Milling a pocket is not a continuous move?

Longbow wrote:

I wonder if the problem lies in the way that nativeCAM builds up the program code (g-code) subrutines?

Hi

I think the problem is not with NativeCAM.

I tested a normal circle and velocity does drop some times then return back to normal.
I tested again with G61 and G61.1 with same result.
I also tested with G64 commented from default with still same result.

I wrote gcode for a simple circle and added it to custom gcode after deactivating previous circle with the same result :


Here is the code with comment where feed drops :
This code could be hand-writen by anyone and I hope someone can test it and tell us about the result he gets.

I do not know why it does this but will keep investigating.

Regards
Fernand

I'm not a good gcode reader but it sure seem like the TP thinks there are sharp corners (in Z?)
Heres with g64 p0.001 q0.001:



and with g64 p0.005 q.005:

Except bug or not, this is a mix of helical and circular moves so would be expected to slow down
at the junctions (and that's just what you see in the first plot, 5 slowdowns at the circle/helix junctions)

There may be another bug but it does not show up on my test setup (inch, 20 IPS/S accel F100)

I tried with g64 p0.001 q0.001 and it slows to a stop 3 times during a circle (the half circle where Z axis is ramping down)

I tried without G64 and it slows down to a stop only once - when the Z axis has finished ramping down to the new Z step.

Could you share your ini file? The only slow downs I see are the expected ones at the Z acceleration/deceleration points.

What if you tried the move as a true full helix, rather than half helical and regular arcs?

I tried your Gcode and it moves in a helix with no pauses! - I just had to add a feedrate because there was none specified.

Here is a video: ufile.io/mqaz8
(I modified the code a little to mill a larger circle)

Here is a video of a circle milling using nativecam code:
ufile.io/jvhdn