I learned a big lesson this weekend. And this thread was the answer i was looking for.
I have had trouble with small features till now. I just ignored it and compensated elsewhere including hand finishing. It was getting annoying having to widen joinery holes and such.
So with the info from this thread i decided to do a test cut to examine the width of cut. Turns out my issue was the same as the OP in that my tool width in CAD was larger than the actual cut.
After currecting the tool width in CAD, my small features are much closer to the design.
Thanks.
2 Likes
If you have calibrated your machine it may be your bits are not true to specified size. Measure them they are often off by .005 inches and in some cases more. Out the actual bit dimension in not the factory specified dimension
There are three reasons for inaccuracy after the machine is proper and calibrated:
1 - Actual tool diameter vs spec. (3mm tool may actually be 2.9mm)
2 - Run out, the spindle do not run 100% true, a 2.9mm bit may cut a 3.1mm slot (run out = 0,2mm)
3 - Tool deflection when subjected to sideways (cutting) forces.
Also having this problem with small verses bigger circles. The bigger circles actually provide a better tolerance than the smaller ones. I am currently try8ng the addition of finishing passesin case of tool deflection. Are you seeing any taper to your cuts? I have seen a .004" difference from top to bottom of a 0.5"dp pocket with no finishing passes.
Also noticed in the picture above that you have issues in the corners unless you have it programmed this way? When you hit the end of the machining pass it is like the cutter has a moment to relax and clean the surface better so you get a slight bit of undercut? Are you climb milling or conventional milling?
Iām curiousā¦ why would our steps/mm be different ā I think weāre dealing with the same NEMA 23 motors, with the same distances between the pulley and the trackā¦
Could the feed rate contribute to #3? Iām not a good judge of torque values, but is it possible to have the stepper motors āmissā steps if the feed rate is too aggressive?
For example, I ran a 2 flute spiral upcut .25in bit at 50in/s, and ended up with 1/16th" error on 3/4in holes. Could this feed rate have been too aggressive for the stepper motors on the X-carve?
A stepper motor could in theory miss steps any time it looses sync with the controller.
Example, try setting your max feed rate to 10x your current value => the motor will most likely just whine as it is unable to keep up with what its told to do.
The harder your bit is working the larger the sideway forces will be, it will deflect and ultimately break.
So if tool deflection is the cause inside pockets will usually end up being undersized and outside profiling oversize.
Most stepper motors are 1,8deg (200 full steps per revolution)
The driver can increase each full step into incremental partial steps like half step (1/2), quarter (1/4), eight step (1/8) etc. This is referred to micro stepping.
So I can have the same ACME, same stepper motor but I may have my motor driver set to half step and you 1/8.
I need 200 x 2 to make my motor turn one revolution, you on the other hand would require 1600.
If however we both had exact same settings, but my value would be 49.8 and you calibrated yours to 50.2 the difference will be due to measuring inaccuracy and/or system back lash (positional change during transition from one direction the opposite direction.
Belt tension can also play into accuracy. Since the belt is how the stepper power is transmitted, it can affect it.
