Looks good, the backlash seems to be working ok on the Z otherwise you would have dug into the paper with the pen and tracking ok in the
x & y.
What F did you run it at?
Did you try cv and exact stop?
Actually i modified the PEN a little .. The pen comes with spring and on One click the point comes out and on the other click the tip get inside.. I modified it a little and it worked very fine as you have seen in the picture..
I ran the Gcode with F10 and i used CV mode On...
After running the program did you bring the Z down and see if it returned to the starting point? If it ran well at say one F, just double the feedrate and
run it again since you returned to the starting point, see if it tracks the same line and returns to the center starting point. At some F she will skip
and then you will know the max velocity you can run at. Then limit it to say 50 or 60 % of ( the 50 to 40 % reduction allows for torque required for cutting).
Yes the Program return to its exact start point..and the Z when commanded to go to zero, it exactly went to the zero location... I imported yours Gcode into NCPlot and exported in DXF... I open the DXF in AutoCAD and measured some dimensions and compared it with the pen-tracing... I was within 0.07mm range.. Amazing:)... As it was on the Paper so measurement was not easy by Vernier calipre;)
I just tried if with F10, tomorrow i will increase the velocity and will show you some more results:)... Lets say F10 is my maximum velocity then i have to keep my rapid F5 and keep the cutting 4IPM....Am I right.. My Stepper motors are for 80Volts rating with 4.2Amp per phase..Currently i am only supplying 36Volts DC.. I have connections of 48Volts and 60Volts too.. I will try increasing the voltage...
It is now clear that My Y-axis i.e. the bottom carriage , loosing steps.. This is because it carries the load of X-axis carriage+Workpiece load+Cutting forces.. It is boxway design so having frictional forces too (although i have appled Oil too)...
Then you can use a end mill to do some cutting in some scrap piece experimenting with some cut depths.
After playing around you will have some level of confidence on the F / max velocity you should be using along with anticipated accuracy.
I will sure do so.. I am planning to use 1mm DOC for Aluminum... My Z-axis has some play, in the impeller case the Gcode was 2mm per pass and the first pass of the impeller was 3.5mm.. That was too much for the cutter..... so i will use shallow cut in future...
---------------------
Your original program had short straight line segments instead of the arcs.That can create a problem for YOU depending on the accel values
and mode you are running ( cv or exact stop ).
Yes, today i figured out how to output the gcode in G02/G03 IJK mode.. so in future i will make my Gcode more clean.. I am also planning to Use EXACT STOP MODE for milling... I hope that for better usage of Backlash compensation i should not use CONSTANT VELOCITY MODE...
Let me comment that in your program you clean out the bulk of the material in the first area and then finish the first vane of the impeller , then
the others go on to profile the vanes only. Do as you wish, but, i would have cleared out all the bulk and then profiled the impellers.
Like said before, wood is different than cutting metal, so fool around to gain experince with YOUR machine.
"SPEED WILL COME UNBIDDEN ONCE ACCURACY IS ATTAINED"
RICH
The Gcode for the above impeller was made in ARTCAM.. and I know artcam is good for wax/wood but not for metal... I will use VisualMill for my next Gcode generation projects... I will follow your advice and will post you some good results..
Thanks for being with me and helping me:)
ONE QUESTION:
If my motors do not loose steps.. Then will the Backlash compensation will be OKAY for heavy Cuts?.. I read somewhere that Backlash compensation do not work for heavy cut.. Please enlighten me:)