any friction or slip-stick-effect. The belts also can´t cause this problem because if so, the wave pattern would have a fixed wavelength, which isn´t the case.
Any further ideas?