I have read a lot of posts about circle cutting - no circle but oval, and checked and tried many suggestions but no success.

My machine still doesn’t cut correct circle.

1.  X and Y backlash les than 0,03mm
2.  Steps per unit equal to both axis - the same ratio between AC servo motor and ball screw
3. Following error on both axis during circle cutting near 0 - there are 750W AC servos on light machine - over dimensioned motors.
4. Acceleration/deceleration equal to both axis.
5. When cutting square 100 x 100 mm, both dimensions are in +/- 0,02mm
6. Circle dimension in X and Y axis is also 100mm +/-0,02mm (when cutting circle 100mm)
7. The difference in diameter measured in 45° and - 45° is 0,6mm
8. By cutting circle in CCW or CW direction, diameter difference in 45°and -45° remain the same but vice-versa.
9 Cutting speed doesn’t change result - cutting 150mm/min or 1500 mm/min. gives the same result.
10. I'm using LPT BB

I have no idea what else I can do to solve this issue.
Any suggestion appreciated…

Please post your problem Gcode file.

It's a simple code with IJ:

M3 S12000
G17
G0 X53 Y0
Z2
G1 Z-2 F800
G3 X-53 I-53 J0
G3 X53 I53 J0
G00 Z20
M5
M30

Also with R I have same issue:

M3 S12000
G17
G0 X24 Y0
Z2
G1 Z-4 F1000
G2 X-24 R24
G2 X24 R24
G00 Z20
M5
M30


Cutter is D6

Thanks

There is nothing wrong with the code, maybe it should have a G90 in it but it will still work without.

My guess is something is loose or slipping on a shaft if your steps are set to the correct values.

Yeah, that would probably be most likely, but I checked everything and there’s nothing like that.
The beginning and end of the circle are also completely equal - there is no visible transition at all.
I also tried repeating the code:
0.1
-0.1
0.1
.
.
.
on both axes and when measured with a measuring indicator, it shows a movement of 0.17 mm on both axes (only 0.03 mm difference)

Even after prolonged milling, the machine remains in the same position. If I repeat the program, it goes exactly the same path - so even after one hour of milling, nothing slips.
Even if I push the X or Y axis with my hands in one  or another direction while milling the circle , the shape of the circle itself does not change.

Tomorrow I will try to cut a circle programmed from short lines (g1 only). I will report what the result will be ...

It may be a good idea to run the same circle with a feed of 100mm/min and see if that makes it round.  This may be a CV issue.

Hi,
have you measured how square the X and Y axes are to each other?. Only a very small angular deviation from
90⁰ will have a significant effect on a circular interpolation path.

Craig