Vibrations in circular interpolation

I'll join to the backlash idea. The vibrations seems to be at that 45-ish degrees position when the axis switch from pushing one way to pushing the other way.

I would try a conventional cutter pass that way you will be pushing on the screw and see if the vibration goes away

The conventional milling pass has been a great idea, but unfortunately the result has been the same vibration in the same place.
Today I don't have time to disassemble the axis screw bearings, but tomorrow I can check it.
Thanks to all for the help!

Today I disassembled the screw bearings of the X and Y axes. Everything seemed correct, I assembled and adjusted it with a little more tightening ... I checked it with a gauge, 0 backlash.
The machining test has been the same vibration, but this disappears if in the middle of the arc, when it is already in the direction of the curve, I tighten the axis screw with the hand trying to stop it, it is indifferent if I press the X or Y screw, the vibration disappears (this confused me)
It is the same as cutting a hard or soft material like PVC, it vibrates the same.
I am using an endmill with 3 teeth, maybe try more teeth to see if it induces less pounding.
I have not yet calibrated the measurements in the movements of the machine, this should not affect this at all, right?
Thanks!!!

If pushing on the table affects the vibration I would look at your linear guideway.

Another thing to check, see if the defect changes shape or position if it is milled at different positions on the table, (front right, front left, right rear and left rear. (That would point the finger directly at the table ways.)

Thanks for your help, it will take me a few days to try what you propose.
I will tell you.
Thank you very much to all.

I have the exact same problem with the closed loop Leadshine drives and I think it's down to resonance.

If I cut a large circle say 160mm diameter then at the North, South, East and West ponts ie where one stepper is running slow and the other fast its fine. On the diagonals I get the same chatter as you. I think this is because when cutting at that speed both the X and Y steppers are in their harmonic danger range. These two frequencies combine and amplify each other (I think)

I ran my program without the spindle on in the air and I can hear and see the vibration caused at these points.

I'm trying tuned mass dampers though with out the tuning to try and overcome this. People have had success with 'rattlers' mass dampers .

I have emailed Leadshine about the problem two days ago but they have not yet come back to me.

In the mean time I am adding rattlers and other vibration suppression devices and strengthening the motor mounts to make them super rigid.

Try just a large circle without the spindle and listen and look for vibration.

Jools