Milling precision
I recently finished my CNC mill conversion. Everything seems to be working great except precision. I always get a +0.12 mm difference. (exactly that)
I have done several tests and I am pretty sure I'm not missing steps.
I've measured backlash and corrected it and it works perfect.
I've tested several endmills and triple checked every g-code program.
But when milling anything I always get this error.
I am runnign a hobbycnc board and 200 oz steppers. microstepping 2
Does anybody have an idea?
Thanks!
PS: I apologize if my english is not that good.
Please Log in or Create an account to join the conversation.
Start with the simple things. I you sure the machine is calibrated correctly? What type of error do you get without the backlash? Besides backlash is there mechanical flex? Do you get the same error when making a light cut as a heavy cut? Same problem with a slow cut as a fast cut? A ight finish cut?
Rick G
Please Log in or Create an account to join the conversation.
I believe the machine is calibrated correctly . There is no mechanical flex.
Backlash was determined jogging with a digital dial indicator. Both axis have preloaded nuts, Y axis has no backlash and X axis 0.07mm. When compensated backlash was eliminated while jogging. Made several tests and no steps were lost either.
With backlash compensation milling error is exactly the same for both axes (0.12mm). If backlash is set to 0 for both axis, the error in X axis is lower (0.05mm)
Al tests where made with light finish cut. I just performed a heavy cut and the error was 0.20mm. I will test difference in speed tomorrow.
Thanks,
Matias
Please Log in or Create an account to join the conversation.
If the Y axis has no backlash leave the backlash for Y at 0.
If the light cut and the heavy cut give different results it seems something is flexing or you have lost some steps..Seems mechanical. Could be end mill flex too.
Try a heavy cut, measure then run a second time, better results?
If the error in x is less without backlash then reduce the backlash.
Rick G
Please Log in or Create an account to join the conversation.
It was with a gantry, driven by a lead screw on the one side only. When under load, the gantry would twist slightly for a short period of time and then spring back to it's normal orientation. In my case, increasing the pressure on the bearings towards the rail solved the problem.
Please Log in or Create an account to join the conversation.
If so, you should be aware that both the workpiece and the cutter deflect under load.
So, if machining an internal cavity, it will always come out undersize, if machining the
exterior it will come out larger. You can reduce this by making a rough cutting pass to make
the part to, perhaps, .25mm within the desired size, and then making the last pass at
a slower feedrate, and removing that last .25mm of material.
You can also make the machine move certain amounts and come to rest on a dial indicator,
to see if approaching from different directions affects the position. You can also apply
force manually while the dial indicator is showing position. This allows you to see the stiffness
(or lack thereof) of you machine structure. If your machine is much lighter than a Bridgeport
milling machine, then the stiffness is going to be less, too.
Jon
Please Log in or Create an account to join the conversation.
OK, you also need to know that software backlash correction is a very poor method to fix backlash.Hi Rick,
I believe the machine is calibrated correctly . There is no mechanical flex.
Backlash was determined jogging with a digital dial indicator. Both axis have preloaded nuts, Y axis has no backlash and X axis 0.07mm. When compensated backlash was eliminated while jogging. Made several tests and no steps were lost either.
With backlash compensation milling error is exactly the same for both axes (0.12mm). If backlash is set to 0 for both axis, the error in X axis is lower (0.05mm)
Backlash means that the motor cannot exactly control the position of the machine, it is free to be
moved by external forces over some distance. The backlash compensation only corrects a zero-force
move to arrive at the desired position, but it has no way to correct for external forces moving the machine to one side or the other of the backlash.
However, your measured error was .12mm, greater than the measured backlash of .07mm
So, I think that indicates deflection may well be the larger problem.
If your parts have nearly as much error in Y as in the X direction, then backlash has little to do with it.
Jon
Please Log in or Create an account to join the conversation.