Could this be a hardware problem?
im just trying out a design for someting but its not working out. What am i missing? This is just one cut ran in the middle of the toolpath, cut from the same piece. There are consistent gaps on one side of the arch. Is this some sort of run out or miss calculations.?
Tool / machine deflection and/or backlash primarely.
going to try a 1/4 bit to see. Thanks
1/4" bit excert greater cutting forces (and larger opposite force) = greater deflection.
The harder you go the more will things flex.
As I dont own an Xcarve I cannot say much on how to improve rigidity, hopefully others may chime in 
yep, looks like your on the right track.
@TerryNicholl I created a model to hopefully explain what you’re seeing. By cutting on a path for both sides of the cut, you will never be able to get them to mate completely due to the kerf of the bit. If you want them to mate you need 2 cuts, one on the outside of the path and one on the inside of the path.
ok, ill try that
@TerryNicholl You are right, I was incorrect in my method of coming up with the 2nd cut.
Essentially you need to cut the curve into one piece, then use the void to define what the 2nd cut should be.
@TerryNicholl This easel project was made from the example file that I have been taking screenshots from. Both should need to be outside cuts in order for them to mesh up. Note - No tolerance has been added and these are size for size.
https://easel.inventables.com/projects/b7ExiDFAI198fbJBtS9TvA
ill try it
@TerryNicholl Did you run that from Easel or from your g-code sender?
from easel
@TerryNicholl I think there is something wrong with the imported DXF and it may be cutting on path instead of outside path (even though Easel is saying it is cutting outside of path).
OK, i just cut exactly what you had setup in easel.
@TerryNicholl Do you have a way of processing g-code from a step file?
If you do, here is a step file with 2 bodies in it. One for each side of the curve.
Curve Test.step (56.8 KB)
no sorry, ive got no way to open those files.
