First, I would measure the peak to peak radial distance of the aberration. Then in order to help, we would need some info:
Pin roller diameter
If the roller diameter * pi is equivalent to the period of the aberration, then
that can be a smoking gun.
Is the motor amplifier sinusoidally commutated?
This would tell if it is torque ripple
What amplifiers are used?
What is the resolution of the linear encoder in both cross-slide and bed?
Does it do the same thing every time?
Are you able to look at a plot of the bed axis encoder position while the cut is being made? Do you see this period on the encoder plot?
If you push on the carriage away from the spindle with something near the cutting force, do you feel the servo loop vibrate?
Mike Visit my projects blog at: http://mikeeverman.com/
http://www.bell-evermannews.com/ http://www.bell-everman.com