Machsupport Forum
Mach Discussion => General Mach Discussion => Topic started by: dreamstar156 on March 25, 2010, 09:14:32 PM
-
Hello all,
I recently finished my homemade CNC 3-axis mill and have been experimenting with Mach 3 with some success. However, when I program a circle or an arc with simple G2 or G3 code, the result is not round but rather ovalish in shape. I did the axis calibration for a second time but the results were similar... Does it mean, I am not calibrating the axis properly, or is there anything else in the configurations that could cause that?
thanks in advance
M.
-
It could be your configuration process, it could be backlash, it could also be your code. If you post your code it would be easy to rule that one out.
-
M.,
If the steps per unit are correct then it is probably a mechanical problem.
I don't know the extents of out of round you are experiencing. Here is a example for thought:
A rolled ball screw is used and it has a little more gain in one spot as compared to how the steps per units were
set based on say a long move....thus you get a little more inaccuracy from the ball screw, maybe a little backlash, maybe a little inaccuracy
in belts , table movement at spot, etc and summing al it up creates some ovality. So it is a matter do degree of accuracy of the whole machine.
RICH
-
Thank you guys.
The test program is:
%
G0 G17 G20 G40 G49 G80
G91 G28 Z0
T1M6
G0 G90 G54 X-0.5 YO S5500 M3
G43 H1 Z1.
Z.1
G1 X0 Z-.1 F10.
X0.5
G3 X0.5 I0.5 J0
G1 X0
X-0.5
G0 Z1.
G91 G28 Z0
M30
%
What is the best way to check for backlash?
Thanks
Max
-
What is the best way to check for backlash
Move the axis in one direction some amount, then put an indicator to measure the next move in the same direction ( say 0.500"), now move the axis
in the opposite direction. The indicator should show a travel of 0.500", if it reads 0.498" then the difference is a backlash of 0.002".
The non-movement on change of direction is backlash and can be due to thrust bearing play or the ball nut.
Do the same for the other axis as well.
RICH