Tried to mill an outline of Australia, but the route kept going off the rails.
The “View” simulation showed the correct result.
Is this a controller error (buffer overflow maybe) or a GCode error or Comms error?
How can I correct this? Slow the feedrate down? Any suggestions are appreciated.
I am using a:
32 bit ESP32 GRBL expansion board MKS DLC32 control card
Thanks
Is this the first time it went crazy and you have previously made good parts with MillMage? What version does it show at the upper left of MillMage (a screen snapshot would be ideal).
Can you upload the .mage design file so we can investigate?
Millmage has worked fine previously, although I had not done anything complex.
Just drag it into the Reply box. Or use the little arrow thingie at the top of the Reply box.
Mike,
the mage file.
Australia.mage (194.0 KB)
JohnS
try updating to this as it is the latest version
https://forum.lightburnsoftware.com/t/millmage-0-8-01-patch-release/188656
set up a new device and retry file , do it on the damaged piece so u dont waste more material 
just looked at your file ur ramp is 4xs faster than your feed rate , i run slower than feed with ramp
material looks like particle board, cabinetry type .i run mdf at 600mm/m with a 3mm chip board around 800 mm/m on particle board
why is your start depth 0.51 mm and not 0
i changed your operation setting and ran job ran sweet
Australia 2.mage (194.0 KB)
just my opinion , id do it with a 20 deg v bit will give you better detail on the finer points and curves
Thanks for jumping in! I just now got back to this.
If the file @wanderingtruckie provided does not cut like his, or like you expected, then [1] reinstall MillMage (no need to uninstall first) or [2] flash the controller board, in that order.
make sure you back up your bundles of your tools and operation library b4 new install sso you can import them
Terry,
thanks for the input. The material was a piece of scrap as I was verifying the operation initially. I am also a bit of a novice at this and am still getting a feeling for what numbers (rate etc) work best, and as such appreciated the feedback.
I will also try the latter version as suggested earlier. From your results, looks like it is all very feasible and promising.
I like the program and the simplicity it offers going from concept to machine (obviously a lot of complexity behind the scenes).
Thanks again for the help.
JohnS
ur welcome , im still a novice also , yes MM is an awesome program
