[jpedrofer]

Hello everyone!
I'm relatively new to Milling and I'm having trouble milling round holes in even small soft woods. Holes are coming out oval alongated along the Y axis(the motors are the same, both l). The path generation is done in Fusion360 and executed in a generic mill using Mach3. Somehow only small holes are having trouble (bigger than the cutter yet). Positioning is perfect, when i drilled 2 holes the distance appeared to be exact, but in cutting, it doesnt seem accurate at all. Any test I could run to measure accuracy? Any procedure I might be doing wrong?
Thank you!
I'll post pictures of the result.

[peteeng]

Hi JP - two things to do 1) run a very shallow pocket and measure it, you can use say a 10mm drill as a pin in the hole to check 2) check you axis calibration, generally its your axis calibration that needs adjusting. If the calibration is correct then you have to start looking for mechanical issues. If the light cut is correct then do a deeper one until you get a hickup. This shows the machine has some issue mechanically... Peter

[routalot]

The OP mentions milling and I would expect mills to have leadscrews or ballscrews-neither of which should permit too much backlash and that which may be present can be compensated for in some control systems.If on the other hand he has a router with belt drive,there could be a mechanical issue behind the problem.Its also possible that the problem is present with larger holes but because of the gentler curvature it may be less visible.Perhaps cutting the largest hole that can be measured accurately would determine if this is the case.

[dazp1976]

If it's a home built machine with single nut 1605 ballscrews then it's nothing new.
Calibrate axis first (moving in the same direction every time for adjusting / measuring). Watch a few videos on it.
Second. You WILL have backlash in a machine home build.

I've had nearly 0.1mm in in a mini mill before which was a combination of wobbly ballnuts and Y carriage lift/twist. Circles were awful and I never fixed it.
A double nut unit would have solved it but it wouldn't fit well.

[jpedrofer]

Hello everyone, thank you for your replies. I will definitely run axis calibration today and do some testing. As far as i could tell, x and y were leadscrews.
I'll update as soon as I can.
Thanks!

[dazp1976]

Hello everyone, thank you for your replies. I will definitely run axis calibration today and do some testing. As far as i could tell, x and y were leadscrews.
I'll update as soon as I can.
Thanks!

'Leadscrews'. Not good.

[jpedrofer]

'Leadscrews'. Not good.

To be honest I'm not entirely sure. They could still be ballscrews, i forgot to check that yesterday

[jpedrofer]

I did some calibration and it turns out X axis was perfect and Y was a few centimeters off. But even after calibration, the circles were coming out pretty bad. I tried to compensate the Y axis until I could get the best possible circle, but that meant the Y axis would come out allongated with everything else. Could it be that the motor current is low? Or maybe they are too weak, all 3 motors are NEMA 23 with 3 N.M. I guess it is lacking power to do the hardwork since its almost perfect in moving outside the workpiece.

[jpedrofer]

Best results I could get

[routalot]

How about a picture of the machine?It might be flexing rather than calibration.

[jpedrofer]

How about a picture of the machine?It might be flexing rather than calibration.

I'll take a picture today! But with the research i've done this weekend I'm starting to believe the machine probably has a lot of backlash